PDA

View Full Version : Scanning Slides



Steven Ruttenberg
8-Apr-2018, 17:10
I am finding that a slide that is underexposed, while it looks fine on the light table, is a pita to scan and then adjust. Seems you lose the crispness of the focus during the scan due to it being underexposed, plus it is hard to bring it up in ps or other software. Wonder if this is just a limitation of the scanner (V850 and calibrated for focus) using Vuescan, creating a raw file so that I can eventually run it thru colorperfect and also for archiving.

Looks like the Fuji Provia 100F needs like 2 or 3 stops over what is metered for a strongly backlit or dim scene. Is there any advice on this?

Ted Baker
8-Apr-2018, 19:17
Try doing separate scans at different exposures. low res is fine and quick

1. Do the first scan raw scan with automatic exposure.
2. Do a second scan where you adjust the exposure gain to blow out the highlights, this will destroy them

Work on the two scans separately, see if you can get a pleasing result for each "half" of the tonal scale. Then investigate multiple exposure.

Remember that neither the monitor nor a print is capable of the contrast ratio of backlit or projected transparency so a tone curve with some compression somewhere will be required to "recreate" what you see on the slide.

Contrast definitely has an effect on our perception of sharpness, the un-sharp mask works on this very principal.

Hope that helps. Hopefully you will get some more useful tips from slide shooters.

Alan Klein
8-Apr-2018, 19:28
You can't get blood from a turnip.

Steven Ruttenberg
8-Apr-2018, 21:12
Try doing separate scans at different exposures. low res is fine and quick

1. Do the first scan raw scan with automatic exposure.
2. Do a second scan where you adjust the exposure gain to blow out the highlights, this will destroy them

Work on the two scans separately, see if you can get a pleasing result for each "half" of the tonal scale. Then investigate multiple exposure.

Remember that neither the monitor nor a print is capable of the contrast ratio of backlit or projected transparency so a tone curve with some compression somewhere will be required to "recreate" what you see on the slide.

Contrast definitely has an effect on our perception of sharpness, the un-sharp mask works on this very principal.

Hope that helps. Hopefully you will get some more useful tips from slide shooters.

I will try what you suggest. I already tried multiple exposure with not much success even with setting gamma to 2.2 in photoshop. Will let know results.

Steven Ruttenberg
8-Apr-2018, 21:13
You can't get blood from a turnip.

Or a vampire :)

Ted Baker
8-Apr-2018, 21:55
I will try what you suggest. I already tried multiple exposure with not much success even with setting gamma to 2.2 in photoshop. Will let know results.

One aspect of what I trying to convey, is reduce the problem to small chunks. I am only guessing what your problem is because I can't see any example.

The other thing to consider is why are you using a raw scan for color transparency? Is it so you can use colorperfect? Otherwise I can't see what you hope to gain, and believe you may just be causing yourself some unnecessary confusion, and problems...

Steven Ruttenberg
9-Apr-2018, 01:27
I am using ColorPerfect. I may try viewscan doesn't have a profile for fuji and the target is the comes with is on a different slide film, kodak I believe. I would need to make a profile specific to fuji. I have the xrite color checker, but need to see if scanner reads the profile it makes. I suppose I could also apply profile in camera raw as well. I am still messing with a workflow for slides and bw negatives. I am good with color reversal. I will post up an image or two of slides Monday sometime. Next weekend I will scan and save as a non raw tiff and see how far that makes a difference. With my shot of Cholla I saved as raw, changed gamma to 2.2 in ps and it worked perfect. So shall see. File type may just depend on each image.

Ted Baker
9-Apr-2018, 03:50
I am using ColorPerfect. I may try viewscan doesn't have a profile for fuji and the target is the comes with is on a different slide film, kodak I believe. I would need to make a profile specific to fuji. I have the xrite color checker, but need to see if scanner reads the profile it makes. I suppose I could also apply profile in camera raw as well. I am still messing with a workflow for slides and bw negatives. I am good with color reversal. I will post up an image or two of slides Monday sometime. Next weekend I will scan and save as a non raw tiff and see how far that makes a difference. With my shot of Cholla I saved as raw, changed gamma to 2.2 in ps and it worked perfect. So shall see. File type may just depend on each image.

If your using ColorPerfect all well and good, you need to provide the correct input for that program.

However for straight to PS if I understand you correctly, then RAW scanning is not helping you. A raw scan has no matrix color transformation applied as well as no gamma curve applied. I don't use PS so I don't know all the details of its processing pipeline, but it sounds like a fudge.

Steven Ruttenberg
9-Apr-2018, 11:59
It is a fudge. The raw has the matrix already applied from the scanner, it is just that, the raw image, no recipe applied. So they generally look ugly till you apply a gamma of 2.2. In this case, it "may" be better to apply the recipe first. I will see how colorperfect does and compare.

Ted Baker
9-Apr-2018, 15:04
The raw has the matrix already applied from the scanner,

It was my understanding that no matrix has been applied to the raw data.

Steven Ruttenberg
10-Apr-2018, 00:19
That is possible, but if that is the case, then you should not get a color image with just raw data, yet I do. I would think the scanner has a bayer filter on it so it is already baked into the raw data. That would have to be the case, otherwise even the simple act of just inverting a raw negative in ps would not give you color. If you mean demosaicing or debayering, then an actual raw converter like camera raw would be needed, but I suspect (could be wrong) that the scanner does this automatically. This is still raw image data, from that standpoint. But, not knowing much about what the scanner is doing, I could be wrong. Color perfect does not do demosaicing/debayering that I know of. I am trying some things out, but definitely the slide was underexposed.

Ted Baker
10-Apr-2018, 05:46
That is possible, but if that is the case, then you should not get a color image with just raw data, yet I do. I would think the scanner has a bayer filter on it so it is already baked into the raw data. That would have to be the case, otherwise even the simple act of just inverting a raw negative in ps would not give you color. If you mean demosaicing or debayering, then an actual raw converter like camera raw would be needed, but I suspect (could be wrong) that the scanner does this automatically. This is still raw image data, from that standpoint. But, not knowing much about what the scanner is doing, I could be wrong. Color perfect does not do demosaicing/debayering that I know of. I am trying some things out, but definitely the slide was underexposed.

I am not referring to demosaicing/debayering, I am reffering to matrix algebra. The process of adjusting the values of each RGB triplet to a known colorspace like sRGB. For example if you had raw RGB values of 65535,0,0, you would have red, but what kind of red would it be?

Peter De Smidt
10-Apr-2018, 06:16
Personally, for slide scans I scan with as few software adjustment made in the scanner as I can. When bringing into Photoshop, I apply a custom icc profile made with Wolf Faust's IT8 targets, and then convert the color space of the file to my editing working space. It's important to use an IT8 file made with the type of film that you are scanning.

Steven Ruttenberg
10-Apr-2018, 08:42
I am not referring to demosaicing/debayering, I am reffering to matrix algebra. The process of adjusting the values of each RGB triplet to a known colorspace like sRGB. For example if you had raw RGB values of 65535,0,0, you would have red, but what kind of red would it be?

Okay, that makes sense. There is no color space assigned to the raw file so I do that in PS.

Steven Ruttenberg
10-Apr-2018, 08:44
Personally, for slide scans I scan with as few software adjustment made in the scanner as I can. When bringing into Photoshop, I apply a custom icc profile made with Wolf Faust's IT8 targets, and then convert the color space of the file to my editing working space. It's important to use an IT8 file made with the type of film that you are scanning.

I will need to photograph my xrite colorchecker target that I use for my digital camera. I can create a "camera" profile for each color slide film type I use and then apply that in camera raw or Lightroom.

Peter De Smidt
10-Apr-2018, 09:24
That is not as good of an approach as using an IT8 reference slide. Such a slide is meant to cover the full range of the emulsion. A color checker doesn't do that.

Steven Ruttenberg
10-Apr-2018, 11:16
That is not as good of an approach as using an IT8 reference slide. Such a slide is meant to cover the full range of the emulsion. A color checker doesn't do that.

Okay. So then I would need to get an it8 target to photograph as you mention above.

mijosc
10-Apr-2018, 11:21
http://www.targets.coloraid.de/

I think doing a proper IT8 calibration with a target of the same film stock you are scanning is critical to getting good color from the scanner. When I first started scanning, I had a terrible time getting decent color from Provia slides when my scanner was calibrated with the Kodak target. Also, I have tried raw scans of transparency film into ColorPerfect and have had less than acceptable results, although I do like it for negative film.

My workflow with transparency film is:
1. Make sure scanner is calibrated with the film stock I will be scanning.
2. Make sure all automatic adjustments and sharpening is turned off.
3. Make sure I'm scanning at 48-bit (16-bit per channel.) (I don't bother with infrared cleaning since I prefer to spot manually in Photoshop later.)
4. Scan film at highest resolution scanner will allow (6400dpi on the Epson V7xx/8xx)
5. Bring the scan into Photoshop
6. Downsample by a factor of 3 or 4 (2133 or 1600 DPI on an Epson V7xx/8xx) using Bicubic (Smoother). This gets me close to the actual resolution of the scanner.
7. Do some quick spotting for dust (how much I do depends on how much I like the image.)
8. Save the image.
9. Import into Lightroom
10. Apply a preset created from a previously scanned IT8 target with exposure, black/white points, sharpening, and color temperature set.

However, even with the most rigorous scanning workflow, underexposed slides are really tough to work with on most commercial scanners. They just don't have enough DMAX to really resolve details in the shadows (this is especially true with Velvia.) I have some darker slides with look ok on a light table, but just turn to muddy mess when scanned. I've tried multi-exposure, but just still can't seem to get there with my equipment. I think, if the image is worthy, those are likely good candidates for a professional Flextight or drum scan.

Steven Ruttenberg
10-Apr-2018, 12:50
http://www.targets.coloraid.de/

I think doing a proper IT8 calibration with a target of the same film stock you are scanning is critical to getting good color from the scanner. When I first started scanning, I had a terrible time getting decent color from Provia slides when my scanner was calibrated with the Kodak target. Also, I have tried raw scans of transparency film into ColorPerfect and have had less than acceptable results, although I do like it for negative film.

My workflow with transparency film is:
1. Make sure scanner is calibrated with the film stock I will be scanning.
2. Make sure all automatic adjustments and sharpening is turned off.
3. Make sure I'm scanning at 48-bit (16-bit per channel.) (I don't bother with infrared cleaning since I prefer to spot manually in Photoshop later.)
4. Scan film at highest resolution scanner will allow (6400dpi on the Epson V7xx/8xx)
5. Bring the scan into Photoshop
6. Downsample by a factor of 3 or 4 (2133 or 1600 DPI on an Epson V7xx/8xx) using Bicubic (Smoother). This gets me close to the actual resolution of the scanner.
7. Do some quick spotting for dust (how much I do depends on how much I like the image.)
8. Save the image.
9. Import into Lightroom
10. Apply a preset created from a previously scanned IT8 target with exposure, black/white points, sharpening, and color temperature set.

However, even with the most rigorous scanning workflow, underexposed slides are really tough to work with on most commercial scanners. They just don't have enough DMAX to really resolve details in the shadows (this is especially true with Velvia.) I have some darker slides with look ok on a light table, but just turn to muddy mess when scanned. I've tried multi-exposure, but just still can't seem to get there with my equipment. I think, if the image is worthy, those are likely good candidates for a professional Flextight or drum scan.


I agree, dark slides suck when scanned. Almost impossible to pull them up. I am going to go ahead and perchase a target and take some shots and create some profiles for different slide film. I scan at 3200dpi (easy division for using with Canon iPF6400 printer. Could go up to 3800dpi. At 4K file was 1.8gb in size, I can only imagine the size and time for a 6400dpi scan, even if down sampling. I prefer though not to down size and just scan at the native resolution I want for the master file.

Ted Baker
10-Apr-2018, 14:55
Okay, that makes sense. There is no color space assigned to the raw file so I do that in PS.

Yes but what colorspace would you assign? :o

Ideally you would calibrate your scanner for the spectral response of the material your scanning, then would know what the raw values in your files are in CIE XYZ colorspace, which you can use to transform to any other colorspace like sRGB.

However in the case of viewscan, when you save anything other than RAW, it is supposed to use a generic transparency profile, i.e. use matrix algebra to transform the raw values into the correct values for the chosen colorspace, the default being sRGB.

chassis
10-Apr-2018, 17:06
mijosc, does step 6 introduce loss or errors, in your view? The workflow seems like a good one. I am curious to hear about your experience on downsampling.

mijosc
10-Apr-2018, 17:32
In my opinion, no, it does not introduce errors. I do it to size the image to the “real” resolution of the scanner and to reduce scanner noise in deep shadows. I’ve played around with different methods of downsampling (nearest neighbor, bicubic sharper, and bicubic smoother.). The differences are subtle, but noticeable. I’ve settled on bicubic smoother, for the moment; I think it minimizes grain.

Of course you could just scan at 2400 or 1600 dpi and get an image that’s 98% there. But where’s the fun it that!


mijosc, does step 6 introduce loss or errors, in your view? The workflow seems like a good one. I am curious to hear about your experience on downsampling.

Peter De Smidt
10-Apr-2018, 17:56
I agree with Mijosc. I usually use bicubric smoother for reduction, and then sharpen as needed at the very end, and depending on use, final size....

Ted Baker
10-Apr-2018, 17:59
In my opinion, no, it does not introduce errors.

I thought your summary was excellent. I use a slight variation on your approach (I am not scanning slides either BTW), but I get improved results, i.e. it looks better and it has a sound technical basis

Remember you can't actually change the sample size of these scanners other than switch lenses, (which is what happens when you go from the full size to the transparency area).

Steven Ruttenberg
10-Apr-2018, 18:59
If I scan at the highest resolution of the V850, what size file does that give me for 4x5? at 4k it is 1.8gb so anything larger than 4gb and you can't write out a tiff or dng file. I do 16 passes to minimize the noise and currently at 3600 dpi it is a file between 1.37 and 1gb depending on how I crop the slide/negative or if I convert to 16bit gray for final output after scanning at 48bit color, this file is about 450-500mb.

Steven Ruttenberg
10-Apr-2018, 19:02
I assign it AdobeRGB1998 in photoshop since it initially comes with no embedded profile. So far, that seems to work. I would also apply the icc profile outside of the scanner only because I am doing raw scans. I would want to compare by applying in the scanner and outside of the scanner to see which is better if either.

Peter De Smidt
10-Apr-2018, 20:15
You apply the icc profile and then covert to adobe98 or ektascan.

Steven Ruttenberg
10-Apr-2018, 20:32
You apply the icc profile and then covert to adobe98 or ektascan.

Then if I bring it to camera raw or lightroom, I apply the icc profile first, then assign the color space? I suppose that would be the same as applying it at the scan stage. I would only do this if creating a file other than a raw file. Colorperfect which I am getting familiar with says not to use an icc profile on the raw image, but to just assign the color space in ps prior to bringing into colorperfect. I know for my digital stuff, I apply the icc profile in lightroom at the beginning before any adjustments. Same with camera raw and work on them in a particular color space. When I output the files then I assign the adobergb1998 or srgb, etc embedding that in the photo.

Ted Baker
10-Apr-2018, 20:35
I assign it AdobeRGB1998 in photoshop since it initially comes with no embedded profile. So far, that seems to work. I would also apply the icc profile outside of the scanner only because I am doing raw scans. I would want to compare by applying in the scanner and outside of the scanner to see which is better if either.

What your, effectively doing here is saying that the raw data from you scanner matches AdobeRGB1998 which is won't but at the same time its not that far away to be a disaster.

So in my example if you had the raw value RGB 65535,0,0 you could now plot that exactly in XYZ space or any equivalent and transform it from there to anywhere in other colorspace. LR AFAIK will use Profoto as its internal model for example.

When you output a standard sRGB jpeg it will transform RGB 65535,0,0 to the appropriate sRGB value. Which BTW will probably result in a clipped value.

Not saying what your doing doesn't work, just trying to be of some assistance.

Steven Ruttenberg
10-Apr-2018, 20:42
What your, effectively doing here is saying that the raw data from you scanner matches AdobeRGB1998 which is won't but at the same time its not that far away to be a disaster.

So in my example if you had the raw value RGB 65535,0,0 you could now plot that exactly in XYZ space or any equivalent and transform it from there to anywhere in other colorspace. LR AFAIK will use Profoto as its internally model for example.

When you output a standard sRGB jpeg it will transform RGB 65535,0,0 to the appropriate sRGB value. Which BTW will probably result in a clipped value.

Not saying what your doing doesn't work, just trying to be of some assistance.

Thank you for the help. I never say I know it all. The day I do, is the time to quit. I always want to learn a better way or the correct way, etc. Being self-taught leaves a lot of room for improvement. I have produced a lot of horrid images that when I go back after learning something they turn out much better.

I can see what you are saying. By applying the icc profile first, then the colorspace, the color values will be more correct than color space first then icc profile. For lightroom, I believe you have a choice of the working space (not sure and will check this tonight)

Peter De Smidt
10-Apr-2018, 21:36
Why are you bringing it into lightroom or camera raw? Just load it into photoshop. Assign the profile. Convert to your editing space. Edit. You're making this harder than it has to be. Raw converters are designed to work with files from digital cameras, as that's what they're overwhelmingly used to do. Their use of profiles is not straight forward.....But Photoshop was developed when high end scanning was the norm, and the workflow has been time-tested for decades.

Steven Ruttenberg
10-Apr-2018, 22:18
I prefer camera raw and lightroom for most editing. The tools for adjusting are much better and easier to use. I use PS for files that are difficult or require more processing than possible in camera raw. Lightroom is my database for my photos and is where I do most editing for files that don't need PS and its additional tools. Same for Camera raw. I use things like luminosity masks, gradient filters, and more in PS. For profile use in LR or CR, I simply choose the camera profile I want and done. Don't get me wrong, I like PS I have been using it since 2001 and it is the king. But I also like simple straight forward. I try to have as many of my photos as close to correct as possible to limit post processing, same with film. Anyway, I do and use whatever method gets me the final image I envision. Sometimes it isn't a straight forward process and I may work an image several different ways before I get something I like.

I do appreciate all the advice.

Ted Baker
10-Apr-2018, 22:41
Raw converters are designed to work with files from digital cameras, as that's what they're overwhelmingly used to do.

Yes that is true, however for raw conversion with the exception of the actual demosacing, what you are doing by assigning an ICC color profile, IS essentially the same. There is only a few different ways to arrange three colours on four spots... LR etc also adds a tone curve to the resultant picture, and make a few tweaks as well AFAIK, but that is not strictly the RAW processing.

This link may help demystif the demosaicing a little bit. https://github.com/codeplaysoftware/visioncpp/wiki/Example:-Bayer-Filter-Demosaic

Of course the way things are explained is very DSLR centric, so I agree is not necessarily all that straight forward.

Steven Ruttenberg
11-Apr-2018, 00:33
Here is where I am at on the slide image of Aspens at sunset in Flagstaff, AZ. This was not an easy scene to shoot. Extremely backlit, no filters. I metered on the Aspens, then adjusted I believe 1 stop, maybe two to the right for zone VI or VII so they would appear white. Even at that, there was still a blue cast I had to fix. I've noticed this a lot with the film both color neg and slide, the blue seems to be way up on the right like zone VI and up. Can't figure that one out unless it is a nature of shooting at sunset or sunrise. I am not happy yet with the forground and it seems a bit too much contrast on the shadow/dark areas. But notice the tops of the trees on the left and see how they are lit up by the sun. Also on the right.

https://ia801503.us.archive.org/17/items/201711040016/20171104_0016.jpg

Steven Ruttenberg
11-Apr-2018, 00:39
Yes that is true, however for raw conversion with the exception of the actual demosacing, what you are doing by assigning an ICC color profile, IS essentially the same. There is only a few different ways to arrange three colours on four spots... LR etc also adds a tone curve to the resultant picture, and make a few tweaks as well AFAIK, but that is not strictly the RAW processing.

This link may help demystif the demosaicing a little bit. https://github.com/codeplaysoftware/visioncpp/wiki/Example:-Bayer-Filter-Demosaic

Of course the way things are explained is very DSLR centric, so I agree is not necessarily all that straight forward.

Yuk, computer code! Good explanation though. RawTherepee and some other sites also explain it pretty well. I spent the better part of last year researching this to decide whether or not to convert one of my cameras to black and white by removing the bayer filter array. I think I will do that later this year or early next. Right now need to get 4 lenses for my 4x5. The ones I am borrowing will be going up for sale soon.

Steven Ruttenberg
11-Apr-2018, 22:56
Here is the scanned slide. 3200dpi, V850, Better Scanning Wet Mount, saved as a raw tif with generic slide settings using Vuescan.

https://ia601508.us.archive.org/7/items/201711040016RawGamma22/20171104_0016_Raw_Gamma_22.jpg

Peter De Smidt
12-Apr-2018, 04:47
Why are you working on ironing out your scanning process using this picture? There are color shifts all over the place. Notice that the center of the sky is much more cyan the the corners. Lower scene value are all on the toe of the film, with poor separation and non-linear color. The sky is blown out. The image lacks good midtones.... Why not start with a well-exposed scene of normal contrast, preferably one with a range of colors, and it's obvious what the colors should be, along with fresh film professionally processed..... Nail that down before moving on to more challenging film.

Steven Ruttenberg
12-Apr-2018, 08:17
I do have the process down for good images. I am trying to work on the more challenging because this is typical of scenes I shoot, but not normally this badly exposed. There is like a 6 stop difference in sky and foreground and using a grad nd wouldn't work well. Given that, there are very little mid-tone in this picture due to the scene and how it is exposd. I had a choice in keeping the foreground better exposed or the sky properly exposed so I took a sorta middle road knowing it would result in washed out sky and darker foreground than I wanted. I see no color shifts as my final image matches well with the scene and other images on different films and even my phone images taken at the time. The sky is and isn't blown out just the corners you get weird stuff due to vignetting.

Appreciate the inputs. Can always learn something. And that is point of this thread, to learn how to work with a difficult images.

mijosc
12-Apr-2018, 08:50
You might be able to squeeze a bit more out of the image by using multi-exposure, but that introduces it's own set of challenges (image alignment.) Ultimately, the scene above is quite high contrast and not ideal for transparency film (especially when scanned with a consumer-grade scanner.) If this image is typical of scenes you shoot, you may be better off shooting on negative film and figuring out that workflow.

Just my 2 cents as someone who also loves transparency film.

Peter De Smidt
12-Apr-2018, 10:10
Why not get the basics down first before moving on to extremes?

Steven Ruttenberg
12-Apr-2018, 10:38
I have basics down. I started with film almost 35 years ago. Scanned thousands of 35mm before switching to digital. Now I am back. This shot just happened to be a bad one and I was curious to see if there is/was a way to salvage it. I got close, but hard to get blood from a turnip as someone stated. Given that, I am not perfect, nor expect to be an award winning photographer however, I am not a beginner. I agree that this scene is not well suited for slide film. The colorneg and bw I took came out much better.

As I stated earlier, I appreciate all comments.

Pali K
12-Apr-2018, 10:50
Steven,

I used an Epson V700 before moving to some of my pro scanners and I can tell you that you are being challenged because darker portions on a slide is what gives the consumer level scanners the most challenge. I don't want to spark another scanner debate but you will get much better shadow detail with a great deal of less headache if you simply used a scanner that was capable of resolving the shadows properly. If you know someone with a pro scanner, have them scan your challenging slides and let Epson take care of majority of the well exposed images.

Pali

faberryman
12-Apr-2018, 10:52
The purpose of scanning is to faithfully reproduce the image that is being scanned. One scanned, you can manipulate the image in LR/PS. They are two separate processes. How closely do your scans relate to the original image?

Steven Ruttenberg
12-Apr-2018, 12:26
The purpose of scanning is to faithfully reproduce the image that is being scanned. One scanned, you can manipulate the image in LR/PS. They are two separate processes. How closely do your scans relate to the original image?

As a raw image with gamma at 2.2 similar but not as good in dark areas. I have not tried creating a file that is not a raw file from scanner. I was considering paying money to have the image scanned with a drum scanner at the camera place I go to in Tempe and see how much better if any it is from the v850. Also for my other shots that are well exposed. Sorta create a benchmark if you will.

sanking
13-Apr-2018, 09:10
I own a Howtek drum scanner and there is no question in my mind but that a drum scanner in good condition will give far better results than any consumer scanner. However, those who own a full frame DSLR or mirrorless camera can come very close to the results of a drum scanner by digitizing their slides by copying with the camera. The effective size in mp of most 35mm color slides maxes out at around 20 mp. Some high resolution FF sensors are capable of twice as much details as there is in a scan, 36-50 mp, and with plenty of dynamic range. For best results, however, you will need a high quality lens optimized for 1:1.

Sandy



...
I used an Epson V700 before moving to some of my pro scanners and I can tell you that you are being challenged because darker portions on a slide is what gives the consumer level scanners the most challenge.
...
Pali