Is DSLR scanning "better" than flatbed scanning?

[brainmonster]

I have an Epson v550 that I got on the cheap for $50, but I've been considering getting a macro lens for doing DSLR scanning. Some people claim the scans are better but is this true?

I also want to do 120 scans...lots of question marks on the setup which has been delaying me from doing it.

Also some people claim it's faster, but I've talked to some that say they don't like the workflow. Inverting the colors and removing the color cast was too tedious for them.

What do you all thing? Do you prefer it and is the effort worth a vastly increased IQ with DSLR scanning (I've got a full frame camera but no macro lens at the moment) or is this a myth?

 

[runswithsizzers]

There is no doubt in my mind that dSLR scanning has the potential to make better digital captures of film than some modestly priced flatbed scanners. Putting together a d.i.y. copy rig does require more research and resourcefullness than just buying a scanner off the shelf. Per dollar spent, I think a modestly priced d.i.y. digital camera copy rig can compete with some of the higher priced flatbed scanners - assuming you already have a decent digital camera, that is. Which workflow you prefer is going to be a personal choice.

I have been playing with both a d.i.y. digital camera copy rig and an old film scanner for a while now. Still haven't made up my mind for sure, but for slides and b&w I think I prefer the digital camera method. But for color negative film the software I use with the scanner (Vuescan) may help me get better color than I've been able to achieve from the digital camera files. I have been mucking about with manual color inversion/corrections in Lightroom and Photoshop. I have not yet tried any purpose-built plugins for inverting color negs.

On my website <here> I have some comparisions between my APS-C mirrorless camera + enlargng lens + bellows --- and my old Minolta film scanner (which is probably a step up from an entry level flatbed). I have also collected some resources provided below.

In <this blog> the author demonstrates superior results with a DSLR and a macro lens compared to an Epson 700V, though he is using a somewhat advanced and tedious camera process.

Discussion of camera scanning and color conversion using the Negative Lab Pro plugin for Adobe Lightroom <here>

Discussion of camera scanning and color conversion using the Color Perfect plugin for Adobe Photoshop <here> by the developer of that software.

Discussion about building a d.i.y. digital camera copy rig <here>

Forum thread about digital camera "scanners" <here>

A long thread on Photrio discussing various methods of inverting color negatives <here>

 

[Anon Ymous]

Hi

I have a Minolta Scan Dual III, not a flatbed scanner. My scanner is better than a typical flatbed and I'm fairly pleased with it. I've used it to scan bw, C41, E6 and Kodachrome film, always saving raw files. I don't use Vuescan for negative film inversion and do it manually in Photoshop using curves. This setup gives me ~10MP files with the 2820dpi resolution it provides. Flatbed scanners may supposedly have much higher resolution, but reality is totally different and fail to even reach 2820dpi.

I also have a Canon 6D (20MP) and an EF 100mm f/2,8 macro lens. I've used it mounted on a tripod facing downwards to the film, which is placed in the film holder of the aforementioned scanner. The holder is placed slightly above an old mobile phone, which is used as a backlight. The camera is connected to a computer and I shoot in live view mode, with x10 magnification. I'm trying to focus on grain. This setup gives me more than 17MP after moderate cropping and easily beats my scanner.

That's the good part and this generally looks promising. The bad part is that it's tedious and has some shortcomings. The film has to be perfectly flat and parallel to the sensor plane, which is easier said than done. It's very tough to have everything in focus. Focusing at the center of the frame will make the edges, and especially corners, out of focus. Stopping down is required, but you may need to stop down more than you'd have wished. f/5,6 might be ideal in terms of resolution, but f/8, or f/11 might be required to have everything in focus, but diffraction will reduce sharpness. So, some sort of jig is required to keep everything parallel and simplify the process.

 

[brainmonster]

Yeah I thought about using the Nikon ES-1 for DLSR scanning, but it seems like it doesn't work with 120 film unfortunately. If you use 35mm it might be a good option although you'd need to do some adapting with your 100mm lens, I'm not sure it would work without some major modification.

That's good to know that it's more difficult than it sounds. This video features a guy who recommends a specific film holder which keeps your negatives flat, might be some help:



But yeah, I'll have to think about it and some way to keep it perfectly level. Thanks for your input as it sounds more difficult in practice than in theory.

Hi

I have a Minolta Scan Dual III, not a flatbed scanner. My scanner is better than a typical flatbed and I'm fairly pleased with it. I've used it to scan bw, C41, E6 and Kodachrome film, always saving raw files. I don't use Vuescan for negative film inversion and do it manually in Photoshop using curves. This setup gives me ~10MP files with the 2820dpi resolution it provides. Flatbed scanners may supposedly have much higher resolution, but reality is totally different and fail to even reach 2820dpi.

I also have a Canon 6D (20MP) and an EF 100mm f/2,8 macro lens. I've used it mounted on a tripod facing downwards to the film, which is placed in the film holder of the aforementioned scanner. The holder is placed slightly above an old mobile phone, which is used as a backlight. The camera is connected to a computer and I shoot in live view mode, with x10 magnification. I'm trying to focus on grain. This setup gives me more than 17MP after moderate cropping and easily beats my scanner.

That's the good part and this generally looks promising. The bad part is that it's tedious and has some shortcomings. The film has to be perfectly flat and parallel to the sensor plane, which is easier said than done. It's very tough to have everything in focus. Focusing at the center of the frame will make the edges, and especially corners, out of focus. Stopping down is required, but you may need to stop down more than you'd have wished. f/5,6 might be ideal in terms of resolution, but f/8, or f/11 might be required to have everything in focus, but diffraction will reduce sharpness. So, some sort of jig is required to keep everything parallel and simplify the process.

 

[Anon Ymous]

Yeah I thought about using the Nikon ES-1 for DLSR scanning, but it seems like it doesn't work with 120 film unfortunately. If you use 35mm it might be a good option although you'd need to do some adapting with your 100mm lens, I'm not sure it would work without some major modification.

That's good to know that it's more difficult than it sounds. This video features a guy who recommends a specific film holder which keeps your negatives flat, might be some help:



But yeah, I'll have to think about it and some way to keep it perfectly level. Thanks for your input as it sounds more difficult in practice than in theory.

Yes, an approach like the ES-1 is good, my first thought was to DIY something like that. But then, this would probably be considerably heavier and my macro lens has plastic filter threads. Somehow, I don't feel too comfortable about it.

Now, regarding the video you mentioned, this guy has a more firm/stable setup than mine. This alone makes things easier for him. The tripod head he uses is easier to set correctly and vastly firmer than mine. That said, he crops heavily. This means that the focusing distance is farther from the film plane and this obviously gives him a fair bit of depth of field more, so he can get away with f/5,6. He also discards the edges and corners of the image he captures, which are the most problematic parts, both in terms of lens sharpness and proper focus. He gains in practicality, but loses in image size. Still, he probably gets better output than from an Epson flatbed. By the way, if scanning BW film, you most definitely don't need a plugin like negative lab pro. You can convert to BW and simply use curves to get the job done. To be honest, I very much doubt about the usefulness of any plugin like negative lab pro, but to each his own.

 

[GLS]

It's a topic with various factors to consider, however if quality in the end result is paramount I would say unequivocally yes, provided you take the time to do it properly and have a reasonably good DSLR and macro lens. I have detailed my DSLR "scanning" setup on these forums in the past (https://www.photrio.com/forum/threads/neg-scans-grain-with-fp4.157955/page-3#post-2137561)

Things become tricky when you start to attempt multi-shot stitches (of medium or large formats) at (or near) 1:1 magnification. At that point the miniscule DoF involved means film flatness and film & sensor plane alignment become absolutely critical, as does any traces of field curvature in the copy lens you are using. However if you keep things limited to around 1:2 or 1:3 magnification or less then it is not hard to maintain even focus across the frame. I myself routinely only take a single digital shot containing the whole of my 6x6 or 6x7 film frames, and with the D810 I still get files of around 20-22 MP when subsequently cropped. This is easily good enough for most uses, unless you want to make giant prints.

 

[Chan Tran]

Compare to your V550 of 4800dpi it would yield 30MP image from a 35mm film. More on 120 film. If you use a camera with more than 30MP you can have higher resolution for 35mm film but it's hard to beat the scanner with 120 film in term of resolution. Using a camera, setting it up takes time and it can be difficult to hold the film flat and you would also need a very good lens for this.
If you scan slides then the DSLR is much faster if you have many slides and you set the system up in a way that changing the slides will not make you refocus or framing. You need very little PP for slides. (except if you want to remove dust in post). For negatives the inverting of color, correcting color and contrast is quite involved.

 

[runswithsizzers]

Hi
[...] It's very tough to have everything in focus. Focusing at the center of the frame will make the edges, and especially corners, out of focus. Stopping down is required, but you may need to stop down more than you'd have wished. f/5,6 might be ideal in terms of resolution, but f/8, or f/11 might be required to have everything in focus, but diffraction will reduce sharpness. So, some sort of jig is required to keep everything parallel and simplify the process.

Agree with just about everything Anon Ymous said.

My set up works vertically - that is, my camera is above the light table, pointing down. To align the film and camera, I use a bubble level - the round kind, like a bullseye. When the light table and the back of the camera are both level, they are parallel.

Personally, I have not noticed corner-vs-center focus problems due to the slight curvature of 35mm film. I use the slide / negative holders that came with my film scanner to hold the film on my light table, and they keep the film somewhat flat. I did some depth-of-field (dof) testing, and I started to notice a slight loss of fine detail sharpness due to diffraction starting to become noticeable at f/16. Diffraction blurring was objectionable to me at f/22. So I set my lens to halway between f/11 and f/16. When using an APS-C camera to photograph 35mm film, the magnification required is about 0.65X. Using the Macro Depth of Field Calculator on <this website> shows the following dof numbers at these apertures:

@ f/11, dof= 1.81 mm (APS-C)
@ f/16, dof= 2.63 mm (APS-C)​

... so my DOF is about 2 mm, which is challenging, but manageable.

When copying a 35mm film with a full frame digital camera, the magnification is about 1X, and the dof is a little more challenging:

@ f/11, dof= 1.41 mm (full frame)
@ f/16, dof= 2.05 mm (full frame)​

Not only is the flatness of the film in play, but some lenses have a flatter field of focus than others. Generally speaking, I believe most quality macro lenses were designed to have a flat field of focus. Enlarger lenses are also designed for a flat field of focus.

Copying 120 film requires lower magnification and is much more forgiving of curved film, careless alignment, or lenses with moderate field curvature.

 

[mshchem]

Old School

 

[markbau]

Just curious if a high quality enlarging lens could be used with a DSLR, assuming you could somehow fit an enlarging lens onto a DSLR?

 

[Anon Ymous]

Just curious if a high quality enlarging lens could be used with a DSLR, assuming you could somehow fit an enlarging lens onto a DSLR?

Yes, it has been done and the results can be excellent. I have a Schneider Kreuznach Componon-S 50mm f/2,8 which I'd like to use. They can be used with either a focusing helicoid, or bellows and most of the times the best results are obtained when the lens is reversed. This also requires the use of reversing ring, combined with any step up rings.

 

[Alan Edward Klein]

I have an Epson v550 that I got on the cheap for $50, but I've been considering getting a macro lens for doing DSLR scanning. Some people claim the scans are better but is this true?

I also want to do 120 scans...lots of question marks on the setup which has been delaying me from doing it.

Also some people claim it's faster, but I've talked to some that say they don't like the workflow. Inverting the colors and removing the color cast was too tedious for them.

What do you all thing? Do you prefer it and is the effort worth a vastly increased IQ with DSLR scanning (I've got a full frame camera but no macro lens at the moment) or is this a myth?

Monster: Your scanner is fine for the web. I have a similar V600. You can see my results for 120 and 35mm on my Flickr page below. What do you want to do with the results? If you want to make a beautiful blowup of a few, it might be better to send those out to a pro scan place. The rest for the web, could be handled by the flat bed. I've had no experience with using a DSLR to scan. So I can't comment on that. Good luck.

 

[Chan Tran]

Just curious if a high quality enlarging lens could be used with a DSLR, assuming you could somehow fit an enlarging lens onto a DSLR?

I use the enlarging lenses for close up work all the time. I don't have macro lens.

 

[markbau]

I use the enlarging lenses for close up work all the time. I don't have macro lens.

How do you attach it to a digital camera? I'd like to try this with a Pentax K mount. I have a spare LPL 6700 enlarger head which I'm assuming would be a perfectly even backlit light source.

 

[mshchem]

I use the enlarging lenses for close up work all the time. I don't have macro lens.

Absolutely, I have a 105mm "Bellows Nikkor" which is basically an enlarging lens. With the Nikon PB-4 bellows you can even do very slight "corrections" . All the Nikon Coolscan units had a ED glass enlarging lens. Use a nice LED light source, it would go fast. Of course this is for 35mm only.
Personally if I have a scanner I would use it.

 

[GLS]

Compare to your V550 of 4800dpi it would yield 30MP image from a 35mm film. More on 120 film. If you use a camera with more than 30MP you can have higher resolution for 35mm film but it's hard to beat the scanner with 120 film in term of resolution.

You can set 4800 dpi on the scanner, but its meaningful (i.e. true) resolution is not much better than 2000 dpi, and that's assuming the scanning lens is perfectly focused on the emulsion (in practice this is virtually impossible to achieve, or maintain, whereas it is trivial to do so with a DSLR). Setting anything higher than 2400 dpi is just generating larger files for no real resolution benefit. Also, I disagree that a flatbed scanner is better for 120 film. I have never once seen a result from one which rivals what I can achieve with my DSLR setup, and I have inspected many such flatbed scans (including ones from professional labs). Many articles I have read comparing the two also come to the same conclusion.

I think in terms of pure convenience a flatbed is more relevant when it comes to scanning large format films. Of course, one could still do so with a DSLR and reap the benefits, but it would require more effort (especially for 8x10).

 

[Mainecoonmaniac]

I think DSLRs are a good alternative for scanning film. I think it's important to get the right lens. Flat field lenses, which some are enlarger lenses are a good choice. I scan with both a mirrorless Sony and a flatbed scanner. My Epson V-700 came with horrible film holders that didn't work well. I a darkroom and I used my old negative carriers to hold the film in place. I use my Paterson Benbo tripod over my light table.

 

[jim10219]

It's a topic with various factors to consider, however if quality in the end result is paramount I would say unequivocally yes, provided you take the time to do it properly and have a reasonably good DSLR and macro lens. I have detailed my DSLR "scanning" setup on these forums in the past (https://www.photrio.com/forum/threads/neg-scans-grain-with-fp4.157955/page-3#post-2137561)

Things become tricky when you start to attempt multi-shot stitches (of medium or large formats) at (or near) 1:1 magnification. At that point the miniscule DoF involved means film flatness and film & sensor plane alignment become absolutely critical, as does any traces of field curvature in the copy lens you are using. However if you keep things limited to around 1:2 or 1:3 magnification or less then it is not hard to maintain even focus across the frame. I myself routinely only take a single digital shot containing the whole of my 6x6 or 6x7 film frames, and with the D810 I still get files of around 20-22 MP when subsequently cropped. This is easily good enough for most uses, unless you want to make giant prints.

That's pretty much it in a nutshell.

Here's what I recommend. You need a good copy lens. One with a flat field a focus. Not all macro lenses make good copy lenses. Enlarger lenses attached to bellow work well. Second, get a neutral light box. By that I mean one with a good CRI and color temperature. I've had issues with light boxes with good color temperatures but bad CRI's, and that made getting a decent shot of a negative where the blues aren't blown out and the reds are still properly exposed difficult. Fluorescent and some LED light sources will have issues with this. Don't worry so much about brightness, as you can always extend your exposures. Third, come up with a good mask system to block out stray light from getting in around the negative. Not only can this cause glare, but it can skew your histogram and make exposures more difficult. I use a standard enlarger negative carrier for this. It has the added benefit of lifting the negative off the light box so I don't get Newton rings. Fourth, do it in a dark room, or cover the copy stand from stray room light. Glare bouncing off the backs of the negatives can cause issues. Fifth, get yourself a bubble level and level both the camera and the film you are shooting. That makes alignment much easier. SIxth, take your time and watch out for dust.

It'll take a long time to learn how to control every step of the process. And it'll take a while to learn how to process the shots in software after you've taken them with your DSLR. But once you've mastered the process, it's the best way to scan film short of a good drum scanner. And even then, your primary advantage to a drum scanner is ease of use and not having to stitch multiple photos together in software. I can easily resolve the grain on any film with this setup. I rarely take it to that level, because of the amount of work involved, the huge file sizes, and the time it takes. But it's nice having a setup where I can do multiple negatives at once as a proof sheet, a bunch of negatives individually for small prints or digital uploads, or stitching multiple shots from one negative for the ultimate level of quality for large prints.

 

[Chan Tran]

How do you attach it to a digital camera? I'd like to try this with a Pentax K mount. I have a spare LPL 6700 enlarger head which I'm assuming would be a perfectly even backlit light source.

You have a DSLR with Pentax K mount? If so get a bellow for Pentax K mount. Like this one http://kmp.pentaxians.eu/wp-content/uploads/bojidar/misc/macro/auto_bellows_slide_copier.jpg.
You would also need a 39mm screw mount to Pentax K adapter like this one.
https://www.roxsen.com/index.php?route=product/product&product_id=1320
I looked at the fotodiox website but they don't have one.

 

[runswithsizzers]

Just curious if a high quality enlarging lens could be used with a DSLR, assuming you could somehow fit an enlarging lens onto a DSLR?

My setup uses a A Schneider-Kreuznach Componon-S 5.6 / 100mm Enlarger Lens adapted to Pentax bellows <image here> I think my total cost for the lens, bellows, and two adapters was around $150.

 

[Anon Ymous]

BTW, coinimaging.com has quite a bit of useful information about lenses that could be used for film scanning.

 

[GLS]

I rarely take it to that level, because of the amount of work involved, the huge file sizes, and the time it takes

Yes, it can be frustrating and time consuming to do multi-shot stitches, because of the issues I mentioned above. The results, however, speak for themselves. The following is an experiment using a 9-shot stitch of a 6x6 frame of Copex Rapid, which gave a 121 MP file (a 100% portion from the red box is shown):

The day a flatbed can do this is the day I eat my shoes.

 

[David A. Goldfarb]

DSLR scanning is what we used to think of as standard copy work, but now it’s digital and you can stitch for higher resolution. I use mainly enlarging lenses, but for multi-shot stitches of 35mm negs or slides, I’ll use a Canon FD 35/2.8 Macrophoto lens. I used to have an older Zeiss Luminar of the same focal length, but found the Canon was sharper. The latest version of the Luminar might be better than the Canon—I haven’t tried.

Vertical alignment of the camera and the copy work are important, as is film flatness, just like in standard film-based copying. I have an Omega enlarging level, but a laser alignment system is better. I usually use glass neg carriers.

I’ve found that horizontal alignment of the copy work is not important, because most stitching software is designed to correct misaligned panels, as long as you provide sufficient overlap.

Like some of the above posters mention, I often find a one or two frame image sufficient for most uses. Occasionally I’ll do a 9 or more panel stitch for a large print. If I use higher resolution, it’s usually for the purpose of cropping.

 

[Alan Edward Klein]

Yes, it can be frustrating and time consuming to do multi-shot stitches, because of the issues I mentioned above. The results, however, speak for themselves. The following is an experiment using a 9-shot stitch of a 6x6 frame of Copex Rapid, which gave a 121 MP file (a 100% portion from the red box is shown):

View attachment 232852

View attachment 232853

The day a flatbed can do this is the day I eat my shoes.

Curious how your stitching method compares with a drum scan?

 

[GLS]

Curious how your stitching method compares with a drum scan?

I have as yet not commissioned any drums scans. It's on my to-do list at some point though, as I'd also be very curious how it compares.