Consumer flatbed scanners like the Canon 9000F are not going to give you good results for 35mm. Effective resolution for this scanner is only 1700 ppi, so scanning at 9600 ppi is just adding a lot of bloat with no more information. You’ll get much better results with using a DSLR or mirrorless with a macro lens. Test report Canon CanoScan 9000F flat bed scanner transparency unit: Description of the device, performance data, 35-mm image, medium format, large format, fotos
Hi,
Thank you for your replies. I scan the whole film with 200 dpi and jpg and only the good ones in high resolution. So diskspace is not an issue but if there is not much benefit in using 9600 dpi over 1700 I’ll safe a lot of time as well. So I’ll give it a try. I didn’t try to process them yet but my x220 could have some difficulties with the huge files : D. Thinkin about buying a new laptop but it’s the best one i ever had. Had a mac book pro for a while but sold it so i could use my old one again.
The Scanner is just borrowed from a friend but I thought of buying one for negatives. So, do you have advice for a good 35 mm Scanner with vuescan? Are there also some capable of scanning medium Format as well?
Greets
Hm, this is a huge topic … My thoughts are:
- Stay away from cheapish film “scanners” that are using a bad camera for digitizing. I used such a ~ $50 device to get an initial overview before doing “real” scans, but only because I got it used from an uncle.
- Upper class flat-bed scanners are reasonable for medium or large format scans, if you don’t want to get the entire resolution of the film. Also, they can include framing/sprocket holes, which may be a reason going this route. However, I am talking about scanners such as the Epson v850 pro, or other scanners in the > $500 range. But for optimal results, it may be necessary to add special film adapters that raise the film to the ideal focal plane of the scanner, and you can also read stories about wet mounting.
- Cheaper film scanners such as some from reflecta/pacific image (e.g. the one I use, crystalscan 7200) are in the range between $200 and $500 and they are doing the job. Depending on the actual model, you are a bit limited by the features (no autofocus/focus setting, no multi pass with different brightness, manual film transport, etc.), but the most important features are available and functional (e.g. infrared scan). The results are great compared to flatbed scans, but of course there are pricier options that deliver more quality.
- There are a bit pricier options from the same manufacturers for special purposes, such as automatic slide magazine scanners. I think the same as above holds for these.
- I don’t exactly know about the plustek brand, but I know some people here have such a device. These have a reasonable price and are looking very interesting. Post-pandemic, I really want to ask them about a comparison session, scanning the same negatives and compare the results.
- The best quality you may get with one of the coolscan scanners from Nikon, however, only the latest generations have USB and these are extremely pricey as they are no longer built for a while already. Expect > $1000 for a 35 mm scanner (coolscan V, coolscan 5000), and $2500 to $5000 for a medium format scanner (coolscan 9000).
- The Pacon f-135 (plus) is a very special device as it “only” scans uncut 35 mm film, it’s typically < $1000, but offers very limited resolution and needs special software.
- Drum scans are another option, but the hardware is extremely pricey and also requires special software. Typically, the film is wet mounted, which reduces the impact of scratches a bit (but I personally think infrared scanning and automated inpainting gives even better results and more control). Resulutions and dynamic range are insane, as the cost
. - Hasselblad’s flextight scanners are another pricey option. Due to special optics, they approach drum scan quality without wet mounting, with reduced impact of dust, but they offer no infrared scanning. I am not sure about software choices here.
- Digitizing with a camera and a light source is the final option that comes to my mind. Big drawback is missing infrared scanning, and extreme dependency on the light source (here on pixls.us it was discussed that less-white spectrum (more separated RGB colours) may give better results), lens, camera, etc. Results can be great anyway, but or me personally, it does not feel right for 35 mm, but feels reasonable for pocketfilm and medium/large format. But that’s my weird brain …
Hope that helps.
I’ve had several slide scanners and flatbed scanners over the years, using Vuescan Pro. The last ones I owned were an Epson 4990 and a Minolta Dimage Multi Scan Pro. The Epson did a credible job on medium format, but not on 35mm. The Minolta was an excellent scanner. When I started experimenting with using a DSLR and macro lens and carefully compared the results to the Minolta I sold the Minolta.
Part of the advantage of DSLR or mirrorless camera scanning is the software. Shooting raw and processing with RawTherapee of darktable gives you so much more control than Vuescan. As well, you can digitize any film size as you’re not limited to the scanner design. You should be able to pick up a good used Nikon D800 or D810 plus macro lens for under $1000 US.
Thank you @chris for your extensive reply. I didn’t want to spend too much money on scanners. But I had a look at the plustek-scanners and I think these or similar ones could be interesting and in my budget. I almost forgot about the option with macro-lens und camera. I already have a XT-3 and if I buy the 80mm macro I would have a macro and a film-scanner-option. So this could be nice as well. Propably I’m gonna rent this lens and test it. Imagine it to be a bit more complicated to get it flat on the light-plate. But you don’t need an extra-scanner. Thank you @troodon.
The best way to align the camera to the film is with a mirror. I covered a bit about camera scanning in this thread from 2019. I’ve made a few modifications since then, including avoiding having to use a Manfrotto Super Clamp. DIY copy stand for DSLR scanning
I wonder what is stopping you from doing a “raw” scan in vuescan (scanning to tiff in maximum quality, but not using the “raw” flag, but that’s another story) and doing the post processing in the software of your choice? At least that’s what I do and it works pretty well .
Edit: I checked and I am currently using the raw flag without any issues.
[/quote]
I wonder what is stopping you from doing a “raw” scan in vuescan (scanning to tiff in maximum quality, but not using the “raw” flag, but that’s another story) and doing the post processing in the software of your choice?
[/quote]
I used to do that, and still have a pro licence of Vuescan. When using Vuescan I’d scan to raw (which is apparently not really a raw file like raw files from a camera sensor), and save as 16-bit tiff to work on further. If the original was a colour transparency I’d also add in the infrared channel to the raw so I could play with different levels of infrared cleaning, but that of course didn’t apply to B&W film.
The reason I switched to camera scanning and RawTherapee is for greater quality and control than I can get with Vuescan.
But this way you loose the ability to remove defects by means of the IR-channel.
Hermann-Josef
Yes that’s true. Without IR cleaning, it means I take more care to physically clean the film before digitizing. It takes me 15-20 seconds to press on 3M Magic Tape to both sides of a 35mm slide which lifts off almost all embedded dirt, but it’s well worth it. It also avoids the slight softening of the image from IR cleaning and issues that IR causes on fine detail such as faces. After I’ve digitized an image I do the spotting of an remaining dirt and scratches in two stages. First is just for screen resolution, which can be done quickly. Then, if I need to print an image I’ll do a further cleanup at a resolution suitable for the print.
And, of course, IR cleaning doesn’t work for silver-based B&W films.
How does this Raw-scan work in Vuescan. I Scanner all in TIFF without the raw checkbox. Is there a big benefit to use it? Aparently I thought I’m done but now I think about rescanning in raw?!
Raw scanning in Vuescan allows you to “re-scan” from the digital raw file using different settings. It’s much faster than scanning the original film again, and allows you to experiment with different settings for things such as colour and IR cleaning. But of course you can’t use IR cleaning on B&W silver-based films.
BTW, Vuescan will also process raw files from a camera. I did some tests using DSLR scans of colour negatives (Nikon .nef) and found Vuescan did a credible job converting the colour negatives. The quality varied, but it was better than I expected. I still prefer using RawTherapee for this, however.
Ah thanks. I’ll try that.
Do you mean the vuescan implementation? Because if you are saving the IR channel, you can use a much better algorithm than what vuescan provides.
I am wondering for a while already if it would work doing an IR channel when photographing the negative with a regular camera. The questions are:
- Is e.g. a flash capable of emitting enough light in the relevant IR spectrum, or what light source that is easy to get would work?
- What kind of IR filter could be used (which spectrum is needed)?
- Would an IR filter be best placed between light source and film, or between film and camera?
- Is the different optical behaviour of the lens at the different frequency an issue? I guess a slight blur would be acceptable for the IR channel.
- Is enough light received by the camera in this particular spectrum, given a reasonable light source? The requirements on dynamic range are probably much more relaxed for the IR image than for the visible color image …
At least my cellphone’s camera is very sensitive to the IR light of a TV remote, but figuring out good parameters for the setup described above may require a bit more effort …
I remember that there were some issues with the checkbox when I researched that a couple of years ago, therefore I only scan to tiff and leave the checkbox unchecked. However, I do not remember any more what the issue was …
In theory I suppose it’s possible to have an IR channel on a scan from a camera, but if I’m not mistaken you’d need it on both sides of the negative.
There have been a few historical B&W negatives I’ve worked on where there were a lot of small dirt specs and small holes in the emulsion. I’ve had some success using G’MIC, Repair, remove hot pixels. It does cause some softening so I really only found it useful for large, even areas such as sky. I used 2 layers, the top without the G’MIC and the bottom with it applied. Then I used layer masks to brush away the sky in the top layer. But this is really only useful when there’s a sky with hundreds of these small specs.
Here’s an example of what 3M Magic tape can do for small dirt specs. I scanned this for a client from a 645 transparency. On the left was just blowing off the dust with a bulb blower and the right was after using tape.
Not sure if I understand this correctly. I guess the scanner only does one side illumination, and scanning from the other side. I don’t think it has two ccds.
I did a comparison of different inpainting algorithms in this post:
The whole technique I used is described a couple of posts before.
As only the missing parts of the image are inpainted, there is no overall reduction in sharpness. Furthermore, there’s at least one inpainting method for which I would attest that the result is not distinguishable from regular image content. Fair enough, my example does not show scratches at areas such as faces where the content is more crucial, but from my experience with old family photographs, the results are really excellent and the manual effort is close to zero, only some initial mask adjustment (which is necessary only one time per film stock).
That’s really impressive. However,
- I have mainly to deal with scratches, not so much with dust (k, dust as well …), as some of the old family photos are not in a very good condition,
- given the already weak condition, I don’t want to stick a tape onto the negatives, and
- probably it’s difficult to get the same tape you have been using where I live. There is some tape called “magic tape” from 3M sold under the “scotch“ brand, but even if it looks the same, it is highly likely that it is produced locally and maybe different from what is elsewhere sold with that name. At least the description says “highly adhesive” such that I wonder if it will rip off the emulsion layer from the carrier layer of the film … Picture here.
Still interested to learn more .