Using a sigma to digitize slide film (maybe negatives too)

What are your thoughts on this? Seems like with no bayer interpolation this would be the best possible technical solution, no?

I am wanting a small compact camera for other reasons, but I am wondering if this would also work for slides..

DP3 Merrill Compact digital camera Refurbished | Sigma Corporation of America

I was thinking the dp3, as this I think is a 1:3 macro lens on apsc. Is this not close enough?

My purpose in digitizing them is to simply create contact sheets so I can easily pick the few (if any) I want scan at 4000dpi on my coolscan. I would like this sigma for other reasons, but if it would serve double duty that would be awesome. I may sell the coolscan soon also, and in that scenario this camera would have to do some more heavy lifting in a more dedicated set up copy stand setup where I would care more about quality/resolution. But it would be nice to know that option is there.

I honestly dont need 4000dpi scans, I just want to share at 20mb on the web...with most of my enjoyment take from printing and projecting...unplugged.

1:3 is not great for a full frame, but perhaps less than 1:1 would be ok on an aps-c sensor? I dont know how this works....

Thanks

I'll go so far as to say any interchangeable-lens digital camera with a reasonably high megapixel count can give excellent digitised copies of slides or negatives. It doesn't need a Foveon sensor.

This Foveon versus Bayer thing has been raised in another thread on camera 'scanning'. It's a completely bogus argument in my view, since it seems based on a misunderstanding of how Bayer de-mosaicing works.

What definitely is needed for decent slide/negative copying is a good macro lens, some way of holding the film flat and parallel at a fixed distance from the lens, and a consistent light source. Those things are much more important than the technology used in the camera sensor.

If you look around, you can find front-of-lens devices consisting of a telescoping tube with a slide/filmholder attached to one end and a filter thread at the other. This is by far the simplest and most reliable method IME. The light source can be simply an on-camera flash pointed at a white reflector in front of the camera and copier.

Further to my post above. Here's an example from an old 35mm colour negative digitised using the sort of rig I described above on a Nikon 24 Mp D7200.

All inversion of tones and colour correction was done using the free GIMP image editor. It took maybe 5 minutes using the curves tool. That's probably less time than waiting for a filmscanner to do its thing, and then still having to tweak the colour by hand.

"as this I think is a 1:3 macro lens on apsc. Is this not close enough?"

Not really. You need a lens that focusses to 1:1.5 with good quality and a flat field.

It's pointless worrying about sensor resolution if you use a poor lens and only half-fill the frame.

Further to my post above. Here's an example from an old 35mm colour negative digitised using the sort of rig I described above on a Nikon 24 Mp D7200.

 

[ATTACH=full]1274241[/ATTACH]

 

All inversion of tones and colour correction was done using the free GIMP image editor. It took maybe 5 minutes using the curves tool. That's probably less time than waiting for a filmscanner to do its thing, and then still having to tweak the colour by hand.

 

"as this I think is a 1:3 macro lens on apsc. Is this not close enough?"

 

Not really. You need a lens that focusses to 1:1.5 with good quality and a flat field.

It's pointless worrying about sensor resolution if you use a poor lens and only half-fill the frame.

Thank you for this. I cant help but always wonder while I sit there and wait for the coolscan to do its thing (which seems to take long enough to make me annoyed even at 600dpi) that I could process 1/2 a roll in the time it takes me to do one, And even then, the scanned image of the slide is never quite matches the light box...then comes Photoshop. Which is fine, but I think I would likely get a much better depiction if I just put a camera lens up to the thing and took a picture.

What would be very attractive to me would be to adjust the color on the live view until the jpeg looked like the slide does on the light box and when I have that nailed, burn through the rest and live with the ones that dont capture perfectly. Only the perfect exposures I spend more time on, the rest either get tossed or added to my carousel.

I have had a constant love/hate relationship with color-negative film. I like the moody film look, but hate the process of checking the roll. Before I even start thinking about how to scan these three types of film in higher detail, I first need to sort out a quicker process for making a digital contact sheet. Because really, this scanning step for me is really just about finding a process that works for proofing that is manageable. Once I have that down, I can start to refine the scanning/workup process to maximize quality and detail.

I will try the GIMP tool. I have avoided manual inversion thus far as I can never seem to be able to remove the orange mask fast enough or well enough.

If there was a digital camera that could shoot an inverted black and white image that would be fantastic! This way I could quickly proof my black and white negative and save my darkroom paper and chemical for actual prints.

The 'trick' with colour negative is to get rid of the orange mask at an early stage of post processing. I copy the negative RAW, which gives me 14 bits/channel to play with. I set the camera white balance to Tungsten and illuminate the film with flash. Therefore the over-blue WB does much of the job of overcoming the amber mask. The job is finished by finding a neutral tone (if possible) in the RAW editor and selecting it with the WB eyedropper. After doing that, the colour pretty much comes right after tone inversion - not the contrast or tone curve mind, just the colour balance.

The next stage is to use the curves tool to boost the contrast and usually apply an 'S' shaped curve. Any colour imbalance is usually slight at this point, and can be corrected using the individual RGB channel curves. It sounds more long-winded and complicated than it is.

The very reason I used to shoot negative film is the flexibility in colour and tonality it allows. The wide subject brightness range allows almost any level of shadow transparency I like, and the fact that the colour isn't 'fixed' allows me to interpret the colour the way I imagined it, and not how the maker of a slide film designed it.

Jiggling with the colour in-camera and expecting a perfect JPEG to pop out is a bit of a pipe-dream IMO.

The 'trick' with colour negative is to get rid of the orange mask at an early stage of post processing. I copy the negative RAW, which gives me 14 bits/channel to play with. I set the camera white balance to Tungsten and illuminate the film with flash. Therefore the over-blue WB does much of the job of overcoming the amber mask. The job is finished by finding a neutral tone (if possible) in the RAW editor and selecting it with the WB eyedropper. After doing that, the colour pretty much comes right after tone inversion - not the contrast or tone curve mind, just the colour balance.

 

The next stage is to use the curves tool to boost the contrast and usually apply an 'S' shaped curve. Any colour imbalance is usually slight at this point, and can be corrected using the individual RGB channel curves. It sounds more long-winded and complicated than it is.

 

The very reason I used to shoot negative film is the flexibility in colour and tonality it allows. The wide subject brightness range allows almost any level of shadow transparency I like, and the fact that the colour isn't 'fixed' allows me to interpret the colour the way I imagined it, and not how the maker of a slide film designed it.

 

Jiggling with the colour in-camera and expecting a perfect JPEG to pop out is a bit of a pipe-dream IMO.

I was not going to use flash, as I have a 5000K color-corrected fluorescent lamp in my light box. Same?

I will give process a go soon....

ES-2 Film Digitizing Adapter Set from Nikon

ummm...this looks handy

ES-2 Film Digitizing Adapter Set from Nikon

 

 

ummm...this looks handy

- Yes. There are several devices of this type that screw to the front of a macro lens. The rigid assembly that moves as one with the camera body almost completely eliminates vibration.

However, there are a couple of caveats with this type of device. The distance between slide/film holder and lens attachment should be variable, and should also be long or short enough to allow use with whatever sensor size and macro lens you intend to use.

The ES-2 is designed to be used with a specific Nikon macro lens, and on a full-frame sensor camera. I'm not sure its range of adjustment would allow use with a smaller sensor, or with a different macro lens.

Personally, I bought an obsolete Nikon ES-E28 kit of similar design. It had a 28mm screw thread to fit to an old Coolpix digital. I bored it out and fitted a 42mm camera flange. This allows me to use M42 extension tubes between the copier and macro lens, and to get a large range of magnifications from the device.

My setup and a result from it is shown in this thread.

Read the entire thread, since it's all pertinent to what you want to do.

There's also a current thread in the medium format forum discussing camera copying.

I don't recommend using a continuous source lightbox; not unless it's rigidly fixed to, or placed on a copying stand. Getting the camera square on to a lightbox with a detached tripod is painstaking, and there's a strong possibility of vibration ruining the definition of copies.

I'm new to this. I did a lot of slide coping with a bellows & slide hold back in the 80's so I understand the idea.

What is the MP count that is needed for this to work? I like 11x14 prints so what camera would be a good choice?

What is the MP count that is needed for this to work? I like 11x14 prints so what camera would be a good choice?

- Anything over 12 megapixels will give you a decent 11" x 14" print, and a 24 megapixel camera will get all the detail there is to be got from a 35mm film frame.

A higher megapixel count won't necessarily get you better digital copies.

The most important part is the lens used to make the copy. Either a good quality prime macro lens or a 6 element enlarging lens on a bellows are what I'd use.

I have huge experience with trying to get a proper color inversion from a negative photographed with a digital camera. The only thing that comes close to doing it properly is the color negative module from Colorperfect but even this has it's problems.

The Photoshop inversion is linear and what is required is an algorithmic inversion that creates a curve similar to RA-4 photographic paper. Neutralizing the orange mask before inverting in Photoshop helps but it's hit and miss and still creates an inversion with color crossover.

As evidenced in Rodeo Joe's sample, the sky in the upper right is approximately proper sky blue but the foliage is yellow as are the rocks which should be far more neutral in tone. Added to that, the shadows are blocking up and have no detail. Pretty normal for a negative photographed with a digital camera. Sorry Joe...I don't mean to rain on your parade and what you have here is probably good enough for most people but it won't compete with the original posters Nikon Coolscan.

Bottom line is don't sell your Coolscan but do as you had originally suggested. Use the digital camera to create quick and easy images that can be used as a sort of contact sheet. Do your best images in your wonderful Nikon scanner.

Too bad that “wonderful Nikon scanner” is an orphan product that Nikon hasn’t supported with parts or service for years. Whip in OS issues and the tedium of requisite work-arounds and the pricing of top-shelf Coolscans starts looking just a bit absurd. Suspect help is on its way with software that will cure the color neg inversion blues. Current production civilian-grade scanners haven’t seen performance upgrades since when?

I've only ever used the Nikon coolscan 9000 HD and everything you say is true. I hated the workflow and the fact I needed to use an ancient computer with it but it did produce good scans...a lot better than an inversion from a digital camera. I think Silverfast has software for them but I was using the unit on lone to see if I wanted to buy it or not so I just dealt with what I had. I didn't buy it because other than producing a good scan, the experience was soul sucking. Now I'm using a Creo iQsmart with older OSX but while there are some issues, it's no where nearly as clunky as the Nikon Coolscan and the scans are a bit better. Anyway...I felt my post was getting a bit negative and needed some positive levity...I get your point.

As evidenced in Rodeo Joe's sample, the sky in the upper right is approximately proper sky blue but the foliage is yellow as are the rocks which should be far more neutral in tone.

- Oh, you were there when I shot that scene were you?

No, the gorse was a yellow green colour, probably more yellow than I've rendered it there, and the lichen-covered rocks weren't neutral grey. Otherwise they'd be the same colour as the white and neutral grey cloud that you see right next to them.

Look, I've used a number of film scanners to scan colour negatives, and none of them come close to a decent colour rendering of colour negs straight from the scanner, and nearly always need hand adjustment. They also have a tendency to blocked shadows.

Let's see some examples of colour negative scans from you can we?

Then we can pick holes in your colour rendering and tonal range.

I have tried a LOT of different methods of scanning film and slides.They all work, sort of, depending on what you are going to do with the scans.

The use of a copy camera seems a good idea on first consideration, but my personal experience is that if you want to archive images in the highest possible quality there is really nothing that works as well and as efficiently as an honest-to-pete film scanner. 4000 ppi scanners like the Canon CanoScan FS-4000US, (cheaper if you can find one that works) and the Nikon SuperCoolScan 9000 (still priced very high) work the best. I have scanned more than 664 GB on disk for over 95,000 images. I have tried most of the various scanning methods (I started out back in film days with a Repronar).

Various reports here on P.net include the following,

not in any order:

Which macro lens to get in order to scan 120 film with a Nikon D600

What scanner for 35mm slides & prints up to A4?

Setup for slide scans

I need a new flim/slide scanner

NIKON SCANNER 5000ED

CanoScan 9000F vs. CanoScan FS 4000US

Huge Scanning Job Finished (yet again!)

Maybe we should recall the OP's question and the purpose of this thread before turning it into a scanner versus camera-copy war?

I honestly dont need 4000dpi scans, I just want to share at 20mb on the web...with most of my enjoyment take from printing and projecting...unplugged.

- There's no doubt in my mind that digitising using a camera is quicker, easier and usually requires less outlay, than using any sort of dedicated scanner.

The question was whether a Sigma Foveon camera would be a good choice or offer any advantage over a conventional Bayer sensor. That question remains unanswered, I suspect because of the low popularity of Sigma cameras.

I'd be interested in the answer myself, but until someone comes along that's used both types (Foveon and Bayer) of camera in a similar resolution for digitising film, then we might never know.

- Oh, you were there when I shot that scene were you?

No, the gorse was a yellow green colour, probably more yellow than I've rendered it there, and the lichen-covered rocks weren't neutral grey. Otherwise they'd be the same colour as the white and neutral grey cloud that you see right next to them.

 

Look, I've used a number of film scanners to scan colour negatives, and none of them come close to a decent colour rendering of colour negs straight from the scanner, and nearly always need hand adjustment. They also have a tendency to blocked shadows.

 

Let's see some examples of colour negative scans from you can we?

Then we can pick holes in your colour rendering and tonal range.

Sorry man...I didn't mean to offend but just expressing my experience that your sample precisely demonstrates. Send me the negative if you like for me to scan and you can compare. No use in posting anything unless it's apples to apples. Once I do that, I'll send you my results and you can compare and see the difference. Unless you'd want me to post them in this thread.

Please, also check out the color negative module offered by Colorperfect. It has actual profiles for different film types your photographing and does an ok job of the inversion. It will save you a lot of time.

Too bad that “wonderful Nikon scanner” is an orphan product that Nikon hasn’t supported with parts or service for years. Whip in OS issues and the tedium of requisite work-arounds and the pricing of top-shelf Coolscans starts looking just a bit absurd. Suspect help is on its way with software that will cure the color neg inversion blues. Current production civilian-grade scanners haven’t seen performance upgrades since when?

if it does not break it does not need support. I have been using it constantly for 5 years with no issues. Any new scanner I bought would be off warranty by now. If it breaks, someone in the world fill fix it. I will just ship it to them. I would have to do the same even if nikon still serviced it

I've only ever used the Nikon coolscan 9000 HD and everything you say is true. I hated the workflow and the fact I needed to use an ancient computer with it but it did produce good scans...a lot better than an inversion from a digital camera. I think Silverfast has software for them but I was using the unit on lone to see if I wanted to buy it or not so I just dealt with what I had. I didn't buy it because other than producing a good scan, the experience was soul sucking. Now I'm using a Creo iQsmart with older OSX but while there are some issues, it's no where nearly as clunky as the Nikon Coolscan and the scans are a bit better. Anyway...I felt my post was getting a bit negative and needed some positive levity...I get your point.

I use the coolscan with the free nikon scan software on my brand new windows laptop with whatever operating system it came with. 20 minutes of googling and I got it to work. in fact, its working right now as I type this

it is slow, but I have kept it. I might retire it for color negatives but keep it for slides (easier to preview). Ill probably but a pakon instead of a digital camera for creating contact sheets and call it a day.

If I get a new phone that shoots raw I may experiment with a phone macro lens for digitizing with negative lab pro

Send me the negative if you like for me to scan and you can compare.

Thanks Greg, but there's absolutely no need.

I went back to the camera RAW 'scan' and re-processed it to address all the 'faults' you accused it of.

As you can see, it lacks the punch and drama of my intentional interpretation of the scene. No shadows are blocked, but why anyone would be interested in peering under a rock or between twigs of gorse, I have no idea!

Here's another camera 'scan' from a quite difficult neg. It refused to print other than flat and quite lifeless. Digitising it allowed me to put a bit more vibrance into the colours.

Again, there's no sign of the blocked shadows you accuse digital camera copies of always having.

Here's a flatbed scan of a wet-print for comparison.

And I will get around to cloning out the ugly overhead wires one day!

I must add that I've used a procession of different scanners, and that I started scanning nearly 20 years ago. I'm not a novice at the game. So I can say with some authority that digital camera copying loses no tonality if shot RAW and processed properly, and is far quicker than any scanner I've yet to use.

It took me no longer than 20 minutes yesterday to digitise an entire roll of film. Most of that time was spent loading strips of film into the film-carrier. Of course it'll take much longer to digitally process and adjust them, but that task is much the same with a scanner workflow.

I used colour negative for its flexibility; in the darkroom and to a much greater extent, after digitising.

To me the final image should be, as Ansel Adams said, the 'performance' from the negative 'score'. Not necessarily giving a colorimetric perfect rendering, nor showing needless detail in highlights and shadows. If the shadows or highlights need to be sacrificed for the sake of extra contrast or aesthetic effect - so be it.

I will give Colorperfect's software a try, but I'm doubtful it can automatically deal with the million different shades of contrast mask out there, and the billion-and-one different lighting conditions that can be encountered.

Thanks Greg, but there's absolutely no need.

I went back to the camera RAW 'scan' and re-processed it to address all the 'faults' you accused it of.

[ATTACH=full]1279975[/ATTACH]

As you can see, it lacks the punch and drama of my intentional interpretation of the scene. No shadows are blocked, but why anyone would be interested in peering under a rock or between twigs of gorse, I have no idea!

 

Here's another camera 'scan' from a quite difficult neg. It refused to print other than flat and quite lifeless. Digitising it allowed me to put a bit more vibrance into the colours.

[ATTACH=full]1279976[/ATTACH]

Again, there's no sign of the blocked shadows you accuse digital camera copies of always having.

 

Here's a flatbed scan of a wet-print for comparison.

[ATTACH=full]1279977[/ATTACH]And I will get around to cloning out the ugly overhead wires one day!

 

I must add that I've used a procession of different scanners, and that I started scanning nearly 20 years ago. I'm not a novice at the game. So I can say with some authority that digital camera copying loses no tonality if shot RAW and processed properly, and is far quicker than any scanner I've yet to use.

 

It took me no longer than 20 minutes yesterday to digitise an entire roll of film. Most of that time was spent loading strips of film into the film-carrier. Of course it'll take much longer to digitally process and adjust them, but that task is much the same with a scanner workflow.

 

I used colour negative for its flexibility; in the darkroom and to a much greater extent, after digitising.

 

To me the final image should be, as Ansel Adams said, the 'performance' from the negative 'score'. Not necessarily giving a colorimetric perfect rendering, nor showing needless detail in highlights and shadows. If the shadows or highlights need to be sacrificed for the sake of extra contrast or aesthetic effect - so be it.

 

I will give Colorperfect's software a try, but I'm doubtful it can automatically deal with the million different shades of contrast mask out there, and the billion-and-one different lighting conditions that can be encountered.

Cool...let me know how it goes with Colorperfect. This is what that module was designed for. If nothing else and you don't get superior results you'll get quicker results of equal quality to what you're getting here.

Cool...let me know how it goes with Colorperfect. This is what that module was designed for. If nothing else and you don't get superior results you'll get quicker results of equal quality to what you're getting here.

Thanks Greg, but there's absolutely no need.

I went back to the camera RAW 'scan' and re-processed it to address all the 'faults' you accused it of.

[ATTACH=full]1279975[/ATTACH]

As you can see, it lacks the punch and drama of my intentional interpretation of the scene. No shadows are blocked, but why anyone would be interested in peering under a rock or between twigs of gorse, I have no idea!

 

Here's another camera 'scan' from a quite difficult neg. It refused to print other than flat and quite lifeless. Digitising it allowed me to put a bit more vibrance into the colours.

[ATTACH=full]1279976[/ATTACH]

Again, there's no sign of the blocked shadows you accuse digital camera copies of always having.

 

Here's a flatbed scan of a wet-print for comparison.

[ATTACH=full]1279977[/ATTACH]And I will get around to cloning out the ugly overhead wires one day!

 

I must add that I've used a procession of different scanners, and that I started scanning nearly 20 years ago. I'm not a novice at the game. So I can say with some authority that digital camera copying loses no tonality if shot RAW and processed properly, and is far quicker than any scanner I've yet to use.

 

It took me no longer than 20 minutes yesterday to digitise an entire roll of film. Most of that time was spent loading strips of film into the film-carrier. Of course it'll take much longer to digitally process and adjust them, but that task is much the same with a scanner workflow.

 

I used colour negative for its flexibility; in the darkroom and to a much greater extent, after digitising.

 

To me the final image should be, as Ansel Adams said, the 'performance' from the negative 'score'. Not necessarily giving a colorimetric perfect rendering, nor showing needless detail in highlights and shadows. If the shadows or highlights need to be sacrificed for the sake of extra contrast or aesthetic effect - so be it.

 

I will give Colorperfect's software a try, but I'm doubtful it can automatically deal with the million different shades of contrast mask out there, and the billion-and-one different lighting conditions that can be encountered.

considering you have extensive experience with scanning, I would be interested in what you think about negative lab pro as well. If you do end up trying both, please share