I bought PacificImage Electronics many years ago, along with SilverFast SE Plus 9 :: LaserSoft Imaging. They have most if not all color profiles for film emulsions and scanners, and support multi-exposure HDR scan along with infra-red dust removal. Scanning is not exactly fast, but there is a lot less fiddling around than trying to hack your DSLR and fight everything at once.
1 Like
I guess we are all a product of our experiences. My experience is that in 2013 I had my wife spend what is for us a considerable sum of money on a dedicated film scanner (with bundled Lasersoft) for my birthday. I thought it was going to be ‘the answer’. It wasn’t.
After 106 scans I concluded that I just wasn’t getting the shadow detail from many of my slides. Stretching the dark areas in post processing quickly produced a nasty reddish purplish ‘mud’ of noise. Admittedly some were underexposed but some just had a lot of dynamic range. I had a second try in 2018 and spent more time adjusting the exposure settings but it made very little difference - the scans got brighter but also more noisy. My conclusion about the scanner was that it just wasn’t able to cope with the density of the dark parts of Kodachrome slides and no amount of calibration or post processing was going to fix that. As I understand it this experience is not unique to me - the dark parts of Kodachrome are very dense but that does not mean that they don’t contain detail.
So for me the film scanner has been an expensive learning experience. Maybe there are scanners out there that can cope better with Kodachrome but I don’t intend to spend more money in that direction.
On a more positive note I discovered Darktable during my second attempt to improve the scanner output and I have not regretted that discovery at all.
I think it was when googling the issue with scanning Kodachrome shadows that I found some articles about using a DSLR to digitise slides. I already own a DSLR (a present for a significant birthday) so I saw no harm in trying it. My initial attempts with a secondhand bellows and my 1980s Minolta lenses on some slides with difficult shadows went well so now I have put together a better setup using an 80mm enlarger lens (secondhand again). Neither the bellows nor the lens has involved a great deal of expense and no birthday credits have been used in the process.
So, I understand what you are saying but please understand that there are financial reasons and past failures which have sent me down this route. I have had some good results already but it’s not ‘job done’ for every kind of image. I have also learned a lot more about image processing from your writing and videos while doing it so thank you for that.
1 Like
I digitized this transparency about 5 years ago, before I was using the camera WB method. Still came out okay by eyeballing the color. Kodachrome 3x4 inch transparency from 1946, taken by my father. One of the wonderful things about Kodachrome is how well it keeps its color over so many years.
4 Likes
Nice! I’ve been sorting through my slides today. I’m fairly sure there are no film stars hiding in my collection.
My earliest (inherited) ones go back to around 1965. Most look OK but a few films have suffered - maybe from age or maybe they were not processed quickly enough after exposure. A few others have not been kept away from daylight and have suffered in that way. Most look usable though, excepting camera and operator errors and some really pointless subjects.
1 Like
My experience is different. I’ve had numerous film scanners with the most recent being the Minolta Dimage Multi Scan Pro, which is an excellent scanner. After starting DSLR scanning I realized that the DSLR scans were not only faster, but more detailed and with better color. Part of the advantage was being able to manipulate those beautiful Nikon .NEF files with great software such as RawTherapee, compared to scanning software (Vuescan) that felt like it was stuck in 1995. I could even get rid of chromatic aberration that was in the original transparency. I sold the Minolta, at a profit, and haven’t looked back. Since I’ve been a professional photographer for a while, I’ve also been able to compare my current DSLR scans to drum scans that were done earlier, and they’re at least as good.
Any scanning workflow, whether using a scanner or DSLR, has a learning curve, and you get better with time and experience.
2 Likes
I know DSLR scanning is the new kid on the block but I agree with @anon41087856 here in that you’re compounding problems doing it that way. I’ve never understood the appeal vs a dedicated piece of hardware like a scanner that is properly calibrated and setup.
I’ve had OK success with my Epson Perfection scanners over the years (works with Linux and VueScan, a nice bonus) doing 35/645 scans. Before that I used the NIkon scanner in the school’s media lab but I honestly didn’t like the results as much as I did the Epson after some calibration. Could be the software or lab build they had on those Macs. Either way, I’m happy with the results from the V600/V700.
Mamiya 645 1000s with Portra 160 from my Epson with some adjustments made in GIMP.
2 Likes
I’m curious about what problems you’re referring to.
Hm, what prevents you from editing the scanner output in the software of your choice? I am using vuescan to capture a “raw” tiff file, which can easily be processed in e.g. darktable.
I am going to test the camera “scanning” method in future though, since there are film formats that do not fit into my scanner. However, if you do own a scanner and the software is the only issue, maybe you can give some details why it does not work for you and maybe we can help solving the issue …
1 Like
I feel like they were discussed earlier in the thread by others, hence my “I agree with …” but the short version: more variables to contend with vs a scanner. You’ve got the profile of the sensor, whatever light source you’re using and the film to contend with. There are also convenience factors to using a scanner like infrared dust removal, DPI adjustment and being able to do more than one frame at a time for smaller formats.
Can you even scan larger formats with a DSLR? The kit I’ve seen for Nikon only does 35mm IIRC. I can see DSLR scanning being handy in scenarios where space or budget don’t allow for a dedicated scanner.
I’m not terribly worried about 100% matching the look of whatever film stock I shot (most will need some correction with age anyway) but I like to get in the ballpark.
Much depends on whether you’re using a dedicated film scanner such as the Nikon Coolscan series or Minolta Dimage, or flatbed scanners such as the Epson V series. The Nikon and Minolta scanners have much better quality than the flatbeds for scanning film. When the Minolta Dimage Multi Scan Pro says 4800 ppi, it means it. An Epson V700, on the other hand, only delivers a true resolution of about 2300 ppi, no matter what the official specs claim. The film scanners mentioned use very sharp lenses and can focus. With the Epsons use fixed focus and require fiddling around with spacers to get the best focus.
With a good DSLR or mirrorless camera, you can use a high-end, very sharp macro or enlarging lens. Each slide is individually focused.
You’re right that the DSLR system doesn’t have IR cleaning. This means I take an extra 20 seconds to properly clean the dirt and dust off the slides. Not a big deal.
I scan 35mm, 6x6, 6x7, 6x9, 3x4 inch, 4x5 inch and 5x7 inch film with my DSLR and macro lens. I don’t use the ES-1 or ES-2 adaptors. When a client needs a very large print from a medium or large format negative I shoot multiple overlapping shots and stitch them together.
You can also take a camera raw file, such as a Nikon .NEF, and process it in Vuescan. I did some testing with this for processing color negatives, and while it’s not bad, I got better results from RawTherapee’s film negative module.
I sold all my scanners after I compared the results to what I was getting with DSLR scanning. I much prefer the quality, speed, and versatility of DSLR scanning.
I’d really like to do a round robin test with the people who are interested here on the forum to compare different methods and hardware. But I guess I need to shoot some images on different film stock first to start such a test 
.
My film scanner does IR dust removal. The manual told me not to use it with Kodachrome. I gather that this is due to the density of the slides being sufficient to block the IR and be misinterpreted as dust. I use compressed air but some dust seems more firmly attached to the slides.
Good luck doing that with Kodachrome 
I do think that Kodachrome adds an extra level of difficulty. I suspect that had I been using other film then I’d have had good results from the film scanner 8 years ago… either that or I’d be asking about correcting faded colours. 95% of the slides in my collection are Kodachrome.
I suspect that the non Kodak films were brief experiments and all three of the people involved went back to Kodachrome after trying the others. It wasn’t necessarily about image quality. I recall complaints from my Dad about plastic slide mounts sticking in projectors. The familiarity of the prepaid processing was probably also a factor.
I’ve scanned - for professional and private use - with pretty much everything there is. From cellphone camera against an opaque food container up to 11k dpi drumscanners.
All those methods have their pros and cons.
In the end it boils down to the classic requirement triangle:
Quality - Time - Money
Pick two of those to optimize. The third one will have to fall in place.
Does not work well with Kodachrome and most of the black&white films due to the infrared properties of the dyes/crystals. In my experience it only works reliably enough with negative film. I opted to turn it off always because in the end nothing beats a manual retouch.
Only limited by the area you are able to light evenly.
2 Likes
It works well on the non-kodachrome slides I inherited from my parents.
I know that there is only approx.1 ± 1 lab left in the world that can process kodachrome. It was meant as a more generic approach. For kodachrome, probably some old “scrap” slides could be used instead of fresh material. The point for me is that there is a lot of opinion about the quality that can be achieved with the different methods, and I really would like to understand how these really compare. A round robin test seems the only reasonable method.
Why am I interested in this? I want to understand if I can do better than what I currently do.
In order for an accurate test to be made, one would have to Compare a scan of the same slide/neg using a variety of different methods: drum, lab, dedicated high end film scanner, dedicated low end film scanner, dslr, and flatbed. Although probably of most interest in this thread would be dedicated high end film scanner vs dslr - the others being too impractical or too low quality. Then with dslr, there is comparisons of the different backlight types. (Not to mention the post processing methods…) If anyone has made such comparisons on the same slide/neg id love to see. I don’t know if resources with these comparisons exist elsewhere.
I never saw something that is at least a little comprehensive. filmscanner.info has a lot of tests (in German language) which include resolution measurements. However, I cannot get practical insight as they are not using the exact same material for testing the different scanners, and they do not provide raw data. Furthermore there’s nothing on other methods than dedicated scanners. Plus, to me the tests feel a bit biased, but that’s my personal opinion.
1 Like
Of course you can make a science out of it. But there are so many factors to control! Some are really difficult to handle, e.g. the aging of the film. So personally I would just go for something pleasing:
Your Lumix photo ssems a little yellowis to me and does not remind my of Kodachrome.
2 Likes
In practical terms the lab count for processing Kodachrome in colour is now zero. I think one very dedicated private individual has had some success in attempting the complex chemistry since Kodak discontinued the chemicals and the final lab stopped at the end of 2010.
Plus there is the issue of finding usable stock of a film that has not been made for more than 10 years.
But if you can find some old slides and enjoy a challenge then I say go for it!
I’ve seen several comparisons but usually they are asking a more limited question and each comparison tends to look at a different question that the author wants to answer. One that I recall was using a (mirrorless IIRC) camera setup and comparing against results from a professional lab scan. Very interesting but not the same question.
Very probably. It’s pretty much an impossible task to convey an experience of the real slide over the Internet. The main point that I was trying to get across is that the greens on that slide are definitely not muted. Why the DSLR raw has flat greens but saturated blues and reds is something that I don’t understand.
I’ve looked again through my OpticFilm scans for comparison. Most are taken in built up areas and green does not feature in the vast majority but in those few slides that have small areas of foliage the green looks pretty muted. I’ve always assumed that this was the slide or just the effect of the hot summer weather but now I am not so sure.