Best Scanner for Medium/Large Format Film?

[denniswms]

<p>

 

<p>I have used the <em>dreaded</em> Epson scanners for a decade- everything on my web site as example- and none of my clients from corporate to individuals have ever registered any dissatisfaction-and I keep working. </p>

 

</p>

[Ed_Ingold]

<p>The Epson is not <em>dreaded</em>, just a distant second place to the Nikon scanner. From my perspective, it is equivalent to getting 35mm performance from medium format film.</p>

<p>We have all used flatbed scanners for film. Sometimes good enough is, well, good enough. However, if you or have never used or seen the results from anything better, you don't know what you're missing.</p>

[q.g._de_bakker]

<p>What Edward said.</p>

 

[philip_wilson]

<p>While not wishing to get into digital vs film you can get better results with large MF film cameras and a good scanner than digital. However it takes a lot more time and effort and most audiences will not see it. If I compare my Fuji GX680 images scanned on a Nikon 9000 to the best images I can get form my Canon 5DII and the best Canon glass - the scanned film has more detail. I think the issue is that the 35mm glass is getting pushed very hard with the 21MP sensor - this is confirmed by my 7D where the resolution falls a lot as the pixel density of the 7D sensor is much higher and thus resolves well beyond the capability of the lenses (you can see this in the DXOmark and similar lens+sensor test results).<br>

For medium format the Nikon 9000 is a great investment if you do not want to change system and shoot MF only as a hobby / cash source on the side. If you shoot MF professionally get a digital MF system. If you what to stay with LF than you are probably best sending out for scan or working with a wet darkroom - I still do my B&W the wet way and used darkroom equipment (even big enlargers like the Omega XL5) is very cheap on eBay etc...<br>

I owned an Epson scanner (700) and found it was very difficult to get a really good scan - the Nikon 9000 makes a great scan relatively easy although it is still very time consuming.<br>

If you want to spend $20K then go with a digital back solution - I think you will regret spending $20K on a scanner. At $50 for a top quality scan you would need to do 400 to pay for the scanner - assuming you get two good images (you need scanned) from each 120 film this equates to 200 rolls of 120 - another $1000 in film costs and $1000 to $2000 in processing. Finally a top scan of a larger MF image is anywhere from 300 - 600 MB (my Fuji GX680 16 bit 4000dpi Tiff files are 500 - 600 MB each) so you need lots of processing power and storage to handle them.<br>

In my own experience 645 film images scanned on a Nikon 9000 are about the same quality as my 5DII images although I personally prefer the Mamiya 645 images.</p>

[rob_piontek]

<p>I find it hard to believe that the V750 is not good enough for nearly everyone's 4X5 needs. At 2000 DPI you should get a pretty nice 80MP image. Even for MF are you actually making prints and seeing the limitations of the scan - or scanning at 6400 DPI, looking at them on the screen at 100%, and wondering what's wrong? This is a serious question as I'm seriously considering a V700 in the near future.</p>

[q.g._de_bakker]

<p>The thing with 4x5" is that the only other (and yes: better) options for scanning the stuff are very expensive. So for most people a flatbed, quite simply, <em>must</em> do.</p>

<p>As Edward said earlier, not all pixels are created equal. The quality of a scan, even at the same resolution, is considerably better using one of the dedicated (and more expensive) film scanners.<br />But that does not automatically mean a flatbed cannot be good enough. it depends on what you need the scan for and/or what will be good enough for you.<br />And it certainly does not mean that flatbed scanners are no good at all. On the contrary!</p>

[philip_wilson]

<p>Rob - my personal problem with the Epson 700 was that it took a lot of time to get a result I was happy with. Setting up and scanning with the Nikon when using 35mm or 120 film is a fairly quick and easy process (for the 120 film you need the glass holder which is a $300 extra). With the Epson I found that it took lots of failed scans and that I had to get into a wet scanning process with after market holders to get an image I was happy with. this is a personal view but I found that I can get the results I want with a Nikon 9000 very quickly. For someone who has time the Epson 700 / 750 is a great bargain - since I am perpetually short of time paying more for the Nikon 9000 is not something I regret.<br>

With large format the Epson is the only real alternative to commercial scanning. The bigger negative should not cause the same difficulties as the 35mm and MF films create.</p>

[f_ph1]

<blockquote>

<p>I find it hard to believe that the V750 is not good enough for nearly everyone's 4x5 needs.</p>

</blockquote>

<p>You find it hard to believe because it simply is not true. There are many who simply believe the more money you throw at a problem, the better and more completely it is solved. Otherwise, I am sure the local photographer I talked with recently would have to adjust his $1000 a pop fee for corporate architectural photography gigs. Then again, he was only using a 4990 last time I talked with him. The enlargements from 4x5 looked pretty good to me.</p>

<p>I recently viewed some images first hand (i.e., not on the web) made with a Rollei 6006 and 16 mp digital back. The images were detailed, very nice actually. But they were clearly inferior to what I and others have been able to yield using 120 film, good lenses, and Epson scanners. I truly believe that most people who drop upwards of 10k on a piece of photo equipment want it to be superior so badly that they convince themselves of the same. You can use whatever charts or graphs you want; my eyesight is slightly better than 20-20, and I know what I see. Is the Nikon 9000 a better scanner for MF than the 4990/v700/v750? Yes. Significantly better...? Not really. And yes, I've owned not one, but three Nikon 9000 scanners in the past. Got some really good Zeiss glass with some of the money I got when I sold it...</p>

[polka]

<p>My concept of "good glass" is "of wine"</p>

<p>Paul -> did you kwow I was french ?</p>

[DWScott]

<p>MicroTek still make and sell film scanners.<br>

<a href="http://support.microtek.com/products.html">http://support.microtek.com/products.html</a><br>

I would consider them to be one step above the Epson, because they allow you to scan film glasslessly (is that a word?) Scanning glassless renders moot the question of wet mount or oil mounting the negative on glass in order to reduce the number of glass/air interface points. Of course, it is potentially not as flat.<br>

I would be curious to hear if anyone has compared the newer Epson V7xx scanners to the Microtek units. I have a Microtel Artixscan 1800F which I like, but haven't compared it directly to the newer Epsons.</p>

 

[AlanKlein]

<p>Just curious. I only scan MF 120 and 35mm. I have a V600. Would I really find any advantage of going to a V700 or V750? Thanks. Alan</p>

[philip_wilson]

<p>F Ph - as someone who has owned an Epson 700 and has two Nikons (a 5000 and a 9000) I find that the difference between the Epson and the Canon is not truly the quality (at least with 6x7 or 6x8 - although with 6x4.5 I could see a clear Nikon advantage). The big difference for me is the time it takes to get the result. Once you have the Nikon glass holder the process always took me between 1/2 and 1/3 of the time with the Nikon. This for me makes it worth 3x the price as I can earn the difference in the price very quickly.<br>

I am not sure which Rollei digital back you used but in my experience the H3Ds and newer PhaseOne backs produce results that are better than my Fuji Gx680 scanned on a Nikon 9000. They are not a lot better but considering the lenses are all Fuji and the GX680 has almost 2.5 times the image area the H3 is very impressive.<br>

It may be that the back on the Rollei 6006 you used was based on the old Kodak 37mm x37mm chip (this was 16MP) and thus it was a small image area and a faily primitive device (CCD and 12 bit colour I think). Take a look at the current generation of Digital backs - they are getting very impressive. As I said earlier my EOS 5DII produces 645 film quality results so it is only a matter of time and money before larger MF film resolution is well below MF digital. It is also possible that the Rollei solution (like my GX680) never produced good results with digital. </p>

[f_ph1]

<p>Philip:</p>

<p>The back and the Rollei belong to a friend, and it was his pictures I looked at. And yes, it was the 16mp back, the one which requires the image bank for shooting.</p>

[Ed_Ingold]

<p>The only thing wrong with using a flatbed for 4x5 is that you only get medium format quality for all your time and effort. That's no small consideration. On the other hand, a 4x5 camera can do a lot of things not practical on a medium format camera. If you don't need anything larger than a 16x20 inch print, a flatbed may be all you need. As Q.G. says, a scanner with capacity for 4x5 film at LS-9000 quality exceeds $22K. You can get a lot of drum scans for that kind of money.</p>

<p>Is there a difference between 2000 ppi on a flatbed and 4000 ppi on an LS-9000 - you doggone bet there is! If you can't see a difference, you aren't being very careful taking the picture, or not very demanding of the results. Selling an LS-9000 to buy better MF glass seems like a variation on an O'Henry story ;-)</p>

<p>(q.v., "Gift of the Maji")</p>

[ray .]

I got the LS-9000- and it gives great performance with a glass carrier for medium format. A friend uses

the V-700 and does well too.... I've made beautiful 10 inch prints from a few of his scans. Not sure if it

blocks up the shadows more than the Nikon; it would be interesting to see the same negatives scanned on each.

[noah]

<p>I've owned a Nikon LS9000 (actually two of them, and one LS8000), An Epson 4990 and my current scanner is a Howtek HiResolve 8000. In addition I have free access to an Epson V750 and I've had Imacon scans done at a local pro lab.</p>

<p>It's true people tend to get defensive about gear they own, but this holds true for Epson flatbeds as well as Nikons, Imacons and Drum scanners. </p>

<p>I'm a documentary photographer. As well as doing scans for my agency and for offset repro, I do large exhibition prints (40x50in. or bigger). I've been shooting mostly 6x7cm. negs, though I'm starting to shoot 4x5 as well.</p>

<p>I recently did a rather unscientific comparison between a 6x7 negative (shot with a Mamiya 7II and 80mm lens) and a 4x5 negative (shot with a modern Apo-Sironar lens and Arca Swiss camera). The 6x7 was scanned on my LS9000 and the 4x5 on the Epson V750. The results were interesting. The Epson/4x5 scan was less grainy but also less sharp. The LS9000 scan resolved the film grain clumps sharply and resolved more total detail. The epson could simply not resolve the grain. Since the neg was so big you could pull a decent print from the Epson scan at moderate sizes, but the scan clearly wasn't recording as much detail as the LS900 scan was.</p>

<p>Now I'm not saying that shooting 4x5 and scanning on an epson is not a good way to work. For many uses it's fine. I'm just starting to shoot 4x5 again and I think the Epson scans would be good enough for many uses, even some commercial uses. But there's no way it captures the true potential detail in a piece of film. </p>

<p>The LS9000 is probably the best deal going for the price. It comes very close to drum-scan quality, though the dynamic range is lacking somewhat, and even with the glass carrier the negs aren't perfectly flat. </p>

<p>Now that I've switched to the drum scanner I've been doing some test scans. The dynamic range is better than with the Nikon or Epson, and the results are truly sharp from edge to edge. Perhaps most importantly, a drum scanner allows you to change the scanning aperture to control the look of the grain. The workflow also suits my needs, since I can load up 9 negs on the drum and let it batch scan unattended.</p>

<p>If you're making 16x20in. prints from 4x5, then perhaps an epson is good enough. But if you're really pushing your films to the limit then there is a huge difference between a consumer flatbed and a dedicated film scanner, and even more difference when you step up to a drum scanner. That doesn't mean the Epson isn't good enough for a particular use. Many photographers never print above 16x20 and their work is mostly seen online. The epsons are perfect for small prints and web scans.</p>

<p>Before I got the drum scanner, I did look into going digital with an MF back. I made test prints and came to the conclusion that I'd need to go with the P45+ or better to get what I was looking for. The P45+ files, when processed carefully, are very good and can easily be printed at large sizes. But I still prefer the process of working with film, as well as the color I can get from the Portra films. The digital setup would have easily run $25k or more. That would give me no backup system which is a problem. And it's a lot of kit to be carrying to the rough corners of the world where I often work. For less than half of that price I got a Mamiya 7II kit (two bodies and four lenses), a 4x5 kit (Wista vx) and the drum scanner. </p>

 

[timlayton]

<p>Noah, thank you for your detailed reply. This is exactly the type of input I was hoping to get when I started this thread. <br>

I am currently doing some 4000 dpi scans on the Epson V750 from my 6x7 slides and negatives from my RZ67 Pro II and Mamiya 7. The TIFF files are HUGE... around 700-800mb with the options I have selected. My plan is to soft proof a couple of these images and then send them to my printer/lab for printing. The printers max effective pixels on either side of the image is 10,000 so I have to keep that in mind when selecting scan resolutions. My plan is to make several different size prints starting a 8" and working my up to about 34" or so, so that I can see the differences with my own eyes. I am going to shoot the same subject with my Nikon D3S DSLR and see how the prints compare. For this type of work I am doing floral and botanical macro images in a controlled studio setting so it will be interesting to view the comparisons. <br>

Thanks again for your input and feedback.</p>

<p>Tim</p>

[noah]

<p>Tim--No problem. If I get around to it perhaps I'll post some tests here. I have some 4x5 test films at the lab. When I get them back I can scan them on the Howtek and Epson scanners and perhaps cut the neg and stuff it in the Nikon. </p>

<p>When you do your test prints, which I advice strongly instead of judging based on screen viewing, you don't need to pay for a large prints. You can just print smaller, at-size sections to see what a larger print will look like. At least for preliminary testing, this is a great way to save ink (or money). Just be sure to type the details (scanner, resolution, print size, etc.) into the image so you know the details for how each print was made.</p>

<p>I don't think anyone has ever tested the Epson or Nikon scanners to provide anything near their stated resolution. The Nikon is said to be about 2500-2700dpi and the epsons are lower than that. So sometimes bigger scans don't get you much more in terms of detail. Just something to keep in mind as you do your testing. </p>

<p>Whatever you do, the only test that matters is if you can get the print quality you're looking for at the size you want to print (or at the size you may want to print in the future).</p>

<p> </p>

[polka]

<p>Just to add my grain of "silver" salt :</p>

<p>I am an amateur and do not worry to get superlatively best results, neither to scan too "slowly" ; and I used Epson "dreaded" flatbed scanners for a long time.</p>

<p>Some years ago, I was rather simultaneously buying a 150mm/2.8 lens for my Pentax 6x7 and switching my 2450 scanner for a 4990. So I decided to test first what I could optimally get with my 150mm lens hand-held and second, to compare outputs from my 2450 and 4990 in that context.</p>

<p>On my test film with the 150mm, there was a picture with a distant street plate that seemed very sharp (circled by a black ellipse on the almost uncropped big picture at the left). I scanned this detail both with my 2450 at 2400PPI and with my 4990 at 4800PPI and the results are shown respectively in the upper right and lower right vignettes.</p>

<p>But then, to check what was really on the neg, I put it under a microscope with a digital camera attached : there (in the middle right vignette) you can see the "real" silver grains.</p>

<p>The three vignettes show the plate at the same scale : this plate is 0.6mm wide on the neg. and the bars of the letters and borders must be under 0.02 mm wide. Viewing it on my computer screen, the plate is 6cm wide, so I see it about 100x enlarged. A print of the whole uncropped neg. should then show me the same sharpness at 5.6mx7.2m ; this satisfies me...</p>

<p>Paul</p><div></div>

[rob_piontek]

<p>That's very interesting, Paul. </p>

<p>I just came across this PDF on-line:</p>

<p>http://www.thedambook.com/downloads/Camera_Scanning_Krogh.pdf</p>

<p>It's an article about scanning 35mm negs with a high res DSLR (16-22MP Canon 1DS) and make the claim at the end ("Centerstage Theater" section) that 24" prints from the DSLR are actually 'more pleasing' than Imacon scans. Pretty interesting.</p>

<p>I think it's clear that there are higher quality alternatives to the Epson flatbeds, and they all cost quite a bit more. But I think unless you're already seeing the limitations of the flatbed in actual prints you're making from those scans there's no point in spending more money on it. Or as noted you could save quite a bit of time.</p>

[gareth_williams2]

<p>Back again! Just to clarify what I meant to convey in my previous post. I believe the Epson to be very good value for money but may well need tweaking in areas of focus and film flatness to deliver of its best. It is certainly possible to spend a lot more to gain a comparatively small improvement but before upgrading check you are getting the best out of what you have got! My Epson was not as good as I had hoped out of the box but a little experimentation showed it was capable of far better and it now meets my requirements- tweaked up it might meet yours too!</p>

[noah]

<p>Ok here goes. I scanned a 4x5 negative on my Howtek HR 8000 at 4000dpi, 16 micron aperture, 16 bit. I scanned the same negative on a V750 with a custom neg holder, shimmed to the sharpest position, at 4000dpi and also 16 bit. The files were both 1.6GB. If printed at 240dpi, these files would give a native print size of 63x70inches. At my normal 40x50 print size the files give about 360dpi, which is perfect for the Epsons I normally print on, though probably a bit of overkill for large prints.</p>

<p>I sold my LS9000, and I don't know when or if I'll get a chance to scan this neg on one. </p>

<p>I printed some test sections to see what the prints would look like at 16x20, 32x40 and 40x50, my normal edition sizes. Prints are a much better way to look at quality than web crops (unless your final output goal is the web, I guess). The quality difference visible in these crops is similar to the difference seen in the prints, at least at larger sizes.</p>

<p>In terms of resolution, it's really no contest. I suppose the 16x20in. print from the v750 was good enough that I would use it for exhibition. The others sizes were not up to par. And even at the small sizes the drum scans had a crispness that was lacking in the Epson scans.</p>

<p>The tonality and separation of tones was better with the drum scan. Software was Trident for the howtek and Vuescan for the Epson. This was an easy negative to scan. It was exposed properly, shot under flat light and had relatively low contrast.</p>

<p>I think what this shows is that the Epson is a very capable machine for large-format scanning, but it's not up to the level of a drum scan (even one done by someone who has only had the drum scanner for a week or two and is far from an expert). This should come as no surprise since the Howtek is a professional machine that cost $30k when new. I would not hesitate to use the Epson for all but the most demanding commercial, editorial and agency uses. </p>

<p>I've heard there is quite a bit of sample variation in the Epson scanners. I only had access to one V750 to test.</p>

<p>A used drum scanner with drum and mounting station can be found for a price similar to what the used Imacons are going for. So even for a serious amateur, they're not necessarily out of reach. The OP threw around numbers in the $20k range for a digital system, so spending $3-8k for a used drum scanning system may be a good alternative. Drum scanning is kind of fun. Even though it's digital, it kind of reminds me of the type of hands-on work of my darkroom days (fingerprints and all!). Mounting takes some practice but it's not all that hard. If you shoot several formats, the drum scanner is ideal since it can give you the quality you need to make large prints even from smaller formats but it can also handle larger formats too. You can mix and match, it's no problem to load the drum and batch scan with different formats, different film types, etc. </p><div></div>

[noah]

<p>And..</p><div></div>

[StuartMoxham]

<p>Hope you don't mind Noah but I had a go at tweeking your Epson scan to get it a but closer to the Howtek.</p><div></div>

[m_dawg]

<p>Noah, how do you find the difference in sharpness between the Mamiya 7 and the Arca with a sironar-s lens? I shoot 6x7 (rz67) and 4x5 (rodenstock lenses) and have found the large format lenses to be softer than the sharpest 6x7 lenses. The Mamiya 7 lenses are alleged to be the sharpest available for the format. 4x5 should have better tonality and less grain (and, importantly, lens movements), but do you find it offers any noticeable resolution advantage?</p>

<p>Nice scans, btw. Makes me regret missing out on a 4990 this week--even though the drum scan is clearly better.</p>