Where would film technology be now?

[laser]

It’s not constant, like the film image is not constant.
One reason why film holders, esp. high quality ones like the Flextights are densely black and flogged to cut down extraneous light.
Sure you can do “something” in post, but completely fix it? No.
Nikons scanners are in many ways the epitome of the macro camera approach. But they where never cheap and I highly doubt they could pull all stops of range out of negative film. At least not in any files I’ve seen.
Again as I wrote earlier, it should be pretty easy to manufacture something at least as good as the Coolscans at a fraction of the original (or used) price today, if some bigger manufacturer should decide to but their weight behind it.


You essentially have to do the equivalent of dodge and burn in camera.
The much maligned Cokin filters and holder is a good cheap entry to do that.

I once had the idea if it would be possible to make a transparent LCD screen filter, that with the image from a simple digital camera could create a mask over the film plane, that would allow slide to capture more detail in contrasty environments.

 

[laser]

It’s not constant, like the film image is not constant.
One reason why film holders, esp. high quality ones like the Flextights are densely black and flogged to cut down extraneous light.
Sure you can do “something” in post, but completely fix it? No.
Nikons scanners are in many ways the epitome of the macro camera approach. But they where never cheap and I highly doubt they could pull all stops of range out of negative film. At least not in any files I’ve seen.
Again as I wrote earlier, it should be pretty easy to manufacture something at least as good as the Coolscans at a fraction of the original (or used) price today, if some bigger manufacturer should decide to but their weight behind it.


You essentially have to do the equivalent of dodge and burn in camera.
The much maligned Cokin filters and holder is a good cheap entry to do that.

I once had the idea if it would be possible to make a transparent LCD screen filter, that with the image from a simple digital camera could create a mask over the film plane, that would allow slide to capture more detail in contrasty environments.

While working at Kodak I built a liquid crystal/photo conductor sandwich in 1973. When combined with a color slide it produced an imagewise mask that was used to reduce the contrast when printing the slide onto Ektachrome Paper. I installed it is a 5S printer and made masked Ektachrome Prints. It worked beautifully. It was never commercialized.

 

[feral_capitalist]

My biggest gripe with film and technology is LIGHTING.

All the LED's china is pumping out looks horrible on film. There is NOT ONE color film out there balanced for magenta-white crappy china LED's that people are putting in their homes. Makes my daylight balance look all blue or cyan, depending on LED. All the LED's will never be the same like tungsten.

If color film is to stay alive, it needs to be resilient to all this lighting technology.

 

[Mr Bill]

While working at Kodak I built a liquid crystal/photo conductor sandwich in 1973. When combined with a color slide it produced an imagewise mask that was used to reduce the contrast when printing the slide onto Ektachrome Paper. I installed it is a 5S printer and made masked Ektachrome Prints. It worked beautifully. It was never commercialized.

Cool. I recall that Agfa presented a related idea, probably in the late 1990s, in one of the IS&T (formerly SPSE) symposiums related to photofinishing. What they had done was to use such an LCD screen, in an optical printer, to produce an unsharp mask over the negative. The idea was, as I recall, that one-hour lab systems were beginning to use scanners as film was loaded into the printer. (These were crude scanners, capable of "index print" quality, but mainly intended to allow advanced estimates of printing exposure/color balance.) So the scan was good enough to produce the image for an unsharp mask. The transparent LCD screen was far enough from the negative to be defocused, and since it only attenuates the light source it doesn't degrade image quality.

When I first saw this I expected it to become a mainstream minilab feature, but for some reason it didn't. It was not too much longer - perhaps a half-dozen years(?) - before high-quality scanning and digital printing exposure went mainstream and there was no long any point to the LCD mask.

 

[Photo Engineer]

While working at Kodak I built a liquid crystal/photo conductor sandwich in 1973. When combined with a color slide it produced an imagewise mask that was used to reduce the contrast when printing the slide onto Ektachrome Paper. I installed it is a 5S printer and made masked Ektachrome Prints. It worked beautifully. It was never commercialized.

Bravo Bob! Sorry it never made it.

 

[Helge]

While working at Kodak I built a liquid crystal/photo conductor sandwich in 1973. When combined with a color slide it produced an imagewise mask that was used to reduce the contrast when printing the slide onto Ektachrome Paper. I installed it is a 5S printer and made masked Ektachrome Prints. It worked beautifully. It was never commercialized.

Glad to hear someone did it!
In the off chance that you should have patent numbers, images, more descriptions, I’d love to see them.

Put one of these inside a camera and you’ll have practically no trouble with dynamic range with slide (or negative for that matter, even if that’s less of a problem).
You'd have to have the ability to flip the sandwich away since the polarizers would rob you of light in low light situations.

You could do essentially the reverse with a concurrent flasher in camera. That is, only flash the shadows to bias them.
That would make concurrent flashing (à la Illumitran or Gerry Turpins Colorflex) useful in contrasty daylight situations too.

 

[Mainecoonmaniac]

Cool. I recall that Agfa presented a related idea, probably in the late 1990s, in one of the IS&T (formerly SPSE) symposiums related to photofinishing. What they had done was to use such an LCD screen, in an optical printer, to produce an unsharp mask over the negative. The idea was, as I recall, that one-hour lab systems were beginning to use scanners as film was loaded into the printer. (These were crude scanners, capable of "index print" quality, but mainly intended to allow advanced estimates of printing exposure/color balance.) So the scan was good enough to produce the image for an unsharp mask. The transparent LCD screen was far enough from the negative to be defocused, and since it only attenuates the light source it doesn't degrade image quality.

When I first saw this I expected it to become a mainstream minilab feature, but for some reason it didn't. It was not too much longer - perhaps a half-dozen years(?) - before high-quality scanning and digital printing exposure went mainstream and there was no long any point to the LCD mask.

Does anybody remember Minit mask?

 

[Helge]

Does anybody remember Minit mask?

Seems like you almost might as well use film, at least for 135 and 120.
Then you can take your time printing and have the exact same mask for future prints.
Anyhow not a very useful technology for in camera use or for dynamically and quickly adjusting during printing.

 

[138S]

Unless Fuji licensed it from Kodak, then NO. However, 2e is not directly related to LIRF. RF in general is related to another heavy metal.

PE

Thanks

 

[138S]

There are absolute limits to what kind of resolution you can extract from the scanner.
No amount of fiddling with holders, Newton ring free glass or liquids is going to change that.

Also there are limits in the negative: One thing are lab tests with flat targets with a tripod and another thing is real photography.

In real photography:

> the scene is usually in the DOF and not in the perfect plane of focus, so most has some defocus,

> Many often use a wide aperture for selective focus and the lens is not in peak performance, or sometimes we close a lot for DOF having some diffraction

> handheld shots always have some shake

> textures have no 1000:1 microcontrast, so film does not deliver 150lp/mm but 40lp/mm

So most of the shots are very easy to scan because basic scanners are outresolving what real shots have, there are some excetionally sharp shots that may deserve a drum scan, for sure, but not many, say 1/30.

Anyway, considering cost, it's always cheaper to shot a larger format than paying for an expensive scanning, a drum scan of a 35mm frame may cost $20 to $50, while a MF shot is less than $1, and that MF shot scanned with any decent consumer scanner will be many times better than the drum scanned 35mm shot.

The same happens in LF, if 4x5" has not quality enough for you in the epson then shot 5x7 ! you may have to split 8x10" sheets, but today four 4x5 drum scans pay for the IR googles allowing a convenient splitting.

No doubt that a drum is a way superior machine, the question is when it makes sense or not.

Going back a bit to this thread, it would be interesting to see what evolution film would have had for better scanning.

For the same film speed there is a balance between cloud size and and cloud overlaping, to not generate color noise in the scanning, a color is often mapped in the three color layers, so how clouds overlap may concern noise in the discretization.

perhaps color film could have been re-engineered to increase resolving power from smaller clouds while ensuring a good overlaping. My guess is that if money had been there for R+D then this would have been a field for improcements.

 

[Mainecoonmaniac]

Seems like you almost might as well use film, at least for 135 and 120.
Then you can take your time printing and have the exact same mask for future prints.
Anyhow not a very useful technology for in camera use or for dynamically and quickly adjusting during printing.

I use both but favor film because I’m an old dog in photography. I was thinking of using a scan of a negative and printing on OHP film for masking negs. My idea I would use magenta and yellow colors to local contrast. I’m sure it’s done before.

 

[138S]

I use both but favor film because I’m an old dog in photography. I was thinking of using a scan of a negative and printing on OHP film for masking negs. My idea I would use magenta and yellow colors to local contrast. I’m sure it’s done before.

This is the Alan Ross selective masking technique, published 10 years ago. He has used that way for 20 years at least, he has made a number of workshops about that and I know very good printers using it, one in China, some made intercontinental trips to attend those workshops. It was published in International Photo Technik magazine, IIRC, in his web site that article can be purchased in pdf.

https://alan-ross-photography.myshopify.com/collections/pdfs/products/selective-masking

Amazingly that technique was mostly overlooked, but it's one of the most powerful techniques we may use in darkroom printing, IMO.

PM if you are interested in that way, I've some experince with it, I'd be happy if I can help.

 

[Photo Engineer]

Great reference ! Thanks for posting

The amazing thing is that D-76 stock was used in this test !

Not really. D76 was and probably still is, the release test developer for all B&W films.

PE

 

[138S]

Not really. D76 was and probably still is, the release test developer for all B&W films.
PE

It looks D-76 is the gold standard (except for Foma, sadly). Anyway that curve with 18 to 20 stops is impressive, I guess that TMX latitude extension in the highlights comes from the cubic fine grain layer under the main emulsion. One question, is that emulsion ortho?

That layer has to sport crystals of an amazingly "low iso" (ISO 0.0001 ?)... or is perhaps the main emulsion over it casting a shadow ?

 

[Helge]

Also there are limits in the negative: One thing are lab tests with flat targets with a tripod and another thing is real photography.

In real photography:

> the scene is usually in the DOF and not in the perfect plane of focus, so most has some defocus,

> Many often use a wide aperture for selective focus and the lens is not in peak performance, or sometimes we close a lot for DOF having some diffraction

> handheld shots always have some shake

> textures have no 1000:1 microcontrast, so film does not deliver 150lp/mm but 40lp/mm

So most of the shots are very easy to scan because basic scanners are outresolving what real shots have, there are some excetionally sharp shots that may deserve a drum scan, for sure, but not many, say 1/30.

Anyway, considering cost, it's always cheaper to shot a larger format than paying for an expensive scanning, a drum scan of a 35mm frame may cost $20 to $50, while a MF shot is less than $1, and that MF shot scanned with any decent consumer scanner will be many times better than the drum scanned 35mm shot.

The same happens in LF, if 4x5" has not quality enough for you in the epson then shot 5x7 ! you may have to split 8x10" sheets, but today four 4x5 drum scans pay for the IR googles allowing a convenient splitting.

No doubt that a drum is a way superior machine, the question is when it makes sense or not.

Going back a bit to this thread, it would be interesting to see what evolution film would have had for better scanning.

For the same film speed there is a balance between cloud size and and cloud overlaping, to not generate color noise in the scanning, a color is often mapped in the three color layers, so how clouds overlap may concern noise in the discretization.

perhaps color film could have been re-engineered to increase resolving power from smaller clouds while ensuring a good overlaping. My guess is that if money had been there for R+D then this would have been a field for improcements.

Sure, sharpness is a bourgeois concept. But that is a whole load of confirmation biased relativism.
It's not that hard to get a "perfect shot" on 135 that warrants a scan that is far better than a flatbed.
Sunlight, a steady hand and a good lens like Rokkor 24mm 2.8 at f8 and you are there.
Put the camera on a tripod and you just eliminated most of the chances of blur even for long exposures.
I shoot most of my night shots at f8 to 16 to get good depth of field. Bokeh just often isn't of importance and doesn't really look that good at night, unless you are going specifically for neon light bokeh balls.
Here a digital camera would struggle too without a tripod.
4x5 quality would be nice everywhere but there are blindingly obvious reasons why people stopped shooting it that much.

 

[138S]

Sure, sharpness is a bourgeois concept. But that is a whole load of confirmation biased relativism.

Yeah, HCB was a die hard anarchist , and he was right, sharpness alone is nothing in art.

It's not that hard to get a "perfect shot" on 135 that warrants a scan that is far better than a flatbed.

IMO a bit depends on particular film, color negative film it's very easy to scan, but extracting the beautiful grain structure of a classic film it's always challenging, even for powerful scanners.

Think that 40lp/mm are 0.012mm dot's on film, if the image moves 0.005mm on film while exposing (trepidation, mirror/shutter) then that shake allows a decent consumer flatbed to outresolve the film image, it's really difficult that a handheld shot is that steady, if a flash not used.

An improvement is VR lenses on say a F80, or a monopod.

Yes, in 35mm a drum or a dedicated roll film scanner is better than a decent consumer flatbed. As the format increases the decent consumer flatbed it's a way better choice. At 6x7cm an V700 extracts 42 effective optic MPix, and this fills 20 Full HD monitors, or 5 4k monitors. At 4x5" the V700 takes 150 MPix effective from film, while a drum at 4000dpi extracts 320MPix effective, so depending on print size the drum may be necessary if wanting to inspect a very big print from reading distance, if the shot is sharp enough, many LF shots have less than 40Lp/mm resolution, because of DOF or because of diffraction, in LF we usually have a tripod, but also difficulties to focus.

 

[Helge]

Yeah a dedicated 135 scanner would be pretty bad for medium format. ;-)
It's setting your bar pretty low if you are content with a mere 42MP from a 6x7. It's also setting your bar low if you use viewing on a monitor, as any kind of metric of quality.
Sure it's nice for getting an idea of the image, like a contact sheet, or for quickly sharing and communicating an idea. But that is another kind of photography, for which simple 135 and even 16mm still cameras where made.

Regarding motion blur, much depends on your focal length and how far away the subject is. VR is really of limited use for longer focal lengths where it is most needed. And I'd say of limited use period. I'd much rather use one of the many alternatives (string pod, mono/tri pod, mini pod, flash etc.).

 

[138S]

It's also setting your bar low if you use viewing on a monitor, as any kind of metric of quality.

I was not saying filling 1 monitor, but filling 20 minitors !!! 42Mpix effective are good for an excellent 50x60cm print inspected at reading distance, this is the bar at the good height, IMO.

Human eye (excellent sight) can't see no more than effective 60 MPix (not moving head but moving eyes), 40 MPix saturates the eye capability if the print is at confortable distance to see all the print without moving head too much.

Of a 6x7cm, a 8000dpi drum scan takes around 80 MPix effective from film, so you would enlarge 40% more than with a V700, to 85cm instead 60cm with same IQ, if the negative is sharp enough.

Still best quality would be enlarging in the darkroom, in that case you put on the print all through the negative quality which amounts around 160 MPix effective for a 6x7cm BW negative, twice than with the drum, again if the negative is sharp enough and the wet printer knows what he does.

Hybrid is very good, wet is superb. Problem is that tonality control is a hard matter in the darkroom, it requires a true photograper in command that is a master of his tools. In the hybrid... tonal adjustments are pure fun, in the darkroom that is an intensive job that not many master well, at least me I don't master it like I'd want.

 

[Helge]

The human eye is a scanner, not a still camera. We build a scene over seconds in hour minds eye, when we scan with seccades and head movement.
It can resolve any resolution you could throw at it, under the right circumstances.

One of the main strengths of large and medium format is readiness for posterity, it's hedging your bets. For the what if moment when you need to print really large or you need to crop.
42MP is going to run into problems with grain aliasing and won't print as large (whatever that is) as a scan that does justice to the format, or crop cleanly.

 

[138S]

The human eye is a scanner not a still camera. We build a scene over seconds in hour minds eye, when we scan with seccades and head movement.
It can resolve any resolution your could throw at it under the right circumstances.

Well, if not moving the eyes or the head then we see around 8MPix, only that, most is concentrated in the fovea, if not moving the head then with seccades we see around 60 MPix, in the field of view. To need more MPix we have to go to reading distance with a big print, but usually this is not the way art is shown. No problem if wanting a 2m print looking perfect with the nose on it...

42MP is going to run into problems with grain aliasing and won't print as large (whatever that is) as a scan that does justice to the format, or crop cleanly.

A V700 does never generate aliasing when scanning at top dpi because it is an optically limited system, its optics resolve the half than sensor, you find aliasing in systems that are limited by the sampling density, but of course you will find softer grain from a V700 than from a 8000dpi drum scan.

 

[Helge]

Exactly, peoples heads are not in a vice when viewing images (that might be an idea when I start a slide show though ;-). I usually walk up to the pictures and study detail when I'm at a gallery or museum, or even on stuff I have on the wall at home from time to time. And that is something I've observed in most other people, if they are genuinely interested, or try to feign interest. It's a natural instinctive action, and a big part of the enjoyment of the fotos.

You don't have to perfectly resolve grain and grain clusters to have interference patterns happening from the line or matrix of the sensor.
Here on the left is a classic crop from a scanner with very similar resolution to the Epson V7/8XX line. Around 2500 dpi on a good day. On the right is a scan of an enlarged wet print flatbed scanned:

 

[138S]

You don't have to perfectly resolve grain and grain clusters to have interference patterns happening from the line or matrix of the sensor.
Here on the left is a classic crop from a scanner with very similar resolution to the Epson V7/8XX line. Around 2500 dpi on a good day. On the right is a scan of an enlarged wet print in a flatbed:

This was badly scanned, sensors have pixel binning for lower resolutions, voltage from several contiguous pixels in the row are averaged before being fed to ampli and A/D, less conversions is faster.

If you see that noise it's "transversal", in rectangles, I guess this comes from the pixel binning, (I guess that the image was scanned 90º rotated).

This comes from a bad practice, not from a bad scanner, you may also obtain funny effects in a drum, if the drum operator is clumsy enough.

Nobody mentioned before that this "classic crop" may have pixel binning effect from a bad practice?

 

[Mainecoonmaniac]

This is the Alan Ross selective masking technique, published 10 years ago. He has used that way for 20 years at least, he has made a number of workshops about that and I know very good printers using it, one in China, some made intercontinental trips to attend those workshops. It was published in International Photo Technik magazine, IIRC, in his web site that article can be purchased in pdf.

https://alan-ross-photography.myshopify.com/collections/pdfs/products/selective-masking

Amazingly that technique was mostly overlooked, but it's one of the most powerful techniques we may use in darkroom printing, IMO.

PM if you are interested in that way, I've some experince with it, I'd be happy if I can help.

I think this technique has a steep learning curve though I've never bought Mr. Ross's PDF. The masking technique is useful for production silver gelatin printing. If this technique is used in the darkroom, do you think this would make each print pretty much the same? Individually dodging and burning prints makes each one unique. Or does it matter anyway?

 

[138S]

If this technique is used in the darkroom, do you think this would make each print pretty much the same? Individually dodging and burning prints makes each one unique. Or does it matter anyway?

You are free to solve with masking what share you want, and sparing the amount you want for genuine manual crafting. My personal approach is solving the curve tonal manipulation with masking and later elaborating the print manually in what is worth.

I think this technique has a steep learning curve though I've never bought Mr. Ross's PDF.

Not at all, it's a kids game, see this:



From an scan you use a grading map in Photoshop to assign a color to each gray level, you make the negative+diffuser+color mask and that's all, without needing a single brush stroke in Ps . With a bit of experience you nail the mask easily.

Make coarse contact copies of the negative + mask until you get the right balance, then you align well the mask and yo enlarge.

By manipulating the grading map manipulate effective sensitometric tonal curve of the paper, just like when you bend the curves in Ps in digital image edition, you give what extension you want to the toe/shoulder, beyond paper grade.

Me, I see it like a way to have the right paper for the scene or for the negative, I don't want that masking to be intrusive in my manual printing, just I solve the toe/shoulder extensions/gradient in that way.

Citing Alan Ross: "keep at it so long as it serves your purpose! Beyond that, everything is like cooking: personal preference! The options are endless and none are right or wrong! "

Way Beyong Monochrome also speaks a bit about that, and Ross pdfs give straight practial instructions. Just the grading map concept is a natural way to continously distribute grade across densities. This doesn't solve totally the print, but it easily solves toe/shoulder compressions, you simply get the paper you want for the job.

Imagine you project a 5x7" negative on the wall, you are projecting an insane amount of IQ, if you also nail the tonality... this rocks...

 

[Helge]

This was badly scanned, sensors have pixel binning for lower resolutions, voltage from several contiguous pixels in the row are averaged before being fed to ampli and A/D, less conversions is faster.

If you see that noise it's "transversal", in rectangles, I guess this comes from the pixel binning, (I guess that the image was scanned 90º rotated).

This comes from a bad practice, not from a bad scanner, you may also obtain funny effects in a drum, if the drum operator is clumsy enough.

Nobody mentioned before that this "classic crop" may have pixel binning effect from a bad practice?

How would you have scanned better?
I see no evidence of pixel binning. What I do see, is the usual transversal artifacts of a line sensor being dragged mechanically over the subject.
It has its own wholly unique brand of artifacts that are different from bayer interpolation.
Again, flatbeds are far from optimal film scanners.