PDA

View Full Version : Digital Camera R&D...



Bobby Sandstrom
16-Dec-2005, 12:47
Does anyone know if any company is working on a digital camera that will produce photos that have more depth... like the depth you get from film. I think most would agree that digital files look flat or 1 dimentional when compared with film. Please, this is not a digital vs film flame starter question. I'm sincerely interested in having my question answered as directly as possible.

Thanks

Bill_1856
16-Dec-2005, 14:55
Your premise is flawed.

Jim collum
16-Dec-2005, 15:11
i'd agree as well. i can take perfectly flat /boring images with either film or digital. and in the hand of someone who knows what they're doing, you can get vibrant images with incredible depth from either one

Brian Ellis
16-Dec-2005, 16:24
"I think most would agree that digital files look flat or 1 dimentional (sic) when compared with film. "

I think you think wrong.

Jorge Gasteazoro
16-Dec-2005, 16:26
I think the problem is not the capture but the oversharpening one usually finds in digital prints. I dont mean sharpening artifacts, but the edge to edge impossibly sharp images that IMO take away the feeling of depth found with analog, specially with contact printing.

With PS not only can you do overall sharpening but by using layers you can sharpen small areas. With film, even if you are doing contact printing, there are always small areas or things which are unsharp (even using the best focusing and DOF techniques.)

IMO those doing digital capture should be very aware of the near/far relationships and DOF, in fact I will dare say they should introduce some out of focus areas. Those doing film scanning should learn to leave well enough alone and know when to stop using sharpening techniques. In fact this is one of the things I look for when I am asked if I can tell whether a print is digital or analog, impossibly sharp prints from edge to edge are 9 times out of 10 digitally manipulated.

David Luttmann
16-Dec-2005, 16:58
Jorge,

That is probably the biggest problem with people processing images. They go WAY overboard with the unsharp masking thinking that if they make the image REALLY sharp, people will be impressed. I for one have never been impressed with the amateurish looking halos and artifacts on ridiculously oversharpened images.

There is a fine balance to aware of. I prefer the natural look that many would say is more "film like" in appearance. With some, they go so far overboard with USM that the shot looks almost like a video image!

QT Luong
16-Dec-2005, 17:05
Jorge makes a good point. Another reason why digital captures are often said to be flat is that when using raw formats, the file is by default more linear than film. But raw files need to be interpreted, the same way as a negative does.

Bobby Sandstrom
16-Dec-2005, 17:28
Like I said, I don't want to argue here. I'm talking photos taken of the same subject one immediately after the other using both digital and film cameras. The results are obvious. The digital file is dimensionally "flatter" than the scanned film file. That's even after tweaking the digital file in photoshop using various techniques including midtone and global contrast to make them match as closely as possible. If someone would tell me how to post a sample I'll be glad to. I believe it will be obvious to all who see it.

David Luttmann
16-Dec-2005, 17:39
Anyone who says they can tell on screen the difference between the two is either misinformed, or blissfully ignorant.

Bobby Sandstrom
16-Dec-2005, 17:53
Dave, I'm neither misinformed nor blissfully ignorant. (Thanks for the vote of confidence though.) However, I'll be happy to prove that of you if someone tells me how to load a jpg without using a link.

David Luttmann
16-Dec-2005, 18:11
Nothing to do with confidence. How about some of us upload some photos and you tell us whether they are digital or film. I've played this game before....and haven't lost yet. Care to try?

Steven Barall
16-Dec-2005, 20:03
As far as the original question goes, I believe you are looking at a problem with user error and not any problem inherent with digital media.

That said, someone out there a lot more smarter that I can comment on whether they believe that the difference between digital capture and LF is the same kind of difference that exists between small format film and LF. It might be that there is nothing wrong with the digital capture but rather that LF has the potential of just being so much better.

Have a nice weekend all.

Bobby Sandstrom
16-Dec-2005, 20:47
No problem dave. You put up 2 photos taken at the same time with the same lens on both digital and analog cameras and I'll be happy to tell you which is which. Now, if someone would tell me how to load a jpg onto the site without a link I'd be happy to post such a comparison and I would venture to guarantee that anyone with half an eye will be able to tell the difference. In fact, the file is from an article that touts the resolution disparity between 35mm film and a digital capture stating the digital file has higher resolution. Maybe so, however the analog photos have much more apparent depth.

In fact dave, I would venture to guess you've never seen 2 files side by side of the same subject taken at the same time with both digital and analog cameras. If you had, you'd not be arguing this. That goes for any other person in this thread that claims there's no difference.

Bobby Sandstrom
16-Dec-2005, 21:38
Here's a picture that clearly demonstrates my point. This image shows 3 shots of the same building. The first is with the digilux 2 the second with an m6 shooting PanF and Agfa 100. If you can't see the difference in apparent depth between the first(digital) and the 2nd and 3rd, then lets just drop the subject and forget I ever asked my initial question.

I apologize for linking to this photo as it is from a photo.net article/review. But, it illustrates my point perfectly even though the photo is used to compare the resolution differences between the 2 fomats. If you want to check it out you simply go to photo.net and look for the digilux d2 review.

http://photo.net/equipment/leica/digilux2/100BWfilmvsD2.jpg

tim atherton
16-Dec-2005, 22:31
Not perfect by any means - working with crappy little 72dpi images, but:

http://gallery.photo.net/photo/3955931-lg.jpg

the file the digital image comes from may or may not go through various stages in either the Leica software and/or photoshop. Similarly, the scanner software will sharpen and adjust the scan, probably based on a canned profile for that film. So it's taken what would lookj like a pretty flat/linear raw version of that file and applied some curves and contrast to it.

The result is that an image from the digicam that hasn't been tweaked in photoshop by someone who at least vaguely knows what they are doing will probably come out looking a little flat (especially as most of the auto software is geared more towards colour than B&W), whereas one from a scanner will most likely be using a profile that gets it looking closer to the film "look"

But it's pretty simple to take a digital camera file and - if you know at least some of the basics about converting rgb top greyscale properly, applying curves etc, very quickly get away from the "flat" (or at least more linear) look you complain about and more like the one from the film scanners.

Aside from issues of grain, I could pretty much guarantee you that if I had hi-res files of all three, I could get the digicam file looking close to identical to the other two in terms of contrast and what you call "depth"

tim atherton
16-Dec-2005, 22:42
Bobby, if I can make it clearer.

It's just a difference between what adjustments the scanner softare and the camera software have applied to the files.

Each of those two types of images (scnner vs camera) have had (or not had as the case may be) different things doen to them.

You are comparing bananas and plantains as it were..... not two radically different things, but you aren't conmparing images wehre all other things are equal

Jim collum
16-Dec-2005, 23:15
http://www.jcollum.com/fm/100BWfilmvsD2.jpg

as tim said... a quick bit of work on a 72dpi image

David Luttmann
16-Dec-2005, 23:19
Bobby,

I use high end digital, MF & LF. All of my film work is Imacon scanned by myself. Now you're example is just the type of amateur stuff I'm referring to. Aee you seriously comparing a digilux point and shoot with full size DSLR chips? That's beyond funny my friend.

Here's what I suggest you do. Go out and do some tests with REAL gear rather than playing around with cheap consumer digital stuff. Figure out what you're doing wrong with your digital workflow that is giving you flat prints. Because if you honestly think that a screen size image will be discernable between a DSLR (not your Digilux toy) and film, than you indeed have much to learn.

I won't waste anymore more time with this. I believe others in this thread have pointed out the same.

Best of luck on your journey.

David Luttmann
16-Dec-2005, 23:24
Jim,

you beat me to it. It appears Bobby doesn't quite understand the differences between scanner and digital workflows.

Thanks for posting the cleaned up image. Now even the Digilux doesn't look to bad against the M6.....and the Digilux is junk compared to a decent DSLR.

tim atherton
16-Dec-2005, 23:36
darn I can never get these things to show up...

here's the link

http://gallery.photo.net/photo/3955931-lg.jpg

http://gallery.photo.net/photo/3955931-lg.jpg

Bobby Sandstrom
16-Dec-2005, 23:41
Dave, I'm insulted by your demeaning comments. I have a Canon 1ds, an M6, an F5, a 501CM, and 4x5, 8x10, 14x17 view cameras. I know photoshop and proper digital workflow better than most. Don't judge me my friend. You don't know who you're talking to. My important film work is drum scanned when not printed in the darkroom. If you really knew what you were talking about you wouldn't be so quick to boast of your Imacon. I have 7600 and 4800 printers dedicated to matte and glossy type papers respectively. I provided an example that is indicative of the same disparity that using a full frame DSLR would show. I know because I've done tests myself with my own gear. I asked a simple question to begin with but somehow always seem to attract a bunch of yahoos whenever I post. I think I'll refrain from doing so in the future. Thanks for the turn off.

Jeff Moore
16-Dec-2005, 23:52
You sure do a lot of arguing for a guy who claims of not wanting to argue. Also, referring to others who disagree with you as yahoos isn't very appealing either.

"I think I'll refrain from doing so in the future."

Thanks; much appreciated.

tim atherton
16-Dec-2005, 23:54
Booby,

you can get into all the mien is bigger than yours arguments all you want, but it still seems fairly clear from what you say that you still don't seem to quite understand the difference or recognise how to adjust the different types of images correctly.

I work with this all the the time and, as I say, apart from such things as differences in grain, I can get fairly easily get the same image from say a hi-res scan from 35mm and a 12mp camera (or whatever) to look close to identical in terms of contrast, "snap" or what you call "apparent depth", tone etc - it isn't really that difficult.

One quick question - how are you converting your RGB digital camera files to greyscale?

tim atherton
16-Dec-2005, 23:55
"Booby" really was a typo... it's late :-)

Bobby Sandstrom
17-Dec-2005, 00:03
Tim, I didn't see your post before my previous one. Believe me, I see the contrast differences between the images. I know that contrast is what creates the illusion of depth. I'm saying that starting with high quality images with careful photoshop work including tweaking contrast, sharpness etc. so the images of a digital capture and a scanned piece of film match as closely as possible the digital image still looks flatter. It's subtle, but it's noticeable. I know that most scanning software will automatically sharpen an image some even when you think you have sharpening turned off. I know that RAW captures absolutely need to be worked... I've worked hundreds of them. I have enough experience to level the playing field as much as possible between the two formats to make an honest assesment, yet, to my eye, there still remains a difference. I know some very respectable photographers that feel the same way. I'll stop posting now as I don't want to beat a dead horse with this subject. Thanks to anyone in the thread that tried to answer the question in a respectful way.

Jim collum
17-Dec-2005, 00:04
don't understand why offering you an explanation for what you are seeing classifies me as a yahoo. the work really was a matter of about 3 minutes. the grain doesn't match.. that would take a little longer.. but still can be done

Bobby Sandstrom
17-Dec-2005, 00:29
Jim, I'm not calling you a Yahoo. Actually I'm referring to Dave Luttmann and his demeaning post in this thread as well as a couple of other folks from different threads in the past that I've posted questions to. My bad. I should have been more clear. It really ticks me off when you get guys that assume you're an idiot and mouth off without knowing the capabilities of whom they're slamming. You did a fine job with your touchup and I appreciate your constructive reply.

tim atherton
17-Dec-2005, 07:08
Bobby , again, I'd be itnerested to know what method you are using for converting rgb to greyscale?

Matt Miller
17-Dec-2005, 08:16
"I think most would agree that digital files look flat or 1 dimentional when compared with film."

For 90% of the digital work I see, and 100% of my own limited digital work, I agree with you Bobby. With my few hundred dollars in film gear I can easily produce photographs with depth, with my few hundred dollars in digital gear I can not. They all look flat. If I had thousands to spend and many hours to practice, maybe I could produce photos with depth in digital. I don't though, so I can't.

Frank Petronio
17-Dec-2005, 08:25
I mix my digital and film work together and I defy you to tell which is which. If you want to argue about the depth of a toned fiber B&W print versus an inkjet print I'll conceed that there is no digital medium that can match the traditional print, but for comparing scanned film versus digital captures, it's all Ones and Zeroes, and I can move those numbers around anyway I please. Ever hear of a "S" shaped curve?

David Luttmann
17-Dec-2005, 08:31
Bobby,

Post your 1Ds sample then. My apologies if you're getting insulted. However, I do know who I'm talking to. This is a common issue that a lot of digital users complain about when they are first starting out with their Photoshop work. I teach a RAW workflow workshop at a local college a few times a year. I find that the people who mention this issue no longer have the problem once they understand how to process their RAW images. B&W conversion is actually more difficult than color. How are you converting your color work to B&W?

Your issues are a matter of Photoshop workflow, not "depth" differences. Because each film has a particular curve characteristic, I believe what you're trying to do is mimic the light response curve you see for each film and find a digital counterpart. Because you haven't found it, you're equating this difference to a lack of depth. As well, I believe that you appear to expect some grain in an image and by not seeing it, you feel the digital source may be "flat."

The depth difference you're describing sounds like nothing more than contrast difference because of this curve. I know you're disagreeing with this, but considering you appear to be the only person on this thread experiencing this particular problem, I would say that this issue is in your workflow, rather than an inherent depth difference between film and digital.

Let us know how you converted yor B&W. As well, post a 1Ds sample. Maybe a color & B&W explaing why you think it lacks depth.

Bets regards,

Bobby Sandstrom
17-Dec-2005, 09:54
Ok. Let me try and clear this up. I've been using photoshop for probably 8 or 9 years. I understand all of the benefits of doing whatever work deemed necessary in the RAW stage to minimize data loss. I know that every move made including simply switching color modes produces a loss of data and thereby decreases image quality to some degree. I use adjustment layers to minimize this effect by giving me the option to go back and make non distructive changes. To convert my images from Color to B&W, typically I first look at the individual channels then go into the channel mixer and play with the percentages to find an appropriate balance or simply use 1 channel if I like it. If I wish I can mask different parts of the image and blend different mixes. I have local and global contrast control using masked curves. I often play with midtone contrast in the shadow/highlight dialog box and sometimes apply either globally or locally via masks. If I wish I can apply grain either scanned from film or generated via the noise filter and applied wih a blend mode. All of the above subtley controlled to taste with opacity. I know a slew of ways to shapen an image, from very advanced methods with ultimate control to simple USM. I think anyone with any taste quickly learns the perils of oversharpening. If anything, I might be guilty of undersharpening. I can go on but I think you get my point. I really do know how to control my images. All this said, I may have been misconstrued because of the simplistic way I posed my question to start with. I'm not talking general global or even macro contrast. I believe there is something on a micro contrast level that I get from film that creates a subtle yet to my eye noticeable difference when I view a scanned piece of film vs a straight digital capture. Maybe it's the way light bounces off the individual pieces of grain, I don't know. However, at this subtle stage things can get subjective very quickly. I'll take responsibility for the confusion because of the way I asked the question and for the very obvious posted photo. I should have been more clear. I was hoping to hear of a newer capturing technology, something other than ccd and not as impractical as a scanning back. Something along the lines of a Foveon or what have you. Sorry for the heated thread.

All the best

Bob

Jim collum
17-Dec-2005, 10:09
one of the things i've found is that grain itself gives an added impression of depth and additional resolution. when you photograph an object (even with 8x10 or 11x14 film), there remains detail that is not captured. If you look at a digital image at 100% (at least with the bayer pattern sensor), that detail stops, and there's a feeling of smoothness. if you look at the original two film images vs the digilux, you'll see the grain adds an additional impression of detail to the surfaces of the buildings, while the digital building looks smoother. adding grain/noise to the digital image, adds the impression of that detail (although a 'false' detail in both cases). in almost all of my images, i add a very small layer of noise (masking shadows, highlights, and decreasing the effect for water and sky).

bglick
17-Dec-2005, 13:18
Bobby, there is a lot of very experienced people on this list, it's unfortunate members resort to name calling or insults when someone asks a question. But it's an open forum, and this how some people react, but the good news is, there is still very many generous and knowledgeable members, so try not to be discouraged. It's been this way since I started lurking in 1999.

As to your question....I shoot digital and film of all formats also. And I think a previous poster somewhat touched on my response. The obvious difference in depth you see in these images is the lack of contrast in the Digital image. I will expand though, as I see many variables in this equation.

Here is my take on the subject. Digital sensors record RGB, nothing else. And even then, since each set of 4 pixels only records 4 data values, i.e. 1 blue, 1 red and 2 greens, you are starting with 4 data values, whereas the other 8 need to be "interpolated" in software to acheive the 12 data values for the 4 pixels, so with no uprezzing, the interpolated raw file is at best 33% recorded values. Of course if the subect colors do not contain green, the numbers can fall apart very fast. (more below)

Now, this shortcoming can be "helped" by bombarding the image with more pixels, as the more pixels, the more information to work with. Hence why digital sensors keep growing. So far, this explains color digital, not B&W digital. Now, taking these "weakened" digital files and converting them to B&W, is a second form of interpolation you are asking the software to perform. Quite a task indeed, vs. B&W film.

Next, you are comparing this finished "heavily worked" digital file to B&W film which is specifically designed to record B&W, no interpolation required. Also, unlike comparing color digital to color film, B&W film vs. digital cpature further favors film, as B&W film records much higher resolution then color film. So in your comparison, digital capture has been severly handicapped, yet, still held up fairly well. This demonstrates just how powerful digital software is, as its working with so little to begin with. Therefore, in my opinion, this is a bad comparison to begin with. For a more fair comparison of the above, use color film and coverting it to B&W, then compare to B&W digital capture file.

You have to use the right tools for the task at hand. Of course B&W film is perfectly well suited for B&W output, however, if there is reasons you want to shoot digital for B&W, there is some some techniques / products that can help your digital capture perform better... and under the right circumstances, even supercede B&W film. Which of course would, create the feel of more depth in the image, amongst other benefits. In order of priority, here is the arsenal of products / techniques I would employ to boost digital B&W capture....

1. Use more pixels, either more MP on the sensor or shoot a montage and stitch em. How big? This varies based on output size.

2. Avoid shooting subjects void of green, such as pure red and pure blues as it will knock your recorded MP's in half, and reduce your recorded "after interpolation" values to 2 / 12 = 16%. Yep, before uprezzing, 84% of the file is "guessed", not a good starting point for B&W conversion.

3. Fine tune the digital workflow in the conversion. This may include building profiles from the point of capture to your output. Quite time consuming, but if you do enough work in this area, it may justify the time invested.

4. Use digital capture that records in 16 bit, vs. 12 or 14 bit. As you are trying to squeeze as many levels in the B&W as possible.

And the obvious, don't rule out using B&W film, when possible.... I hope this sheds some new light on your question.

Paul Kierstead
19-Dec-2005, 17:16
I think most would agree that digital files look flat or 1 dimentional when compared with film.

I think most would disagree. I would also hope you actually meant 2-D; 1D would truly be kinda sad.

To even have a slightly sane discussion about this, I would think that:
- What this depth is; at the very least, clear examples containing depth and not containing depth. It doesn't matter if they are digital or film, as surely the medium of film does not automatically grant depth.
- From those examples, draw some kind of hypothesis of what trait(s) provide that depth.
- Take pre-existing examples of photo's and try to add and/or remove depth byt altering those traits. Or, at the least, take multiple pics of the same subject with and without those traits, in the event it cannot be altered post-exposure.

Once you have verified what depth is, what it looks like and how it is achieved, you can have some kind of discussion. Not everyone will agree with you definitions, etc. but at least you will have the *basis* of a discussion. Instead you have just spoken for "most" and determined a medium wipes out a trait which is ill-defined at best. This is why you get no satisfaction. When everyone else looks like a Yahoo, it is time to look at yourself.

Paul Kierstead
19-Dec-2005, 17:17
All this tie and I still do not format lists properly. My apologies for the horribly formatting.

Bobby Sandstrom
19-Dec-2005, 19:58
Paul, read all of the posts!

Paul Kierstead
19-Dec-2005, 20:16
I certainly read all the posts. I see nothing even close to a definition of what constitutes a definition of the term. The provided examples look like 100% crop and certainly are a poor illustration of any aesthetic, including depth.