Best B&W film/developer for scanning (that's not bad for optical prints either!)?

[Noah A]

While I've done my share of B&W photography, I've been shooting color for the past few years. And before that I was shooting digital at my newspaper job. So I'm a bit out of the loop when it comes to the latest and greatest films and developers for B&W. The films haven't changed much since I last did serious film testing in college. But developers have. For an upcoming project I may want to shoot 4x5 B&W as a departure from my other large-format work that has all been shot on C-41.

My goal would be to make great drum scans on my Howtek for large inkjet prints. But I don't want to optimize for scanning to the extent that I can't also make excellent darkroom prints. Going back to B&W is a bit of an experiment so I don't know for certain what the final product will be.

I'm not too worried about film speed for my static subjects. I also like lots of acutance and I like the look of nice, sharp grain. (In my photo-j and street photography days, my all-time favorite was tri-x in Rodinal!) But I don't want exaggerated grain for this work, just a nice texture. I love the tonality of Tri-X but for this work I might want a bit less grain. My initial thought was to try FP-4, but I'm open to suggestions.

Of course I'll do some testing, but any recommendations for starting points? Everyone seems to love the pyro developers these days, but how do those negs scan?

[Richard Wasserman]

Noah,

I can't really help with the scanning aspects, there are others here who know more than I do about that, but can say that I like FP-4 very much. All my Chicago River images were done on it. I am currently doing another long term project for which I wanted a faster film, and am using TMY-2 which appears to me to be a bit sharper (higher resolution and acutance) than FP-4. The 2 films have a different look—they are both very nice, so you might want to try them both and see which you like. I find Pyrocat to be the best developer for both films, although XTOL is very close, especially diluted 1:3.

How big will you be printing?

[Jim Cole]

Noah,

I get really nice scans with TMY-2 and DD-X. Smooth tones and great highlight/shadow detail.

[ROL]

Here's (http://www.largeformatphotography.info/forum/showthread.php?84942-Non-Pyro-Developer-recommendations&p=823877#post823877) a comparison of PMK Pyro and XTOL with FP4+. (Caveat: These are scanned prints, not cheated scanned negatives.)

[Armin Seeholzer]

In my opinion the best compromise is XTOL!

Cheers Armin

[polyglot]

Any of them. The only real sin when scanning is overdeveloping; you can underdevelop a bit if you like but there's not a lot of point as it will make the dust stand out. Since you want to wet-print these eventually, just make properly-exposed and -developed negatives and they will scan nicely. While overdeveloping (to the point where the neg spans about 3.0D) will make your scan a bit grainy and unwieldy, it would be basically unprintable on normal paper, so don't worry too much about it.

Any decent scanner (especially drums!) will have a Dmax of about 4.0D, whereas a good neg will have a dynamic range of about 1.5D that you put onto the print; you'll find you get a lot more shadow and highlight detail in a 16-bit scan than you can fit onto paper.

Short answer is: a good scanner is way more forgiving than paper, so don't worry about it.

(like apparently everyone else above, I too really like TMY2 in XTOL for its excellent speed and fine grain. But nearly any other combo can be made to work for you; there are no magic bullets)

[Noah A]

Thanks for the tips! I'm not looking for a magic bullet, just some starting recommendations that I can test since I've been out of the B&W world for a while.

I never liked the look of the t-grain films when I was shooting 35mm, but then again that style of work is a whole different game from what I'm doing now. I really like the look of my old standby combination (Tri-X in Rodinal) but I want smaller grain.

I think I'll try out TMY-2 and FP4, maybe I'll test Tri-X too since I have a few sheets. Perhaps I'll try a pyro developer as well as xtol.

I'm pretty sure my drum scanner can do well with B&W negs when it comes to capturing the tonal range, what I don't know about is how the staining developers will affect the look of the grain when the negs are scanned and enlarged by a good amount. I don't mind grain, but if it's visible I like it to be nice, tight, sharply-defined grain.

Frankly I'm not sure how big the prints will be. Maybe I'll just decide to do silver contact prints for a real change of pace. But most likely I'll end up printing in my normal two edition sizes of 20x25 and 40x50.

[jp]

If you like grain, TMY2 isn't ideal as it is really fine grain and LF doesn't show grain much; if you make it big, it's tight and sharp though. It's a good film in that it can capture a massive range of light and dark to an extent usable on the computer and far beyond what is printable. It's also good quality like Ilford and can be used with many developers like Ilford.

Scanning will provide more flexibility than optical printing, so something good for optical printing will scan fine. Something too thin for optical printing will also scan fine.

Xtol is a good choice of developer for common films like have been described. Pyro (like pyrocat-hd in glycol) is also excellent over-the-counter choice but the stain might hide some grain making it smoother rather than crispy (and that could be good for people photos). It all depends on how big and contrasty you make things. It still has plenty of detail, versatility, scanability, and printability.

As far as printing on inkjet, the quality of the scan is more important the grain of the film. A bad scan or not high enough res scan will show square soft pixels (from upsizing or lack of res) rather than actual grain. On my epson, a scan shows a clear difference between big grain fomapan100 and little grain tmy2. It's not high-end enough to show a difference between tmy2 and tmx though. (And I consider tmx needlessly fine for most needs)


It sounds like it's an aesthetic choice for the OP as we've pointed out some good options to try.

[Ari]

Noah, fwiw, I like Acros and HC-110.
Scans and prints wonderfully.

[ROL]

(like apparently everyone else above, I too really like TMY2 in XTOL for its excellent speed and fine grain. But nearly any other combo can be made to work for you; there are no magic bullets)

I know you're just being general, but for the record, as one of the 'above', though I have used them, I do not prefer TMAX films in large format. But then I don't hard copy fine art work digitally either.

[false_Aesthetic]

I've had the luxury of working with a bunch of differnt negs in differnt soups on different scanners (howtek, scanmate, v750, 10000xl, imacon, 4990, tango).

Morrell/Nixon: tri-x + hc110
Bosworth: tri-x + d76
Gohlke: tri-x + xtol
Johnson: delta 3200 + xtol
Perrott: tri-x (@ 100) in xtol
vitone: hp5 + d76
Me: adox 25 + pmk; tmax + tmax; tmax + f76+; hp5 + xtol; fp4 + xtol; tri-x + pmk

They're all printable in the darkroom and, if you know what you're doing, can be scanned with good results.

[turtle]

I would steer clear of traditional emulsions and go for Deltas/Tmax/Acros. You can always alter the curve in PS.

As for developer, thats simple: Xtol 1+1. It is great with all of the above, has fine grain but is not mushy.

TMax 400 II is far finer grained and higher resolving than FP4+ and I would not recommend FP4+ anyway as it has quite sandpapery grain when heavily enlarged. Its not like TriX in pattern, but somehow less pleasant.

I find the Delta grain shows up more on drum scans than the comparable Tmax films. It has harsher, crisper grain and drum scanning seems to accentuate it.

I'd go for Acros in 100/Tmax 100 and Tmax 400 II for a faster film.

[Edwardv]

Would a two bath developer, like Vestal, be idea for black and white negatives to be scanned? :)

[Ken Lee]

Any decent scanner (especially drums!) will have a Dmax of about 4.0D, whereas a good neg will have a dynamic range of about 1.5D that you put onto the print; you'll find you get a lot more shadow and highlight detail in a 16-bit scan than you can fit onto paper.

A few days ago, I scanned a Stouffer Step wedge. My Epson V700 scanner shows a range of around 16 steps, each of which is 1/2 stop. That's an 8-stop range, which in log 10 is 2.4

Epson claims a dMax of 4.0 which equates to around 13 stops. Not this machine anyhow. But then Epson claims a lot more dpi than any of us have measured too. :cool:

[turtle]

Neg scanners do a lot better than Epson flatbeds and with drum scanners, Dmax does not seem to be an issue, ever. The great thing about drum scans is one can develop the negs such that they will yield good silver prints too.

[Noah A]

I'm not worried about Dmax as much as about what the grain looks like, especially with Pyro since I've never used it. I'll try out the suggestions above. Thanks again!

[Jay DeFehr]

I'm not worried about Dmax as much as about what the grain looks like, especially with Pyro since I've never used it. I'll try out the suggestions above. Thanks again!

Old fashioned pyro developers and old fashioned films made for grainy negatives, but modern films developed in modern pyro developers can be at least as fine grained as any general purpose developer with the same film. I spent this morning scanning Tmax 100 negatives developed in 510-Pyro, and there just isn't any grain, to speak of.

[Heroique]

My story is simple.

My best Epson 4990 scans of 4x5 b/w film are T-Max 100 (in T-Max rs).

I suspect that the optimal lighting of these compositions + my perfect exposure choice + wise (tray) development technique + informed scanning habits have a lot to do with it.

But as long as I achieve all these things together, it still seems that T-Max 100 + T-Max rs is my combo that wins & wins again for scanning purposes.

[nolindan]

The best film for scanning is color negative. Convert to black and white in software.

I've always had much better results scanning a print than scanning a B&W negative.

YMWV.

[Lynn Jones]

The best modern b/w film I have ever used was the recent Plus X Pan, but of course Kodak discontinued it.

Your best choices today would be Ilford Delta 100 (18X with barely discernable grain), Fuji Acros 100 (barely discernable grain at around 20X), TMX (at least 24X), all of these with Rodinal or its copy. Of course TMY and TriX are still the favorite 400 speed films although I really like Fuji Neopan 400 with a true fine grain developer.

As for a true "work horse" film in the 100 speed group area, my last and discontinued choices were VP and PXP, when Ilford changed their FP and HP the the "+" version, I quit using them because there were less sharp and more grainy in my tests.

Lynn

[sanking]

My favorite medium speed B&W film (ISO 100) is Fuji Acros. In terms of grain and resolution it is at least as good as the two main competitors, Delta 100 and T-Max 100, but what sets it apart from these films is its very low reciprocity failure. Delta 100 and T-Max 100 are a lot better than traditional films, but Acros is in a league by itself in my experience. In low light conditions you will often find that because of the low reciprocity failure of Acros actual exposures time will be less than with ISO 400 films.

I use Across with Pyrocat-HD, 1+1+100 at 72F for 15 minutes and get great resolution and virtually no grain at 20X. However, Acros is a robust and easy to use film that should also give great results with a wide range of both staining and non-staining developers. In my opinion Acros is easily the best medium speed film on the market if you consider ease of use, grain, resolution/sharpness, and low reciprocity failure.

Sandy King

[John Henry]

Sandy is right about Acros and he's also right about using different developers with it. I've had really fine results with Ilford DD-X and Rodinal (semi-stand). I haven't tried Xtol or Pyrocat-HD, but plan on it shortly since I just received my chemical order and Freestyle just shipped my back-order of Fuji Acros. One other benefit of Acros is it is damn cheap compared to the Yellow Gods film or even Ilford. I love Delta 100 and TMY2, but for an all-around B&W film Acros can't be beat. The playing field is getting thinner when it comes to film, but it's even worse when it comes to paper. JohnW

[Nathan Potter]

If the concern is mainly grain delineation then any of the fine grain films are OK. But consumer grade scanners aren't going to see that anyhow.
I find that the dynamic range is more important, so you need to consider the type of scanner that will be used before you develop your film. I mostly use an Epson V750 or a Nikon. I generally find what Ken has mentioned above from a Stouffer density wedge; that is maybe 8 stops decently resolved.

I then want to get all the subject brightness values within at least 8 stops on film, say logD 2.0 to 2.3. This can reasonably be handled by the two scanners I use. For fine drum scanning I'll push this to maybe 2.6 to 2.8. To control development (limit density) I mostly use Diafine unless excessive N- or N+ development is required; then I usually use D76 at 1:2.

Lately I am seeing that maybe what I have been using is a bit too high for density range and am inclined to reduce the film Dmax to 1.6 or so. I need to get a better delineation of tones at the high end to achieve better separation. But much depends on the nature of the negative.

Nate Potter, Austin TX.

[Jerzy Pawlowski]

It is dirt cheap in 120 rolls format that I am buying for roll film holder, but in 4x5 size I would say it is the most expensive one. May be because it comes in 20 sheet boxes only.
(it is regarding Acros and John Henry post)
Jerzy

[SergeiR]

One other benefit of Acros is it is damn cheap compared to the Yellow Gods film or even Ilford.
No its not. 100 sheets of Acros will set you off about 5x35$ - 165, whereas 100 sheets of Ilford Delta is about 104$...
While i like it for 120, i don't see it to be of reasonable price for 4x5.

As of the original topic - i am sticking with UFG and Adox(low speed)/Efke(middle speed)/Ilford (high speed). Scans on regular flatbed pretty well. But then what do i know.. i am just experimenting so far ;)

[Keith Tapscott.]

For the finest grain, a true fine-grain developer and a high resolution film like T-Max 100, Delta 100 or 100 Acros would be a wise choice.

http://www.digitaltruth.com/devchart.php?Film=Neopan+100+Acros&Developer=Perceptol&mdc=Search

[John Henry]

No its not. 100 sheets of Acros will set you off about 5x35$ - 165, whereas 100 sheets of Ilford Delta is about 104$...
While i like it for 120, i don't see it to be of reasonable price for 4x5.

As of the original topic - i am sticking with UFG and Adox(low speed)/Efke(middle speed)/Ilford (high speed). Scans on regular flatbed pretty well. But then what do i know.. i am just experimenting so far ;)

Stupid me! Yes, I was referring to Fuji Acros in 120. I guess jumping around to different forums can be detrimental to your memory. In 4x5 and 8x10 I have only used TMX, TRI-X, and cheap Chinese stuff. Never tried Acros! I do know it Acros scans beautifully. I processed a roll of 120 shot with a Rollei/3,5 Planar, in Rodinal 1:100 semi-stand, and detail was amazing. I scanned it on a Nikon LS8000 with glass carrier and you can't even find the grain. So, 4x5 or 8x10 should be the cats meow, but like has been said, Fuji in large format is more expensive. I have always wondered why that was, but have refrained from asking Fuji direct. The reason I'm afraid to ask Fuji is that they'll probably say, "Oh yes, well we'll take care of that" and then instead of lowering the 4x5 sheet cost they raise the 120.

[Jerzy Pawlowski]

The reason I'm afraid to ask Fuji is that they'll probably say, "Oh yes, well we'll take care of that" and then instead of lowering the 4x5 sheet cost they raise the 120.
I am sure you are right. Probably it will materialize one day, whether you will ask or not. Till then lets enjoy 120 rolls at least.

[Bruce Watson]

My goal would be to make great drum scans on my Howtek for large inkjet prints. But I don't want to optimize for scanning to the extent that I can't also make excellent darkroom prints. Going back to B&W is a bit of an experiment so I don't know for certain what the final product will be.

I'm not too worried about film speed for my static subjects. I also like lots of acutance and I like the look of nice, sharp grain. (In my photo-j and street photography days, my all-time favorite was tri-x in Rodinal!) But I don't want exaggerated grain for this work, just a nice texture. I love the tonality of Tri-X but for this work I might want a bit less grain.

I researched the heck out of this more than five years ago, and the results of my research are posted in various threads on this site. My work was with a ColorGetter 3Pro drum scanner, and 5x4 Tri-X. The results carried over to TMY-2, my only B&W film these days.

What it comes down to is that all scanners, drum scanners included, will do a fine job on film that's optimized for silver gelatin printing in the darkroom. If you are ever going to print in the darkroom, optimize for the darkroom. I'm not kidding, that's the best thing to do. Really.

If you in fact are never going to print in the darkroom, optimize for scanning. This usually means less density and therefore less graininess. In Zone System terms, a good place to start is a Zone VIII density of around 1.0. Such a negative would scan easily, but would be a PITA to darkroom print on silver gelatin paper. I don't have proof for why this works for scanning, but my personal theory is that drum scanning is just as susceptible to Callier Effect as darkroom printing is, and lowering the density means less light scatter and therefore better separation in the highlights. But I could well be wrong. All I know is that it works for me and my negatives.

That said, if you are going to optimize for the darkroom, you can still do some things that will improve scanning results. One of those things is to use a small grained film. Fuji Acros, Ilford/Harman Delta, or Kodak Tmax films scan better for me on my drum scanner than older emulsion films like Tri-X and HP-5.

Finally, don't think that any scanner, not even a drum scanner, is going to image the physical film grain. It's not. The shape of film grain is fractal (see page 19 of Tim Vitale's excellent paper (http://cool.conservation-us.org/coolaic/sg/emg/library/pdf/vitale/2007-04-vitale-filmgrain_resolution.pdf) on this); a drum scanner can't give you enough resolution to resolve that shape. What you get from a scanner scanning B&W films is a square pixel that contains exactly one shade of gray. That pixel is featureless and perfectly symmetrical, and about as far from looking like film grain as you can get. So the noise you get in your digital print isn't film grain. But the vast majority of people won't be able to tell. Not even guys like me who know the difference.

So... bottom line? Use the film you want, developer you want, optimize for the darkroom, and don't sweat it.

[hendrik faure]

If you are not in a hurry, you find a lot of intersting classic films at fotoimpex, berlin
http://www.fotoimpex.de/FILME/FILME.html
they speak english, some catalog information is english too
http://www.adox.de/english/FilmDeveloping/startpage.html
for your purpose maybe one of the "spur kit"
http://www.fotoimpex.de/cgi-bin/shop/de/index.cgi?func=searchdo&sfields=&sbeg=spur%20kit&x=5&y=6
regards,
Hendrik

[turtle]

I agree Bruce. Drum scanners have no issue with negs made for wet printing, even very dense ones. Personally, I hate the idea of making negs that are so flat they suit only scanning (presumably on scanners with low Dmax). At least with darkroom optimised negs you will retain both options long into the future and I strongly suspect darkroom printing will be around every bit as long as scanners capable of dealing with transparent materials.

The last prints I made from drum scans were from 6x7 negs and printed to 40". There was grain, but as you say, it was slightly digitised grain if you like. Not as organic as real grain, but a lot closer to that than to pixels or other digital artifacts. The look was very pleasing and only if you sniffed the print could you see that it was not the grain you would expect from a wet print, but a little different.

[Jay DeFehr]

Turtle:
"There was grain, but as you say, it was slightly digitised grain if you like. Not as organic as real grain, but a lot closer to that than to pixels or other digital artifacts."

Turtle,

How is a pixel generated by a film scanner different than one generated by an in-camera image sensor? Is it just smaller, or is there some other distinguishing feature of an otherwise featureless pixel that distinguishes one generated from a film scan from another generated by an in-camera sensor? It's a sincere question -- I don't know much about scanning or digital cameras, but I'm trying to learn.

[turtle]

Jay,

I'm not talking about pixels, but more how a finite number of pixels manages to resolve grain.

My impression is that the resolution of the scan (for 300dpi output at 40") was easily sufficient to show the grain, but not high enough to resolve the precise shape, edges etc of the grain. The result is that the grain does not look as crisp and 'natural' as it would on a wet print. Printed to a smaller size, I suspect this would be harder to see, but as a roughly 15x linear enlargement, it is noticeable if you look closely enough. I am guessing that the resolution required to show all the nuances of grain at large print sizes would be stratospheric. Bear in mind I am talking about Delta 100 grain here, so it is very small. In shadow area it was very hard to see this, but in mid and upper mid-tones where there were contrast increases made, it was most apparent. The grain just looked a little 'blockier/blobbier'. This is the sort of stuff only photographers would see and not important, but the result is still clearly a film based image. You are left in no doubt about that.

I am sure that messing around with scan resolution and output would alter things, but I was gobsmacked at the quality of the prints after working with my digital printer (the person, not the machine) for a few hours and this issue was more of a passing observation than a criticism. Up on the gallery wall, the images look closer to wet prints than I thought possible with all the benefits of being able to work with the files in LR/PS (one of which is absolute hell to print conventionally at any size)

[Bruce Watson]

Turtle,

How is a pixel generated by a film scanner different than one generated by an in-camera image sensor?

A digital sensor is deterministic. The sensor sites are usually the same size, and usually uniformly distributed. This uniformity gives rise to digital artifacts like moire.

What a film scanner does is somewhat different. It lays a deterministic grid over the film, which in turn is stochastic. That is, the film grain is more or less randomly distributed, and the scanning grid is uniform. Film grain varies widely in size, while the scanning pixels are exactly one size. The scanner then "looks through the holes in the grid" and measures what it sees of the film through that hole. If it sees nothing by metallic silver, the resulting pixel is solid black (a digital zero). If it sees nothing but film base, the resulting pixel is solid white (a digital 256, or 4096, or 65,536 or whatever max your scanner has). This is obvious, yes? But what happens when it sees just a portion of a film grain? It does just what you'd think -- it averages the part that's blocked by metallic silver with the part that isn't, and the pixel gets a value in between zero and max.

So the scanner doesn't necessarily follow the exact pattern of the film grain. But on the plus side it tends to even out the tonal changes. Combined you can get a really excellent representation of the image. Highly detailed, and amazingly true to the negative. But it's not an exact copy. Think of it as an extremely high quality second copy (scene you photographed > film > scan). Just sayin'.

[Heroique]

...The scanner then “looks through the holes in the grid” and measures what it sees of the film through that hole...

Bruce, nice explanation ... but this part is just a little bit creepy.

I’ll never think of my Epson in the same way.

:eek:

[Jay DeFehr]

A digital sensor is deterministic. The sensor sites are usually the same size, and usually uniformly distributed. This uniformity gives rise to digital artifacts like moire.

What a film scanner does is somewhat different. It lays a deterministic grid over the film, which in turn is stochastic. That is, the film grain is more or less randomly distributed, and the scanning grid is uniform. Film grain varies widely in size, while the scanning pixels are exactly one size. The scanner then "looks through the holes in the grid" and measures what it sees of the film through that hole. If it sees nothing by metallic silver, the resulting pixel is solid black (a digital zero). If it sees nothing but film base, the resulting pixel is solid white (a digital 256, or 4096, or 65,536 or whatever max your scanner has). This is obvious, yes? But what happens when it sees just a portion of a film grain? It does just what you'd think -- it averages the part that's blocked by metallic silver with the part that isn't, and the pixel gets a value in between zero and max.

So the scanner doesn't necessarily follow the exact pattern of the film grain. But on the plus side it tends to even out the tonal changes. Combined you can get a really excellent representation of the image. Highly detailed, and amazingly true to the negative. But it's not an exact copy. Think of it as an extremely high quality second copy (scene you photographed > film > scan). Just sayin'.

Thanks for the explanation, Bruce. It still seems to me that there is no grain in a scanned image, just pixels, and that the scanner converts a stochastic image to a deterministic one. I get moire patterns and noise when scanning film, for instance, so I still fail to see how the pixels of a scanned film differ from those captured in-camera, except in location, ie, digital sensors tend to exhibit noise in low values, while film tends to be grainiest in high values. I'm not trying to be pigheaded about this, I'm really trying to understand the issue. Thank you for your patience.

[sanking]

No matter how good a scanner you may have it will not resolve individual silver grains. However, higher quality flatbed and drum scanners are definitely capable of imaging clumps of grain, and this may or may not be visible as granularity in the scan and/or resultant print. In my experience Epson flatbed scanners and the like are not capable of imaging clumps of grain. Interestingly, this lack of definition sometimes results in smoother, and more pleasant looking, scans and prints.

One of the very big advantages I find in digital capture is the lack of noise/grain at normal and medium ISO.

Sandy

[Jay DeFehr]

No matter how good a scanner you may have it will not resolve individual silver grains. However, higher quality flatbed and drum scanners are definitely capable of imaging clumps of grain, and this may or may not be visible as granularity in the scan and/or resultant print. In my experience Epson flatbed scanners and the like are not capable of imaging clumps of grain. Interestingly, this lack of definition sometimes results in smoother, and more pleasant looking, scans and prints.


One of the very big advantages I find in digital capture is the lack of noise/grain at normal and medium ISO.

Sandy

Sandy,

I'm just learning about digital capture, but already I agree with you. I have yet t make a digital print, but I'm looking forward to it.

[turtle]

...and of course the presence/absence of grain is either desirable/undesirable depending on what you are doing. My images were documentary shots in Afghanistan and so a silky smooth image were not desirable. Drum scanning of large negs is of course a hybrid process, with hybrid results, but that is its appeal to me in some respects. It allows me to produce very large prints without the enormous cost and difficulty of darkroom work at these sizes, but with a 'feel' to the images that is fare closer to the original film capture than digital.

I have heard of some people scanning film images at a resolution that just begins to show signs of soft grain, the uprezzing them. This supposedly maintains a smoother look but allows large prints. I guess it would work well for some applications, but I tend to want all the detail and sparkle I can get and grain has a role here.

[Bruce Watson]

Thanks for the explanation, Bruce. It still seems to me that there is no grain in a scanned image, just pixels, and that the scanner converts a stochastic image to a deterministic one.

Not quite. The scanner creates a deterministic image from a stochastic image. It's not a straight up conversion. There's some blurring and averaging taking place, but at a very low level. If done well, below the level at which image detail information is recorded on the film. If that's the case, then for all practical purposes a scanner does what you say.


I get moire patterns and noise when scanning film, for instance, so I still fail to see how the pixels of a scanned film differ from those captured in-camera...

The major difference to me is that the digital capture is a first generation copy. A film capture is also a first generation copy. When you scan the film, you are making a second generation copy. The big question though is "does it matter?" As far as I can see, it doesn't, not even when making 10-12x enlargements.

The problem in comparing digital capture to scanned film is that, like most such comparisons, it's an apples to oranges comparison. Both methods have their own strengths and their own weaknesses. So what matters to me may not matter to you, and vice versa. I'm more interested in the images than I am in the technology of the capture. But that's just me.

[Jay DeFehr]

Not quite. The scanner creates a deterministic image from a stochastic image. It's not a straight up conversion. There's some blurring and averaging taking place, but at a very low level. If done well, below the level at which image detail information is recorded on the film. If that's the case, then for all practical purposes a scanner does what you say.

The major difference to me is that the digital capture is a first generation copy. A film capture is also a first generation copy. When you scan the film, you are making a second generation copy. The big question though is "does it matter?" As far as I can see, it doesn't, not even when making 10-12x enlargements.

The problem in comparing digital capture to scanned film is that, like most such comparisons, it's an apples to oranges comparison. Both methods have their own strengths and their own weaknesses. So what matters to me may not matter to you, and vice versa. I'm more interested in the images than I am in the technology of the capture. But that's just me.

Thank you for your patience, Bruce. I'm more interested in images, too, but comparing apples and oranges becomes important when one has a choice between the two. Until recently, I did not, and any interest in such a comparison was purely theoretical, but now it's a practical issue due careful consideration. I appreciate your experienced perspective.

[Bruce Watson]

Thank you for your patience, Bruce. I'm more interested in images, too, but comparing apples and oranges becomes important when one has a choice between the two. Until recently, I did not, and any interest in such a comparison was purely theoretical, but now it's a practical issue due careful consideration. I appreciate your experienced perspective.

There are many more ways to look at the problem, some of which may (or may not) generate some more (better?) understanding.

One is to consider how many film grains it takes to record the smallest amount of actual image information. In order for film to work at all, it has to take a lot of film grains to successfully describe a "unit" of image information. Because if it doesn't, photography has to look like a Seurat painting. So there is a many->one relationship between film grains and image detail, yes?

This in turn implies that there is some resolution that can be used in scanning said film, that will capture all the image detail, but will not capture much, if any, grain detail. Some middle ground as it were. In fact, there seems to be many scanner operators that believe that scanning above around 2400-3200 ppi becomes meaningless because it doesn't capture any more image detail. All it does is capture more grain detail (aka noise). There seems to be some elements of truth to this "school" of scanning.

I say "some elements of truth" because if nothing else, this seems to offer an explanation of why scanning works at all, and why it works so bloody well -- why the scanned copy (second generation) can be such an excellent representation of the film (first generation copy) that it is, for all intents and purposes, an exact copy. What you loose when you scan is some film grain data, but you lose very little of the actual image information (note the distinction between the concepts of data and information). And this is why it doesn't matter that a scanner can't image the film grain itself.