Digitize 4x5 Film Using DSLR Camera vs Flatbed Scanner

[peterbcarter]

<p>Note that there are quite a few people using Epson Scan with great results. What ever you wind up using, there will be a learning curve. You can always re-scan when your skills and / or equipment get better.</p>

[alan_cox3]

<p>I believe my response <strong>was</strong> to the point:</p>

<blockquote>

<p>" .. love digitize. I can't afford drum scans or ... "<br>

" am wondering if any of your guys have real life experience comparing film scanned with flatbed TMA vs shooting it with DSLR?"<br /></p>

</blockquote>

<p>Ethan, I have in fact done <strong>both</strong>, and like many technical problems, many times the solutions require either a lot of money or a lot of effort, or both.<br>

<br>

With a little effort, a flatbed scanner with TMA will give <strong>good results</strong>. With a little effort, a DSLR will give <strong>so-so results</strong>.<br>

<br>

With a lot of expense, drum scans give <strong>great</strong> results, but have been ruled out. And contrary to statements made here, <strong>with a lot of effort DSLR captures can also give great results</strong>. I've done them. (I have a background in both drum scanner and flatbed scanner software development and military image processing and techniques, and used to shoot Tech Pan in my 4x5. So I'm kind of a sharpness freak.)<br /><br>

<br>

Ethan, your eyes do not deceive you. Getting the results you've seen from a DSLR is possible. If you cherish your 4x5 negatives like I do mine, then I'm sure you'll invest the effort to capture them using either device.<br>

<br>

<br>

<br>

</p>

[ethan_haun]

<p>@peter, @Alan</p>

<p>The techniques for scanning and shooting you guys outlined here are equally intriguing. I am committed to learn and try them both. Thank you for sharing.</p>

<p>However, I feel that my main concern has still not been addressed. That is to say, the inherent technical advantage(s) of one method over another. By "technical advantage(s)" I don't mean the techniques being used or time involved or cost or convenience. I mean the nature of the technology themselves.</p>

<p>Maybe a analogy can serve to illustrate my question better. Now, we all know that reversal film inherently has a much shorter latitude (tonal range) than negative film. That's just the nature of reversal film. There are developing and printing techniques one can use to squeeze out more from a reversal film but the result is still not going to compare to negative film, in term of tonal range. That's the natural limitation of the film. This is what I want to find out about the two digitizing methods here. Is there any limitation of the technology itself that is of concern? I've heard that, for example, the D-Max one can archive by digitizing with a DSLR is never going to be comparable to that by scanning (assuming equally top grade equipment in both cases.) But then again, I've heard the exact opposite from some other folks.</p>

<p>Maybe there is none and that the two methods are equally good overall when all things considered. IDK.</p>

<p> </p>

[q.g._de_bakker]

And maybe there is, and have you also been told a number of times already which method produces better results?<br>Use the Epson!

[alan_cox3]

<p>The technology in either one will not be the limiting factor. They are both good enough, such that your technique can overcome any weaknesses in dynamic range, noise, or resolution.</p>

<p>The sensors in current DSLR sensors and electronics are slightly better for light gathering than what's in the scanner. Just like a modern digital camera can capture details in those low light shadows better than film. (I love film, but sadly my medium and large format cameras are sitting idle for this reason.) The analog-to-digital conversion process in the scanner is only 8 bits per color channel. Current DSLRs all use at least 14 bits, assuming you are shooting in RAW.</p>

<p>The risk in DSLR captures is the lighting and setup. Even lighting and correct focus are not always straightforward to achieve. The optics and sensor in the scanner will only get you to around 2400 dpi with the best technique, and that is once you get the height of the carrier dialed in. The DSLR and a good macro lens and/or extension tubes on a copy stand can get you way past the resolution of your film grain. If you are willing to splice 20 or more DSLR frames together, you can achieve around a very real 4000 DPI or more. It just depends on how far "over the top" you want to take it.</p>

<p>If you have a lot of negs to scan, use the scanner on all but your most prized 4x5's. It will still require learning the best workflow and practice for both cases. Good Luck.</p>

[ethan_haun]

<blockquote>

<p>And maybe there is, and have you also been told a number of times already which method produces better results?<br>

Use the Epson!</p>

</blockquote>

<p>It seems to me that there are equal number of people advocating either method. I haven't seen anyone given any reason(s) why one method is inherently better than the other, yet. My impression thus far is that either method can produce outstanding result if it is done right.</p>

[Glenn McCreery]

<p>If you would like to see a direct comparison of images of scans made with an Epson v700 and with a DSLR, for in this case, 35mm Kodak Ektar film, see a post by Gianluca Bevacqua at <a href="/digital-darkroom-forum/00b7Fk?start=30" rel="nofollow">http://www.photo.net/digital-darkroom-forum/00b7Fk. </a>The results support Alan's conclusions, both for the ultimate quality achievable using a DSLR, and the difficulties involved in getting that quality.</p>

[ethan_haun]

<p>@Alan, thank you for the information. Make sense and it's a good place to start. Thank you for sharing.</p>

<p>@Les Sarile, point well taken, but to me, it isn't such a big advantage b/c my negs are in great shape and I usually are very good at keeping all surfaces as dust "free" as possible.</p>

<p>I am planning to convert one of my Omega or Durst enlargers to use for the shooting method. What do you think?</p>

<p>And I will try my friend's V700 again with the tips I get from here. </p>

[ethan_haun]

<blockquote>

<p> ... Coolscan takes it away "magically" with no further spotting needed in post.</p>

</blockquote>

<p>Yes, but, Digital Ice (or other dust-scratch remove algorithm) also soften the image, which is a higher price to pay, IMO.</p>

[ethan_haun]

<p>How about wet-mounting? Does wet-mounting (for dust-scratch removal) soften the scan?</p>

[peterbcarter]

<p>I'm not a fan of ICE, but it is a valid feature. What keeps me on the scanner pile is the amount of detail I can pull from a film; and there are much higher end scanners than mine.</p>

<p>When you look at the cost of a scanner and the quality of scans you get out of it, the price can't be beat. With a scanner you are scanning 1 or so raster lines at a time. The DSLR you have to grab all at once. So to compare the cost of the CCD (or what ever) that would be required would be outrageous.</p>

<p>The way the CCD works on a scanner is more specific to the task. It's a slow exposure allowing better control over noise. On mine, I have real-time over sampling. The lens is also specific to the task and aberrations are easier to control. I work with some films developed with pyro and I can select which gun to sample with (even infra red), as opposed to have the entire spectrum brought in whether I want it or not.</p>

<p>Yes, stitching DSLR images have been mentioned. But your now introduce parallax and other distortions. This degrades the image even more.</p>

<p>Just a couple of my dependable arguments.</p>

 

[peterbcarter]

<blockquote>

<p>How about wet-mounting? Does wet-mounting (for dust-scratch removal) soften the scan?</p>

 

</blockquote>

<p>It does and it doesn't. Every scan requires sharpening, whether it's done in the scanner software or in post doesn't really matter. But when you are in focus, it allows you to need less sharpening because your film is in a perfect plain and held in place by a transparent layer. Look at the detail of the samples I provided, verses the other images in this thread. The 35mm (orwo n74+) shot was developed in thornton two bath stand to maximize edge effects and the 645 (acros 100/120) was developed in plain old d23 on a motor base. The latter should have been softer. I don't think I would get that detail if they were not wet mounted. But the glass of both cameras used were also very good.</p>

<p>So if I were lucky enough to own a 4x5 with moderate to good glass, I could show you a good sample.<br>

</p>

[ethan_haun]

<p>@Les Sarile, I am unsure of what you want to show me by the examples you posted (and I am NOT being sarcastic).</p>

<p>If your point is that ICE does a good job removing dust and scratch, I agree, it does, but I already know that and are not debating it at all.</p>

<p>If your point is that ICE removes dust and scratches without softening the scan, then, I have to strongly disagree. To my eyes at least, in each example, the camera copy version is quite noticeably sharper than the Coolscan versions, even the Coolscan without ICE. The scans with ICE on "Fine" are worse in term of sharpness among the bunch.</p>

[ethan_haun]

<p>@ Les Sarile, I am truly sorry if I have given you the wrong impression. I am familiar with ICE and other dust-scratch removal algorithm and what they do; how they do it and the effect thereof on the image. As I mentioned b4, this is the reason I don't count having dust-scratch removal algorithm built-in an advantage for me as I do the removal post-scan.</p>

<p>I understand every person has his/her own priority. For me? I just can't stand soft image. It's a major beef for me. Sure, I know I can sharpen post-scan, but it is not the same and I'd rather to have an image as sharp as it can be to begin with and apply as little sharpening as I can avoid it. Your priority may be different but that's cool.</p>

<p>Actually, your examples are pretty consistent with other examples I have seen online, which is that camera copy does seem to produce sharper image than with using flatbed TMA or even dedicated film scanner. Do you concur?</p>

[Tim_Lookingbill]

<blockquote>

<p>If your point is that ICE removes dust and scratches without softening the scan, then, I have to strongly disagree. To my eyes at least, in each example, the camera copy version is quite noticeably sharper than the Coolscan versions, even the Coolscan without ICE. The scans with ICE on "Fine" are worse in term of sharpness among the bunch.</p>

</blockquote>

<p>I'm going to have to go with Les on this one. I've processed quite a lot of images both scans and normal landscape captures in Raws off my DSLR.</p>

<p>But Les's somewhat soft look and tonality in the ICE Fine setting of the baby example got me to see how high pass sharpening on a layer in Photoshop would do justice and my suspicions were correct. High Pass sharpening was made for sharpening film scans such as this. </p>

<p>Hope Les will forgive me for doing some editing to show what I'm talking about since I don't have this good a scans of my own. </p>

<p>This is what I did to get the results below. Made a layer set to Soft blend mode and applied a High Pass filter setting of 8. Then I duplicated this layer and adjusted opacity to 50% and applied USM of 50 Amount & 10 Radius. Of course better will be had working on the original than on web browser jpeg.</p>

<p>I would seriously go with the Coolscan with ICE route if possible. </p><div></div>

[q.g._de_bakker]

The thing with using a DSLR is that it has too little pixels to do justice to a 4x5" negative. You can resort to dividing it up in many smaller parts that are copied separately and later stitched to a large file. But even then there is little or no gain for all the extra pain.<br>A scanner's optics can be dirty and give you bad results. It's also harder to get a scanners focus right, if it happens to be off. But all in all, it is the better option. No problems getting an evenly lit box. No problems with light fall off and distortion in the DSLR's frame (nice, that, if you want to stitch. Yes, stitching software can deal with that. But though slight, it will mean loss of image detail). No problems keeping the negatives flat. Less space required by something that is not as much in the way when not in use as a copy stand necessary for the DSLR. And more resolution and - as set out above - more control over things like dust.<br>Had i already mentioned that i think the scanner is the better option? ;-)

[ethan_haun]

<p>@Les Sarile, I don't have a "position" and I am NOT advocating one method over another. I am certainly not debating b/c I have a preference for a particular method.</p>

<p>As mentioned, I have always been a "film guy". I have scanned quite a lot of film before, using both dedicated film scanner and flatbed with TMA. Not a master, but I have also mentioned I rate myself as having mid-level skill in term of scanning. I have not done any digital camera copy before, although I have quite a bit of experence with the analogue copy- and inter-negatives process.</p>

<p>My goal here in this thread is to get as much information as I can before I start my digitizing project. Personally, I only care what gives me the best result within my own limitations and priorities. Other than that, I don't care.</p>

[ethan_haun]

<p>Thank you guys. Very good info here. I'll definitely try the tips given here for both methods. It'll be a while b4 I have the time to rig 1 of my enlargers into a proper and decent copy apparatus for the camera method. But I will report back my experience. </p>

[steven_clark]

<p>Even an old Epson 2450 would do a spectacular job scanning 4x5".</p>