Let me be clear this is not a consumer or technical review, really it is just a bit of fun born out of an opportunity. A few weeks back week I met up with my friend Mark Heaver so that he could give me a large format 4×5″ negative that I had shot using his Chamonix large format camera. At the same time, he had booked the use of an Imacon 949 scanner at a nearby commercial lab.
I was pretty stoked at this double opportunity to see the results of my first large format photo and to use a legendary Imacon scanner to do it. We had booked the scanning suite for an hour and Mark suggested that there could be time to scan a couple of 6×6 frames too. Any of you who have read my previous blog posts will know by now I have a bit of a scanning obsession and the chance to compare my current set up to an Imacon was, of course, irresistible. So, at home later that evening I repeated the scan but this time with my Epson Perfection V800.
Scanners do two things. First, they secure a film negative or slide flat and second, scan/photograph it. This way an analogue photograph is digitised. Scanning hardware, as opposed to software, is designed to primarily do this. With professional drum scanners, the film is mounted to a spinning drum and rotated over a lens. Stable, accurate and fast, drum scanners are hugely expensive and increasingly rare. Very few photographers are lucky enough to own such a device. More often negatives are scanned with dedicated desktop machines and increasingly with DSLR cameras, copy-stand style. All these methods employ various frames and holders that are illuminated and transported through a device or mounted flat on a plane and scanned from below or photographed from above.
The second feature of any scanner is software. The digitised data from a scanner is processed through its compatible software. The accompanying software inverts the negative and displays an automatic starting image. This along with a powerful set of user-friendly tonal adjustment, colour correction and post-processing tools, such as sharpening and dust control make the final scan.
The aim of this experiment was not to look at the software’s performance with regards to colour, tone or contrast. This was partly because I’m no expert using Imacon’s FlexColor software and certainly did not have time to familiarise myself with it. But more than that, for me, any comparison of a scanner should start with the hardware. Scanning software is secondary to a sharp and detailed base scan. So I wanted to see was the difference in the hardware, the film holders capability to hold the film flat and the quality of the optics and resolution of the scanner. Essentially how sharp and detailed the scan would be.
You might be interested in...
The Imacon uses a flexible magnetic frame that sandwiches the negatives film border flat between two surfaces and then by arcing the holder into the machine it ensures flatness over its internal optics. I use a Betterscannng adjustable height film fame with anti-newton ring glass to secure and hold my negatives flat before being placed on the glass surface of my flatbed Epson Perfection V800 scanner.
I scanned the frame at 3200ppi setting both Imacon’s FlexColor and in the case of my Epson, SilverFast to their own Fuji Pro 400H profiles. Then with Adobe Photoshop, I colour corrected and adjusted contrast before finally cleaning up spots and dust by hand. This is what I got at:
Whilst far from a controlled or scientific test the difference between these two scanners is negligible. Given that in real-world scenarios where most photos are displayed electronically or printed at small sizes the base scan quality of both these machines, and I suspect most other machines and methods, is probably more than sufficient. It would seem to me, that it is the capacity of software such as EpsonScan, FlexColor, VueScan, SilverFast or Negative Lab Pro to invert and deliver good initial colour, together with their subsequent ability to make adjustments by using their specialist tools, that truly differentiates ‘scanners’ more than the hardware itself.
Thanks for reading,
Share your knowledge, story or project
The transfer of knowledge across the film photography community is the heart of EMULSIVE. You can add your support by contributing your thoughts, work, experiences and ideas to inspire the hundreds of thousands of people who read these pages each month. Check out the submission guide here.
If you like what you're reading you can also help this passion project by heading over to the EMULSIVE Patreon page and contributing as little as a dollar a month. There's also print and apparel over at Society 6, currently showcasing over two dozen t-shirt designs and over a dozen unique photographs available for purchase.
I have read many comments here. There is a sense of logic that might be applied – and that is not all photographers wanting to scan have the kinds of funds to support an X1, X5, 949 etc, and those that do then choose where they spend their money. If they felt an under AUD 1,000 Epson was the pinnacle of all scanning -then there’d be zero market for the virtual drum scanners. I have an old but super hasselblad 343. It scans 35 minutes at 30mb a year (almost) and I make a scan, and go for a cup of coffee. I dont use it commercially. I have 1.5m prints from 35mm Ektachrome and I absolutely love the quality it produces. That is not to say – a 1.5m enlargement is like an 8/10 contact. its not. I have used Epson too and do not like the method, but I think for those who want to use it – its fine. Someone made comments about the comparison with digi SLR or MF scanning and that is a critical question if res and dynamic range are the obsessions. The Hasselblad or other MF sensors, in particular the RGB individual scanning backs with single pixel offset ‘probably’ would produce results far superior than the scanners – which is why theyre used (frequently/often) in recording art/ static objects directly. Back to the subject – Id be very keen to see the Epson scan with a very wide radius low frequency kicker from photoshop in post – thios would substantially kick fine/local area contrast. AT the end of the day – if you have the money and want to spend it on a 949/X5 X1 its a personal choice. The images on the wall – well, if you can pick the differences – then one might argue youre looking at the wrong part of the image for the wrong kind of information. Stand back and enjoy the experience. Thats all that really matters I guess. Some of the greatest things arent perfect. Im going to have a 4 coffee session. 4 scans. I like that.
I would personally like to see a good comparison between the top flatbed scanners and DSLR scanning. For the DSLSs, I would also like to see what the difference between full frame and APS-C cameras too. I know, that i a lot! But I think there are a lot of film photographers who are struggling with what the difference is.
Here’s one which covers the scanning aspect. To the other, I’ll add something to the queue.
Thanks for an interesting article. It is unfortunate that the unprocessed scans are not shown. But that said, it is pretty clear that the 949 has pulled considerably more from the negative than the Epson scanner. Look at the 100% crops. There is a large amount of fine detail which is simply not there in the Epson file. The Epson file also looks much less sharp (see the glints in the eye). In addition, the texture of the skin is shown beautifully in the Imacon scan whereas in the Epson file the skin looks crude and unreal because the scanner is not capable to remain faithful to the information in the negative.
I also want to add that there are other ways to scan with an Imacon/Flextight when it comes to colour negative films. In my testing Flexcolor does not always deliver the best results and can be downright awful particularly if film profiles are used. In my opinion the best results are obtained when the file is opened and inverted in ColorPerfect to get a good starting point for further colour corrections elsewhere (I used Adobe Camera Raw which is excellent for this).
There is a lot on the internet about how flatbeds can compete with high-end scanners for 6×6 and larger but if one really looks at the details that is just not true. It is plainly evident in this test to. And it becomes even more clear if one needs scans for, say, larger prints. Then flatbeds simply fall away as viable alternatives. But for posting on the internet or on instagram or for not so large prints they are certainly sufficient.
I don’t mean to suggest that this article is not a worthwhile exercise. I think it is because it clearly shows that high-end scanners produce high-end scans. I’d happily buy an inexpensive flatbed if it could give me what my X1 does. But that’s not the case. And the difference becomes even greater at small film formats like 35mm.
With due respect to you, you must be an incredibly difficult person to please. “..crude and unreal..” is quite a castigating comment in comparing the two results. I see the differences, too, but to me they do indicate just how well the MUCH cheaper flatbed has performed in comparison. No, I wouldn’t have expected the Epson to beat, or even equal, the Imacon, but the results, as I see them, speak for themselves and I, for one, would echo an earlier comment “give me the Epson any day.”
The author set out, out of curiosity, to see how well his flatbed would perform set against the horrendously expensive Imacon when scanning 5×4. We now know, from the data spec, that the Imacon is dpi limited, depending upon format, and 5×4 is scanned at a moderate 2040. In comparison, I know that my Canon could scan 5×4 at a true resolution of just over 2,400, so it should in theory, and may well in practise, beat the Imacon. However, the file size would be enormous and there is more to mere dpi in the final quality of a scan.
I believe that the author has shown what he set out to do with a 5×4 film.
I do agree with you that with 35mm flatbed scanners leave a lot to be desired, even the better ones.
Hi Terry, nice to meet you. I always enjoy comments which begin with “with due respect” and then make a personal attack. Makes me smile.
Just to be clear, I have no reason to defend either of the two scanners in this comparison, let alone the author. I take this article for what it is, a fun comparison born out of an opportunity. I happen to use an X1 in addition to other scanners and only chimed in because I am familiar with the quality of the images that such scanners and Flexcolor produce.
It is not a castigating comment at all, it is the obvious truth judging by the photos that the author elected to publish. The Epson image is ‘crude’ in that it fails to reproduce fine image detail and it is ‘unreal’ because it does not make the skin look natural. This would be evident in any larger print of the file. Like I said, I would have wanted to see the unprocessed scans to reduce the effect of post-processing on fine details since that is what needs to be discussed in order to compare the scan quality.
Incidentally the image scanned is 6×6 btw, not 4×5. As stated by the author, the image was scanned at 3200dpi on the 949 which is a higher optical resolution than the V800’s. But the difference in scan quality is mainly due to the fact that the Epson is a consumer-grade scanner.
You may call the Imacon’s dpi limited, but it is absolutely sufficient for most applications, including very large prints from 120 film.
Like I said, I don’t suggest that this is not a worthwhile comparison. I think it is a fun comparison to make. Everyone wants as much of anything for the money, that’s human nature. I also don’t suggest that the Epson is not a worthwhile scanner. It is at that price point. But it is not a professional scanner. Whether the price of a professional scanner is “worth it” is simply a matter of personal opinion and preference.
Hello don’t mean to intrude a few months later…but i have been a professional photographer since 1985…so i have seen many upheavals in the field. In all honesty i have used imacon and epson to make images for the screen and for large prints…for the screen an image will never render what a apo-artar on fine grain film has seen…and for prints, paper and ink quality from the output printer are as important and finicky…ink tends to bleed ever so slightly even on the best papers. In my experience of both mediums ie: digital v/s silver halide the final output is what counts, is it a print for a gallery is it for a straight forward job on a tight budget or is it for viewing on a screen. And there, the price of the scanner will have an impact…i think given the price difference, epson does a pretty amazing job in comparison to an imacon…especially with available post prod software which everyone uses anyway…Of course nothing replaces the acutance of a well processed sheet of film matched to a beautifully calendered silver halide paper! But anyhow i think pretty amazing strides have been made with consumer grade scanners and the margins between imacon and epson are narrowing…Of course at the end of the day imacon is still a winner but not by far!
The flextight has a resolution that increases in an inverse relationship to film size. Hence 35mm is scanned at a true 6300dpi. 120 is scanned at a true 3200dpi. 4×5 is scanned at a true 2040dpi (all from the imacon data sheet. While the scanner does do 4×5, it is designed for 35mm and 120. Naturally a scan at 2040 dpi will be similar to a flatbed scan
in terms of resolution. Compare a 35mm or 120 negative to another scanner, let alone a flatbed, and it’s a very different story.
Interesting article. Give me 20 Epson V850 instead of the Imacon 😝
Trying this experiment now with my Epson Perfection V700. I don’t think I will ever order a drum sca… https://t.co/uO46rI1juV
This is a question of true optical resolution.
No matter what it says on the package.
The Epson V800 has an optical resolution of 2300dpi, everything above that is interpolated. And interpolation means loss of sharpness and detail. With medium format and larger, the loss is limited, but with 35mm images, details are lost irretrievably. I have reference scans from a drum scanner of own medium format negatives, which help me to sharpen my V800 scans in the best possible way.
The true optical resolution of flatbed scanners leaves something to be desired, despite claims by the manufacturers. Canon claims 4,800 optical dpi for the 9950F, with interpolation settings as high as 19,200 dpi. My measurements with Lasersoft’s 35mm slide of the USAF 1951 Resolution Target produced a best result of just over 2,400 dpi with a scanner setting of 3,200 dpi. Anything over the 3,200 setting produced no increase, but at the 4,800 setting there was a visual increase in apparent grain, but this wasn’t the result of any increase in measured resolution.
It is interesting to learn that the much later Epson has a true optical resolution close to that of my Canon. I wonder, does this mean we’re at the technical limit of flatbed scanners?
Thank you so much for this article. Very timely for me, as I am going in circles with my local printer regarding a drum scan and getting the image right for my print. I’m going to put that negative in my Epson V700 this weekend and see how I feel about the two versions of my image!
Cool, I use a flextight, this should be interesting.
Yes, with medium format films it is no problem for the low-cost scanners to compete against the high-end scanners. But what about 35mm? There is already a huge difference between flat bed and special film scanners of the same price range. Unfortunately I can’t use my film scanner under my new operating system.
I can confirm your comment. I have two Canon flatbed scanners that can accommodate up to 5×4. For 35mm scans I’d class them as adequate for monitor viewing or small prints, but for reasons I don’t technically know why, there is a quantum leap in scan quality with my MF and LF negatives and with a max. dpi setting of 3200. But then beyond this, file sizes become huge. Interestingly, I’ve never had an issue using Canon’s own negative carriers.
Really interesting article Tom. I use an Epson V850 with modified negative holders and scan 6×6 120 negatives with Silverfast software and I personally dont have a problem capturing the necessry details. It would be good to see a “How to” scan a negative article as I spent many hours trolling the internet for advise and experimenting to finally be comfortable with the images I was capturing. Thanks gain for the article….
Mark, some years ago when Imaging-Resource reviewed the Epson V600 they included a short course on scanning. It is still on the net and can be found here for anyone interested:
Thanks Tom, excellent article. The Epsom results look very good. When you use your Epsom with a non-Epsom film holder, how do you force it to use the high-res scanning lens? Mine will only use the best lens when using the dedicated Epsom holder. Any other holder causes it to switch to the less sharp general purpose lens. Maybe that is a limitation of Epsom’s own software, and you have got around that by using Silverfast?
Thanks Ray, In EpsonScan Professional mode insure the document type is set to: “Film (with Film Holder)” to use frames, this will set the scanner to use the correct lens that focuses above the glass. in SilverFast they recommend. Note on the Dual Lens System: For transparency mode, the Epson Perfection V800 features two different focus levels. Select scan mode “Transparency”, if you want to use a filmholder. In this case, the focus of the scanner optics is precisely adjusted to the film holder. Use scan mode “Wide Transparency” only, if you want to place transparent originals directly on the flatbed.