I knew I could cut costs even more by scanning the slides myself, but could I get scans as sharp and colorful as Fulltone’s? Also, Fulltone’s scans are smallish at 1024 pixels square. I can easily get more pixels from my Canon CanoScan 9000F Mark II.
I tried it on a couple frames. I thought I’d show you my scans and the Fulltone scans to see what you think. My scans are at least 5100 pixels square — I select each frame by hand in VueScan, so the pixel dimensions vary slightly frame to frame. I shrank them to 1200×1200 for this comparison. WordPress shrinks them further to fit the blog template. I edited them all in Photoshop to my liking — nothing too invasive, mostly stuff like color temperature and exposure.
My scans are first, Fulltone’s are second.
Fulltone managed to bring out far better shadow detail than I could get from the CanoScan, VueScan, and Photoshop flow I use. Their scans look slightly sharper than mine.
But the Fulltone scans have a green cast that I couldn’t entirely erase, a cast that isn’t present on the slide. My scanner captured color that looks a little truer to the actual slide. Also, I was able to capture more of the frame than Fulltone did.
I don’t think there’s a clear winner here. Both Fulltone’s and my scans are fine. It’s a roll-by-roll judgment call whether saving $5 in scanning charges is worth the couple hours I’d spend scanning the roll myself. But when I want scans with large pixel dimensions, it’s very good to know that my existing scanning setup produces good results.
I thought it was a shame I hadn’t shot my Nikon F2AS in a long time, so I put some film through it recently. The meter led me to shutter speeds that seemed slow for the full-sun conditions, out of line with Sunny 16.
I shot four subjects twice, once using the F2’s meter and once using the my phone’s light meter app. The app consistently had me expose two additional stops!
I shot Ilford FP4 Plus through my 35-105mm f/3.5-4.5 Zoom Nikkor lens (which I like less and less the more I use it). I developed the film in Rodinal 1+50 and scanned the negatives on my Canon CanoScan 9000F Mark II using VueScan. I brought the scans into Photoshop where I unsharp masked them all, corrected perspective if that was necessary, and on one shot toned down the highlights, but otherwise left them alone.
There are so many possibilities in any scene, from how you expose it to how you develop it to what you do with the negative in printing or scanning and post-processing. These pairs show it well. The F2 metered shot is first in each pair. In this first pair, I like the second shot more for its better definition in the houses, and the more silvery reflection in the pond.
In this pair, I prefer the second shot again for its rich, smooth tone in the tennis court surface and the better definition in the houses.
In this pair, I like the first shot better for its slightly better shadow detail. The first photo is the one where I toned the highlights down slightly in Photoshop. The path was a little washed out in the original scan.
In this pair, I like the first shot better for its slightly better shadow detail and its better definition in the sky.
What do you see in these photos? In each pair, which do you like better?
I think to some extent what we’re seeing here is the good exposure latitude of FP4 Plus — these are all technically decent photographs. Also, what we all like in a photograph is subjective.
After I finished this roll I checked my F2’s meter under a bunch of lighting conditions and couldn’t reproduce the odd meter readings I was getting. Soon I’ll mount a lens I know and like better, probably my 35mm f/2.8, and shoot this F2 again to validate the meter’s functioning.
Lab scans of 35mm color negatives are miracles. Any lab I routinely use reliably sends me crackerjack digital images.
Getting usable scans from my CanoScan 9000F Mark II via its ScanGear software, on the other hand, is a lot of work involving a number of subjective choices in scanning and post-processing.
I used to think that the colors I got back from the lab were the film’s true colors. I see now how much of that is in the scanner settings, and that I don’t actually know how any film I typically use renders color.
The improvements I made this time were to scan to lossless TIFF files, and to turn off ScanGear’s Image Adjustment setting (which I had overlooked when turning off all the other image-enhancement settings). It helped? I think?
Here’s my scan of a photo I made on Kodak Gold 200 with my Olympus OM-1 and a 50mm f/3.5 Zuiko Auto Macro lens. There’s a little of that mottling in the blue sky that I keep trying to prevent. But it’s not as bad as in previous scans.
Roberts Camera scanned this film when I had them process it. It’s a touch brighter than my scan. The sky has a slight turquoise tint and lacks any mottling. Otherwise, either scan is fine.
Here’s my scan of a butterfly pausing over this flower. Notice how purple the flowers in the background are.
Roberts made those same flowers quite pink, but brought out the detail lurking in the butterfly’s wings.
I also tried scanning some Kodak Ektar 100 I shot in my Pentax Spotmatic F with a 35mm f/3.5 SMC Takumar lens.
Here’s Robert’s Camera’s scan. They got richer colors than I did, although I’d say the sky in mine looks more realistic. The green tint on the right edge of my scan is clearly an artifact of the negative that Robert’s somehow edited out.
I walked over to the building to make this close shot. My scan:
Roberts Camera’s scan got a richer red, but my scan offers better highlight detail.
It was so much easier when I accepted whatever color I got from my lab scans, as if they were the final word on film and lens. Now I’m suspicious of every scan, because of all the choices it represents. Is it possible that the only way to truly know what colors are in a negative is to make a darkroom print?
This, by the way, is the last in this series of experiments. I’ve learned what I need to. I get good enough black-and-white scans now to start processing and scanning black-and-white film, which was my goal. Now that I work Downtown in Indianapolis, eight blocks from Roberts Camera and their C41 lab, I’m likely to have them process and scan my 35mm color negative film. They charge just $10.
I made one more experiment scanning negatives on my CanoScan 9000F Mk II and its bundled ScanGear software. This time I tried scanning black-and-white medium-format negatives.
I hope to start processing my own black-and-white film, especially in medium format, this year. I don’t shoot as much medium format as I’d like because processing and scanning costs about $17. That’s a buck and a half to two bucks per frame! Processing and scanning my own will manage medium format’s costs better.
I went back to 2016 to find some images I made with my Yashica TLRs, a Yashica-12 and a Yashica-D. These cameras have wonderful lenses that make the most of whatever film I put behind them. Here is a scan I made of a scene on the square in Lebanon, Indiana, on Kodak Tri-X.
Here’s a crop of my scan at 100%. I scanned at 2400 dpi, by the way, and applied unsharp masking and other tweaks in Photoshop until the image was to my liking. That’s some pretty good detail right there.
Here’s the scan Old School Photo Lab delivered, after I tweaked it in Photoshop to my liking. Both my scan and Old School’s scan are crops of the original image to the interesting part of the scene. My scans are about 5200 pixels square, give or take, while Old School’s are slightly off square at 4832×4760 pixels.
Here’s my scan of the Boone County courthouse in Lebanon’s square.
And here’s Old School Photo Lab’s scan. Either scan is acceptable. I like the tonality in my scan a little better as it feels more realistic to me. The Old School scan looks to be a bit sharper.
The next two images are from my Yashica-D on Fujifilm Neopan 100 Acros. I could have done a better job of cleaning minor dust marks off my scan, which is below. It’s otherwise a perfectly usable scan.
Here’s Old School Photo Lab’s scan. At blog sizes, they’re hard to tell apart. Both images are crops of the original frame, by the way.
These two images are from the far-northwest corner of Indianapolis, which is quite rural. Here’s my scan of a cemetery that lies along the road above.
Old School Photo Lab’s scan appears sharper — compare the grass in both scans. But either scan is eminently usable for my purposes.
I am pleased with my scans. I would use them for any of my usual purposes.
These experiments, and your comments on them, have taught me some key techniques. First, thanks to your advice I’ve turned off all the built-in image improvements in ScanGear and scan at as close to 4800 dpi as I can. Second, I’ve learned enough about the Amount, Radius, and Threshold settings in Photoshop’s Unsharp Mask filter to sharpen my images acceptably.
I see I’ve still much to learn about how to look at a photograph and see its details. In these experiments I’ve studied my scans in far more detail than I’ve ever studied a photograph, and compared them in depth to the lab scans, and thought about what I like in a scan. I realize I need to study far more photographs to learn how to see their details and decide what I like.
I also now realize just how much the quality of a lab scan might have affected my views of various cameras, lenses, and films, and how excellent scans might have enabled the praises I heaped on particular gear or films.
I feel like a man who thought he’d climbed a mountain, only to find that he had scaled but a foothill to see the real mountain emerge from the mist.
I’m experimenting with scanning medium-format color negatives in my CanoScan 9000F.
I’d shoot more medium format if it weren’t so expensive per frame to get scans. Every lab I use charges about the same to process and scan both medium format and 35mm, around $17 shipped. A roll of 35mm yields 24 or 26 images, while a roll of 120 or 620 yields only eight or 12. If I can get credible scans from the CanoScan without too much fuss it would cut about $5 out of that equation. I might shoot my TLRs, folders, and boxes more often.
I first scanned some Kodak Ektar 100 negatives I shot last year in my Agfa Clack. (Ektar is my go-to medium-format color film.) Old School Photo Lab processed and scanned the film.
Here’s a photo from that roll, scanned through the CanoScan and ScanGear. I scanned at 1200 dpi, the maximum ScanGear allowed to avoid enormous file sizes. This resulted in images 3968 pixels long. I left all image enhancements off in ScanGear. I applied unsharp masking and other enhancements in Photoshop. I shrunk the scans to 1200 pixels long to upload them to the blog.
Here’s a crop of this image at 100%. The Clack is a box camera with a simple lens that’s acceptably, but not exceptionally, sharp in the middle. This is a pretty reasonable result.
Here’s Old School Photo Lab’s scan. It’s 3569 pixels on the long side. I like both scans equally.
Here’s another scan from this roll using the CanoScan and ScanGear.
In this case I like the Old School Photo Lab scan better, as its colors look more true to life. I did the best I could in Photoshop to get better colors from my scan but they just weren’t there. Either scan is acceptable for my usual bloggy purposes.
Next I dug out some Kodak Ektar 100 negatives I shot in 2017 with my Yashica-D and a closeup lens attachment. Old School Photo Lab processed and scanned the images.
ScanGear let me scan at 2400 dpi but no larger to avoid extremely large file sizes. This yielded images of about 5200 pixels square. Again I left all image enhancements off in ScanGear and used Photoshop to apply unsharp masking and other enhancements. I shrunk the scans to 1200 pixels square to upload them to the blog. Here’s my favorite photo from this roll.
Because this scan is so large, a crop from 100% shows only a small portion of the image. But as you can see it’s reasonably sharp and detailed.
The Old School Photo Lab scans are about 2400 pixels square. My scan offers more contrast and a lovely purple in the sky, but the OSPL scan offers a more limited and nuanced color palette.
Here’s another CanoScan/ScanGear scan from this roll.
The Old School Photo Lab scan is flatter and warmer. Both scans have their charms.
Finally, a CanoScan/ScanGear scan of this lily. I made all of these shots in my old house’s front garden, which I sorely miss.
The Old School Photo Lab scan is again warmer. It’s been a while since I’ve seen these lilies but I believe my scan’s purple is more true to life.
Unsurprisingly, the CanoScan and ScanGear do credible work making scans of color medium-format negatives. It was far, far easier to get good enough scans from these negatives than with any of the color 35mm negatives I’ve scanned. When it comes to negatives, there’s no substitute for size.
Scanning isn’t a joy any way you look at it. The act of scanning mostly involves waiting, which isn’t terrible. The real work begins after the scanner produces the files. The worst of it is removing dust marks. Even after gently wiping these negatives with a cloth designed for the purpose, a lot of dust remained on them. It was tedious to remove all of the marks in Photoshop.
I scanned some black-and-white negatives recently with my CanoScan 9000F Mark II and the ScanGear software that came with it, and I want to share the results.
I took much of the advice some of you gave me in my last CanoScan post. Namely, I scanned at 4800 dpi and turned off all of the image enhancements, including unsharp masking and dust/scratch reduction, that ScanGear offers.
My scans were still mighty soft, but what I learned from you is that this is to be expected, and it’s what unsharp masking is for. So I looked up some online information about how to use Photoshop’s unsharp mask tool and fiddled with the settings until I liked the results.
This is the scan I made that I like the most.
Here’s the scan Fulltone Photo made, after I Photoshopped it to my liking. Both scans have their positive qualities. I like the great detail the Fulltone scan shows in the brick foundation of the log cabin. My scan looks good to me and I would happily use it for any of my usual purposes.
Let’s pixel peep for a minute. At 4800 dpi, my scans turned out to be about 6800 pixels on the long edge. There’s minor variability among them in length and width because ScanGear determines each image’s edges individually. The Fulltone Photo scans are all 6774 pixels long. So these are comparable scans. Here’s a detail from my scan of the above image at 100%.
Here’s about the same square from the Fulltone scan at 100%. I’m straining at the seams of my experience here, but at 100% the Fulltone scan looks more usable to me despite its enhanced grain.
But at blog sizes, my CanoScan/ScanGear scans are great.
The Fulltone Photo scan is below. Both scans look wonderful to me.
I made 1200-pixel-long copies to upload here. 1200 pixels is big enough for every blog purpose I have.
Again, my CanoScan and ScanGear scans are, at blog size, in the same league as the Fulltone scans.
One more scan fro the CanoScan and ScanGear.
In this case, I prefer the Fulltone scan. As you can see, my scanner got some ghosting from the sprocket holes. Also, in my scan the barn is softer; its roof slats aren’t as defined as in the Fulltone scan.
I made these photos on Arista EDU 200 with my Nikon FA and 35-70mm Zoom Nikkor, by the way.
I am getting somewhere with the CanoScan and ScanGear. Thank you for your kind and excellent suggestions.
In this same scanning session I scanned more 35mm color negative scans, also at 4800 dpi with all image enhancement turned off. I’ll share results in an upcoming post, but I got mixed results.