A modern scanner for 35mm and 120 film

[MattKing]

A darkroom plus digital output world is the one I enjoy inhabiting, and it much easier and more practical if you shoot film first.

 

[albireo]

Very well said. I often hear "paper print is what matters" argument and my response to that is "use digital then", because you'll get to your beloved paper much faster and lose nothing in the process.

For me, the exposed film is the final product. It is a chemical interpretation (vs accurate digital representation) of a scene. Everything after is irrelevant and boring presentation medium, basically various forms of getting photons to deflect from from film into your retina with minimal quality loss. The same product can be projected on a wall, shown on a OLED screen, printed on a piece of dead tree or even on a T-Shirt or a tatoo. From this perspective, scanning is of paramount importance and ideally I'd want almost molecular level indestructible digital copy, so the fragile negative can be tossed away immediately after scanning.

Old Gregg - related to the above - I've been mulling over the idea of starting (or joining) a serious, science-informed discussion on how to best optimise exposure and development of a negative for scanning, rather than wet printing.

There seems to be a widely accepted consensus on these boards (but also on other ones) that the best negative for scanning is nothing else than 'the negative optimised for printing'. I remain to be convinced. I haven't been able to find any real scientific evidence in support of this - it does sound more like a truism. To me, it sounds counterintuitive that a negative optimised to be fitted into a highly non-linear medium (which is technically inferior to the negative, but often times artistically interesting) such as the paper should also be optimal for a linear CCD scan whose non linearities would be confined to the realm of thermal transistor noise and probably little else (happy to stand corrected).

So if you, or anyone interested in this matter, would like to start a discussion I'd be all ears and be happy to contribute. Apologies to OP for the slight tangent.

 

[mshchem]

I enjoy digital photography. I enjoy film photography. I like the fact that with film you have a instant durable record. I like making AgX prints color and black and white. I love the Cibachrome prints I made. I love ink jet, amazing technology.
I think I would love high resolution digital photos of my film images. I suppose just starting is the logical place to start.

I had a early Microtek film scanner. It made amazing scans. Had trays upto 8x10. So slow, like a dial up modem

I have scans (Nikon Coolscan V) from original ASA 10 Kodachrome (1?) that I've printed on a Canon ink jet 13 inches wide x 17/18? That are stunning.

So now I am going to look at the thread on electronic flash.

P.S. I fiber base AgX.

 

[Les Sarile]

There seems to be a widely accepted consensus on these boards (but also on other ones) that the best negative for scanning is nothing else than 'the negative optimised for printing'.

Back in early 2000, Kodak released the Kodak 100UC - and other color negative films, calling them "scanner friendly".
From Kodak Ektar 100 datasheet https://imaging.kodakalaris.com/sites/default/files/files/products/e4046_ektar_100.pdf

SCANNING NEGATIVES
You can easily scan EKTAR 100 Film negatives with a variety of linear-array-CCD, area-array-CCD, and PMT film scanners. You can scan negatives on desktop scanners as well as high-end drum scanners.​

But yet when these color negatives were released, there were a lot of posts here and other sites that said the results from their film scans were atrocious and never to use the film again. Was it the film, the conditions and/or the scanner? Even more recently there continuous to be posts asking where the problem is. As an older person who used to get optical prints from many color negs, I don't ever recall getting back prints that were so far off. So it would seem to me, the proper translation of how those optical prints were made would greatly help scanning or DSLR copying.

I asked one of the more esteemed colleagues here about what Kodak meant by "scanner friendly" and his response is, "As I understand it, scanner friendly films have no rough retouching surface or other such "pebbly" surfaces. This improves the scan quality by giving the scanner a smooth surface to act on." https://www.photrio.com/forum/threa...range-mask-removal.166696/page-6#post-2172529

Maybe if Kodak and Fuji had standardized scanning it may have helped - if that was even possible?

 

[brbo]

This is just a question of resolution. The reason it looks so good is because I always run a color de-noise step regardless of a scanning method. I despise color grain. But the fact that you did not see this in the scan, and assumed that this is how CN scans come out of a camera simply tells me you never actually tried scanning with a camera. Why comment then? For some reason it's a trend on this forum.

I agree it's a question of resolution. Line ccd sensors just plain and simply resolve better than sensors with Bayer filter at the same pixel density. Only way around it is to increase pixel count and that is happening with area sensors in digital cameras and line ccd sensors are the thing of the past. One day even Bayer sensors will be good enough...

I don't share your opinion that your scan looks good (when looking at 100%). But that's just my opinion. And I have experience at scanning. Digital camera scanning included. I've done it with many light sources, many different lenses, setups... I currently use "tri-colour" light setup with independent power control of R, G, B channels (halogen bulb filtered by dichroic filters), pixel-shift camera (80MP) and a scanner lens that's better than any other standard macro or enlarger lens at the magnifications typically used in camera scanning. Pixel-shifted 80MP is not enough to outresolve (is that a word?) my Minolta 5400 (using the same lens) on USAF 1951 resolution target, but that is mostly irrelevant since rarely a negative will have that many relevant pictorial information to begin with...

I mainly do RA-4 prints now and I have a fairly good idea of how the "structure" of colour negative looks like (at high enlargements). It looks nothing like you scan would suggest (or any other camera scans I've seen or made myself). High resolution line ccd scans come closest to the "real thing". Most drum scanner are actually too harsh on the C-41 film (at least mine is) when scanned at aperture opening optimised for max. resolution.

Ccd line scanner vs. digital camera scan (at 200%):

You don't see colour dye cloud "noise" in camera scan and it's not because, as you say, you despise it, but because Bayer sensors in current crop of digital cameras can't record it. Pixel-shift doesn't help much either, contrary to the popular belief that pixel-shifting removes the limitations of Bayer filter. And I'm not the only one noticing that.

But, if you are after speed of capture, there is currently nothing even close to camera scanning.

 

[Helge]

I agree it's a question of resolution. Line ccd sensors just plain and simply resolve better than sensors with Bayer filter at the same pixel density. Only way around it is to increase pixel count and that is happening with area sensors in digital cameras and line ccd sensors are the thing of the past. One day even Bayer sensors will be good enough...

I don't share your opinion that your scan looks good (when looking at 100%). But that's just my opinion. And I have experience at scanning. Digital camera scanning included. I've done it with many light sources, many different lenses, setups... I currently use "tri-colour" light setup with independent power control of R, G, B channels (halogen bulb filtered by dichroic filters), pixel-shift camera (80MP) and a scanner lens that's better than any other standard macro or enlarger lens at the magnifications typically used in camera scanning. Pixel-shifted 80MP is not enough to outresolve (is that a word?) my Minolta 5400 (using the same lens) on USAF 1951 resolution target, but that is mostly irrelevant since rarely a negative will have that many relevant pictorial information to begin with...

I mainly do RA-4 prints now and I have a fairly good idea of how the "structure" of colour negative looks like (at high enlargements). It looks nothing like you scan would suggest (or any other camera scans I've seen or made myself). High resolution line ccd scans come closest to the "real thing". Most drum scanner are actually too harsh on the C-41 film (at least mine is) when scanned at aperture opening optimised for max. resolution.

Ccd line scanner vs. digital camera scan (at 200%):

You don't see colour dye cloud "noise" in camera scan and it's not because, as you say, you despise it, but because Bayer sensors in current crop of digital cameras can't record it. Pixel-shift doesn't help much either, contrary to the popular belief that pixel-shifting removes the limitations of Bayer filter. And I'm not the only one noticing that.

But, if you are after speed of capture, there is currently nothing even close to camera scanning.

There is clearly more real detail in the above example than what any of the scans got out. Most immediately obvious in the wildly different interpretation of the small text.
Camera scanning really only comes into its own when you go to higher magnifications than 1:1 on current cameras.
Having easy access to a monochrome sensor would be ideal, but you can kind of, sort of fudge an approximation with a bayered filter and filtered backlight. The bayer filter in almost all commercial cameras is rather weak, precisely to allow for better luminance detail, relying on post work to get the colours back to "normal".

 

[faberryman]

I love striving to obtain the best exposed and developed negative I can achieve - what I then do with it after is entirely my business.

It was ever thus.

 

[warden]

Start a new thread! This is a topic I will gladly follow but I am afraid I can't contribute because last time I wet-printed was in high school. One variation of this topic I am interested in is pushing/pulling. I find it utterly pointless to push/pull because altering contrast digitally after scanning is way easier than chemically, and I am too looking for a solid, scientific validation/rebuttal of my personal experiences.

I have good reasons to both push and to pull, but will wait for someone to start that thread. I'm no scientist but I'll pretend to be one just this once. ;-)

 

[Les Sarile]

I just spent a few minutes trying to find a full-res sample which is not down-sampled and de-noised. I think this one is fairly close. Proimage 100. Even though it's down-sampled it hasn't been de-noised as much like the previous image. You can clearly see the tri-colored dye clouds especially in the shadows. Again, I see this as mostly a pointless exercise as it's clear to me the film doesn't hold enough detail to warrant this resolution especially in low/mid contrast areas.

I've scanned thousands of every type of film - except Proimage 100, and can't say I've seen anything resembling the "noise" shown in your image. Judging by the shadows it is probably a dark area not suitable for this much brightening and ii is a smallish JPEG file so compression artifacts are compounding that mottled look.

Closest I have to compare is this Kodak Ektar 100 with deep shadows from a noon day sun but yet clearly the deep shadows are recoverable and doesn't exhibit that mottled noisy look.

Untitled by Les DMess, on Flickr

Could be the film and/or your scan/copy and/or post work. Been meaning to try a real drum scan as I understand the PMT can pull out much more shadow detail . . .

 

[brbo]

There is clearly more real detail in the above example than what any of the scans got out. Most immediately obvious in the wildly different interpretation of the small text.

That must be some wild interpretation of the text on your part then. As nobody else can read the small text in the camera scan or the ccd scanner scan. The information is not there. The resolution test scans don't lie, though (the camera scan was given an advantage with a bit higher magnification (at that magnification the entire 35mm doesn't fit the sensor) and the scanner was used with in grain dissolver mode that takes away a bit of bite/resolution). Guess which one is 80MP pixel-shift camera scan and which is a 5400dpi ccd sensor scan (both unsharpened and resized 200% in PS):

Camera scanning really only comes into its own when you go to higher magnifications than 1:1 on current cameras.
Having easy access to a monochrome sensor would be ideal, but you can kind of, sort of fudge an approximation with a bayered filter and filtered backlight. The bayer filter in almost all commercial cameras is rather weak, precisely to allow for better luminance detail, relying on post work to get the colours back to "normal".

I agree, with camera scanning you can do higher magnifications and stitch. Monochrome sensor would save you a lot of trouble and/or give you the negative "texture" much closer to reality. You can "beat" the Bayer filter, but it's not something that most will be doing.

This is an example of a proper rendition of a C-41 negative even with a Bayer camera. I did 3 separate (pure R, G and B with narrow band light) exposures captured in raw pixel-shift mode. I then "harvested" all the info that Bayer filter "leaked" into wrong channels and then reassembled purified R, G, B channels and properly inverted the negative.

"Regular" scan with 1 exposure with narrow peak RGB light:



"Proper" scan with 3 RGB exposures:



If you click on both crops you can view the full original scan and you can see how much more colour information the proper scan has (there will be some softness to the 3 exposure scan since it wasn't sharpened after reassembling and there will always be some misalignment with multiple captures). It's not perfect but it's much closer to how a real colour negative looks. This it's NOT noise. You may not like how a colour negative looks and you may prefer to "denoise" your scans, but it is not digital noise.

Also, some of those random-colored pixels you enjoy is noise, not dye clouds.

It's not noise. At my example you can see dye clumps in the same places in both ccd scanners. You telling me what is what on my negative when you even don't have the said negative is... nuts. I know what I see in RA-4 wet print at very high enlargement. It's not noise.

 

[Helge]

That must be some wild interpretation of the text on your part then. As nobody else can read the small text in the camera scan or the ccd scanner scan. The information is not there.

I’m not talking about reading the text (though I’d not be surprised). I’m talking about the guesses the de-Bayering takes and the interference beating the shifting and rotation of the pixel matrix over not quite outresolved high frequency information results in.

The UASF 51 chart plays too nice with the sensor. You need to rotate it and colour it and ideally introduced random colour spots.

 

[Helge]

Very well said. I often hear "paper print is what matters" argument and my response to that is "use digital then", because you'll get to your beloved paper much faster and lose nothing in the process.

For me, the exposed film is the final product. It is a chemical interpretation (vs accurate digital representation) of a scene. Everything after is irrelevant and boring presentation medium, basically various forms of getting photons to deflect from from film into your retina with minimal quality loss. The same product can be projected on a wall, shown on a OLED screen, printed on a piece of dead tree or even on a T-Shirt or a tatoo. From this perspective, scanning is of paramount importance and ideally I'd want almost molecular level indestructible digital copy, so the fragile negative can be tossed away immediately after scanning.

So much contradiction and lack of deep grokking in this post, I’m flabbergasted
For starters:
There is nothing perfect or neutral about a digital “interpretation”. It’s a quantization of the very analog readout from the sensor, with a hell of a lot of processing to “sweeten” the data.

No indestructible almost molecular level digital medium can exist outside wild science fiction. Signal theory and entropy sees to that.
Your “almost molecular” storage is the developed film.
Nothing beats really good analog storage for ultimate resolution. Neither in theory or reality.

And talk about throwing negatives away is wrong for so many reasons and on so many planes that I don’t know where to begin.
Especially considering the recent horrifying trend of labs doing exactly that, and telling customers to not care.

Paper is fantastic for very hard, concrete reasons.
Paper has a few drawbacks and inkjet a few advantages.
But for photos from a well exposed negative (or indeed positive with the right processing) nothing beats paper.

 

[faberryman]

And I thought digital photographers were pixel peepers. But I guess that scanning film, with a scanner or a camera, is, after all, being a digital photographer.

 

[Auer]

And I thought digital photographers were pixel peepers. But I guess that scanning film, with a scanner or a camera, is, after all, being a digital photographer.

The good part about threads like this one is that they do encourage you to shut off your computer/lay down the phone and head out to just shoot some pictures.

Or just start drinking early.

 

[Les Sarile]

@brbo in your samples and comments you are again, just like Les before you, conflating skills and hardware. To anyone who couldn't get the resolution or color out of their camera, I can only say two things: get a camera with resolution you need, and learn to scan with a camera. I could teach you here, but one can't learn while in adversarial state.

As a test engineer I generally don't make assumptions of things I don't know firsthand. And since I don't have the source negatives you used or observed your workflow, I stated your anomalous results could be a combination of a few things. I simply stated I have made thousands of scans and have never encountered those mottled nosiy results you show on your image as well as provided an example to more clearly show the difference in results. Albeit with a different film stock and workflow. An observation rather then an opinion on the results instead of the person. A community minded approach.

I would encourage you to take a selfless self aware approach of your own work, and consider external verification of the issue you identified so that you may know where the problem(s) lie and can improve on it.

 

[Les Sarile]

But for photos from a well exposed negative (or indeed positive with the right processing) nothing beats paper.

It's unfortunate I got into photography just a tad late to experience cibachrome . . .

 

[faberryman]

It's unfortunate I got into photography just a tad late to experience cibachrome . . .

For some reason you have quoted me as saying:

"But for photos from a well exposed negative (or indeed positive with the right processing) nothing beats paper."

That quote belongs to Helge.

As far a Cibachome goes, I did my share back in the 1980s. I don't miss it. My aesthetics have evolved.

 

[Helge]

Nothing? My digital storage beats your film archive. It also deals with increasing entropy nicely. Speak for yourself.



My screens beat your paper. Again, speak for yourself. Also, it doesn't need to be beaten because it has nothing to do with the product, it's just a presentation medium. I can scan and hex-dump a negative onto a t-shirt and scream "nothing beats t-shirt" and that would be just as absurd.

Speak for myself‽ What does that even mean? Evidently you are not.

Your digital storage is prone to a whole host of threats that film is not.
Hollywood still does separation prints for movies digitaly recorded or not, for long term storage. That should tell you a thing or two.
No museum relies on digital storage for anything important either.

None of your likely storage solutions are immune to entropy.
No storage solution really is. Some are just a whole lot more resilient.
Personal drives and backups are just so obviously problematic, that I won't even bother starting to explain why.

Cloud storage is just off the top of my head susceptible to:
- Paused or stopped payment for shorter or longer periods. Forgetfulness, neglect, death, sickness or other distress can easily lead to payments methods and accounts not being updated for shorter or longer periods.
It can happen now, in ten years or in fifty, or a hundred.
If so, your photos are gone.
Forever.
- Malicious ransom or war attacks.
- Cosmic or other radiation. Even with redundant storage, this is still an issue.
- Corrupted writes and files on your side. You won't realise until you try to retrieve them.
- Cheap cold storage solutions (non wired with mechanical retrieval from, optical physical media), could turn out to be a little too cold. The cost of retrieving the data could turn out to be more expensive than anyone is willing to pay for partially or wholly forgotten photos. Photos that would have been enjoyed immensely if they had been casually available.

"Just" a presentation medium‽ It's not too much of a stretch to pull out McLuhans old dictum "The Medium is the Message" (of course one should keep in mind McLuhans definition of a medium as nested inside each other, being bodily extensions of some kind and having a recursive and co-evolutionary nature).
Slide is the presentation medium for instance. Is that absurd?
Photo paper simply has certain qualities that makes it better than screens and ink prints, and... yeah what else is there really? Digital projection?

 

[Alan Edward Klein]

Let's face it. Pictures of relatives are the photos most descendants care about, Are your landscapes and other vacation shots that great and better than others they'll care about them? Todays; youngster don't even care about hard copies of current stuff they took being satisficed to flip open their cell phones.

My suggestion is make enlargements of family shots, the best, and give them to your relatives. Landscapes too. They'll treasure them and actually display them in their homes for you to see when you visit and left displayed when you're gone. They'll be appreciated now so you'll enjoy the thank you's now. Who's visiting your attic today and rummaging through your shoe boxes looking at your old photos?

 

[Alan Edward Klein]

No matter how much you backup, print, or otherwise secure your work for the future, is there really anyone around who will care enough to even look at it? I'm currently going through all my photos to come up with enough pictures to make a tabletop book. Even I'm going crazy with this project going through all the pictures. And it's my work! Who else in the future would care enough to spend the time? It seems like a hopeless supposition.

We all tend to think we're more important than we really are. Maybe we are. But our photos? Are they really that good; really that important? Better off scratching your initials and date on some rocks in the Grand Canyon if you're looking for immortality.

 

[Helge]

@Helge The marginal cost of having a digital copy is astronomically lower than a physical copy. And making copies is how you beat the odds. Come up with N doomsday scenarios and I'll have N+1 copies. Same goes for paper vs screen vs projection. You can have all of the above from a quality digital "source of truth". My point is that simply designating a single source of truth and separating it from presentation is important. Because your handling of the single source of truth should be different vs presentation copies. Presentation mediums come and go, there's variety of papers, after all. I am firmly in "high quality scan is the source of truth" camp. An actual negative is slightly better, but too fragile and prone to loss. And yes, when you talk about data loss, speak for yourself, because it's entirely dependent on one's habits and data management hygiene.

My negatives cost me nothing in the few binders they occupy. And I can move them with me wherever I'd like.
Do you know where your copies/backups are? Exactly?
It's not the cost of storage. It's the cost of loosing it I worry about.
Copying your files, I guarantee you that at some point you will forget. And then it's too late. The flash cells will have de-charged and corrupted your compressed data.
I have both. But I'll never have to worry about my negatives as long as I keep them out of fires and floods. I'll always have access to the "pure source".

 

[Helge]

No matter how much you backup, print, or otherwise secure your work for the future, is there really anyone around who will care enough to even look at it? I'm currently going through all my photos to come up with enough pictures to make a tabletop book. Even I'm going crazy with this project going through all the pictures. And it's my work! Who else in the future would care enough to spend the time? It seems like a hopeless supposition.

We all tend to think we're more important than we really are. Maybe we are. But our photos? Are they really that good; really that important? Better off scratching your initials and date on some rocks in the Grand Canyon if you're looking for immortality.

99 percent of photos are of little value to anyone. I have access to lots of old photos, some of them over a hundred years old. Most of them are just plain boring to look at, and tell very little of story or about history.
But here is the thing: A small percentage of those photos I think are pointless, are potentially very, very valuable to someone somewhere. And which ones that are valuable changes from person to person, and through time. That is why you should just save everything. Or at least as much as possible. And in the form of film that can always be read and realised as what it is.

 

[Helge]

Well as I said, it comes down to personal data hygiene. Funny you mention this, but yes, I know exactly where most of my copies are. Maybe not the room/rack/blade, but the datacenters. We also have vast datasets on failure rates, which allow us to calculate probability of data loss with high precision. That gives me far greater confidence in my practices vs hypotheticals you have to deal with. Fire? Fixer left in emulsion? Rat piss? Batshit crazy girlfriend throwing them out? Care to assign a number on these?

I have 9 copies of everything in 5 physical locations (two of them are mine), with automatic duplication and automatic restore procedures running weekly and bi-monthly with email notifications on completion or failure.

But yes I still keep the negatives simply because why not. But I am very deliberate about treating my scans as my single source of truth.

"Data hygiene"? Sounds like hours of fun.
And yeah, data-centers and flash drives... where you will have to sink thousands of dollars into new drives and fees every decade. Until you or someone forgets.
How about no?
You can never completely guard yourself from maliciousness or force majeure. Stay away from insane people, bottom shelves and firetraps though, and there is a good chance your negatives will be there in a hundred years, without any more effort than being lifted into and out of a box or shelf every other decade.
I have albums, plates and negatives from the 1890s and the 1920s that are in pretty much perfect condition.

 

[faberryman]

And then you have to ask yourself whether any of the images are even worth saving. And for whom. And why.

 

[Helge]

From 1890s? That's adorable, Helge. One moment you're applying immersion lithography ideas to film scanning, and then you proceed to estimate data loss chances based on a dataset of one. It's like there are two identities behind that online alias Frustrating but fun. Always enjoy reading your posts.

Dataset of one? Personally I own a few boxes of them, and I know for a fact hundreds of thousands more exist in Copenhagen alone. Can't imagine it's much different generally globally.
Immersion lithography? How about plain microscopy. It dramatically cuts down on diffraction and dispersion and is used routinely.
Wouldn't be that hard to do. Just a small flat bottomed "stuck tire track" pool of scanning liquid and a glass pressure plate on top.