PDA

View Full Version : E-6 to C-41: A Transition



rguinter
8-Mar-2011, 10:57
Greetings and a thread starter:

I'm looking for suggestions.

After TPI in Fairfield, NJ closed its doors at the beginning of this year, I lost the last available lab in my normal travel area that processed E-6 films.

So now I'm in the process of switching over to C-41 films for most of my photo work that I can still get processed locally. I like the idea of being able to hand deliver and pick-up my work, paying out-of-pocket, without having to resort to mailing and charge-cards.

Now I've always been an E-6 film fan and I have my favorites. Many of them are discontinued but I've managed to stockpile quite a supply so that is not an immediate issue.

The problem is that no matter what C-41 films I've used to date, I simply don't seem to get the same results at sunset and after dark with these films that I get with my favorite transparencies. And yes I've tinkered with all the colors, saturation, histogram, brightness, and other controls that are available to me with my Epson 4990 scanning software.

To now I've tried a lot of different C-41 films including Ektar 100 and Portra VS but still no real comparison to RVP, RDP, RTP, and Ektachrome transparency films I've used extensively for so long. I've also tried cross-processing a few and had some moderate success with CDU II which give me some nice sunset colors.

So, I'll post a few examples later but for now I'd like to hear others' suggestions about their success with C-41 films during those witching hours of sunrise and sunset.

Bob G.

Juergen Sattler
8-Mar-2011, 11:13
Sounds like you like very saturated colors. Kodak Ektar 100 is a good place to start. After scanning (with minimum adjustments in the scanner software) I would then treat your scans in Photoshop or similar software - you should be able to get very close to what you are looking for - there are even actions in Photoshop that allow you to simulate various E-6 looks. I believe Nik software is offering a bunch of those.

David de Gruyl
8-Mar-2011, 11:17
I guess Taylor (Princeton) is not in your travel area?

More importantly, where are you getting large format C-41 done that you can't get E-6 done? (I can't help with the original question).

David Higgs
8-Mar-2011, 11:23
for not much money you can process E6 at home, all my work is home processed as there is nowhere left locally. The chemicals cost around 20 cents a sheet. Its not a hard process and I've always had great results.

Drew Wiley
8-Mar-2011, 11:30
Chrome and color neg films are just different. I find myself using both, but for different
kinds of subject matter per hue pallette, contrast range, etc. The saturation peaks of
the dyes ultimately have to somehow wiggle around that orange mask, and there's
an inevitable effect, though improvements in C-41 films have certainly been dramatic
in recent years (unless you preferred the old muddy Vericolor look).

rguinter
8-Mar-2011, 11:52
I guess Taylor (Princeton) is not in your travel area?

More importantly, where are you getting large format C-41 done that you can't get E-6 done? (I can't help with the original question).

David:

There's a lab in Clifton NJ, Pro-Lab that has been doing my LF C-41 with great results. Very professional.

But they don't do E-6 or B&W and send it out. Same as I could do myself.

Bob G.

rguinter
8-Mar-2011, 11:54
Sounds like you like very saturated colors. Kodak Ektar 100 is a good place to start. After scanning (with minimum adjustments in the scanner software) I would then treat your scans in Photoshop or similar software - you should be able to get very close to what you are looking for - there are even actions in Photoshop that allow you to simulate various E-6 looks. I believe Nik software is offering a bunch of those.

Yes I really like the Kodak Ektar 100 and it gives great results during the day. I stocked up on it a few weeks ago.

But so far (at least the few times I tried) it's not a good comparison to E-6 at sunset or after.

Bob G.

rguinter
8-Mar-2011, 11:57
Chrome and color neg films are just different. I find myself using both, but for different
kinds of subject matter per hue pallette, contrast range, etc. The saturation peaks of
the dyes ultimately have to somehow wiggle around that orange mask, and there's
an inevitable effect, though improvements in C-41 films have certainly been dramatic
in recent years (unless you preferred the old muddy Vericolor look).

Actually I have lots of Vericolor VPS and VPL in stock and I use these regularly for the grainy-muddy look that you suggested.

Sometimes after dark they do well but the graininess intensifies.

Example attached on VPL.

Bob G.

rguinter
8-Mar-2011, 12:00
for not much money you can process E6 at home, all my work is home processed as there is nowhere left locally. The chemicals cost around 20 cents a sheet. Its not a hard process and I've always had great results.

Good suggestion and I've thought of that.

Unfortunately after a costly divorce I'm homeless and live with family members. Not my house and I've been reluctant to consider home processing until I retire and invest in my own home. Probably about 7-years down the road.

So probably stuck with lab processing until then.

Bob G.

Eric James
8-Mar-2011, 12:02
Have you considered mailing your E-6 exposures to a reputable lab? I use AgX. Rumor has it, Fuji and Kodak will continue to provide chemistry for the next two years, and I imagine that the few remaining E-6 labs will stockpile a couple year's worth of chemistry. If you prefer E-6 emulsions your only obstacles for the next few years are: 1) dealing with the wait, and 2) dealing with the worry of loss in the mail. I've never had film lost or damaged in the mail, but it always makes me a little nervous. Using the USPS, round trip postage is about $12 for a one-week turnaround, Seattle to Sault Ste Marie.

rguinter
8-Mar-2011, 12:04
I guess Taylor (Princeton) is not in your travel area?

More importantly, where are you getting large format C-41 done that you can't get E-6 done? (I can't help with the original question).

I used to work for AT&T at a time when we had a facility in Princeton that I covered. Used to travel there regularly.

Now that I live (and work) in North Jersey it would be quite a haul and directly opposite direction from my daily commute. So no real possibility of getting there on weekdays.

But a good thought for compiling my E-6 films and taking down there once or twice a month. I'll keep that in mind.

Bob G.

Bruce Watson
8-Mar-2011, 14:31
I suspect that the "problem" here is that the subject brightness range (SBR) of your scene is limited, down around 4-5 stops. This makes it a "better match" to tranny films in general -- IOW, the tranny will be more contrasty and "rich" while the negative will look somewhat flat and "pale" in comparison. This is because you are using most of what the tranny can do, but a much smaller portion of what the negative can do.

A secondary problem may originate in the fact that negatives are more color accurate than trannies. This according to a retired Kodak engineer who worked on such things (he's PE on APUG). That's what the orange mask is all about -- improving the color accuracy. But just because it's more accurate doesn't mean that you'll like it better. Really -- there's often a difference between what's accurate and what looks good.

Either or both of these things can easily be "cured" in a scanning and photoshop workflow. Considerably more difficult IMHO if darkroom printing.

If you really like what E-6 gives you, it might just be worth finding a lab and mail ordering your processing. But if you are scanning, C-4 can do what you want if you're willing to climb the learning curves in photoshop. But really, you should do what works for you.

rguinter
8-Mar-2011, 14:47
Have you considered mailing your E-6 exposures to a reputable lab? I use AgX. Rumor has it, Fuji and Kodak will continue to provide chemistry for the next two years, and I imagine that the few remaining E-6 labs will stockpile a couple year's worth of chemistry. If you prefer E-6 emulsions your only obstacles for the next few years are: 1) dealing with the wait, and 2) dealing with the worry of loss in the mail. I've never had film lost or damaged in the mail, but it always makes me a little nervous. Using the USPS, round trip postage is about $12 for a one-week turnaround, Seattle to Sault Ste Marie.

Eric: Yes I've considered the mail-order process. But was hoping to avoid it as long as possible for the reasons you mentioned.

Bob G.

Ben Syverson
8-Mar-2011, 14:52
It's easy to get an E6 look from C41 if you're scanning. But you'll need to do some things in Photoshop that you've been taught are "bad," such as clipping blacks and highlights. That and some Saturation adjustment should give you the punchy look of E6.

Essentially you need to throw away the extra information contained in negatives.

Henry Ambrose
8-Mar-2011, 15:17
It's easy to get an E6 look from C41 if you're scanning. But you'll need to do some things in Photoshop that you've been taught are "bad," such as clipping blacks and highlights. That and some Saturation adjustment should give you the punchy look of E6.

Essentially you need to throw away the extra information contained in negatives.

^^ This. ^^

You absolutely can get most any look you want from color negative film. It can be liberating in that you're going to get to interpret the scene to your vision without being held hostage to some transparency's version of reality. Just accept that the transparency's version of reality was not truth anyway and you'll be alright.

Make your pictures look like what you had in mind when standing there tripping the shutter.

Drew Wiley
8-Mar-2011, 19:30
Henry - if you can get anything you want from a color neg, you're the first person
on the planet to accomplish that and should probably patent your method. I've never met anyone who could make a neg look like a chrome, by any method - scan
& PS, photoshop, dye transfer - you name it. Every type of film has its limitations,
and each type is in fact only an impression of reality. One gets used to a certain kind of film and workflow and begins seeing the world like that; but that doesn't make it even close to reality. Films are engineered to do certain things. Chrome films were first engineered to look good on a light box or slide show. Neg films were made to be easy to print and give good skin tones, and evolved from there. But there's still a distinct gap between what negs do and chromes do. Still a big leap
from Ektar to Astia, and that's about as close as the gap gets.

Roger Cole
8-Mar-2011, 19:48
Have you considered mailing your E-6 exposures to a reputable lab? I use AgX. Rumor has it, Fuji and Kodak will continue to provide chemistry for the next two years, and I imagine that the few remaining E-6 labs will stockpile a couple year's worth of chemistry. If you prefer E-6 emulsions your only obstacles for the next few years are: 1) dealing with the wait, and 2) dealing with the worry of loss in the mail. I've never had film lost or damaged in the mail, but it always makes me a little nervous. Using the USPS, round trip postage is about $12 for a one-week turnaround, Seattle to Sault Ste Marie.

What are these rumors?? That would imply they're also planning to stop making film some time before that, and while selection is certainly down particularly in sheets, there's really no shortage of E6 film in 35mm and 120, nor have I heard any real rumors of its demise. I guess we all suspect it will go sometime, but I never thought E6 would disappear entirely within that time frame, certainly not in 35mm or 120.

Kirk Gittings
8-Mar-2011, 19:57
When I was still shooting color negatives for commercial work, I got very good at shooting and scanning color negs to look like chromes. I didn't find it that hard. My situation was that a few times I sold stock articles to magazines based on my earlier set of transparencies that i shot for the architect. Then later when I was mainly shooting color negs a magazine would want the story but with some additional new shots with the family in the kitchen etc. So I got good at giving the new neg scans the same "look' as the earlier chromes.

I'm too tired right now to remember the guys name, but there was/is a landscape photographer in the Midwest. His name started with a B. Who swore by scanned negs. I saw some scans that he had done for another photographer and I got over my prejudice against scanning negs.

Eric James
8-Mar-2011, 20:07
What are these rumors?? ...

A friend told me that he had heard it from the manager of the lab he uses, and the manager heard it from the respective film reps. I'm certainly not trying to stir anything up by passing this along, and other than my friend's integrity I know nothing of the truthfulness of the rumor. IIWII

Henry Ambrose
8-Mar-2011, 20:55
Henry - if you can get anything you want from a color neg, you're the first person on the planet to accomplish that and should probably patent your method.

snipped.......



Certainly not the first and hardly the only.
Color negative is wonderful for scanning.
If you know how you can make it look like what you want.
Which is a big part of photography for some - coming back with what you "saw".
CN is a very flexible material.

Read Kirk's post.

Drew Wiley
9-Mar-2011, 09:27
HENRY - insects etc see certain things in UV or infrared which we can't see; similarly
there are certain hues we can see that one film or another can't. I've run into mineral
colors in the desert and fluorescent algae out along the coast that no film can seem
to record (although the old style side-by-side grainy Agfachrome sometimes worked, or
a tricolor b&w white system might work). If trannie films are relatively limited in scale,
color negs films have distinct limitations in hue gamut. If you're able to reproduce those missing colors in PS then you are inventing them and not duplicating the results
of the film per se. Maybe dithering like heck. More likely, you're just used to a particular film and have learned to see the world the way it does, which is not realistic. Doesn't mean the results aren't potentially pleasing, but it's a myth that any
kind of film begins to capture light or color as we really see it. And it's all really a game of illusionism, making our media look convincing when its actually artificial. I've
spent my whole adult life trying to get certain rare hues to reproduce in print, and I don't think I'll ever succeed. Neg films just make it harder; and PS is limited by the
output media. Inkjet certainly has its gamut limitations; and Lightjet etc is just std
C paper. My last resort is dye transfer, which is not perfect either, but a more
accurate gamut than anything else I'm aware of. It's a bit of a revelation when these
things are put side by side.

Henry Ambrose
9-Mar-2011, 09:52
Drew,

Read Ben's post again -- "if you're scanning" -- is the part that guided my further comments. Your darkroom experience is impressive but it does not apply to "if you're scanning".

I've spent a large part of my life transferring color digital images to paper and showing others how to do the same. It used to be hard and fairly complicated but today its a well known and fully developed technology. I can say with certainty that scanned color negative film can be rendered to give a wide variety of looks including the look of color positive film. Its up to the user to decide what they want, then learning how to get it.

I looked at your site - I bet your prints are impressive in person.

Drew Wiley
9-Mar-2011, 10:13
Thanks for the compliment, Henry, but I think my website stinks, and there are very
few images on it printed less than twenty years ago. Low priority. But it illustrates what I said, getting used to a particular medium, in my case, the idiosyncrasies of
Cibachrome and how to deal with it. Negs and PS offer different options (I also do
color neg optically printed, for its own pallette). But nothing can be called "universal".
Makes me very jealous of what real painters can do with watercolor etc. I love the
complex neutral gray-greens and dusty gold and earthtones we get in our California
summers, and haven't found any kind of photographic output that will match them.
Anybody can take a box of Crayolas and work with saturated colors; it's the neutrals
which test the system - and color neg film still dumps certain things toward skintone.
I'm working with masking corrections, both with Type C and dye transfer printing, which is a tactile option to what PS does. Ironically, got my best neutrals with Ciiba,
but that was due to years of experience bludgeoning it with masks and formula tweaks. Can't do much with DT till I formally retire from my day job, and then have
just so long till the supplies run out. Can't do it all. Have to pick and choose your
battles. I wonder what the old-school Vericolor guys think, after they've formed entire
bodies of work around palletes which are now discontinued, with color neg film "improved" for guys like me.

rguinter
9-Mar-2011, 10:45
I suspect that the "problem" here is that the subject brightness range (SBR) of your scene is limited, down around 4-5 stops. This makes it a "better match" to tranny films in general -- IOW, the tranny will be more contrasty and "rich" while the negative will look somewhat flat and "pale" in comparison. This is because you are using most of what the tranny can do, but a much smaller portion of what the negative can do.

A secondary problem may originate in the fact that negatives are more color accurate than trannies. This according to a retired Kodak engineer who worked on such things (he's PE on APUG). That's what the orange mask is all about -- improving the color accuracy. But just because it's more accurate doesn't mean that you'll like it better. Really -- there's often a difference between what's accurate and what looks good.

Either or both of these things can easily be "cured" in a scanning and photoshop workflow. Considerably more difficult IMHO if darkroom printing.

If you really like what E-6 gives you, it might just be worth finding a lab and mail ordering your processing. But if you are scanning, C-4 can do what you want if you're willing to climb the learning curves in photoshop. But really, you should do what works for you.

Bruce:

Some very good points. To-date I have avoided doing much of anything with Photoshop other than tidying up and removing a few dust specs. I've always liked to work with exactly what I can get (in-camera) with E-6 films and the great variety of filters I've collected over the years.

I have worked with all the adjustments that can be done with the scanner software and I do agree that the C-41 negatives can be made to look nice. But they are so much different than the slides that that was the reason for starting this thread.

The lab I use does farm-out the E6 that I've been doing and the lab they use has been giving me some good results. But I miss having the control over where the film is going and when. I may just have to live without that control and start mailing my own to one of the bigger labs.

I think your comments about SBR are probably right-on regarding sunset/evening photos.

The Ektar 100 and Portra films do look beautiful during the day and are becoming favorites so I've stocked up on both.

Bob G.

rguinter
9-Mar-2011, 11:00
... I love the complex neutral gray-greens and dusty gold and earthtones we get in our California summers, and haven't found any kind of photographic output that will match them...

I think Drew is right.

I'm not debating others' comments above that scanning negatives can make the output look like chromes. But so far for me it hasn't and I've tinkered with all the controls.

This sequence just back from the lab and scanned yesterday. Film was Ektar 100 4x5-inch and cropped a bit at the bottom to remove irrelevent foreground.

All are nice sunset photos but totally different from what I would expect with E-6 scanned with default settings.

Bob G.

Ben Syverson
9-Mar-2011, 11:39
All are nice sunset photos but totally different from what I would expect with E-6 scanned with default settings.
It's the phrase "default settings" that I have a problem with. Like anything worth doing in life, scanning requires a little care and attention. Think about wet printing. If you slap a negative in an enlarger with "default filtration," and develop the paper with "default times," you really can't draw any conclusions about the film itself. You're looking at one narrow, lowest-common-denominator interpretation of the negative.

Obviously no one does that; if you're spending the time to make a wet print, you carefully adjust your process to produce the best image. Exposing the negative in the first place requires your care and attention. Why should scanning be any different?

I just don't buy the argument that "E6 gets me the image I want and I don't have to mess with the software." If that's the workflow you want, do what Ken Rockwell does—shoot JPEGs on a DSLR with saturation set to +10.

Decisions about how to render your image are going to be made by someone or something. I believe it should be you. That decision-maker could be the technician scanning your film, your DSLR's image processing chip, Photoshop's "Auto Levels," your scanner's default settings, pure chance, or you. I believe photographers should not absolve themselves from making those critical decisions.

Attached is a fairly raw scan from Ektar 100, followed by an "E6 version" made in Photoshop. This is my best approximation of what I get with E6—deep blacks, punchy highlights, and a slight overemphasis in the blues (makes skies look great). The saturation has not been adjusted. If grabby color is what you want, it's quite easy to achieve.

Drew Wiley
9-Mar-2011, 11:47
I'm just getting started with Ektar and haven't done any sunset work except over at
the Hana coast in Maui (yeah, a real hardship!) - generally very pleased, but of course
not the same saturation level as a chrome film. The question is the blues. The turquoise blues of those tropical waters responded wonderfully to Ektar; but I have a
serious question how deep true blues will do. Eager to get fiddling with this film but
have a lot going on right now. Still fiddling with formal Macbeath chart test and
contact sheets of various formats, plus masking protocols, especially versus the performance of Portra films. What I just mentioned about neutrals, however - I can
take a high-quality set of watercolor pigments and reproduce difficult hues in moments
that become a nightmare photographically. PS lets you eyedropper and "mix" or dither
too, but with nowhere near as much ease and precision. Maybe in my old age I'll end
up where I began, painting watercolors. Every medium has its limitations. Just about
the worst gamut I've ever seen was with color Fresson and the old alizarin crimson
and chromium oxides pigments it used - but it would print gold buttons that looked
like metallic gold instead of yellow! Even the dye transfer prints which Eggleston made of Elvis' gold piano look yellow not gold. Limitation is the name of the game.

Bob McCarthy
9-Mar-2011, 11:50
My issue has always been, "I" cant see through the orange mask and reversed colors on a light table.

I completely agree that scanning and PS editing can come very close to replicating most any film.

So does anyone make a color reversal film w/o the mask. Isn't the mask about printing directly on paper.

a)or is it better to remove the mask with an evedropper to grey and then reverse

b)or is it better to do a reversal and they remove mask.

I would love Ektar to be my film of choice, but its always a challenge to convert them to a positive. I use method a) but get results all over the map, sometimes great, sometimes a challenge.

some help?

bob

Drew Wiley
9-Mar-2011, 12:04
Ben gives an excellent example of tweaking the curve of Ektar on PS to simulate chrome film. But that doesn't solve the excess cyan which drifts into certain neutrals,
not as bad as old-time color neg films, but still there, and not in fact as neutral as what a chrome film has. The orange mask does in fact affect the gamut and is not
simply related to contrast control. It has to if you think about it, and it effects cannot be totally nulled out. Looking at a color neg on a lightbox is a totally different subject -I'm just used to it, so no big deal. Or you could buy one of those old Kodak video analyzers, but if you scan anyway, just have your monitor adjusted to see positive.
The other thing I find disappointing is the generally poor punch in the blacks of digital
color output in general, at least in the color versions of it. Even Fuji Supergloss is dull
compared to Ciba or DT. But in DT you can separate the emulsion layers separately
and mask for each to correct the final printing dyes, for example, the cyan issue, and reassemble (a lot of work, because interpostives have to be made even before the separation negatives.) This could hypothetically also be done by scanning RGB individually and PS correcting each film dye this way rather than overall. Maybe some of you advanced PS technicians already do it this way.

Ben Syverson
9-Mar-2011, 12:04
some help?
Once you invert the negative (so that the image is no longer a negative), it's usually a three part process for me. First, adjust the black levels in Red, Green and Blue, one at a time so that black is neutral (doesn't need to be pure black, just neutral). Second, adjust the white points in Red, Green and Blue so that the image is in the right ballpark in terms of color, without clipping any channel. Third, introduce some curves in the RGB channel to restore contrast, and do simple curves in the color channels to get your final color.

It takes longer to type it out than to do it.

Ben Syverson
9-Mar-2011, 12:09
Ben gives an excellent example of tweaking the curve of Ektar on PS to simulate chrome film. But that doesn't solve the excess cyan which drifts into certain neutrals,
not as bad as old-time color neg films, but still there, and not in fact as neutral as what a chrome film has.
Every slide film I have shot goes very blue across the board. You may get neutral gray to be pretty close on a Macbeth chart in full sun, but for an overcast, shaded or sunset shot, I end up with shadows that are a vibrant blue, and midtones that are at the very least cooler than neutral.

I 100% guarantee that if we shot E6 and C41 of the same scene side-by-side, I could make the C41 a complete match for the E6. There is nothing mystical about E6 materials or the colors they can capture; E6 just has a sharper response curve.

Edit: Here's an example. This is Astia, supposedly one of the most neutral E6 emulsions out there. The buildings and cars on the street aren't getting direct sun, so they go purpley blue. This scan is pretty true to the original slide.
http://farm1.static.flickr.com/163/353981566_8110fac551.jpg (http://www.flickr.com/photos/bensyverson/353981566/)

Bob McCarthy
9-Mar-2011, 12:12
Once you invert the negative (so that the image is no longer a negative), it's usually a three part process for me. First, adjust the black levels in Red, Green and Blue, one at a time so that black is neutral (doesn't need to be pure black, just neutral). Second, adjust the white points in Red, Green and Blue so that the image is in the right ballpark in terms of color, without clipping any channel. Third, introduce some curves in the RGB channel to restore contrast, and do simple curves in the color channels to get your final color.

It takes longer to type it out than to do it.

So is the clear film area outside the exposure (that has nothing but pure orange mask) your white point after reversal. So your white point is the clear film.

Black point is a black if there is one available.

I'm assuming your scanning as a transparency. My software is ColorGenius (not named as well as I'd like) and I can scan as a color negative, but I need to see if I can find a canned profile or figure out how to build one myself.

UGGG

But I suspect it's worthwhile, but after 30+ years of positive film, I'm in uncharted waters (for me).

bob

Ben Syverson
9-Mar-2011, 12:29
So is the clear film area outside the exposure (that has nothing but pure orange mask) your white point after reversal. So your white point is the clear film.

Black point is a black if there is one available.
If you're scanning as a positive, then yes (but that's before reversal)—I try to get the orange base as close to white without clipping as possible. I don't set the black point, just in case there are some sun glints or something that are quite dark on the negative.

Once the images is inverted, the clear area with just the orange mask is obviously your black, so at that point you're trying to sink it to your desired black level. And you're not likely to see any clipped whites, so from there it's just pushing the highlights to where you want them.

Personally, I fundamentally disagree with the whole concept of scanning profiles. The theory is that you can set up a profile for your scanner and a particular emulsion, and then after that, all you have to do is press "Scan," and you'll get a perfect neutral image. That might be seen as a benefit for high-volume wedding shooters who don't have time to tweak their images. For anyone who's going open up Curves anyway, profiling is a complex waste of time.

Ben Syverson
9-Mar-2011, 12:57
I just made a short video that shows the process. It starts with a well-scanned negative (in this case Portra 400 NC) scanned as a positive, and ends just over a minute later with a final image:

http://www.youtube.com/watch?v=5oqA298hSMM

Curves are your friend!

Bob McCarthy
9-Mar-2011, 13:22
Thank you,

Appreciate your help, im going to try to get out of the office early just for this.

bob

Drew Wiley
9-Mar-2011, 13:24
That's a nice little video, Ben. But here's the conceptual problem: just check out the
spectral sensitivity graphs for Ektar (or any current color neg film) vs Astia (or any
chrome film). The yellow sensitive layer of a neg is basically a plateau, while on chrome
film all three layers constitute a spike. This means there will be a reproduction error
which merely nulling out the orange mask cannot correct. You might be able to simulate E-6 in terms of contrast and saturation, but there will still be distinctions in
certain hues. Maybe OK sometimes, maybe intolerable other times. So I don't see how
anyone can say equal results are possible. The best gamut I have ever seen comes
from the tricolor filter black and white film method, which is a lot more practical for
studio still life than in the field. Beyond that, every film has a bias, and to some degree
its apparent right in the spectral sensitivity charts.

Ben Syverson
9-Mar-2011, 13:37
Hmm, don't know what to say—never had a problem reproducing yellows.

Spectral sensitivity is nice, but exposure latitude is even nicer. :)

http://farm2.static.flickr.com/1338/5124299402_a9f69f8a68.jpg (http://www.flickr.com/photos/bensyverson/5124299402/)

http://farm6.static.flickr.com/5128/5383446976_1c20d37ac2.jpg (http://www.flickr.com/photos/bensyverson/5383446976/)

http://farm3.static.flickr.com/2291/1681817820_d8aecd8c2c.jpg (http://www.flickr.com/photos/bensyverson/1681817820/)

Drew Wiley
9-Mar-2011, 13:49
It's not just saturated yellow, Ben. That's easy (though it certainly wasn't prior to Portra-generation films). Try cleaning up an old Vericolor neg (which had an exaggerated version of this same kind of problem) and see how clean a palette you can create, just to illustrate the point. That's something inherent to an entire color
layer in the film. A spectral plateau is not a spectral point. I'm getting painfully aware
of it, and that's why I find myself traveling with both E100G and Ektar, maybe Portra
too (which is a painfully expensive strategy in 8X10). All wonderful films, but each
has its pros and cons with respect to color reproduction per se, not just contrast.

engl
9-Mar-2011, 13:49
All the tweaking in the world won't be enough to work around the metamerism issues inherent to right about every color recording system. Objects in the real world do not have "color" that can be represented with RGB components. They have properties that result in different wavelengths of light being reflected (or emitted). A spectral reflectance curve describes this reflection. What reaches our eyes depends on the spectra of all light hitting the object, as well as the spectral reflectance of the object.

This is where things get complicated. Our eyes can not uniquely identify each frequency of the incoming light, the spectral power distribution. What we have is three types of cones, S/M/L, with different spectral sensitivity, and some amount of overlap. What we see as color is the stimulation of these cones.

This is where metamers come in. There are many spectral power distributions resulting in the same stimulation of the cones. We simply can not tell these spectral distributions apart, and consider them to be the same color even if the light has different characteristics.

Film, and digital sensors, also suffer from metamerism. However, since the spectral sensitivities of the layers (or sensor sensels under a CFA) are not identical to human eye cones, film and digital has different metamers. How much of a problem this is will depend on the specifics of the spectral sensitivities, wide or narrow, the amount of overlap, the shape.

This may even result in two objects which are identical to our eyes to have different colors in a photograph. The spectral power distributions would be metamers to our eyes, resulting in identical S/M/L cone stimulation. When photographed, the medium would not have identical stimulation and could separate the objects. How we'd then get to see the photo with the differently colored objects depends on the medium, but they would be mapped to colors that are not metamers to our eyes.

The point I'm trying to make is that the specifics of the recording medium will affect the final result, and the colors from one medium can only be achieved through another medium that has the same metamers. No amount of Photoshop or darkroom trickery will change that. I know more about metamerism in image sensors, but given the differences between the dyes and layers of E6 and color negative film, I'd never expect to get the same look from these two film types under a wide range of conditions.

Drew Wiley
9-Mar-2011, 14:25
Well stated! I used to get very well paid sorting out metamerism issues with architectural coatings, paints, etc. Don't have the energy to run around doing color consultation any more. However, metamerism is just one of the issues separating specific films or films types. Then you've got a repetition of the same kinds of issues
at the output end - the print media itself. Contrast and saturation are the easy part.
Some films just can't "see" certain hues; and certain print media can't duplicate them.
That's why we need these choices.

Ben Syverson
9-Mar-2011, 15:01
With alllllll due respect, I think you guys are overthinking this.

If you can generate an image that you like, you really don't need to worry about spectral response peaks and such nonsense.

There is no perfect image sensor, and there is no perfect display system. Whatever you do, you will never get a 100% perfect reproduction of the colors your eyes saw at the moment of exposure.

And honestly, who cares? Unless you're photographing artwork or corporate colors, the goal is to wind up with a pleasing image, not an objective record of the light reflected by the scene. Even objective XYZ color measurements are relative, since they're specified in relation to a reference whitepoint.

Yes, I could make the yellows in my images more saturated (though you might want to check your monitor first—they look quite saturated on my calibrated display). That's one slider adjustment in Photoshop. However, a yellow taxi should not be represented as 100% pure yellow. It's out of step with reality.

I wouldn't judge anyone who pushed their colors that far, but claiming that it's more photometrically accurate is just complete hogwash.

engl
9-Mar-2011, 15:03
I think you could get paid very well for doing that kind of work again if you wanted to, there seems to be no lack of people who have grown up with computers and think colors are as simple as the RGB values they get with their eyedropper tool in Photoshop :)

And yes, there are many issues beyond metamerism at the capture stage. The stimulus values might be stored within the layers of the film, capable of separating two objects, but they then have to be converted to a usable form. This conversion causes metamers for a system that reads all the layers combined, even with slide film, since the spectral sensitivities are not the same as the spectral density (captured with one spectrum, resulting in a dye with somewhat different transmission spectrum in development). If scanned, the scanner will have its own set of metamers. And all this is just about separation, nothing about accuracy. If you let the blue sensitive layer of slide film turn red in development, you could still in theory separate the same colors by slicing the film and looking at one layer at a time. If you looked at the slide with your eyes, it would look nuts.

So its all really just black magic, and I agree, we need many ways of working with these imperfections.

engl
9-Mar-2011, 15:11
With alllllll due respect, I think you guys are overthinking this.

If you can generate an image that you like, you really don't need to worry about spectral response peaks and such nonsense.

There is no perfect image sensor, and there is no perfect display system. Whatever you do, you will never get a 100% perfect reproduction of the colors your eyes saw at the moment of exposure.

And honestly, who cares? Unless you're photographing artwork or corporate colors, the goal is to wind up with a pleasing image, not an objective record of the light reflected by the scene.

Yes, I could make the yellows in my images more saturated (though you might want to check your monitor first—they look quite saturated on my calibrated display). That's one slider adjustment in Photoshop. However, a yellow taxi should not be represented as 100% pure yellow. It's out of step with reality.

I wouldn't judge anyone who pushed their colors that far, but claiming that it's more photometrically accurate is just complete hogwash.

I never claimed you need to know about these things to do photography. I just tried to explain why some films can separate things that other films never can, no matter what you do in Photoshop. This thread is about using one medium in the place of another, and not being satisfied with the result.

I'm not talking about "accuracy" or "truth", just mediums having different characteristics.

Ben Syverson
9-Mar-2011, 15:32
there seems to be no lack of people who have grown up with computers and think colors are as simple as the RGB values they get with their eyedropper tool in Photoshop :)
As someone who has indeed grown up with computers, I take specific, personal exception to this.

Not that it's a prerequisite to taking a picture, but I happen to have written conversion routines between RGB, XYZ, Lab, HSL, YUV, YCrCb and Cineon. I've bought and paid for more standards documents than I can recall. I learned two things from that experience:

1) Maintaining color calibration through the chain is like keeping dust off of negatives. You can do a perfect job and still get some contamination. Even the best CCD sensors do not have perfectly linear response, and basically any kind of photochemical film is a complete crapshoot. There are variations in emulsion, processing and literally every step of the scanning process. People who care about color accuracy above all else are simply not shooting film. It's a bad medium for repeatable results.

2) Perfect calibration is actually not desirable. Well, at least as an end result. I've gone through scientific-level calibration using custom profiles (created using custom electronic charts, actually, but that's another long story), and you wind up with a very, very boring looking image. You always need to apply some color correction to get a pleasing image. The only conclusion you can draw is that reality itself looks a little flat and dull. We all want drama. I mean, E6 is 100% drama.

There is only one (one!) situation where calibration is desirable. If you have multiple cameras or scanners, especially of different types, AND you need to match them exactly, it CAN be helpful. But even then, calibration isn't necessary to get them in the same ballpark. Very rarely do you need even a 90% match between two sources. One of the things I had to learn about 3D LUTs for calibration is that the fewer entries, the better. That's why Hollywood uses 17x17x17 LUTS. It would be trivially easy for them to use 32x32x32 LUTs instead, and the charts will match up better, but you'll start to see undesirable artifacts in skintone as the tonal curve is stretched like putty to match the "ideal."

The point I'm trying to make is that this entire line of argument is bull$#!t. As soon as photons hit an imaging sensor, the ball is in the photographer's court. Don't worry about your end result emitting similar number and constitution of photons. Just worry about making the best image you can.

Edit: The whole line about "separation" is also questionable to me. It smacks of superstition. I have yet to scan a negative and think "hey, this color and this color should be different, but they're scanning the same."

Roger Cole
9-Mar-2011, 16:09
My issue has always been, "I" cant see through the orange mask and reversed colors on a light table.

I completely agree that scanning and PS editing can come very close to replicating most any film.

So does anyone make a color reversal film w/o the mask. Isn't the mask about printing directly on paper.

a)or is it better to remove the mask with an evedropper to grey and then reverse

b)or is it better to do a reversal and they remove mask.

I would love Ektar to be my film of choice, but its always a challenge to convert them to a positive. I use method a) but get results all over the map, sometimes great, sometimes a challenge.

some help?

bob

Rollei Digibase 200, but they don't make it in sheets and I've no idea how good it may be:

http://www.freestylephoto.biz/811201-Rollei-Digibase-CN200-Pro-C-41-Process-120-size-20-roll-Pack?cat_id=1103

Drew Wiley
9-Mar-2011, 16:31
Ben - maybe you're right hypothetically 60% of the time; the other 40% its the difference between a masterpiece and a bellyflop, the difference between a Stradivarius and a fiddle at a corndog eating contest. "A man's got to know his limitations"(Clint Eastwood), and this certainly applies to color. To abstract this a little
distance from PS, film, etc, I'll give you a common analogy about metamerism etc.
Try taking a sample book of Pantone ink colors and go to a really high-end pro paint
store and see what percent of these can actually be matched. Or even take printed paint chips from that very store and see what percentage can be precisely matched
using actual paint pigment. Or even take paint from one brand and try matching it in another. Now I'm not talking about going to Home Depot and talking to some gofer, but to a pro paint matcher with a high-end spectrophotometer and decades of skill. It ain't that easy. Now take a color neg film made to render really nice skin tones and see what it does with every analogous hue in nature. (And that's why Ektar probably isn't a very good portrait film; but it ain't a chrome-balanced film either). To me this
subject is not overkill, it's the ABC's. Film or digital, PS or darkroom, art or physiology, whichever way you choose, the same principles are still encountered, just with different specific connotations.

engl
9-Mar-2011, 16:40
Sigh, I give up...

Ben Syverson
9-Mar-2011, 16:54
Now take a color neg film made to render really nice skin tones and see what it does with every analogous hue in nature. (And that's why Ektar probably isn't a very good portrait film; but it ain't a chrome-balanced film either).
Both E6 and C41 will reproduce that hypothetical color as some combination of dye clouds. They're both "wrong." They're both inaccurate.

I still can't see your point, which seems to revolve around colors not reproducing exactly as they appeared "in situ." Well, guess what, they never will. Even the best printer out there can't get a 100% match on a Pantone spot color, even with Pantone brand inks and the world's best scale.

And failure to perfectly reproduce observed photons does not represent a "belly flop." Name any color image you admire, and I'll show you a complete belly flop in terms of objective color accuracy.

Drew Wiley
9-Mar-2011, 17:05
No need to give up. Just carry more than one film, and learn more than one printing
process. Or like most good photographers, subconsciously select just the images which best match your chosen medium. Can you imagine Ernest Haas and his Leica using anything other than Kodachrome, and not printing on dye transfer? Give him Vericolor L like Meyerowitz or Misrach, and visa versa, and what kind of horror story endings would you have? Photoshop doesn't change any of this. It gives you "control"
but controls within its own set of parameters, which are still bottlenecked by the
characteristics of the printing inks or whatever. Being restricted is actually liberation.
Maybe I'm old-school, but I really like the "signature" of a specific color film. Even
digital cameras don't escape that kind of limitation. They have their own issues. A
skilled craftsmen will recognize these limitations and use them to specific advantage.
Maybe this happens methodically, maybe subconsciously, maybe both. But it has to
happen to achieve optimum results. Trying to make a given medium all things to all people is a formula for mediocrity. Watercolor isn't acrylic, isn't oil paint, isn't tempera,
etc. Good to have choices.

Ben Syverson
9-Mar-2011, 17:08
Drew, it is good to have choices—I just bristle when I see people spreading Fear, Uncertainty and Doubt about C41.

If you're a skilled photographer looking to move your hybrid (ie, scanning) workflow from E6 to C41, you are absolutely not giving anything up. In fact, you're gaining flexibility.

Drew Wiley
9-Mar-2011, 17:14
Ben- you're right, no color film is perfect or even close to it. Neither is any print medium. Neither are our eyes, which are subject to simultaneous contrast, successive contrast, afterimage, all kinds of interesting things. Digital receptors have noise etc. Color inks and dyes have their issues. The "real world" is changing all the time and highly subjective psychologically as well as physiologically. But this isn't about matching color per se, but about what works best for what. Traditionally, chrome films
were chosen for color accuracy in product and nature photography, color neg films for
portraiture and scale. Now the gap has narrowed a bit, but it still exists and there is
still a need for both. But if one or the other disappears from the market, we just adapt
as best we can. I would classify Cibachrome as a terrible portrait medium, though a
number of people including myself have made wonderful portrait prints with it. I just
wouldn't market that way. Similarly, I don't think of C-prints and color neg film as
"ideal" for landscape work, though many folks have done compelling landscape work with it. Misrach did wonderful muddy, mushy sunsets. Now you could hype the same
piece of film in PS, but that wouldn't be what he wanted. Someone else may need it.

Drew Wiley
9-Mar-2011, 17:19
Ben - why would I spread "fear" about C-41. I just topped off my freezer with the damn stuff. I'm clearing a slab in the backyard to put up an outbuilding with a 40" RA4processor. But I don't have any illusions that this will be a replacement for chrome film. Nor will PS and inkjet be a replacement (not better, not worse - just a different set of options - a different medium). I'm not saying anything that the tech sheets don't say.

rdenney
9-Mar-2011, 17:27
I think it's reasonable to distinguish between detailed requirements for subtle hue representation and the general look of an image. I have no doubt that Drew is extremely subtle in his observation of hue, and in his standards for his own prints. But the OP was talking about getting the "look" of transparency film using negative materials. I don't equate "look" with "exact hue representation". Negative materials provide useful densities over a subject brightness range about twice what transparency materials do, and they do so because they are not required to look realistic on direct viewing. The narrower density range in negatives representing a wider SBR makes it easier to scan and maintain tonal separation over that range.

The problem is that after scanning, we are loathe to chop off a couple of stops of SBR on each end of the histogram. Once we have it, we want to use it. But if we really want to edit them to have the "look" of transparencies, that isn't that hard. I doubt the OP's issue was related to the subtleties on which Drew is basing his comments.

Rick "differentiating between gross effects and fine effects" Denney

Ben Syverson
9-Mar-2011, 17:31
Well I have to say, for wet printing, I can appreciate that there are very real practical differences between E6 and C41, and of course between different emulsions within those processes. I guarantee if I was still printing wet, I would be looking at those same tech sheets.

But for those who are just scanning their film, I think the tech sheets often do more harm than good. Even if you're pretty knowledgeable, they can throw you off, because they often specify data in unconventional units or very odd logarithmic scales.

Bruce Watson
9-Mar-2011, 18:52
I have worked with all the adjustments that can be done with the scanner software and I do agree that the C-41 negatives can be made to look nice. But they are so much different than the slides that that was the reason for starting this thread.

Two things. First, much depends on the scanner software you are using. Nearly every scanner ever made was first optimized for trannies. Some are actively hostile to negatives. Others are quite negative friendly. I'm just saying that some scanners / software are better for negatives than others, but most any scanner is good for trannies. Clearly, this can have an an effect on a photographer new to negatives. Just sayin'.

Second, images from negatives can look quite different than the same image from a tranny. What I've found when I've seen this is that neither is actually right. Both were wrong, but in different ways. And this is to be expected considering what they are and how they are made.

The question is, which one do you like the most?

For me it turned out that negatives, a good negative friendly drum scanner, and some knowledge of photoshop is just the ticket. But what works for me won't necessarily work for anyone else. No reason to think it should. Again, just sayin'.

rguinter
9-Mar-2011, 19:05
...It's the phrase "default settings" that I have a problem with....

...I just don't buy the argument that "E6 gets me the image I want and I don't have to mess with the software..."

Ben:

Well in reference to your comments above, I've scanned several thousand of my slides from 35-mm Kodachromes all the way up to 8x10-inch Fujichromes and Ektachromes.

I'm looking back at them all with a WYSIWYG mentality.

What I see on the slide is just about what I get when I use the default settings on my Epson 4990 and scan them at the best resolution the scanner will do. And that is essentially the meaning I was alluding to about the E-6 "look" that I used to start this thread.

Granted, enough posters here have mentioned all the adjustments that can be made with the scanner software and Photoshop after the fact. But that is not my goal.

If I have to do all that fussing around downstream I would much rather be out and about shooting more E-6 photos and letting the scanner do its thing without hours and hours of tinkering trying to get the "look" I get naturally with the slides.

Although I do have to admit, the look is different with Ektar 100 sunsets that I posted above... not like the same scene I've done before with Fujichromes. But nice all the same. And the software tinkering was minimal to get them.

Bob G.

Drew Wiley
9-Mar-2011, 19:15
Ben -logarithmic units are standard with densitometry. On a large format forum like
this one I feel it is perfectly valid to bring up these kinds of technicalities and that
they are potentially useful to people conversant with them. But this certainly doesn't
mean that excellent results always require this kind of reference. Scanning doesn't do anything to erase the signature of a particular film. Afterwards your workflow may indeed provide shortcuts which are more convenient than in the darkroom, and if you are judging only from the endpoint of the process, that in itself is perfectly appropriate. No quarrel. But there are times when a knot needs to get untangled.
This can occur at the film level, at the scanning software level, at the equipment
level, and obviously with the printer and its software. I just got my truck back from
getting the engine rebuilt. They obviously didn't need to rebuild the whole thing,
just what wasn't working. Analagously, maybe just tweaking the contrast curve in
PS and a few minor adjustments will prove adequate in many situations to bring
color neg film within chrome printing expectations. Then a pallete situation might hypothetically arise someday which requires looking at the idiosyncrasies of the film itself, and then its helpful to learn how to interpret the film graphs and curves.

Ben Syverson
9-Mar-2011, 21:52
Ben -logarithmic units are standard with densitometry.
Right, but to a working photographer, Log 10 and densitometry are not exactly second nature.

Take this graph for example, the characteristic curves for Ektar:
http://img810.imageshack.us/img810/2430/ektarcurves.png

A lot of photographers will look at the X axis and think "oh, log exposure, so each whole number is one stop." But stops are Log 2, whereas this chart is Log 10. That means that one photographic "stop" is roughly 0.3 units on the X-axis. Not exactly 0.3. And not one third. More like 0.301029995663981. So -2 is not minus two stops, it's minus ~6.6 stops. This chart is actively hostile to photographers.

If you're knowledgeable and somewhat handy with a calculator like us, you can look at curves in Log 10 and convert it into some useful knowledge. But this chart practically dares the photographer to misinterpret it.

Not only that, but it stops short over 1.0, when it's clear to anyone who shoots C41 that those curves would level out at some point around or past 2.0. So they're underselling the highlight latitude.

James Hilton
10-Mar-2011, 03:06
I 100% guarantee that if we shot E6 and C41 of the same scene side-by-side, I could make the C41 a complete match for the E6. There is nothing mystical about E6 materials or the colors they can capture; E6 just has a sharper response curve.

I have to agree, I've seen this done many a time. Getting the well exposed C41 is key. Getting 90% of the way there is easy, the real skill is doing that last 10% and you often have to work at it.

Matching of different film stocks and film stocks to digital is going on all the time around the world. It could be considered the bread and butter work of a good, professional colourist working in tv or film post production, and also people who are restoring old films who are having to combine scans of release prints and original camera negatives because parts are missing or damaged etc etc.

rguinter
10-Mar-2011, 05:16
...Two things. First, much depends on the scanner software you are using. Nearly every scanner ever made was first optimized for trannies...

You know... that was something I never intuitively realized. That scanners were optimized for transparencies.

Thanks Bruce for pointing that out. May in fact be the real reason I see such initial differences.

Immediately when I started scanning negatives I began working with the histogram adjustments to optimize the scan. And not at all realizing that, what I was actually doing, was reducing the wider dynamic range of the negatives down to something that more closely matched that of the transparencies I'd been doing for so long.

It makes sense.

Bob G.

Dave Jeffery
10-Mar-2011, 07:11
Getting 90% of the way there is easy, the real skill is doing that last 10% and you often have to work at it.


Thanks! Any recommendations or further information that you would be willing to share would be greatly appreciated.

Ben Syverson
10-Mar-2011, 07:50
Some scanners have bugs that limit their Dmax in negative mode. My 35mm film scanner is one. That's why I have to scan 35mm as a positive.

However, the vast majority of scanners, including the Epsons a lot of us use, are not hobbled in this way. The Dmax of slide film may teat the limits of your scanner, but the Dmax of C41 is much lower. In that Ektar chart I just posted, the blue record barely scrapes 3.0... Just about any scanner can handle a Dmax of 3.0.

It's a bit counterintuitive, because C41 holds more latitude, but compresses it into a smaller density range than E6. E6 is further at a disadvantage, because its shadow areas test the limits of scanner CCDs, which can lead to clipped or noisy shadows. C41 has the advantage of being inverted, which means shadow areas are captured as information-rich highlights (remember ETTR?). And the lower Dmax of neg means you generally don't need to worry about clipping dark areas (highlights) of the negative.

Sometimes the software will be more geared to slide scanning, but the scanners themselves do much better with negs.

Bob McCarthy
10-Mar-2011, 08:53
When I kick ColorGenius (Screen Cezanne) into negative mode, it appears to provide templates for a number of films, many which are not current. I'm would assume these are profiles of a sort. Some of the manual adjustments are greyed out.

Any suggestions? I run this scanner 16 bit rgb transparency for E-6 and all b&w.

I'm open to any/all suggestions on where to begin.

bob

rdenney
10-Mar-2011, 09:24
Vuescan provides built-in support for a range of color negative materials, many of which are not current. But I find that they are close enough for me to let my eyes do the work after that, as long as I don't clip the scan with other settings.

Working from negatives requires us to know what we want. Slides, on the other hand, give us such a realistic visualization that we may be tempted to allow the slide to define what we want.

A lot of scanning profiling has been done in support of the printing industry, to provide at least some calibrated way of maintaining fidelity to the slide, with the assumption that the slide is the photographer's statement on the characteristics of the image, and the printing technician's job is to carry that out to the extent possible. That represents a completely different paradigm than working from negatives, where the negative can never be a final statement, and the photographer must remain involved through to the final display medium.

Matching color to a slide is a process with a clear goal, while manipulating color and contrast of a negative is a matter of matching it to our murky visualization of the image. That will exercise muscles not normally flexed when working with transparencies.

I wonder if all the manipulation we do of black-and-white prints would have developed the way it did if the film had been direct-positive from the start. The science of the processes as they developed meant it would be negative, but when I think about it that seems an arbitrary more than anything.

Rick "with lots of experience with Fuji Reala in 120 rolls, but experimenting with transparencies of late in to prepare to use up a large stock of Velvia and Provia Quickloads" Denney

Drew Wiley
10-Mar-2011, 09:28
Ben - implying that LF photographers are incapable of understanding standardized log
curves reminds me of back when school teachers said American students are incapable
of learning the metric system. You don't need a calculator; you just need to make a
visual comparison between respective charts. And if you have your own densitometer
or sensitometer, it does the work for you, and reading in log density units automatically. A helluva lot easier than spending three weeks reading thru a DLSR
manual as thick as a phone book. If you do need to transition from log units to convention, like adjusting seconds of exposure, all you need is a $15 high school math
calculator. You just push a button or two. Even I can do that.

Ben Syverson
10-Mar-2011, 15:14
I never said incapable--just that a lot of photographers might misinterpret the graph unless they really spent some time with it.

I would venture that 99% of LF photographers don't own a densitomiter or sensitometer, and would have some difficulty converting logarithm bases.

But all of this is a far digression from the original topic, which we've already flogged to death. :)

Drew Wiley
10-Mar-2011, 16:23
Then what's the point of the tech sheets, which have been standardized in this respect by all the major mfg as long as I've been around? Maybe there are many many photographers out there who don't know what an f-stop means either, but I'd venture
to say they don't work in LF or any pro level whatsoever. The quantification has already been done. One merely has to compare the respective charts to see that the
distinction between a chrome film and color neg film is inherent in certain ways that
PS can't modify. This distinction begins at the capture point. If you shoot a duck with
buckshot instead if birdshot, there's simply going to be less duck to eat. This is science. Maybe you can "paint", dither, and add things with P&S that weren't even
there in the exposed film in the first place, but that kind of function is quite different from contrast adjustments which operate on the premise of what the film does contain,
which are merely brought out. Just like someone airbrushing a print in the old days.
Spectral sensitivity graphs are just as basic and H&D curves. They aren't a substitute
for actual live-world testing and printing, but do quickly inform us of certain basics.

Drew Wiley
13-Mar-2011, 16:49
Hope I don't get everyone pissed off again, but that plateau in the yellow forming layer of the film affects the accuracy of colors containing blue, not yellow. There is no way of correcting this with a white-light scan. It is inherent. You'd have to separate each spectral wave individually with very sharp-cutting filters or lasers and truncate any overlap with the adjacent wave, then recombine them. Of course, you
can simply shift the color balance, but then some other hue gets knocked off its block. These films are pretty remarkably engineered to do certain things well, but at
the expense of other things. It's useful to know their palette or signature even when
various signficant printing controls are available. With Ektar, there's a distinct problem with blue reproduction.

Ben Syverson
13-Mar-2011, 21:39
Anyone who has shot Ektar can tell you that the only "problem" with blue reproduction is how beautifully saturated it is.

Drew, you say you're not spreading fear about C41, but implying that you need lasers to scan Ektar is truly outlandish.

Honestly, with all that's going on in the world right now, let's all be thankful we're around to shoot anything at all. It seems almost insulting to be worried about the linearity of spectral response at this moment.

Kirk Gittings
13-Mar-2011, 21:44
Unless I've had my head up my ass for the last few years (always possible).......I've had no problems selectively correcting blue accuracy with any scans of any type film or digital capture. What am I missing here Drew?

Ben Syverson
13-Mar-2011, 21:48
What am I missing here Drew?
A Ph.D, apparently.

Drew Wiley
14-Mar-2011, 09:43
I pulled out one of those old Kodak color film books from the mid-70's and was reminded at what a good job they did in those days of explaining all this, and why specific films are optimized for different applications, how to interpret the graphs etc.
A whole section was devoted to explaining why certain color reproduction errors are
inherent to the compromises in film design. That was back in the heyday of Vericolor
types S and L, Ektachrome 64, and Kodachrome 25. The films choices have obviously
changed, but to underlying rules haven't. But now they just give you some marketing
oriented hints and the tech sheets. For example, the magenta sensitive wave for
Vericolor S was conspicuously chopped flat, and indeed this film had a terrible time
reproducing greens, which were routinely contaminated with cyan, just like the blues.
To balance this the yellows always turned pumkin - but the end result gave wonderful
skin tones, and guys like Stephen Shore made compelling images based upon the
idiosyncrasies of the palette. By contrat, the Ektachrome 64 graph shows all three
layers with distinct spikes, but a bit too much overlap at a certain point, which right
away would clue you that the greens were a bit too contaminated with red - something which was well known to advanced color printers. (I have to mask differently for E64 than for E100G, for example, or for Fuji chromes). Scanning doesn't
change any of this. Digital output might give you a slightly different tool box for dealing
with it, if that is what you want. I feel it is often better just to learn the signature of
the film and live with it, shoot images which take advantage of it, and sometimes do
a tweak or two to bring it within high-quality printing parameters. But no way you're
going to turn the hue palette or gamut of a neg film into that of a chrome.

Ben Syverson
14-Mar-2011, 22:07
The films choices have obviously
changed, but to underlying rules haven't.
Actually, the rules have changed completely.

When Portra was introduced (1998?), the majority of the target market (wedding photographers) was still printing traditionally. So they built in a substantial difference between Portra VC and NC.

Then the whole industry moved gradually from the darkroom to the scanner. So with each new revision, the difference between the two Portras got smaller and smaller, until they were barely differentiated at all. Finally, it became ridiculous, so we now just have "Portra."

What we are left with is a portfolio of C41 films (Portra 160, 400 and Ektar 100) that are all fairly general purpose, and scan easily. I assume they print well too, but that was clearly not a primary design goal.

The days of a certain film having a very peculiar or distinctive "signature" are long gone. Kodak is well aware that you can get those looks after scanning, so it has wisely focused on making neutral, balanced emulsions of outstanding technical merit. Any differences between them will seem like hair splitting to anyone who made wet prints with the old wild assortment of films.

Drew Wiley
15-Mar-2011, 09:29
Ben, you couldn't be more wrong. Like I said, the inherent signature of a film is apparent on the tech sheets, then its up to us what to do with it. This might mean simple visual comparisons rather than graph analysis, but it means something concrete. What you're implying is analagous to saying, now that we've got e-books and laptops, nobody needs to learn how to ready anymore. Why on earth would Kodak at this point in history market two different color neg films fairly close in ASA if there weren't different optimum applications anticipated? They've got a Portra fairly similar
to the previous NC that optimizes skin-tone neutrals like a traditional portrait film, but
with certain improvements of course; then they've got a more saturated film which
sharper spectral peaks which is kind of a hail-Mary pass to bridge the gap toward
chromes (but no touchdown yet). The manner in which neutrals are formed has a lot
to do with the way the dye sensitivity waves overlap. When you gain something you
give up something else. There is no such thing as "one shoe fits all" film. But at least
there has been a remarkable evolution of such films. The difference is, that there are
standardized ways of hinting at how a film will perform.

Henry Ambrose
15-Mar-2011, 10:03
This is getting pretty silly.

Drew, you may know a ton about wet color printing but I don't think you have the standing to comment on what a scanner can or can't do with C41 film. I'll believe most anything you tell me about making cibas but I'm not buying your digital views.

Do you have any volume of personal hands-on experience at all scanning E6 or C41 film and subsequent use of PS? Have you really worked through this enough to be able to make the claims you make? It doesn't look that way to me.

I have spent a large part of the last 20 years wrestling this digital color dragon and I know Ben, Kirk and others have put in their time too. We're not making this up, its first hand experience talking.

Its time to stop being a broken record and share your actual experience at scanning and manipulating color digitally. A 1970s book from Kodak or any number of tech sheets or paint matching experience has little, if any, bearing on this discussion.

Drew Wiley
15-Mar-2011, 10:56
Scan what???? Film or something imaginary? Scanning doesn't erase the neutrals inherent to how the film is designed. Maybe you guys have learned how to violate the laws of physics? Don't argue with me. Why not argue with Kodak and kindly inform them that they have just wasted millions of dollars of R&D and PR developing several new films when it's totally unecessary. Why not argue with all those photographers who are buying one film or another for specific reasons and would scream like hell if only one option became available. Scanners aren't immune to basic color theory - they're based on it. Like I said, you could go in with very narrow-response separate scans and try and reassemble them selectively. That's been done before, and the result is usually something very harsh, because all this does is select the spikes and chop those overlaps in the curve which so effectively create the flesh tones etc.
Current Kodak films are part of a long and very sophisticated evolution; yet every one
of them represents some kind of compromise in color gamut. But apparently there are
folks on this forum who know a helluva lot more about this than Kodak engineers,
so what's the point? Go ahead and turn a neg film response into a chrome, then go
patent it, because no one else has.

Ben Syverson
15-Mar-2011, 11:59
Drew, this is not aerospace engineering. When someone says they want the "look of chrome" with negatives, they are not talking about a laboratory-level match in every nanometer of the color spectrum.

They're photographers. They're basically saying they like the saturated colors and poppy contrast of chrome. That's easy. And nothing worth getting worked up about!

Drew Wiley
15-Mar-2011, 12:34
I agree with you, Ben, but only about 75% of the time. For many photographers just
tweaking the curve, color balance, and saturation levels will be perfectly adequate to
deliver attractive prints. And if a particular negative doesn't work, just move on to the next one. There are other printmakers who might labor a whole week in PS over a single image, and wrack their brain with the gamut details. I know a few folks like this. There are yet others who treasure the signature of a particular film, work with it for years on end and master the creative potential of the very idiosyncrasies other find
annoying, and then panic when that film is "improved". But what might be a perfectly acceptable liberty for creative landscape work might turn out as a horror story to a professional portrait studio who needs very reliable neutrals, and is less concerned with spike saturation. I'm still struggling with the palette of Ektar myself with respect to certain hues which I consider important, but which might not matter to someone else. Clearly, I don't want all my eggs in one basket. For all the Kodak bashing which goes on around here, they still deserve credit for the level of quality and smart engineering present in the film products they still do offer.

BetterSense
15-Mar-2011, 13:00
I agree with Drew that there is no way you can magically change the color response of a film post-capture by the scanning process. The scanning process gives a lot of liberty but not liberty to change the past or the physical properties of the film.

Ben Syverson
15-Mar-2011, 13:24
You guys make it seem like these films all have really weird idiosyncratic color responses or drastic response curves that make them impossible to match.

Well, they don't.

They have very predictable, neutral color reproduction and characteristic curves. Kodak has done a great job.

It's trivially easy for me to match even my Fuji negatives to my Kodak negatives. I can match either against Fuji chromes. The single biggest difference I find between neg films is grain level.

This whole notion that every film is a beautiful snowflake—that each film has a unique palette impossible to achieve with another film—is very romantic, but not accurate if you're scanning.

Henry Ambrose
15-Mar-2011, 13:32
Here, watch this:

http://www.youtube.com/watch?v=kQFKtI6gn9Y

Drew Wiley
15-Mar-2011, 13:54
Yeah, Photoshop let's you walk on water ... been tested for colorblindness yet?

Ben Syverson
15-Mar-2011, 14:26
Yeah, Photoshop let's you walk on water ... been tested for colorblindness yet?
So in order to match two emulsions, you need to be Jesus?

Bruce Watson
15-Mar-2011, 14:37
This whole notion that every film is a beautiful snowflake—that each film has a unique palette impossible to achieve with another film—is very romantic, but not accurate if you're scanning.

+1.

Kirk Gittings
15-Mar-2011, 14:55
Guys, having tried a few times to shine some light into the cracks in Drew's granite like belief systems...............:confused:

Drew Wiley
15-Mar-2011, 15:10
Hey - what the heck would some old fossil of a darkroom worker know about all these
new hi-tech gadgets? I've gotta admit I've never been around any scanner/colorimeter
that cost more than six million dollars operated within a timed vault; but that's what you wanna do if you charge forty grand per sample. It too is based upon standard
1920's 4-dimensional color-mapping, just like your Epson printer. You could even plot the results on log graph paper, but it would probably take about twenty years per sample - so they do it with proprietary application software no two people on earth are allowed to know the full formula for, though otherwise its just basic analytic geometry (nice high school stuff worth forgetting). But hey, I'll pass along a good word and tell them that now they can go out and get something at Office Depot that will let them reinvent the properties of color physics, just like in those alternate universe movies. ...

Drew Wiley
15-Mar-2011, 17:01
Well, Kirk ... my granite mentality (yes, I did grow up surrounded by granite - love the
stuff!) just saved me hundreds of dollars of film and weeks of testing time... by reading the published Ektar graphs I essentially bagged a family of correction or reproduction masking curves on the first try. Still some fine-tuning to go, which will depend on the exact RA4 output paper. But even in the inkjet realm, you can cut to the chase by understanding these things in order to save some wasted time, printing ink & paper.

rguinter
15-Mar-2011, 17:01
Drew, this is not aerospace engineering. When someone says they want the "look of chrome" with negatives, they are not talking about a laboratory-level match in every nanometer of the color spectrum.

They're photographers. They're basically saying they like the saturated colors and poppy contrast of chrome. That's easy. And nothing worth getting worked up about!

Ben:

I couldn't have said it any better in post #1.

You captured the essence of my question in very few words.

I'm surprised at how far this thread has progressed but pleased because (apparently) many others are interested in (and concerned about) the transition that is slowly being forced upon us.

Bob G.

Drew Wiley
15-Mar-2011, 17:14
Hence the "overkill" of the subject, Bob. I'm not doing this just to be academically correct. I want options if tranny film &/or the relevant printing materials are lost; so
I want to explore just how far negs will realistically go (which is indeed a moving target
as the products and processes undergo further refinement). Plus color neg photog has
distinct advantages of its own - not as an outright substitute for chromes, but as a
different palette, luminance range, etc - a different ethnic restaurant, so to speak.
Then maybe later some of our fine-tuning experiements will indeed be helpful to certain individuals with more specific needs. A lot can change quickly if the rug gets
pulled out from under us. The direction Kodak is going with with bifurcation of Ektar
vs Portra is in itself a kind of hint of where they might like to position themselves in
the color film market - getting around E6 and chromes altogether would save them
overhead; but it would also be a risky experiment until high-end digital capture becomes more convenient, affordable, predictable, and precise concerning it own
gamut limitations.

Ben Syverson
15-Mar-2011, 17:53
If what you want are options, color negative is the material to be using.

I've taken the liberty of overlapping E100G's response curve with Portra 400's. I inverted E100G, so its Y axis is inverse density. First of all, notice how incredibly uniform they are. Neither one is weird enough to be considered unique or idiosyncratic:
http://img823.imageshack.us/img823/1808/characteristiccurves.png

Next, I normalized the curves at the black and white points. This simulates what would happen if you scanned both with a very linear, neutral scanner, and then ran Auto Levels on both. Normalizing is overly simplistic, but it's pretty impressive how close this gets you:
http://img816.imageshack.us/img816/8961/e100gandportra.png

From there, you would probably want to reduce the Blue curve and boost the Red curve on both materials. But looking at the above normalized curves, it should be clear to anyone who has used Photoshop what your "C41 to E6" curve would look like. Something to this effect:
http://img580.imageshack.us/img580/46/negtochromecurve.png

...so as far as overall contrast (response curve) matching goes, it's a real cakewalk.

Ben Syverson
15-Mar-2011, 17:54
Now that I've handed you the tools to correct for contrast disparities, let's consider spectral response (color).

http://img854.imageshack.us/img854/5845/spectralsensitivity.png
First, here is the spectral sensitivity for E100G overlaid on top of Portra 400. I adjusted the curves for E100G vertically, so the Y axis becomes a relative rather than objective measure. But even so, we can see that the curves align quite well. All you need to do is check out the Spectral Response of digital sensors (http://www.maxmax.com/spectral_response.htm) to see how much these curves can vary yet still produce a similar image. The fact that Kodak's E6 and C41 materials have such a similar spectral response is an absolute feat of engineering. They are both incredibly, incredibly neutral.

The E6 film's very slightly "sharper" individual spectral curves mean it has slightly more natural color separation (saturation), but this is obviously easily replicated. For absolute precision, you could subtract the E100G response from Portra's, and apply those curves to the ab channels of a Lab image, but honestly, it will be pretty indistinguishable from a few targeted HSV adjustments.

Next, consider the spectral dye density curve for Portra 400:
http://img268.imageshack.us/img268/2325/deltacurve.png
I've taken the liberty of subtracting Dmin from the Neutral curve, to create a combined relative density curve of a neutral subject (the red line). The big dip in the orange part of the spectrum is actually intentional (that's the orange mask). Normalizing the curves the way I instructed in my previous post would bring that section of the spectrum back up

Finally, let's overlay the combined E100G curve with Portra 400's:
http://img219.imageshack.us/img219/6112/deltawithe100g.png
Hey, what do you know. They are extraordinarily similar. The biggest difference is that dip in orange response, which is again because of the orange mask, and is corrected out either in scanning or printing.

So what have we learned? Kodak's C41 emulsions should be trivially easy to match to their E6. The most important adjustment will be adding a nice S-curve to the extremely linear response of the negative film. In terms of color reproduction, they should already be quite close, but targeted HSV adjustments should help individual areas of the spectrum match closely.

Drew Wiley
15-Mar-2011, 18:25
Well, I like what you're illustating, Ben. But what you regard as very minor differences in the spectral sensitivity curves I perceive as pretty significant. One has
to get used to interpreting these things and what they mean in practical terms.
Basically, once a neutral is in effect, you can't remove it. Makes no difference whether we're talking film or industrial colorants - same principles. Portrait-style color neg films are particularly adept at creating neturals which mimic skin tones, but at the same time will "dump" analogous hues in nature the same direction. The question is, what other colors can you salvage-engineer in terms of priority? Significant progress has been made, first with Portra VC, and possibly now in Ektar, to invade territory formerly held only by chromes. The curves are getting more
similar. And making a fuss about it has its benefits, because this problem isn't
going to go away. Maybe a good percent of the time, things will be easy to correct;
but if LF chrome film is outright unobtainable, then that other 10% of applications
becomes very important to de-bug too.

Drew Wiley
15-Mar-2011, 18:37
Your comparison graphs themselves, Ben ... It's logarithmic, so the difference between the E100G vs Portra 400 show a curve overlap fully two stop of density further into each spike - that's a gigantic difference in terms of how much the inherent neturals or cross-talk between the respective dye layers begin. (With Ektar it's obviously less). To an important degree, your own graph overlays disprove the point you are trying to emphasize. Those neutrals are there. You can change your color balance, contrast, and overall curve, but there is no way to remove those neutrals or any related hue contamination from an adjacent dye layer (that is, unless you want to "paint" or dither or whatever, applying information which never existed in the film itself). But in terms of making these particular neg films look "more" like E100G, what you have done is certainly a useful illustration, because many indivduals will want to stick to basic PS controls. Yet what I'm implying doesn't
outright negate what you are attempting to do, because we are using different tools
for an analogous problem. I like putting a bit of fire under this debate, because folks
like you do come up with very useful tidbits of image mgt.

Brian Ellis
15-Mar-2011, 18:41
Guys, having tried a few times to shine some light into the cracks in Drew's granite like belief systems...............:confused:

:)

Roger Cole
15-Mar-2011, 18:43
Hence the "overkill" of the subject, Bob. I'm not doing this just to be academically correct. I want options if tranny film &/or the relevant printing materials are lost;

Print materials for them are pretty much lost already with type R paper gone and Ilfochrome now an exorbitantly expensive special order item.

Back to the regularly scheduled fracas. ;)

Ben Syverson
15-Mar-2011, 18:58
Well, I like what you're illustating, Ben. But what you regard as very minor differences in the spectral sensitivity curves I perceive as pretty significant. One has
to get used to interpreting these things and what they mean in practical terms.
Basically, once a neutral is in effect, you can't remove it. Makes no difference whether we're talking film or industrial colorants - same principles. Portrait-style color neg films are particularly adept at creating neturals which mimic skin tones, but at the same time will "dump" analogous hues in nature the same direction.
This, I think, is not supported by the data. Saying that Portra "dumps" colors to neutral is nonsense, when you can see for yourself that the discrete spectral curves for Portra 400 are well separated. E100G's color response is marginally steeper, which means marginally less cross-color contamination, but the differences we're talking about would be measured in fractions of a camera stop.

What it means in English is that a particular green on Portra may have a quarter stop too much blue when compared to E100G. Talk about splitting hairs. If you can even see it, it would be eminently correctible with one targeted HSV adjustment.

I'm honestly starting to get insulted on behalf of the Kodak engineers who slaved to make these films as phenomenally accurate as they are.

Drew Wiley
15-Mar-2011, 19:23
Ben - the difference in green is very frustrating. I spent about two years trying to
tame it and decided either it's a no-go and back to chromes, or I'm going to have to
try some advanced masking controls or even printing from sep negs. The problem
isn't Crystal Archive paper - which is the same thing used in many digital printers.
It's pretty amenable. And getting nice bright green-greens is pretty easy with Portra.
It's when you get neutral-inflected greens; they get rather cyan contaminated, just
like in the old Vericolor, just not quite as bad. As Roger noted, Ciba is getting dreadfully expensive, and it has serious gamut problems of its own, while Type C
color paper is quite affordable, especially if direct optically printed. Supergloss is
a lot more money, but still way below Ciba. Chromes are still far superior in gamut
(that is, within their intended scale - don't tell me about blues in the blacks which
are clear off the deep end). And don't tell me you can take the contamination out
with a scan, not unless you can overturn certain laws of color which are behing
everything out there, including your scanner and the way your printer inks have
been programmed and balanced. If you want to arugue fine, but you're going to
have to start your fight even before VanGogh arrived. Cross-talk across the color
wheel results in neturals (brown, beiges, greiges, etc - if that is what a film is
engineered to do, you can't stop it!)

Drew Wiley
15-Mar-2011, 19:27
Another minor note - don't tell my problem is my darkroom. I don't use white light,
and my system is very analagous to the narrow-band lasers used in high-end digital
printers. The "mud" in a neg image comes mainly from the film itself. Scanning chromes and outputting onto Crystal Archive is one method of working cleaner; inkjet from chrome scans is another, but with a lot more potential for metamerism.
Ciba is downright goofy, and requires massive mask bludgeoning. The silver bullet ... there ain't none, but dye transfer is about as close as it gets.

Drew Wiley
15-Mar-2011, 19:34
Ad infinitum ... but I could plop down a Fuji optical print made directly from a standardized Portra neg of the Macbeath chart, and you would say the color reproduction is damn near perfect ... well, at least until I set down a significantly
masked Ciba from a chrome of the same chart... the Fuji print would win the skintone and netural patches, the Ciba everything else.... But then if I set down a
dye transfer version made from color separations from the same chrome, well then,
at that point the color neg Crystal Archive version starts looking pretty muddy. If it
had been made with a conventional additive enlarger, it would be even dirtier, but
nowhere near as bad as an old Vericolor S print on Ektacolor paper.

Ben Syverson
15-Mar-2011, 19:48
Chromes are still far superior in gamut
(that is, within their intended scale - don't tell me about blues in the blacks which
are clear off the deep end). And don't tell me you can take the contamination out
with a scan, not unless you can overturn certain laws of color which are behing
everything out there, including your scanner and the way your printer inks have
been programmed and balanced.
I don't know why you insist that C41 films are designed to have cross-color contamination. They're not. And it's really not an issue. Believe it or not, Kodak film is actually designed to reproduce colors accurately. The reason why Portra films are labeled "portrait" films is precisely because they have such long, linear response curves, NOT because of some funky desaturated color response.

And to claim that Chromes have better gamut "within their intended scale" is sort of like saying cellphone cameras have better resolution than DSLRs "within their intended print size." It is literally just complete nonsense.

You can think of Gamut as a 3D map of what colors can be represented by a format. C41 film has much greater latitude than E6 film, yet has similar spectral sensitivity, so it's objectively true that C41 has a larger gamut. If you mapped Portra 400's gamut in 3D, E100G's would fit neatly within it.

Honestly, you can't just write things and will them into truth. If you want hard data, I've given you the hard data. If you want real-world empirical evidence, several of us have chimed in with our personal experiences. I'm sure there is no hope of convincing you of anything, but please stop posting disinformation to this forum, because that type of folksy anti-wisdom spreads far too easily. We've all seen it happen.

Drew Wiley
15-Mar-2011, 19:59
Latitude is not gamut. Two different variables, at least until the point the chrome
bottoms of tops out, which is beyone its intended usage anyway. Gamut more specifically applies to what specific hues will reproduce within that hypothetical sphere. When primaries intersect, a neutral is formed which disallows distinction of
similar hues. We have a problem here with basic vocabulary and color theory. Mine
is based upon standards which cut across many industries, and which have been in
place almost a hundred years. Neutrals aren't necessarily bad, but when they are
predominant, as in portrait films, something has to give in terms of saturation and
clarity of the other hues. The nature of the difficulty hasn't changed since color neg
films were first introduced, though progress has been significant. But if you want really clean reproduction, neither color negs or chromes are the way to go - just
build a sharp-cutting tricolor camera using black and white film (not exactly a point
n' shoot option, however).

Drew Wiley
15-Mar-2011, 20:04
Should have added, that's why modern computerized color mapping requires a four
dimensional model, with a separate axis being used for tone than for tint (white vs
black content). It's easier to write that way. And in this respect, white is not just
brightness; it's an ingredient and essentially contamination of each layer, like you
get in subtractive color systems (not additive, though I accidentally reversed the
terms earlier). But about all you have to do is go to Kodak's marketing images to
see how they've gotten pros to highlight the flavor of specific films, with the new
Portra leaning heavily toward portraiture. Films all still have a signature.

Ben Syverson
15-Mar-2011, 20:07
Latitude is not gamut. Two different variables, at least until the point the chrome
bottoms of tops out, which is beyone its intended usage anyway. Gamut more specifically applies to what specific hues will reproduce within that hypothetical sphere.
Wrong. Re-read this. (http://en.wikipedia.org/wiki/Gamut#Representation_of_gamuts)

"Gamut" encompasses brightness—and "latitude" is another way of saying "range of brightness."

Latitude is in fact a key component in gamut. Imagine you could only represent colors at one stop—Sunny 16, EV 15. If EV was 16, the output would be pure white. If EV was 14, it would be pure black. But this hypothetical sensor could reproduce any color under the sun, as long as it was EV 15.

The gamut chart for such a sensor would look like a deflated beach ball floating in 3D space next to even sRGB.

Drew Wiley
15-Mar-2011, 20:12
Ben - I haven't willed anything into truth. This is old,old, very routine stuff to color
specialists. Basic theory, and it amazes me that a little bit of gadgetry can make one
think it's obsolete. I've sat down and had respectful conversations with the head of
the International Color Council and discussed pros and cons of spectrophotometer
design, and we understood each other. I can talk to X-Rite engineers and we perfecty understand one another. Why is this conversation so difficult?

Ben Syverson
15-Mar-2011, 20:19
I can talk to X-Rite engineers and we perfecty understand one another. Why is this conversation so difficult?
I don't know—you tell me. I've given you a mountain of information about why you're mistaken, and you've basically told me to make some wet prints and judge them by eye.

If you insist that transparency films can do amazing feats of image capture that negatives can't touch, I'd say the burden of proof is in your court, my friend.

Drew Wiley
15-Mar-2011, 20:54
Ben - I've got to do something more exciting right now like work on taxes. I apparently just don't have the skills to teach Kindergarten color theory to kids more
interested in eating their paint.

Ben Syverson
15-Mar-2011, 21:06
Drew, I have no doubt that you know more about color theory than I do. I know this because you write it so often, and with such passionate conviction.

In fact, you seem to be so scholarly that you don't even feel moved to demonstrate or exercise your knowledge. I can only hope to reach that level of enlightenment.

I suppose at your level, one transcends worldly trappings such as "data," and simply divines Color Truths from the Gods.

But thanks anyway for the ad hominem.

rdenney
16-Mar-2011, 09:04
Sheesh.

I'm a systems engineer, in addition to other things. I don't measure things against an absolute standard, because nobody in the engineering world can define let alone afford absolutes. I think in terms of fulfilling requirements, and requirements emerge from goals and needs, not the other way around.

Drew's empirical evidence is in the coloration of a few specific minerals in the work that he does. He is also making very specific color separation exposures for making dye-transfer prints. The requirements that emerge from those activities, particularly in the context of the intended processes, are considerably different from those that emerge from the activity implied in the original post.

It has been said and said again that it wasn't the OP's intention to fulfill Drew's exacting requirements, but rather the much more general requirement of achieving the "look" of E6 when using C41 materials.

Much of the argument is based on words like "quality" and "look". But those words have no testable definition, and can't be used as requirements in the practical world (and, Drew, it was you who mentioned the practical world).

The biggest mistake people make is imposing their requirements on everyone without checking to be sure they trace to the activities of others. I write policy for systems engineering compliance, and have to devise ways to lead people through systems engineering processes without tell them what their needs and activities are, so I know how tempting this mistake can be. Many engineering projects--and color matching is an engineering project even if the standards applied to it are based in artistic objectives--run off the rails because the requirements are not nailed down based on what their owners actually do.

Ben's standard is "reasonable", and Drew's is "exacting", and each is an appropriate requirement given what each do and hope to achieve. To say one necessarily applies to the other makes the mistake I warn against.

For constructing a model of what everyone should do, it's easy: If you will do what Drew does and apply Drew's standards, then you have much work ahead of you to determine which process fulfills your requirements. If you just want to achieve a general look of transparency film using negative materials, then Ben has outlined an approach that will fulfill those requirements.

The next step in these arguments is for one person to say that the other person's requirements are wrong, and then when they are traced accurately to what they do and their needs, then the attack shifts to them having the wrong goals. I have presided over many, many meetings where the conversation had gone that way and it was up to us to bring it back. Thinking another person's goals are invalid because they are not our goals is where the insult is.

Rick "it ain't about the logarithms" Denney

Ben Syverson
16-Mar-2011, 09:47
Rick, I think part of the problem is that there are actually two arguments in this thread: one is that E6 and C41 films are impossible to match, and the other is that E6 films can capture colors that C41s simply can't.

Obviously I take exception to both, but I would not impose my standards on anyone else. I think you should test for yourself. My problem is precisely that Drew repeatedly tells people "you can't do that" on this thread, without backing it up.

Drew Wiley
16-Mar-2011, 11:49
No ... my problem was with trying to communicate with sheer visual illiteracy. The data
was right in front of your very nose but you don't understand how to interpret it sensibly. I gave a subtle hint. Just go back and study some of the old color guide books and they will gently introduce you to the subject. The Kodak guides weren't written for engineers. But the response was, this is 1970's stuff and obsolete. No. The graphs by Fuji and Kodak are still made the same way and are still just as relevant if applied to current products. You don't have to analyze things that in partiuclar manner, but this is a shortcut which simplifies actual film testing. Imagine going into an elementary astronomy class and stating that you don't have to learn the significance of Copernicus or Galileo because their idea are hundreds of years old and therefore now meaningles. Guess what, the earth still revolves around the sun.

Ben Syverson
16-Mar-2011, 11:54
Never said anything from the 1970s was obsolete. That seems to be you projecting, Drew.

All I want is for you to substantiate one single claim you've made.

Drew Wiley
16-Mar-2011, 13:37
Begin with the basics, Ben. Get some ABC color theory books from an art store or something before you try to understand a CIE profile. Do a little homework, and then
we'll have something meaningful to chat about. You're arguing about things which every color professional in the world takes for granted and operates according to on a routine basis.

Ben Syverson
16-Mar-2011, 13:45
Begin with the basics, Ben. Get some ABC color theory books from an art store or something before you try to understand a CIE profile. Do a little homework, and then
we'll have something meaningful to chat about. You're arguing about things which every color professional in the world takes for granted and operates according to on a routine basis.
Actually, I understand CIE standards quite well. I've implemented XYZ and Lab conversion routines by hand in GPU assembly code.

Writing that "every color professional in the world" agrees with you does not make it true. You're not even making a coherent argument anymore. You're basically saying "shut up, I'm right."

Whatever, I have no interest in being "right." If E6 really can do things C41 couldn't dream of, I'd love to hear about it. So get on with it already. Back up your assertions or just get off this thread.

Drew Wiley
16-Mar-2011, 14:04
You might understand CIE in relation to a monitor, but don't seem to have the slightest clue to its relation to gamut in reflected color work or in spectral dye absorbtion. And that's just one of the models available. Every single statement you
have made so far tells me that if you ever applied for a color matching job, they
would instantly know what shape of receptacle to file your application in. If you don't
even begin to understand how neutrals are formed, you can't understand color film
curves either. You look at .60 density difference in curve overlap distinction between a chrome and neg film and call it inconsequential! That's all gamut your neg film cannot
reproduce in the same manner as a chrome. In practical terms, it just depends which
hues count the most and what you're willing to sacrifice. And I'm merely restating this
on the offhand chance that someone else out there is scratching their head about the
nature of the problem. I've given up on you.

Ben Syverson
16-Mar-2011, 17:35
An 0.6 density difference is not inconsequential, but it is meaningless in isolation. You have to look at it in context.

It makes no sense to point to one part of the spectrum and go "There! At 580 nm exactly, there is a difference of X.X%!" A yellow schoolbus is not a pure spike at 580 nm; it gives off a broad spectrum of light with a hump in the yellow area of the spectrum. The sensor (film) records the sum of the reflected light from the yellow schoolbus. That evens out lumpy areas of spectral response. Just look at the vastly different DSLR sensitivities—yet using profiles, they can be calibrated within 1% or so.

As far as gamut, I thought it would be obvious why latitude factors in, but I'll spell it out with a thought experiment. Imagine you have a programmable light source with three dimmers—red, green and blue—which you use to expose a sheet of chrome film. You turn the red and blue dimmers down until they're just barely above black (below D-max) on the film, and the green dimmer up until it's almost clipping (approaching D-min for green).

Okay, now switch the film to color negative. Increased latitude means you can leave the red and blue dimmers where they are, but crank the green way up. The negative is now recording a brighter, purer green than chrome can. Alternately, you can leave the green where it is, and turn down the red and blue. Because of increased shadow latitude, that's another way to record a purer color than chrome. Either way, the negative film can address a much, much greater range of colors than the chrome, which is another way of saying it has a wider gamut.

Talk about ABCs.

Drew Wiley
16-Mar-2011, 18:40
Well now, you're starting to think about those curves, Ben ... Yes, the density difference is downhill from the peak itself, so in the final color, has somewhat less
influence than the peak area. The cross-talk or "mud" (not necessarily a bad thing) or neutrals, whatever you want to call them, are still present however. All this interacts in a very complex manner, and color-coupler chemistry is way over my
head; experts can select different dyes, balance and sensitize them differently etc.
A lot of voodoo and possibly trade secrets. Vocabulary-wise you still have a problem
differentiating additive from subtractive systems, and the distinction between luminance vs tone in CIE three-axis vs 4-dimensional modes. But that can be overlooked. What counts are certain laws of history: it is far easier to destroy a civilization or business to build one - agreed? - and likewise, it is fairly easy to create a netural from different primaries, but almost impossible to take one
apart once the "mud" is made. Again, not necessarily a bad thing - every hue
Van Gogh worked with was deliberately contaminated with a complementary.
(Momentary break)....

Drew Wiley
16-Mar-2011, 18:54
The signature of the film is due to the specific dyes as well as how they overlap in
spectral absorbtion, plus interaction with the specific print medium, of course. Your
problem with the concept of gamut seems to be from looking at these curves in
relief. Once they intersect, that's where the "mud" begins. This mud has a very important function in all films, especially negative films, because that's how these
films are engineered to capture the essence of certain things like skintones, or at
least make people flattered by them, and think this is how they are meant to look.
The area where the curve overlaps in the neg vs the chrome represents areas of
chrome gamut ABSENT from the neg gamut. So in this respect, the chrome has
greater gamut, because it can reproduce certain hues which the neg cannot. The fact
that you can hypothetically position these things within a CIE map is dependent upon
them existing in the first place. You can't place any colored marble you want inside
that baloon - you have to somehow acquire it first. Where the neg film has greater
gamut is where it reaches luminance values above or below what the chrome film
can - but that is a range issue, not a chroma or color purity one.

Drew Wiley
16-Mar-2011, 19:06
Next, your programmable dimmer analogy - good one, except that it applies to things like computer monitors, color RGB lasers, strange additive enlargers like
I use, backlit transparencies or slide projector shows (thought not precisely). With
color film you're talking subtractive color (CMY dye absorbtion, not RGB). You take
white light, selectively remove what you want, and anything left over contaminates
every color layer to some extent. Well, it's not so bad, because that contamination
occurs to a relatively small degree, as you can judge from the curves yourself. So
lets call this a secondary issue. When you routinely think CIE that you can match on
a monitor, your concept is valid. With film, it's upside down. So you decide to go in
and just overlap curves to the extent trannie film might. Seems good at first, but
you are giving up a lot which made your neg film balanced and appealing in the first
place. It's a lot more complicated game than simply tweaking the contrast or curve
shape. And basically everything which falls into that zone where the curves overlap
in one type of film vs another is going to be nonexistent - outside your baloon of
usage. You can map it there as much as you want, but its hypothetical. Crop it out
one way or another, and you find out you have to put in something else to balance
the film. In the real world this obvious depends on the specific colors in the image,
but if you start with a hundred flavors of color, like a comparison of a paint chip rack using two different films, you'll end up seeing a certain percent of them
unrecoverable - the ones on the edge will simply drift into the nearest lane the film
was engineered for, or if you edit, will be crammed and lumped into something harsh. Follow?

Ben Syverson
16-Mar-2011, 19:49
Drew, you're speaking my language, but unfortunately you're not making a ton of sense. My dimmer analogy applies to any type of photographic sensor, because both our eyes and our sensors are only sensitive to big general bands of "Red," "Green," and "Blue" light.

When hit with Red, Green or Blue frequencies, color film forms a latent image using CMY dyes. In the case of reversal, red light hitting the emulsion will cause cyan dye to be removed during development—letting red light through. In the case of neg film, red light allows cyan dye to remain—which after some form of inversion, is red. No matter what, in the end, red is red. Honestly, this is pretty straightforward.

My dimmer analogy stands. If you can represent a range of more saturated colors with one medium than with another, you say it has greater gamut. That's why CMYK printing has a very small gamut when compared to a computer monitor—your monitor can reproduce shades that are simply impossible in process printing. The same goes with negative film, which can reproduce shades which are impossible with slides.

I don't mean to be insulting, but I don't even know how to parse your post about "mud," marbles, balloons and what they have to do with gamut or color reproduction. Perhaps it would help to see the sensitivity curves in linear units, rather than log, which I feel confuses the issue. Here.

http://img822.imageshack.us/img822/7158/portralinearweb.jpg

http://img825.imageshack.us/img825/1605/e100glinearweb.jpg

I'm not sure where you're seeing all that dreaded mud-causing overlap. Portra has about a 5% contamination of magenta dye (green light) in the blue section of the spectrum, but if you'll notice, it's quite uniform. The yellow record (blue light) is discrete enough that all you have to do is remove a small amount of green from blues areas in the final image—or in practical terms, go into Hue & Saturation in Photoshop, select "Blues," and adjust Hue to the right by about 10%. Nothing destructive about it whatsoever, and it certainly does not affect the gamut of the film.

Drew Wiley
16-Mar-2011, 20:25
Well, at least I've got you thinking about this. Don't have much time left to chat this
evening. But to start, take those nice graphs you've just shared, find the point in the
base where the curves overlap each the case of each spectral peak, and draw a line
STRAIGHT UP from each intersection. The gamut on either side of each line outside
the central area of the spike is affected to some degree or another. Exactly how can
be a pretty complex subject and well beyond what even the graphs show. These dyes overlap and interact in many complex ways. But everything not in each respective center column is essentially mud (a common industry term, or "dirt" - not
necessarily evil, but hard to remove - it doesn't take much to have a significant effect). Any "mud" which is present in one curve set and not the other represents
an inherent incompatibility. But my feeling about this is learn how to dance with
what you've got before you try to change it. The problem with PS is that it lets you
do just too many things. Learn to see through the eyes of your film first, dance at
its pace before you lead too hard. (And I don't pretend to know how to dance with
Ektar - we haven't even really had our first date. I'm just beggining to learn Portra).
But go ahead and think about this, and feel free to challenge what I said - at least
you're starting to visualize the issues. You seem like a clever fellow; it will come
bit by bit.

Ben Syverson
16-Mar-2011, 20:55
But everything not in each respective center column is essentially mud
Drew, this is objectively hogwash. Hogwash!

Care to check out the horrible, terrible, overlapping curves of the human eye? I guess everything we see must be "mud."

http://img577.imageshack.us/img577/8302/eyesensitivity.png (http://en.wikipedia.org/wiki/Color_vision)

Henry Ambrose
16-Mar-2011, 21:09
Drew may be right about the mud - but only in his narrow-band "laser" illuminated enlarger I suspect.


I don't use white light, and my system is very analagous to the narrow-band lasers used in high-end digital printers.

Otherwise, not.

Ben Syverson
16-Mar-2011, 21:16
Drew may be right about the mud - but only in his narrow-band "laser" illuminated enlarger I suspect.
Henry, interesting point—but actually his narrow-band enlarger shouldn't be an issue, since the visible spectrum would have already been decimated to just three dye layers (three basic RGB "humps") by being captured on film. In other words, Drew's enlarging system is designed to mask out extraneous frequencies from the film itself, not extraneous frequencies from the original photographic subject.

Drew Wiley
17-Mar-2011, 09:36
I don't know where to begin - apparently you think that because you can fit the
curve structure of the chrome film inside the spectral sensitivity curves of the neg film,
this means the neg film has more "gamut". Exactly the opposite is the case. It means
that a lower percent of potential hues will be correctly reproduced by the neg film.
It is easily to conceptually describe this on an absolute basis, though no film is perfect,
and in the real world we individually make the choices of what is or is not an acceptable reproduction level. As far an 5% contamination of a curve goes, for example, try taking a "pure" green gallon of paint and adding only 5% of an equal
intensity red pigment into it, or magenta - you no longer have something resembling a
process color, but something "dirty" and conspicuously brown-inflected. This is basic
stuff. Take an art class or something. Buy some books. Or if your version of the universe is correct rather than mine, go engineer a colorhead of your own and see
what kind of results you get with it! But by strange coincidence my model happens
to be the same one used by Durst to make proprietary colorheads for the NSA, the
same one that ZBE used to make a me a high-feed "color neutral" easel densitometer and matching monitoring devices for internal controls, and so on. Those guys would
probably just chuckle at your notion that the laws of optics can be overtuned in
Photoshop.

Drew Wiley
17-Mar-2011, 09:41
Ben - at least you got the enlarger thing correct. It cleans up misrepresentation of the film which occurs in conventional subtractive enlargers, where leftover white light spills over to slightly contaminate each sensitive layer in the paper. Although this
spillover might only amount to 3% to 5% (depending on the model of colorhead), it has
a conspicious effect in the final print (I also have a conventional subtractive enlarger
on hand for comparison). Now apply an analogous manner of thinking to how film itself
deals with the outside world, and you will start down the right path.

Drew Wiley
17-Mar-2011, 09:57
There is an interesting thread on this forum about color photography with black and
white film, using sharp-cutting RGB filters, and why this gives such pure color reproduction - meaning a better gamut than color film is capable of. The spectral peaks
in this approach are much sharper and narrower than with any color film. This is also
the approach Technicolor film uses and why it renders such clean color compared to
conventional movie film (the gear and dyes still do exist, and could be put into use
again if someone had the budget and processing facilities for it). So Henry, to be
consistent, you're going to have to call the entire film industry nuts too, and even
the entire history of color photography. Better color reproduction occured way back
when people did lantern slide shows using three arc light projectors, each with a
different RGB b&w white neg in it.

Ben Syverson
17-Mar-2011, 10:17
Take an art class or something.
Dude, you need to cool your jets. I went to arguably the best art school in the country and I spend my days doing image processing.

First of all, about the contamination issue: you're all mixed up. A contamination of ~5% green across the blue spectrum is not a big deal because the blue response curve effectively creates a mask to remove it. You can do this optically or digitally. It has absolutely nothing to do with Photoshop.

Second, about the gamut: amazingly, you're even more mixed up. You mistakenly believe that the tighter those response curves are, the better. In fact, you want those curves to cover the entire visible spectrum, so that every frequency results in some density on the film.

Imagine you had response curves like this:
http://img16.imageshack.us/img16/5075/badcurves.jpg

The output of such a sensor would be wildly unpredictable. A neutral color in daylight would look fine. But certain saturated colors that throw off spikes (not all do) would render far too dark, unless they hit the curve exactly, in which case they would render far too bright. Shooting in fluorescent light would result in extremely dark exposures, because fluorescent is given off in sporadic spikes.

Contrary to your belief, neither trichromie nor Technicolor processes have such sharp response curves. Open up your Lee or Rosco filter book and find the most spiky primary color filter, and it's still got a curve to it.

Look at your eye's response curves again. They overlap greatly, and it does not negatively affect color differentiation.

You seem like a smart guy, but it's becoming more and more apparent to me that you don't have a very firm grasp on these concepts, which are admittedly unintuitive.

Drew Wiley
17-Mar-2011, 10:40
Too bad Timothy Leary still isn't around. He too lived in an alternate universe and had
a very different option for visualizing color.

Drew Wiley
17-Mar-2011, 10:46
Even your scanners are based on the principles I've described, so you better throw
those out too! The reason extremely narrow-band filters aren't used (except for
certain laser applications) is 1)with dichroic filters they shift sensitivity somewhat with
temperature, 2) with things like gel filters they simply can't be made that tightly (but
are VERY tight compared to conventional taking filters (overlap 29,58,&47B and you get black black); 3) you want to make sure a good percent of the desired spectral
peak is covered or you will end up with very slow exposure times; 4) a film is engineered with both idisyncrasies of real world dyes in mind (which are never perfect
in their spectral response) and to deliberately create complex neutral necessary for
certain applications like skintones.

Ben Syverson
17-Mar-2011, 11:00
Even your scanners are based on the principles I've described, so you better throw
those out too!
Drew, you're confusing photographic capture with scanning/enlarging, which is a no-no.

Film has to confront the full spectrum of visible light. But after it's processed, film is composed of just a few very broad dye layers—"Cyan," "Magenta," "Yellow".

The scanner/enlarger doesn't need to have broad spectral response. The film does.

ABCs.

Drew Wiley
17-Mar-2011, 11:10
So again, Ben, why does sharp-cutting true tricolor RGB camera like Devin, Curtis, or Technicolor yield a better gamut than conventional film? Heaven knows what could be done with a modernized version of something like this with an optically coated
beamsplitter, though no one would want to lug one around! This is where color photography began in a practical sense, soon after panchromatic emulsions were invented.

Ben Syverson
17-Mar-2011, 11:22
They don't yield a "better" gamut, just better color separation (higher inherent color contrast, or "saturation"). That was desirable in the early days of printing, especially before sophisticated color tools were available.

It does not equate to better gamut or more representable colors. The filters they used in tricolor cameras still have enough spread to cover the whole visual spectrum. So an NPC camera and Portra can both represent the same color, but it will render differently. However, it's important to understand that neither is "right." Both are based on the "hack" of trying to represent all frequencies of visible light with three big general lumps of red, green and blue. The hack works well, because our eyes use the same hack.

Ben Syverson
17-Mar-2011, 11:32
Here. You seem to appreciate visual aids, so maybe this well help your understanding. Let's put all the pieces of the puzzle together.

First, let's take three emulsions. E100G, Portra 400, and the kind of hypothetical ultra-spiky film that seems to be your ideal.

Let's see what they would do under standard "cool white" fluorescents. (http://en.wikipedia.org/wiki/Fluorescent_lamp#Phosphor_composition) First, the spectral output of our Cool White:

http://img192.imageshack.us/img192/2571/coolwhite2.jpg (http://en.wikipedia.org/wiki/Fluorescent_lamp#Phosphor_composition)

Next, let's convert the Kodak response curves to spectra, and then multiply them by our cool white spectra to see how they should respond to fluorescent light.

http://img822.imageshack.us/img822/8649/coolwhitespectra2.jpg

Next, we'll sum the RGB value of the spectra, and convert them from linear values to sRGB for viewing. This effectively simulates photographing a gray card under "Cool Whites:"

http://img18.imageshack.us/img18/2080/coolwhiteanalysis2.jpg

So E100G and Portra are very similar. The E100G is slightly darker, as a result of its steeper response curves. Either way, it would be trivial to color correct them against each other (the contrast difference is another issue). Your ideal sensor, the UltraSpike4000, is unacceptably dark, because its own spikes never coincide with the Cool White's spikes.

THAT is why you actually want curve overlap, and why all this talk of "mud" is nonsense.

Drew Wiley
17-Mar-2011, 11:53
Tell me something I don't already know. And heaven help us when we can no longer
get incandescent bulbs and have to illuminine our interiors with CFL's. We're right back
to square one and why different films have different inherent signatures - you won't
get much repeat business taking high-school shots with a harsh chrome film, unless
you have a lot of time to airbrush or photoshop out all the zits! You want a neg film
with complex mud deliberately engineered into it. But you don't want to go out and
try exact color rendition of certain product shots with a portrait film. In a still life studio you can select your objects and lighting to match the strengths of the film. Saturation is a slightly different variable than purity of color. Like I said long ago,
there are hues in nature which are extremely easy to replicate with a good set of
watercolor pigments, but almost impossible with any kind of color film or existing print
medium. But that's what we've got - a medium with just three spikes of sensitivity.

Ben Syverson
17-Mar-2011, 12:17
We're right back
to square one and why different films have different inherent signatures
God, it's like you're deliberately missing the point. I've just shown that Portra and E100G have extremely similar responses to even a challenging light source such as fluorescent. That's without any color correction! Even the most basic color correction would bring them even closer.

You don't shoot school portraits with chrome film, but that's because of its characteristic curve (contrast), not because it differs so greatly in spectral response.

The neg and chrome film actually have quite similar spectral sensitivities, which makes matching colors pretty easy. So go ahead and shoot your product shot with portrait film, because you CAN reproduce the full range of colors with negative film, and you'll have a greater gamut than slide film to boot.

If you want to match contrast of neg to chrome, you'll need to apply the chrome film's "S-curve." That's it.

It's an easy, easy, easy match.

Henry Ambrose
17-Mar-2011, 12:39
snipped a bunch.....

The scanner/enlarger doesn't need to have broad spectral response. The film does.

ABCs.

A scanner's or enlarger's spectral response that enveloped the entire gamut of the film would be ideal but not mandatory.

Ben Syverson
17-Mar-2011, 12:41
A scanner's or enlarger's spectral response that enveloped the entire gamut of the film would be ideal but not mandatory.
Not necessarily... That's why Drew built his laser-like system.

Film only has three sets of dye clouds, and they're not full spectrum. So if you can hit the peaks with three lasers, you could still reproduce 100% of the gamut captured by the film.

Henry Ambrose
17-Mar-2011, 12:47
snipped a whole bunch again.......

THAT is why you actually want curve overlap, and why all this talk of "mud" is nonsense.




How hard is this to get?

Drew Wiley
17-Mar-2011, 12:51
Been there, Henry. The problem with designing something for continuous spectrum is not only the extreme expense, along with significant maintenance issues, and very long reading times, but the sheer difficulty of reducing the data mathematically into a discrete set of variables which can be quickly handled by current software programs. Research devices like that once existed. Nowadays the point is to have a rather rapid
and predictable light source and a representative set of spaced readings. The inks for
your printer would have been plotted that way, probably with a 24-point reading converted to a CIE analytic geometry map. The downside is potential metamerism,
but that occurs once you shift the standard lighting source anyway. Nothing is perfect. I could go into some of the technical details, but now that you already think
I'm a rat, I'll confess to it and tell you that it was a fun prank for me to slip color samples from time to time to the operator of one of those enormous IBM units which
I knew the spectrophotmeter couldn't recognize - sorta like the color film and cool
white fluorescent analogy but more stealthy (good thing the sun isn't cool white!)
If you understand how the gear thinks it's pretty predictable to fool - eventually the
programmer slipped a message into the system whenever the software would be
bewildered by the sample, and it printed out, "are you kidding"!

Henry Ambrose
17-Mar-2011, 13:01
Not necessarily... That's why Drew built his laser-like system.

Film only has three sets of dye clouds, and they're not full spectrum. So if you can hit the peaks with three lasers, you could still reproduce 100% of the gamut captured by the film.

Not full spectrum but not sharp cutting either as shown in the curves you posted. There's more info there than can be described by 3 sharp peaks. That it gets left out is where the issue arises. Thing is its ideally not something to be left out or to be interpolated. The sharper the sampling the sharper the omission.

The other part is how clean is color in real life? Not very. Like how humans see "sharpness" they see color as "perfect" but its not. Our blue car looks the same to us most all the time but if we measured the reflected light from our car at different times of the day it'll be all over the place. We adapt to a "muddy" world.

As far as humans can see a well tuned digital color system can make plenty of colors to convince us that it looks just right.

I'll go even further and say that C41 is superior to E6 in that it is capable of capturing more color and more range in a typical scene.

Ben Syverson
17-Mar-2011, 13:10
Not full spectrum but not sharp cutting either as shown in the curves you posted.
The thing is, with printing/scanning, what you're interested in is presence of dyes. You don't get various shades of magenta dye—the dye is either there, or it isn't. No matter how complex the colors in a scene, at the end of the process, it's just some proportionate combination of three dyes which themselves don't have any color range. In that way you can think of it like CMYK printing—the inks themselves only come in one shade, but by halftoning and combining them, you can produce many shades.

But I certainly agree with your other points!

Drew Wiley
17-Mar-2011, 13:17
Digital capture certainly has its problems too. But with film there's at least enough
standarization that we are given certain kinds of graphs and charts as a clue, though
these are never perfect in presenting the complete picture. Another reason one doesn't want an absolutely narrow band peak in the film sensitivity itself is that any tripack output media, such as color printing paper, is not itself perfect in spectral dye response; so your peaks need enough area to cover the difference. When scanning or making color separations in the analog mode there are certain options afterwards to fine-tune the product which are unavailable when printing directly onto color paper. We've already wasted a lot of time just trying to get to first base on this subject, and that's just conceptually. Film dye engineering only gets way more complicated from there. This is a fairly gentle forum. You should go over to the Dye Transfer forum if you
really want to get eaten alive!

Drew Wiley
17-Mar-2011, 13:20
Again Ben, the dyes aren't perfect but have certain secondary features, some wanted
and some probably unwanted, which we might be able to trim out in certain cases.
Trying to tame them completely can backfire. Knew a lab which spent a couple of
million just to do this, plus patent fees, and basically bellyflopped. Concept is one thing, practical results another. I have no doubt that even the best dye and emulsion
chemists scratch their head from time to time and simply settle on what seems to
work as a cumulative result.

Ben Syverson
17-Mar-2011, 13:47
Sure, whatever. Personally I use cold cathode for scanning, which works perfectly fine. There may be certain advantages or disadvantages to using a source with a more focused spectral emission, but as you say, "concept is one thing, practical results another." And it's beside the point.

The point is, I have convincingly shown the following:

Modern emulsions don't have idiosyncratic or specialized "color palettes." Even a "portrait" negative film and a "saturated" slide film have very similar color response. The reason they appear so different is related to contrast (saturation is color contrast), not color sensitivity.
Negative films actually have a wider gamut than slide films. Partially due to the greater sensitivity of negatives, they can reproduce a wider range of colors. Slide films simply do not have secret mystical colors that you can't get with negatives.


Drew has responded with folk wisdom, imprecise generalities, misinformation and condescension. He hasn't quite given us an intelligible, rational argument to support his apparently unyielding beliefs.

Drew Wiley
17-Mar-2011, 14:21
Maybe we better leave it awhile, Ben. Go ahead and have the last word if you like.
You clearly don't even begin to have the background to understand the nature of the
problem, and apparently think you know more than the whole history of the medium.
As long as you get the results you want, that's fine, but apparently you do it with a
horoscope.

Ben Syverson
17-Mar-2011, 14:34
As long as you get the results you want, that's fine, but apparently you do it with a
horoscope.
Haha... This, coming from the guy who has yet to present one coherent, rational argument to back his vehement claims. The guy who sneers and patronizingly insists he knows something that we don't, but refuses to tell us what it is.

There's a word for that kind of unflinching faith in a belief, despite any shred of rational evidence: Religion. Now, I don't have anything against religion per se, but I'm not crazy about Drew's Church of the Supremacy of the E6 Film.

Me? My mind is open. It just doesn't let in nonsense.

falth j
17-Mar-2011, 14:49
holy crap...

I can’t even begin to imagine how these guys’ works, (pontormo, warhol, van gogh, picasso, eakin, titian, reubens, polloc, renoir, monet, cezanne, klimt, michelangelo..) would have turned out had they been born in our age, and had to fret over all the nuances of their paper, pigments, splitting hairs over bristle brushes, electrons, and dye layers.


After reading, and trying to digest the last 145 posts about changing to a different sheet of film, I'm afraid to charge my batteries, change my film type or snap another shot for fear of …. ?


Geez, I hope the physicists don't change the composition of their electrons anytime soon, or my old photo cells and new electrons, won't match the output with those old electrons used years ago in my cameras and scanner?


What’s an old man to do… ?


Whoever said,

too “much ado about everything”?

Drew Wiley
17-Mar-2011, 15:21
How did they do it in the old days??? Manet picked up a sword and got in a duel over
color theory - neither artist was skilled enough to inflict a mortal wound, however.
Van Gogh and Gaugin merely flung paint at one another and each stormed out in a rage. Carravagio actually killed someone, but more for personal reasons. It also helps
to have a little lead and cadmium poisoning, with some absinthe added in.

Robert Hughes
17-Mar-2011, 15:23
Van Gogh and Gaugin merely flung paint at one another and each stormed out in a rage.
Rumor has it that Gaugin, an experienced swordsman, cut off Van Gogh's ear during one of their squabbles, and VG (ear in hand) ran down to his local whorehouse for medical assistance. They thought he was crazy! Go figure...

Ben Syverson
17-Mar-2011, 15:33
Manet picked up a sword and got in a duel over
color theory - neither artist was skilled enough to inflict a mortal wound, however.
Wow, you are just chock full of misinformation.

Manet got into a duel with a friend named Louis Edmond Duranty, who was an art critic, and it was over a perceived snub, not color theory. Not only that, but Manet most definitely drew blood, and the duel "was so violent that both blades buckled."

They remained friends, however.

Drew Wiley
17-Mar-2011, 16:27
Thanks, Ben ... I'll defer to you on that one. I't been a long time since I read the story. But everyone gets the point. Absinthe and art talk don't mix.

Henry Ambrose
17-Mar-2011, 20:46
The thing is, with printing/scanning, what you're interested in is presence of dyes. You don't get various shades of magenta dye—the dye is either there, or it isn't. No matter how complex the colors in a scene, at the end of the process, it's just some proportionate combination of three dyes which themselves don't have any color range. In that way you can think of it like CMYK printing—the inks themselves only come in one shade, but by halftoning and combining them, you can produce many shades.

But I certainly agree with your other points!

Think about Fuji 4th layer films. How did it work if there's no more to get from anything but strict RGB? Scanner chips aren't sensitive to only a single wavelength of light. Population density of dye clouds does show up in the sampling when the sampling is not perfectly sharp cutting - and its not.

Then think about the area under the curves that show what portion of the spectrum the film responds to. Then consider hexachrome offset (or maybe that died a natural death long ago?) Or maybe the 10 color inks in some inkjets.

There's more to be had but it may be moot given what we already have with just RGB alone.

Ben Syverson
17-Mar-2011, 21:30
Henry, you need to be careful to distinguish photography (using an image sensor to capture reality) from reproduction (using an image sensor—such as photo paper or a scanner—to reproduce a film original).

In the first case—capturing reality—your sensor should be sensitive to every frequency of visible light. It may be more or less sensitive to certain frequencies, but ideally there are no big sensitivity gaps. But all those frequencies will get dumped into only three all-purpose buckets—Red Green and Blue.

In the second case—reproducing a film image—your sensor only needs to be sensitive to... three buckets. Because there are only three dye layers, you don't need a full spectrum light nor a full spectrum sensor to capture the "information" the film contains. For example, I made a photograph years ago on color negative that was exposed only with a laser pointer. Even though the laser itself emitted a tiny spike of red, the cyan dye left on the negative is no different than what would be left by a red stage light at a concert or a stoplight. Once it's on the negative, red is red is red.

Fuji's "4th layer" is a bit of a marketing gimmick. AgX is blue sensitive by default, right? That's why the top layer in any color film is blue sensitive. Then you need a yellow filter so that the lower layers only receive R&G light. Then you have a green-sensitized layer. Then Fuji has cleverly thrown in another "normal" blue sensitive layer, which doesn't get much blue light, but does get some cyan light that's not fully filtered out by the yellow filter. That gives the emulsion a tiny kick in cyan response. Then, finally, you have a special red sensitive layer.

Don't tell Drew, but this is an example of a film company actually increasing spectral curve overlap by use of a special layer, in order to cover a broader spectrum!

However, in the end, that cyan layer turns into a very faint magenta layer. You would think it should end up red, but... it doesn't. In other words, Fuji's "4th layer" is only there to aid sensitivity, and in the end, it ends up as plain-Jane magenta dye.

Hexachrome and 15 color inkjets exist to address gaps in the gamut of subtractive print materials, and are a bit of a tangent to this discussion.

NorthSands
18-Mar-2011, 06:02
All a bit heated, I must confess to getting halfway and hitting 'reply.' My main issue is the fact that C-41 films have noticeably more grain than E-6, with my work being mostly long exposures at night I find this a problem. Hopefully there'll be E6 processing available for a good few years yet.

Anyone have any contingency plans for if/when the chemicals dry up, so to speak?

Noah A
18-Mar-2011, 06:40
Interesting conversation...but to the OP, I don't think you should give up on E6 if you really prefer it.

I second (or third) the recommendation for Taylor Photo. It may be a haul for you but it's a good lab and seems to be committed to remaining in the film business. I normally shoot C-41 but i've had great E-6 done there too.

Or--and I understand your preference to deal locally in cash--you could send your film to Samy's. Their prices are very good as is their service. C-41 is $1.50/sheet for 4x5 (E-6 may be a few cents more, I can't remember). They're so cheap that I save money even after paying for shipping.

Henry Ambrose
18-Mar-2011, 07:13
snipped........ but this is an example of a film company actually increasing spectral curve overlap by use of a special layer, in order to cover a broader spectrum!



But this confirms what I wrote. The 4th layer does add to the color information in the film. And it is another container. So its not purely three buckets. The 4th layer gets dumped into more than one of the 3 buckets just as the overlap of the 3 layers does. I do understand what you're saying but what I'm saying is in real world its not so neat about how the color information is translated.

Ben Syverson
18-Mar-2011, 07:35
But this confirms what I wrote. The 4th layer does add to the color information in the film. And it is another container. So its not purely three buckets. The 4th layer gets dumped into more than one of the 3 buckets just as the overlap of the 3 layers does. I do understand what you're saying but what I'm saying is in real world its not so neat about how the color information is translated.
Henry, the "4th layer" does end up in one of the three buckets—it ends up in the Green sensitive bucket. So its net effect is that it changes the green response curve ever so slightly. It actually is quite neat and tidy.

At the end of the day, your negative only has three colors in it: CMY. That's why the spectral response of your scanner does not matter too much.

Again, you have to be very careful when you talk about film, because the process of photography and reproduction have drastically different concerns.

Drew Wiley
18-Mar-2011, 09:39
Anyone who wants to objectively study this can easily do so. Edmund Scientific industrial division (not the hobby division) sells some good manuals on optical design
and color separation. OCLI was a good source of technical data, but they are now
owned by someone else. Plenty even on the web on if you study the history of film,
either still or motion. Entire industries like prepress operate on RGB capture not only
of film, but potentially of actual subject matter (that is, as long as it didn't move).
There is an entire herd of elephants in the kitchen which my friends claim don't even
exist, not just one of them. It helps to have some actual hands-on experience with
this kind of work, or at least a knowledge of its historical importance. What I have
routinely done for several decades based upon firm principles acknowledged by several
entire industries isn't going to be overthrown by some one-upmanship geek talk.

Ben Syverson
18-Mar-2011, 10:20
What I have
routinely done for several decades based upon firm principles acknowledged by several
entire industries isn't going to be overthrown by some one-upmanship geek talk.


Another minor note - don't tell my problem is my darkroom. I don't use white light,
and my system is very analagous to the narrow-band lasers used in high-end digital
printers.

Well, which is it, Drew? Either you agree with what I wrote in response to Henry about not needing broad spectrum light for reproduction, or you disagree with your own darkroom practices. PICK ONE.

This is getting quite tiring. With all due respect, you need to put up or shut up. Let's see some real information instead of insults and posturing.

rguinter
18-Mar-2011, 10:31
holy crap...

...After reading, and trying to digest the last 145 posts about changing to a different sheet of film, I'm afraid to charge my batteries, change my film type or snap another shot for fear of …. ?

What’s an old man to do… ?



Blame me. I started it.

This was the sunset photo sequence on C-41 that had me questioning how to get back to my old favorite E-6 look.

But I am starting to like the new look and will probably shoot twice as much film now... C-41 and E-6 shots back-to-back.

At least until I use up my freezer-full stash of Chromes.

Tis "what an old man's to do."

Cheers. Bob G.

Drew Wiley
18-Mar-2011, 10:32
Ben - what you told Henry is correct. Some of things you've noted about original film
capture itself are not. I've done testing forwards and backwards on these issues, and
really am not going to waste time trying to convert a know-all-it. I have enough friends who are way,way over your leage in digital skills and even hold patents on
some of these devices and programs, and would find your notions strange. You claim
to be able to easily do things they cannot, but you probably wouldn't even understand
their language either. Like I said, I have no intest in converting you to anything. And
I wouldn't even be fishing around on this thread if it wasn't a topic of genuine practical
concern to me. I too am interested in realistically bridging the gap between chromes
and negs. I've done hard experimentation with this already, but it's a waste of time
explaining it unless people understand the basics. You might indeed have a worthy
contribution to the subject, and have already shown how simple curve and contrast
changes are useful. But once you get into hard decisions in the chroma domain, theres's a need to recognize some basics and the need to visually perceive them.
As I said, it's hardly my word against yours. Real facts are available from entire industries.

rguinter
18-Mar-2011, 10:34
P.S. And an old one on E-6 for comparison.

Bob G.

Ben Syverson
18-Mar-2011, 10:58
As I said, it's hardly my word against yours. Real facts are available from entire industries.
The "real facts" (the data) support my argument, and you've done nothing to convince anyone otherwise. The only thing you've done is write repeatedly that I'm so wrong, on such a "basic" level, that you can't even begin to explain why I'm wrong.

I guess the fact that you have friends with patents means I'm wrong. Maybe I'll "go take an art class."

I'm no know-it-all—I'm a "tinkerer" (not my word) and experimenter, and I'm constantly revising and supplementing my understanding of the world. What I like is empirical data—information gleaned from and verified by actual experimentation. What I don't like is someone making bold or theoretical claims and shirking empiricism.

When someone does this, they have positioned themselves against the goal of collective knowledge on this forum. That's bad enough, but when these people have the gall to insult someone who is actively trying to illuminate the discussion, it's beyond uncouth. It's a big F-You to the whole forum.

I know you're better than that, Drew. So please, either contribute to this discussion in a constructive way or at least politely refrain from insulting me.

Drew Wiley
18-Mar-2011, 11:21
This "empiricism" you want has been done over and over and over again. Just go study
it. I'm interested in making prints. You do what works for you; I'll do what actually,
visually, and empirically works for me and many others long before me. Respect for the
formum is also realizing that photography is a historic evolution, and that the rules which have worked for generations aren't universally thrown out just because some new or more convenient tools have been added. "Respect for the forum" also implies a
little respect for techniques which are so taken for granted that they are virtually
second-hand to many of us. And remarks like specific chrome or neg films having no
apparent signature or distinct look are not only misleading to any newcomer in this
business, but would undoubtedly shock anyone in Kodak R&D or marketing. Yes, ultimately our eyes are the most empirical evidence. If you can actually make neg
prints routinely look like chrome prints, just do it. The proof is in the pudding.

Ben Syverson
18-Mar-2011, 11:56
If you can actually make neg
prints routinely look like chrome prints, just do it. The proof is in the pudding.
I can... and do.

You need to scan film to make it happen. You can't do it in a darkroom. There is no way to match the response curve (contrast) with enough precision in the darkroom.

I never wrote that specific chrome or neg films had no signature or distinct look. What I wrote was that modern emulsions behave predictably enough that you can easily match negs to chromes (but not vice versa). That seemingly straightforward assessment is what caused your conniption fit.

Drew Wiley
18-Mar-2011, 12:09
Again, Ben, you're running on a bluff. How would you know what can or cannot be done in a darkroom? How do you know what or what not kind of experiments have
already been done along these lines, or how many people a lot more expert than me
have already thought them through? It is this kind of arrogance which elicits sarcasm.

Ben Syverson
18-Mar-2011, 12:31
Again, Ben, you're running on a bluff. How would you know what can or cannot be done in a darkroom?
Experience (I've spent plenty of time printing RA4) and common sense.

Printing photochemically involves taking one material with a response curve (a piece of film) and applying a second response curve to it (typically paper). That produces a "net" curve that is more complex than either individually.

While you can control that curve to a limited degree with exposure and filtration, you do not have 100% flexibility over it. To invoke an extreme example, you can't turn RA4 into direct positive paper (curve inversion).

It so happens that RA4 has a contrast-producing "S-Curve" that is vaguely similar to the S-Curve of reversal film. That's why making an RA4 print of a Macbeth chart negative will look acceptably similar to a transparency of the same chart. However, as you're all too aware, it won't be an exact match. It's even less of a match if you strike an Ilfochrome from the transparency, because that paper has a curve too.

To match the negative to the chrome, you need to make the response curve plastic enough to manipulate selectively. Anything is possible, so I suppose there might be some way to do this in the darkroom using extremely laborious and unconventional techniques. I would call that "The Hard Way."

Drew Wiley
18-Mar-2011, 12:42
Whether it's the hard way or the fun way, easier or more difficult depends upon ones
preferences, gear, and experience. There are all kinds of options out there - some were once common and are now forgotten (because PS does indeed improve workflow
for many of these task), and more options still being developed. People who work on
these tasks from the standpoint of scanning and digital seem to put in just as much
time per print for high-quality results as I do in the darkroom. Sometimes an image
just falls into place with basic controls, sometimes a lot of work has to be invested.
Sometimes we just set aside a neg or chrome until we learn a new trick or a better
printing option comes along. I know folks who operate digital labs with staggering dollar
investments in them, but prefer to do personal work in the darkroom, the old slow way.

Ben Syverson
18-Mar-2011, 12:57
Hey, I love photochemical printing and analog film. That's really not the issue. No need to pull the "in my day" card. I'm on your side.

The issue at hand is matching negs to chrome. It requires curve control, which is very straightforward in Photoshop. If you can execute that degree of control in the darkroom, hey, more power to ya. The point is, however you get there, you can match negs to chrome using curve control and maybe a dash of selective color correction.

Drew Wiley
18-Mar-2011, 13:12
For a darkroom junkie like me, there is indeed a kind of satisfaction in solving a difficult
problem and making it efficient. Any number of tasks where were once very laborious
in the darkroom I have found shortcuts to, including certan versions of sophisticated
curve control. For the average worker, however, it is indeed far easier is PS. And this
isn't an either/or "analog vs digital" subject. There are innumerable hybrid options to be
explored as well.

Ben Syverson
18-Mar-2011, 13:25
Exactly. The hybrid workflow is precisely what makes it possible for average photographers to do things which would be extraordinarily difficult in the darkroom, such as... matching negatives and slides.

Drew Wiley
18-Mar-2011, 14:04
Again, what constitutes an acceptable "match" is a relative expression. For a certain
class of images of hue families, a fairly straightforward workflow might be perfectly good. But then you up the ante or encounter a new chroma scenario, and the old rules
don't work so well. It was exactly the same situation back when chromes were printed onto RA4 via interneg - fairly easy to get routine commercial-quality results, but a nightmare to finetune, even in those exceptional circumstances the lab was willing to put in the extra work. Getting "snap" reminiscent of a chrome form a modern neg using
PS curve control is easy enough, and getting overall color response isn't all that hard,
but it may take quite a bit of advanced use of the tool kit to match certain hues which might be to particular images. There a pros and cons all over the place.

Ben Syverson
18-Mar-2011, 14:34
Sure. There are also variations in both negative and reversal emulsions, as well as development variations. You can't sit back and push a "match" button. You may have to do some selective hue adjustments in addition to Curve work. I can do a 90% match in under a minute, while a 99.99% match might take an hour or two.

But make no mistake. You can produce two matching prints, to whatever degree of precision you desire.

SCHWARZZEIT
18-Mar-2011, 16:06
How can you match a negative to a slide if the color separation of subtle hue and brightness differences isn't recorded on the neg in the first place?

It's one thing to get a meaningful response from all the patches of a color checker over a range of f-stops. It's another thing to get enough separation from two tones that are very close. It has been my experience in professional drum scanning that if the separation is lost in the statistical dye cloud fluctuations that we see as grainy texture then the scanner captures virtually the same RGB pattern. There is nothing you can do in Photoshop to recover that tonal separation.
In direct comparison reversal films have better color separation than most color negs. Transparency films use a large density range to capture a narrow exposure range. Slight tonal differences in a subject are transferred into larger density differences than on color neg film where a huge exposure range is compressed into a smaller density range. Once the density differences fall below the threshold of the grain fluctuations you're losing tonal separation.
I've seen excellent color separation from Ektar. Portra 400 NC-3 on the other hand was very muddy in that regard. None of these films separate tones as fine as a digital sensor of the same size. But you can increase the probability of tonal separation by choosing a larger format and thereby increasing the capture area.

-Dominique

Ben Syverson
18-Mar-2011, 16:46
How can you match a negative to a slide if the color separation of subtle hue and brightness differences isn't recorded on the neg in the first place?
ARRRRRRRRRRRRRRRRRRRRGGGGGGGGGGGGHHHHHHHHHHHH

Read every page of this thread and then ask this question again.

Drew Wiley
19-Jun-2011, 16:19
I'm resurrecting this thread (minus the semantics debate) simply because I've finished enough definitive testing to come to some answers which might be practical
to most people using current Kodak color neg films. My anticipation regarding the
characteristics of film responses based upon the dye curves allowed me to cut to the
chase very quickly, identify the major potential problems, and start objective tests
for distinct remedies. At the core of the question was whether or not negatives can
be made to resemble chromes simply by contrast tweaking. As anticipated, it all
depends. Even after you null out the orange mask, once you significantly alter the
contrast you begin reshaping the geometry of the three respective dye curves, not
only per individual peak, but also in relation to one another. The way this occurs is
quite a bit different from how modern chrome films respond and is also at a practical level somewhat different between specific neg film. I was particularly
concerned about reproduction characteristics Portra 160VC versus its successor,
Ektar 100. ... (next post) ...

Drew Wiley
19-Jun-2011, 16:27
As I expected, the main problem with Ektar is in specifically how blue becomes
reproduced. The question is definitively not whether this film can render a good blue
or not, but the fact that its complement, the yellow dye layer, is present to some
extent in virtually every hue in an image, and a problem here will affect the image
overall. This is a bit easier to deal with in analog fashion simply because a range of
formats can be more universally accommodated using a standardized set of procedures. With scanning, the smaller the image the bigger the problem, because
your statistical sample of minor but significant changes in the dye curve slope can
be incorrectly interpolated readings versus actual neg content. In other words, with
something small like 35mm the importance of a very high quality scan because much more important than just dynamic range or detail acquisition. I conducted
scanning experiments only so far as to confirm this fact, and will leave it to others
to address PS questions of correction .... (next post) ...

Drew Wiley
19-Jun-2011, 16:34
My way of looking at the name of the game is to identify the sweet spot in how these films perform and optimize in that regard. And again as expected, Ektar is
very impressive once yet hit this sweet spot, and will best mimic the effects of a chrome film, but is more finicky than the more generalized Portra films per se.
In particular, you need to start out on the right foot in the first place regarding blue
balance, because once the relation between the respective dye curves is messed up,
it will be quite difficult to correct either analog or in PS, simply because you've
altered the relative geometry of the curves and not just their exposure density relative to one another. In this respect, modern color neg films are still quite distinct
from the way current chromes respond. The most important practical tip in this regard is that if you encounter a situation where your color temp meter would call
for correction, apply this correction on the spot. Don't expect Ektar to forgive you
like old-school Vericolor. For me it's as simple as packing along an 81A and 81C
filter and using them religiously when needed. The improvement seen in actual
prints, not only in balance but in broad-spectrum color purity, is significant. I hope
this helps someone.

Drew Wiley
19-Jun-2011, 16:42
This is my fourth post or paragraph in sequence, so please read the previous three
to understand the practical implications. Once can certain ignore all this as get from
point A to point B, but having a film like Ektar around is like owning a Porsche and
only knowing how to drive it like a Chevy. With a bit of forethought these newer
color neg films can be optimized to give them an exceptionally useful niche, and if
needed behave much like chromes, but not without knowing some tricks.

Drew Wiley
19-Jun-2011, 18:37
(Fifth post after interruptions between layers of paint drying on the front door): The
importance of simply recognizing the principle that curve geometry changes as you
change contrast (either PS or analog masking) will spare you from a rude awakening
if you happen to establish a protocol for one format and then try to apply it in the
same way to a different size format. With film masking, you might simply have to change the development time of the final mask if you have a straight-line technique.
But digitally, you might have to significantly back up your procedure and establish
some new rules. Just sticking with LF film would make life relatively easy, but if you go back and forth between this and a small camera, you will probably need to recalibrate the entire workflow, starting with the scan. Should be worth the
extra effort. These new films are wonderful.

atlcruiser
20-Jun-2011, 07:53
At first I thought I would learn much from this thread about color and color theory...now I can just lean how to argue on the internet. Still and all educational :)

David Higgs
20-Jun-2011, 09:27
I'm still learning with Ektar, when it works, it really really works, but when it doesn't - I get really stuck with colour balance.

The Portra stuff appears much easier, but you never hit the highs that Ektar can give.

I'm looking at the learning curve and the top seems a long way away!

Drew Wiley
20-Jun-2011, 09:46
David & David - you can pretty much ignore a lot of the semantics debate in this lengthy thread and just skip to the how-to results. Like I suggested, Ektar is like a Porsche with power steering. Very small changes in color balance will steer it rather
abruptly, but when it tracks down the road correctly, it's an incredible performer.

atlcruiser
20-Jun-2011, 14:35
Thanks Drew..i bought a minolta color meter and a ful set of CT and CC filters. With portra 160NC i get excellent results and i have a box of extar100 in 810 that i need the time to mess with

Drew Wiley
20-Jun-2011, 14:49
David - so far I've calibrated Ektar from 8X10 down to 6x7, which basically just involved minor gamma changes in the contrast mask via dev time. Magnification ratio
significantly affects how not only the contrast responds, but internal chroma
too. I'm obviously working analog, but these kinds of corrections should be fairly straightforward in Photoshop too, or even easier.

Ben Syverson
20-Jun-2011, 15:59
I don't regret much in my life, but I do regret the time I've recklessly bled away on these types of inane internet arguments.

At the end of this epic (and epically stupid) discussion, you should know that Drew and I both agree that C-41 can be made to look "much like" E6, though it is "tricky" (in the darkroom, anyway).

Drew Wiley
20-Jun-2011, 16:22
Nothing tricky about it in the darkroom at all, Ben. Rather simple cure to anyone with
elementary sheet film developing skills and a registered contact print frame. Digital or analog same issue - understanding how the film behaves and how to best plan for it.
The operator of a Lambda, Lighjet, or Chromira would face exactly the same issues trying to optimize the look of the film. Just punching up the contrast via PS won't accomplish the same thing because it inherently skews the film curves relative to one
another - you will be effectively reshaping the curves differently by selectively exaggerating just one part of the curve slope versus another, and the three dye curves aren't either identical or symmetrical in the first place. Someone else already
tried pointing this out to you with respect to scanner sampling error. With Ektar these
kinds of errors build up faster because its a narrower-range curve with more saturation. Like I said, you can always drive a Porshe like a Chevy, but it you want to
optimize what this film is capable of, then you need to understand it in a specific sense. And once you start discovering the sweet spot, Ektar is more capable of duplicating the results of a chrome film than Portra films, with the added advantage of
the wider latitude of a color neg. But if you ignore its idiosyncrasies, you're going to
face certain serious reproduction issues that could otherwise be prevented. I have
no interest in complicating your own workflow, but please give others a chance to
learn from my own very specific testing without confusing them with the idea that it's
all a wash. It isn't.

Ben Syverson
20-Jun-2011, 17:56
I'm the one confusing people? Fine. Honestly, I'm done. I would strongly prefer that a moderator delete all of my posts from this thread, but I wouldn't torture them by making them revisit every page of this train wreck.

From here on out, feel free to discuss "color casts" in color negative films (-"Ektar is too blue!" -"No, it's too red!") and other mystical superstitions, but you'll be doing it without me.

atlcruiser
20-Jun-2011, 18:27
David - so far I've calibrated Ektar from 8X10 down to 6x7, which basically just involved minor gamma changes in the contrast mask via dev time. Magnification ratio
significantly affects how not only the contrast responds, but internal chroma
too. I'm obviously working analog, but these kinds of corrections should be fairly straightforward in Photoshop too, or even easier.

I would love that info!

My plan is to contact print either extar or portra and I am leaning towards extar in 810.

I got so frustrated with trying to color correct I essentially gave up clr neg film. There was another thread on here that really opened my eyes to the issue of exposing all of the layers correctly to avoid all of the PS guesswork and frustration.

Now I find that when I meter the color then correct for temp and correction I get negatives that are as close to perfect as I have ever had with clr negative film when scanned.

The next step is to make this happen in the darkroom.

david

Drew Wiley
20-Jun-2011, 19:41
I'm not in this to discredit anyone or disprove this or that. I'm in it for completely
selfish motives. The stimulus of the whole topic over this thread and analogous ones
(like the thread on correct ASA for Portra) has been helpful in gelling a series of
experiments already underway; and I thank Ben for this. What I get out of it is a
dramatic reduction in time, money, and trouble getting to a practical endpoint.
And given the fact that members of this forum have been kind enough to put up with
my often abrupt and unpolished hillbilly manners, the least I can do is share the
results for the sake of anyone who might also wish to save money or grief in an
analogous manner. Second, I'm not really interested in replicating the results I get
with chrome film. Each media has its special idiosyncrasies and look; but what I
am interested in is a step forward with results which are actually better and cleaner
than I typically get printing chromes!

Drew Wiley
20-Jun-2011, 19:50
David - the best way to start (if you haven't already done this) is to make a master
negative out of your MacBeath chart or other standard color reference, expose it either under tightly controlled studio conditions, or outdoors under mild white subdued sunlight using the exact correction specifications of your color meter, along
with precise exposure compensation per exact 100 ASA. Everything we do needs some kind of calibration, and Ektar is fussy about getting off on the right foot to begin with... My own tests were standardized to the latest Fuji CA II paper, which
seems to hit its sweet spot with about a 3X or 4X magnification ratio per the standard neg (though actual subject shots will obviously differ in contrast). I control
this up or down through silver masking, but if you attempt this, the tricks are a bit
different than masking for either b&w printing or Ciba. You might have a problem
with too much contrast if you are contact printing. You can take a step down using
portrait paper (Type P - discontinued, but still plenty around), or take a step up in
contrast using Super C (also discontinued but still available). These papers are not
quite as clean in chroma as the new Type II but are still pretty good anyway. Or you
could just shoot Ektar selectively in more limited light ratio settings analogous to
how we have successfully shot chromes all along. You'll get on the road pretty fast.

Corran
20-Jun-2011, 20:11
I would love that info!

My plan is to contact print either extar or portra and I am leaning towards extar in 810.

I got so frustrated with trying to color correct I essentially gave up clr neg film. There was another thread on here that really opened my eyes to the issue of exposing all of the layers correctly to avoid all of the PS guesswork and frustration.

Now I find that when I meter the color then correct for temp and correction I get negatives that are as close to perfect as I have ever had with clr negative film when scanned.

The next step is to make this happen in the darkroom.

david

Link to thread? Sometimes I get great color negs, sometimes they suck. I am reading all I can so I'd love to see that thread.

atlcruiser
20-Jun-2011, 20:30
Link to thread? Sometimes I get great color negs, sometimes they suck. I am reading all I can so I'd love to see that thread.

http://www.largeformatphotography.info/forum/showthread.php?t=76385&page=3


I am in atlanta if you wanna hook up and play with the toys

Drew Wiley
20-Jun-2011, 21:12
Once you've gotten your standardized test neg, use it to balance your paper exposure under the colorhead. As your tests get closer and closer, it will seem that
every little adjustment is like power steering. At the end, even a 1cc adjustment will
make a visible difference. All the white to black patches of the Macbeath chart should come out almost perfectly neutral under a standard viewing light. Every other
patch should come out clean, vibrant, and without any visible contamination from
another hue. There shouldn't be any visible bias, either blue or red. Sounds impossible for a color neg film? Well, that used to be the case. Not with Ektar,
provided your standard neg is spot on. Enjoy your new Porsche.

Corran
20-Jun-2011, 23:33
http://www.largeformatphotography.info/forum/showthread.php?t=76385&page=3


I am in atlanta if you wanna hook up and play with the toys

Thanks, I will be reading that thread...
I live waaaay down here near Florida, otherwise I would take you up on the offer. What are you shooting? I just started with 4x5 and home development about 2 months ago.

Jan Normandale
21-Jun-2011, 10:49
wow.. what a thread... clearly there are two separate worlds going on. I frankly don't really care all that much because I'm not reproducing image colours accurately and I don't think it's really possible despite assurances by film manufacturers. The whole shooting, ISO rating, lens character, film character, film processing accuracy, printing process using chemicals at varying temps and ratios , human error and idiosyncracies all affect the final image throughout these steps. In my opinion it's "voodoo"

So I'll go back to shooting and developing with questionable accuracy then dealing with scanners that probably should be calibrated over individual films with different planes of curvature, covered with dust and other junk or printing in a darkroom where I have chemical process that is affected by temperature variations and on and on.

Point I'm trying to make is that the image "right" or "wrong" is what people look at and often the most heavily manipulated ones or cross processed ones or over exposed ones are favoured. C prints, darkroom prints or inkjet prints all add one more level of variable.

My thanks to Ben for some very interesting 'chart analysis' and explanatories. It confirms what I suspected but now know thanks to his lucidity. I'm back to my camera now. I'll leave the charts and hope to compose a photograph that will work. FWIW I obsess over lens character.. maybe a thread on that topic next? I'm sure it has the potential to exceed this discussion. Then again maybe not.