Digitizing Color Negative with Camera

[AdHoc007]

Nice work and a cool set up.

I recently found that I could use my recently acquired FF with a 25 year old slide tube duplicator. Before my crop camera was wrong size.

I did this for diffusing the light





The light diffused off the card stock and the TTL of the camera did well with the flash.

I know a good macro lens is better, but I do not have one and this is for old family slides and pictures where the IQ and focus is not the best anyway.

Is there any way I could put negatives in this set up other than cut them up and put in slide mounts?

thanks

whvick

It looks like that white diffuser panel on the front has a clip, does that panel flip open?

 

[whvick]

Nice work and a cool set up.

I recently found that I could use my recently acquired FF with a 25 year old slide tube duplicator. Before my crop camera was wrong size.

I did this for diffusing the light





The light diffused off the card stock and the TTL of the camera did well with the flash.

I know a good macro lens is better, but I do not have one and this is for old family slides and pictures where the IQ and focus is not the best anyway.

Is there any way I could put negatives in this set up other than cut them up and put in slide mounts?

thanks

whvick

It looks like that white diffuser panel on the front has a clip, does that panel flip open?

yes, but there is a big gap between the plane of that diffuser, and the plane of the channel that you push the slide into. I need something that would allow me to slide the row of negatives through. As I said before if I really want one I can cut it and put it into a slide mount. Do you think there is a mount to hold a row of negatives? what would it be called?

thanks

whvick

 

[ZOIP]

Ah, well this is where it gets tricky, as I mentioned in a prevous post, it is not actually that easy to get the correct filter set up. Heres a few issues that must be adhered to if you want optimum results, the lowest ISO must be used and the exposure perfectly set, filtering can cause issues if ) a you raise the ISO to compensate or the exposure time becomes longer. Both will increase noise of course, hence my use of a powerful flash to capture the image.

Next the correct WB setting is not daylight, in fact far from it, the ideal is actually the cameras uni WB setting, which is probably more like 3200k, anything else compromises the exposure. Hence the blue filter is only a rough option.

You absolutely cannot judge anything regarding the quality of the cross channel exposure in anything other than the cameras Uni WB setting, all you are doing is looking at the cooked JPEG histogram with all the cameras in built processing parameters factored in, in Uni WB with all the cameras parameters zeroed out you will be seeing as close as possible what is actually going on across the three channels.

Basically it is like this you factor in the uni WB for the camera, the orange mask of the film you are scanning and the colour temperature of the light source, complex I know and it takes a lot to even explain how to do this, but trust me it will reduce in combination with a suitably powerful light source your chroma noise to virtually nothing, all you will have is residual photon shot noise and of course the film grain.

ultimately as I mentioned previously the correct filter is far more likely to be a mid cyan, if you go too far you will in fact get more noise in the sky as the sky before inversion is orange.....go too far and you will need to gain up the cyan/blue and increase noise.

A jpeg done in camera may show less noise, but it will also be have poorer highligh and shadow edit ability and will likely have some issues with micro banding and innacuracies in the shadow and highlight colour and the built in NR will make the final result look less analogue.

Of course all this is not say a quick and limited approach is not sufficient for web or small print purposes.

I really should write a full article on this, i can see from all the posts there is a fair level of interest and most of the info on the web I have seen is cursory at best and often completely wrong and since it takes a lot of effort it would be nice for people to be able to get good reliable results.

 

[Wayne Larmon]

I really should write a full article on this, i can see from all the posts there is a fair level of interest and most of the info on the web I have seen is cursory at best and often completely wrong and since it takes a lot of effort it would be nice for people to be able to get good reliable results.

Yes there is a fair level of interest. I've been working on this for a long time.

1. I didn't know about the importance of UniWB. What you've said makes sense. I'll withdraw anything I've said about using Daylight WB and starting with a Rosco #3202 filter.
2. That aside, I do have a good handle on doing negative conversions derived from this thread on Luminous Landscape. (There have been other related threads on LL.) Look at what Bart vanderWolf has written on this thread. I translated what Bart said into a step-by-step procedure that is easy to duplicate. "Converting method 1" is closest to what Bart recommended. The other methods aren't as rigorous, but are easier and still produce excellent results.
3. The other difficult part is mounting the film/slides in front of macro lens. One of my earliest camera scanning resources was Peter Krogh's Camera Scanning page on dpBestflow.org The most salient point is there are very few film carriers (for want of a better term) for camera scanning on the market. Peter Krogh's recommendations are almost entirely "find used darkroom equipment on eBay." I used to recommend the Photosolve Xtend-a-Slide, but that is no longer being manufactured. However, there is a discussion going on on a different DPReview thread that may offer a solution.

I get good, reliable results and anybody that reads my collection of camera scanning pages can duplicate my process, because I give step-by-step instructions with copious screenshots and I make the RAW files that I used available for download.

I don't think that I have anything that is completely wrong on my camera scanning pages. I'll cop to "cursory", but I welcome any constructive criticisms.

Wayne

 

[ZOIP]

Hi Wayne, thanks for the link, I have not seen your page before and I certainly was not being critical of anything you have done, rather I just feel there is a lack of truly in depth analysis of the process on the web and how to best resolve the annoying issues.

I think you have done a great job translating Barts info into step by step instructions.

On the issue of mounting the film/slides, I agree this is very tricky and often overlooked. I had to build a custom rig and mounts out of circuit board material which allowed me to tape the film to the mount after stretching it, a lack of flatness reeks havoc with across frame clarity as you would expect.

Equally important is getting everything absolutley parallel, hence my rig has various methods of leveling both the camera and film board, ( I shoot vertically) and even more important is accurate focus, which I resolved by building a micrometer adjuster to adjust the camera up and down, rather than turn the lens barrel, which is far to coarse for repeatable and accurate results, especially once you get into the realm of stitched hi res captures.

Slightly off focus at very high magnifications can actually make the grain of the film look far worse, I suspect this is often an issue with casual setups. My testing revealed that at the really high magnification for stitched film capture you can actually get noise grain structure differences due to which layer of the film you are focused into. Anyhow the long and short of it all is that until I built a purpose built rig I like everyone else struggled to get consistent results and most of that inconsistency was due to micro variations in focus.

For the above reasons I have found bellows useless, (lack of properly parrallel faces), non macro lenses ( field curvature issues), optical viewfinders on DSLRs, ( lack of alignment between sensor and focusing screen and too difficult to micro focus).

Easily the best lens I have used is my 55mm micro nikkor reversed on fixed lenght extension tubes, stopped down to f7.1. Here is another tip, most folk try to compensate for the lack of focus precision and film flatness by stopping down to say f11 or smaller, unfortunately the effects of diffraction will play havoc with the grain structure of the film and make it look more pronounced than it really is, although in normal shooting the small losses from diffraction would be a non issue and could be resharpened out, but with film will just make the grain structure look worse......so back to that old chestnut..... The film must be ultra flat and everything perfectly parallel!

Anyhow Wayne do try to play with getting the filtering working for you, magically if you do exactly nail it all you have to do is invert the file without any difficult adjsutments and then tweak the endpoints and the colour a little on the final file.

One final point, strong reds are the one colour that will cuase the greatest issues, negs are usually weak in the cyan area and poor captures/ conversion will see the resulting reds look noisy and posterised and maybe grainy.

Cheers

Brad

 

[Wayne Larmon]

Hi Wayne, thanks for the link, I have not seen your page before and I certainly was not being critical of anything you have done, rather I just feel there is a lack of truly in depth analysis of the process on the web and how to best resolve the annoying issues.

The process has lots of complications. It should be possible to abstract a lot of the complications away, but this remains elusive. Because there is so much DIY involved.

On the issue of mounting the film/slides, I agree this is very tricky and often overlooked. I had to build a custom rig and mounts out of circuit board material which allowed me to tape the film to the mount after stretching it, a lack of flatness reeks havoc with across frame clarity as you would expect.

The Xtend-a-Slide with my custom negative carrier works fine for most 35mm negatives. Most of my negatives only has a modest amount of curl and my negative carrier suffices to flatten most negatives. It is unfortunate that it isn't being manufactured anymore. I hope that ProfHankD works on his 3D printed "Duplihood" some more to fill the gap.

Equally important is getting everything absolutley parallel, hence my rig has various methods of leveling both the camera and film board, ( I shoot vertically) and even more important is accurate focus, which I resolved by building a micrometer adjuster to adjust the camera up and down, rather than turn the lens barrel, which is far to coarse for repeatable and accurate results, especially once you get into the realm of stitched hi res captures.

I use manual focus using live view at 5 or 10x with my Canon 60D. At 10x I can focus the grain (for consumer grade color negatives. I don't have any fine grain B&W.) I focus each frame individually. The manual focus "throw" on the EF-S 60mm macro lens is barely adequate, but it works. On the consume grade 35mm negatives that I scan.

Slightly off focus at very high magnifications can actually make the grain of the film look far worse, I suspect this is often an issue with casual setups. My testing revealed that at the really high magnification for stitched film capture you can actually get noise grain structure differences due to which layer of the film you are focused into. Anyhow the long and short of it all is that until I built a purpose built rig I like everyone else struggled to get consistent results and most of that inconsistency was due to micro variations in focus.

Have you seen these pages?

http://www.addicted2light.com/2012/...non-5d-mark-ii-vs-drum-scanner-vs-epson-v700/

http://www.addicted2light.com/2012/11/29/how-to-scan-films-using-a-digital-camera/#more-3882

He uses stitching to camera scan medium format film. However, his method involves resting the front of the camera+lens directly on the negative, which make me feel queasy.

For the above reasons I have found bellows useless, (lack of properly parrallel faces), non macro lenses ( field curvature issues), optical viewfinders on DSLRs, ( lack of alignment between sensor and focusing screen and too difficult to micro focus).

This is interesting. I never tried a bellows. Peter Krogh recommended bellows based carriers (on the page I linked to earlier.)

Easily the best lens I have used is my 55mm micro nikkor reversed on fixed lenght extension tubes, stopped down to f7.1. Here is another tip, most folk try to compensate for the lack of focus precision and film flatness by stopping down to say f11 or smaller, unfortunately the effects of diffraction will play havoc with the grain structure of the film and make it look more pronounced than it really is, although in normal shooting the small losses from diffraction would be a non issue and could be resharpened out, but with film will just make the grain structure look worse......so back to that old chestnut..... The film must be ultra flat and everything perfectly parallel!

Anyhow Wayne do try to play with getting the filtering working for you, magically if you do exactly nail it all you have to do is invert the file without any difficult adjsutments and then tweak the endpoints and the colour a little on the final file.

The adjustments aren't difficult--just snug the ends of the histogram for each R, G, and B channel. The amount of chroma noise I get by filtering digitally (instead of optically) isn't bad. I only have to use Noiseware sometimes. The worst offender for noise was Photoshop ACR's default sharpening. Once I realized that that the default sharpening in ACR increased grain and chroma noise (while doing nothing for image detail), most of my noise problems went away.

Well, and after realizing that Rosco #3202 filter made the noise worse, compared to no optical filtering at all.

Do you have any links to Uni WB techniques? Google isn't coughing up much. All I'm getting are links to DPReview forum posts that discuss using Uni WB. But no tutorials. (Google keeps insisting that I really want to search for "Uniball" and only grudgingly gives the few links to DPReview posts.)

One final point, strong reds are the one colour that will cuase the greatest issues, negs are usually weak in the cyan area and poor captures/ conversion will see the resulting reds look noisy and posterised and maybe grainy.

We can also use RawDigger to examine R, G, and B exposures in our RAW files.

My mind is a bit sluggish on all this because I haven't dug into camera scanning for a number of months. I'm glad that you are on board and are contributing to the discussion. I need to think about this for a while and remember other issues. (Illuminant spectrum considerations...)

Wayne

 

[AdHoc007]

I use manual focus using live view at 5 or 10x with my Canon 60D. At 10x I can focus the grain (for consumer grade color negatives. I don't have any fine grain B&W.) I focus each frame individually. The manual focus "throw" on the EF-S 60mm macro lens is barely adequate, but it works. On the consume grade 35mm negatives that I scan.

Makes me glad I use a manual focus lens since I can set it to exactly the magnification ratio marker written on the lens then do a slight adjust on the focus rail and I'm done.

He uses stitching to camera scan medium format film. However, his method involves resting the front of the camera+lens directly on the negative, which make me feel queasy.

I've seen that method as well, and I would not want to do it, just sounds like a way to add scratches to the negative. I also saw someone say they taped their negatives to the light box to remove the curl, no thanks. The best way to get the negative flat I have found is to use a negative holder that is used with film scanners. Only downside right now is that they are too big for me to use with my current slide copier. Ideally I would have a negative holder that when put together is 2 inches high and could fit within the slide copier attachment that I have.

right now my slide copier can copy film negatives fairly well so long as the frame I want to copy is not at the end of the film strip. I have made a crude improvisation to get around that by cutting open a slide mount and using it to hold frames that are at the edge of the film strip. I'm not crazy about the plastic, I'll probably replace it with cardboard in the next few days.

Well, and after realizing that Rosco #3202 filter made the noise worse, compared to no optical filtering at all.

I went and checked what Rosco 3202 does, and from the description of it, and the graph provided on the website, it appears that 3202 suppresses both Green and Red channels.

Here is a graph from the Cyan filter I use. It greatly suppresses just the red channel. There are other cyan filters that allow more red light to pass through..

 

[rjjr]

I'm starting to get into similar work with my A6000 and I got some very good ideas from your setup...Thanks.

Just to see what I could do I put your greener mountain scene though CS6 using a crop, color adjustment using individual color levels and color balance, and a bit of sharpening. Here's what I came up with...

 

[TJ61]

Glad you found this info useful, and nice work on the photo! It gives me more confidence to forge ahead with the digitizing even though I lack the PP skills at this point. I haven't had time to play with this since I got my gel filters, so that's next on my list.

 

[TJ61]

I went and checked what Rosco 3202 does, and from the description of it, and the graph provided on the website, it appears that 3202 suppresses both Green and Red channels.

Here is a graph from the Cyan filter I use. It greatly suppresses just the red channel. There are other cyan filters that allow more red light to pass through..

Glad I reread this just now, now that I have my filters in front of me. I was thinking I'd start with the 3202, but not after seeing these transmission curves.

 

[rjjr]

Glad you found this info useful, and nice work on the photo! It gives me more confidence to forge ahead with the digitizing even though I lack the PP skills at this point. I haven't had time to play with this since I got my gel filters, so that's next on my list.

Thanks, I learned restoration while digitally repairing old photos for a museum and publications. I found a book back then to get me started but now there are tutorials on the 'net about the restoration of old photos, including color correction.

 

[Dreamscape Photo]

WOW great job, i have a negative scanner but am always trying to do it with a better sensor. I have tried doing it a few times but not got it right yet!

 

[Richie Beans]

I've done a similar kind of thing but had a much easier time of it with a legacy macro and bellows set-up. But that was the easy part. Despite successfully processing a roll ASA400 color film, my entire system became useless when processing a different film speed, or even the same film speed, but of a different generation.

In digging around, I discovered why results were so damned hard to come by. As it turns out, the reason why color film negatives are orange is because of paper technology. Color response of print paper requires a projection through film of that color. This is why white light, projected through the unexposed part of a color negative (which is orange) yields a black field on print paper. I forget the reason why print paper was made to respond to that wavelength in such a specific way, but it's worth looking up.

In the end, all of the colors within a negative contain differing responses to that orange, which is not the same as shooting the same image through an orange filter screwed to the front of your lens. If the latter were the case, it would be no trick to simply compensate for that orange filter in the white balance setting, and every image processing software would have a one-click sliding compensator button called "EZ Color Negative Comp".

So, even if you could wrestle decent results out of a roll of color negative film, you'd have to go through the same process and tweak the settings again for that next roll, because, unless you've stored those negatives in vacuum-sealed temperature-controlled environment deep in the belly of an abandoned gold mine, every roll of negative film will be different.

Which is why, in the end, a dedicated scanner still provides the best color renditions for the least effort: those color profiles are built into the scanner software. However, if you absolutely, positively have to scan that 40lb box of negatives, right before Thanksgiving, take it to Costco, have prints made, shoot the prints.

At least, that's been my experience.