Film Photography

More 35mm color negative work from the CanoScan 9000F MkII and ScanGear

The advice some of you gave me in this post helped me get decent black-and-white scans from my Canon CanoScan 9000F Mark II and its bundled ScanGear software. I used the same advice to scan a little more color film.

I made these photos last fall with my Olympus XA2 on Agfa Vista 200. Roberts Camera in Indianapolis processed and scanned them. Their scans are 3130 pixels on the long side. I used ScanGear to scan them at 4800 dpi with all built-in image enhancement turned off, resulting in scans of between 6750 and 6800 pixels on the long side. I resized my scans to 1200 pixels long to upload them here.

I edited scans from both sources as best I could in Photoshop, including adding unsharp masking to the ScanGear scans.

My first test was of this shot of old US 52 and a great abandoned neon sign near my home. It shows considerable vignetting, which I believe is endemic to the camera. While I like the depth of blue in the sky, I don’t like how mottled it is. I tried various Photoshop settings and tools to smooth it out but wasn’t happy with any of the results. I wonder if the film profiles and multi-exposure scanning in Silverfast would resolve these challenges.

The Roberts scan captured more turquoise in a perfectly smooth sky. The Wrecks sign shows far better definition and detail. I suppose the Roberts scan might have a touch of green caste to it. Roberts also reduced the vignetting. I prefer the Roberts scan.


The CanoScan/ScanGear scan of this abandoned farm co-op building shows the same mottled deep blue sky, but plenty of great detail in the corrugated walls. This building is all that’s left of the onetime town of Traders Point, Indiana, by the way. See 1950s film footage of this town, including a brief look at this co-op building, here.

Here’s a crop of the image at 100%. It could be sharper, but it’s fully usable.

In the Roberts scan the colors aren’t as vibrant, and the sky is again more turquoise. In retrospect, I could have helped this photo by reducing exposure a little in Photoshop.


From here on out, the winner isn’t as clear between the Roberts and ScanGear scans. This ScanGear scan from downtown Indianapolis shows a scene that’s changed, as the Hard Rock Cafe has since closed and its signs are gone.

The Roberts scan looks like it got more exposure than my scan. My scan highlights the vignetting the XA2’s lens tends to deliver.

Down Maryland St.

These arches are around the corner from the previous scene. Here’s my scan.

Here’s the Roberts scan. Each has its charms; I can’t call one better than the other.


Still downtown in Indianapolis, I shot this outdoor cafe scene. The day was drizzly and chilly and so not ideal for outdoor dining.

Here’s the Roberts scan. I like my scan’s blue umbrella and the overall color temperature better.

Blue umbrella

Finally, here’s a forlorn building. My scan gives its gray painted brick a bit of a blue caste.

The Roberts scan is more of a straight gray. Like all of the Roberts scans, it got a touch more exposure. Either scan is good enough for my purposes, but I believe I slightly prefer my CanoScan/ScanGear scan.

Gray building

I believe I’ve figured out a good base 35mm scanning technique and can refine it from here. Perhaps I can get a little more sharpness, a little better color. I do have to solve that terrible mottling problem, though; the two scans with blue sky in them aren’t that great.

Next, I’ll try scanning some medium-format negatives with the CanoScan and ScanGear. This is perhaps the most important test, as my goal is to shoot my lovely TLRs and my simple box cameras more often, and process and scan the film myself.

Get more of my photography in your inbox or reader! Click here to subscribe.


23 thoughts on “More 35mm color negative work from the CanoScan 9000F MkII and ScanGear

  1. Dan Cluley says:

    I see your concern with the sky in the first two, but on the whole I prefer the color and or exposure on all of your home done scans over the commercial.

    I’m curious your process for color correcting from negatives. Does the scanner software do this, or is that part of your PS work. That has been the most difficulty I’ve had using my DSLR. Color from scanned slides is mostly just matching the white balance to the light source, but dealing with orange negatives has been a mixed bag of success and disaster.

    • Interesting, you don’t find the greater vignetting in my scans to be a detraction? I am 100% on the fence on this. But I agree with you that the color is better in my scans.

      The scanner software removes the orange mask. It’s witchcraft and magic; any color neg film I put through it comes out with the mask perfectly removed.

      • Dan Cluley says:

        For whatever reason, the vignetting doesn’t bother me. I see it if I’m looking, but then I see it in the lighter scans too.

        I think I’ve just spent too much time looking at Kodachrome slides taken with ’50s rangefinders, so your scans look right.

        If I was doing more with color negs, I would probably have to get a scanner, but fortunately most of what I want to do is slide copying.

    • P says:

      Hi Dan. So you’re digitizing your film with a DSLR, I take it? I’ve never done this, but if I did, and I was attempting to do so with color negatives, then the first thing I would try to do in order to eliminate the orange mask would be to scan a blank frame of whatever C-41 film stock I’m using (shot with a lens cap on, and developed normally), and then use the color data from that frame, which theoretically is just the orange mask, to subtract the orange mask from my actual photos. Then, adjust the levels/curves afterwards in a normal manner once the orange mask has been effectively eliminated. Of course you could also get this data from the film rebates or the blank space between two frames, but I think it would be safer and easier to just use a full blank frame and avoid the possibility of variation. You would probably have to do this for each different film stock you use as I imagine their film bases and orange masks are not identical. This could be done in a variety of ways using layers and masks in Photoshop or other similar photo editing software if that’s what you’re familiar with, or it could also be done with command line software like ImageMagick, which is free and extremely powerful. If you want to simplify things and don’t mind investing in additional software, then there’s a tool for Lightroom called Negative Lab Pro that you may want to look into. It’s primarily tailored to people “scanning” film with DSLRs, and even though I’ve never used it, I think it’s probably one of the better options out there for achieving good results using this method for people who don’t mind spending the money. I imagine it’s basically doing what I described above.

  2. P says:

    Nice work, Jim. It looks like your making quick work of figuring out your new scanner.

    Regarding the mottling, do you have any “grain smoothing” features turned on in-scanner for the scans? If so, disable them. They could possibly be the culprit. Another possibility is if your scanning as JPEG and then pushing the level/curve adjustments too far in post.

    I’m sure medium format negatives will be a lot easier to work with.

    • Grain smoothing is off, but as I looked I discovered I did have something called “image adjustment” set to “Photo” rather than to “None.” Next time I’ll make sure that setting is “None.”

      I am scanning to JPEG, but the mottling is straight out of the scanner. I’ll try TIFF next time. I shy away from TIFF just because of the large file sizes, but I guess large hard drives are cheap these days.

      • P says:

        Sounds like a plan. Good luck! Even if you only scan as 8-bpc TIFFs, you should still get much more flexible files to work with. 16-bpc would be better, but I know the file sizes can get unwieldy. At 8-bpc, your scans would probably only be 3-4 times larger than JPEGs. That’s pretty manageable, I think. Thanks for keeping us apprised of your scanning progress!

        • P says:

          Also, one more thing you might want to try is manually increasing the exposure in the scanning software. With fairly dense and roughly continuous tone areas of negatives, if not enough exposure is given in the scanning process, then noise will undoubtedly be exaggerated. And I think that’s what the mottling your dealing with is, digital noise and subsequent artifacting (it’s clearly not the actual film grain). Afterwards, exposure could be adjusted back to normal in post, hopefully without any clipping. Doing this would also probably help make the tonal gradation more natural and reduce the intensity of the vignetting. You’ll definitely want to scan as TIFFs to perform these actions though. It’s just a thought, in case other avenues fail to help. Again, good luck.

  3. I vote for your TIFF plan. If you save the original scan as a JPEG and then save that as a smaller JPEG for on line display, you are likely losing some sharpness in the end. I scan 35mm at 1600 dpi and medium format at 1200 and save the initial scan as a TIFF file. I make my editing alterations and then save that file as a TIFF with a new name. I then reduce the edited image to the size I want for on line display which is usually 700 pixels in height and save that image as the final JPEG.

    • Useful advice, Mike, thank you. I’ll try again with 35mm color, maybe with the same images, and scan as TIFF and see what happens. I will also turn off “Image Adjustment” in the scanner — I thought I’d turned off all of the in-scanner settings but missed this one. It appears to adjust brightness to the kind of source you’re scanning.

      Good point about losing info each time I resave a JPG. In this case, the mottling was present straight out of the scanner, though to be sure saving the scan certainly could not have helped that.

  4. Jim, I think (for me), what’s more important than the overall colours etc is the consistency. I wouldn’t want to be scanning a single roll of film (or multiple rolls of the same film, shot with the same camera) and the colours and look be all over the place. I guess I just like to feel like if I follow a certain process, I get a certain result, that I can more or less depend on. I’m not sure I really got that when I was scanning myself. How is this consistency aspect working out with your home scanning would you say?

    • So far so good, at least within each roll. I think the key to consistency in scanning is building your techniques and habits, which I’m starting to do with these experiments. I suppose you have to adjust those techniques and habits per film but I haven’t scanned so many film types yet to know for sure.

  5. -N- says:

    What brand of scanner is your Roberts shop using? That could be interesting. Different commercial scanners may have different color skews. I think the two big ones are different from one another what they produce. I vaguely remember reading something about this years ago. As far as vignetting, I use it selectively in post, usually to draw the eye to a specific area or to change how things look. My primary post programs are LR and On1. This is for both analog and digital. As an aside, I am interested to see how you resolve your mottled skies. Myself, I use a Pakon 135 for 35mm and a V600 for 120mm. I am too cheap to pay for scans and have a lot of hardware for my tiff saves!

    • The scans’ EXIF mentions “EZ Controller” which is Noritsu’s software. I don’t know what model Noritsu though. I think all the labs I use have Noritsus. I wonder if Robert’s software detected and tried to remove the vignetting, which resulted in the appearance of slight overexposure.

      You have a Pakon! I’m sort of jealous. Only sort of because I know you need a Windows XP box or emulator to be able to run it and that’s a hassle.

      I should try scanning these same images again as TIFF and see what happens.

      • -N- says:

        Don’t know what your budget is, but you can find them for sale on eBay as well as on the FB Pakon group, if that helps.

        • P says:

          The problem with trying to buy a Pakon today is that people are selling them for truly absurd prices. A few years ago they were reasonable when they were available refurbished with a guarantee for about $250. But the last time I looked I believe they were selling for nearly $1,000, without any guarantees. That’s just ridiculous for what it is. And for all I know the price gouging has gotten even worse. I’ve said it before, but I’ll say it again; I am really getting tired of all the exploitation of amateur film photographers. It is nothing shy of rampant, and it’s just wrong. I know it’s hurting the community and the industry.

        • -N- says:

          AAA Imaging in California periodically has them in stock, but as you said, at high prices. I bought mine for $275 or something, and am happy with it. I use a V600 for 120 but would like to find something better, but not interested in the prices of used good ones. If there is something I really want printed, I will have it professionally scanned . . . but I seldom do.

        • P says:

          You’re lucky you were able to buy one for $275. Honestly though, even that is way overpriced for what the Pakon actually is. There really isn’t anything special about it as a piece of hardware that warrants a high price tag, even moderately so. It’s actually a very simple device, built with very common (and extremely cheap to manufacture) components. There just really isn’t much to it at all.

        • P says:

          -N-, I wish I could. Trust me, if I had the ability to build one myself, I would. Unfortunately, I am not a massive company that has the means to set up manufacturing and procure electronics components in bulk for cheap. I’m still holding out hope that one of the film manufacturers will recognize the need for a Pakon-like scanner for the amateur film community, and will start manufacturing such a device and selling it at an affordable and fair price. For them to do so wouldn’t be difficult, but for an individual to do so on their own is a different story.

  6. Carl Nygren says:

    To adjust your blue sky problems you might try the blue and or aqua luminance sliders in your editing program. I do most of my editing in Lightroom and find these sliders to be very useful!

    • I thought about that after I’d already written this post. I use Photoshop and it has those sliders in the RAW editor. I may try that next time.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.