scan sharpness

Discussion in 'Film and Processing' started by jamie_duivesteyn, Mar 30, 2010.

  1. Previous reader, now first time posting something.
    I'm having some image sharpness problems with my new v600 scanner. It can not even resolve the grain on neopan 1600 :(
    The scanned image was with vuescan 8.6.22 with no sharpening or noise reduction. The second image was inverted, cropped but otherwise straight from my 450d + canon 100mm macro lens.
    I'll try and attach the 100% crops so you guys can see what I mean.
  2. Here's the image from my dslr.
  3. I'm having some image sharpness problems with my new v600 scanner. It can not even resolve the grain on neopan 1600 :(
    If you can't see the grain, the scanner is out of focus and nothing will be sharp. It took me a year to be able to scan 35mm in my 4490 in focus. I had to make my own wet mount holder because no one else did.
  4. We'll need to know what settings you're using. Grain reduction softens the image, as does multiple passes. What DPI are you scanning at?
    Does the scanner have to compensate too much for under/overexposure?
    Do you have dust removal turned on in Vuescan? Are you scanning from the film directly or are you scanning to a file then scanning that file, and if so what DPI is your original scan set too?
    Assuming that you don't have Vuescan set to auto, have you tried auto (guide me)?
  5. 3200dpi and no dust removal since I am under the impression that it is useless for b&w film. Scan straight from the negative. The scene was exposed correct enough. 4 passes, multi exposure, 16 bit grey, no black or white point changes, no "s" curve. Auto level colour balance, GENERIC COLOUR NEGATIVE profile used.
    Just tried the auto guide me mode, no difference.
  6. I would suspect a film holder height issue. Either the software has not told the scanner to focus at the height of the film holder, or the scanner/film holder has a mechanical issue that physically prevents it from focusing at the height of the holder. Read this forum question that I asked: Does any of this sound like it might be the cause? Perhaps you have not selected the proper film holder option? Otherwise try adjusting the heights of the film holder to see if that improves scan focus.
  7. I don't use the Epson software but using it doesn't change how sharp the scans are. I wasn't aware that the v600 could change focus...
    I'll uninstall the software and drivers, do a fresh install and try packing the holder up with some paper tomorrow after work. Now it's time for sleep.
    Feel free to keep the ideas flowing. I hate it when things don't work like they're supposed to. It seems that ever since I started working with film a few months ago I've had dilemma and delay after delay. It's really trying my patience.
  8. Maybe you could try Emailing this info to Epson customer service.
  9. Jamie:
    Somebody might correct me on this (I am not familiar with the scanner you mention).
    In my experience, if you are expecting the same (even nearly the same) quality from 35mm film scans on a consumer-grade flatbed scanner as you get from your DSLR - you will continue to be disappointed and frustrated.
    I initially purchased an Epson 4990 Photo flatbed scanner (which I believe was Epsons top of the line Photo flatbed at the time) and after working at it for months, the best I could do with properly exposed, 35mm slide film (Provia 100) was to achieve barely acceptable results that looked bad if printed larger than 6" x 9".
    There were some differences according to which software I used (Vuescan, Epsonscan or Silverfast Lite) and I got better results if I did not manipulate the image with the scan software (beyond white-point and a bit of sharpening). However, the results were just plain disappointing.
    My 4990 really struggled if the film was under or over exposed. Since I am not that good at reading negatives, I only scanned positive film with the 4990.
    The 4990 film holders do not hold the film flat. If your film holders are similar, this could account for some of your lack of sharpness. I got into the habit of leaving my film strips under a stack of books for a few days before I scanned them. This reduced the amount of film curl.
    It was very important to ensure I cropped-out the "film borders" in the preview window, before I scanned the frame. Scanning those dark areas seemed to throw the scanner exposure out of whack.
    I also discovered that, if I left any of the film holder slots empty, I would get a strong magenta colour cast on the frames that I scanned. I got in the habit of filling the entire film holder and scanning only the frames I wanted.
    The best thing I ever did to improve my 35mm scans was to forget the flatbed and buy a film scanner. I lucked-out and purchased a Nikon Coolscan V ED just before they were discontinued. The difference is like night and day. Even if I just use the default settings in Nikonscan, the results nearly matched the output of my DSLR for prints up to 8" x 12" (the largest size I print).
    Dan Ferrel:
    I was surprised to read that multiple-pass scans degrade image sharpness...I thought that 3 passes improved my images. It was the reason I bought Vuescan. I will have to examine that more closely.
    Cheers! Jay
  10. I can get my v500 to scan 35mm film OK but not great. Here is one from the V500 comparing 100%crops to a Canon IXUS70. The film used was APX100 developed in Agfa Rodinal not really a fine grain combo.
  11. Here is another one from 35mm Trix scanned at 2000ppi. Full frame plus 100%crop.
  12. Here is another from APX100 in Rodinal not the sharpest of negs this one, but then the v500 is not the sharpest scanner, again scanned at 2000ppi.
  13. For 35mm my V500 doesn't get a whole lot sharper than that. I have found that negatives look a little sharper than mounted slides, but overall it's not much sharper than your crop, if any. Outside of experimenting with the holder height you don't have many options. There are some aftermarket adjustable holders, or you can monkey with it yourself like I did, but I never got much improvement even after I added 1mm of height. Make sure your negative is completely flat before scanning too. I'll put mine under a few college textbooks for a couple days before scanning just to make sure.
  14. I'm having some image sharpness problems with my new v600 scanner. It can not even resolve the grain on neopan 1600​
    Sorry, but it's your scanner. I went through two low end Epsons, a 4490 and a V500, and that's about the extent of what they're capable of.
    Don't bother fiddling with film height. I have, and it won't help on these. The optical path just isn't of high enough resolution.
    Actually, the problem isn't just Epson. Basically all consumer flatbeds are good for around 2000dpi. This is about 5MP from 135 format film (but keep in mind this yields a very decent 20MP from 6x7 120 roll film.)
    For example, here's an overview test image:
    This is a crop from the Epson 4490:
    This is the same frame scanned with a Nikon 5000 dedicated film scanner.
  15. Actually what I should have said was "if you have multiple passes turned on try a scan with it off and see if there's a difference". Vuescan relies on the scanner hardware for alignment from what I can tell, so if multipass isn't aligning properly it could give you softness issues. I have this problem with my HP scanner, but I shouldn't have made it out that I was saying that Epson has this problem as I've never used an Epson scanner. It's just a possibility that can be eliminated by doing a scan with multipass turned off.
    If full auto gave you no difference, good or bad, then you may be stuck with what you have. Can you get a pro lab or mini lab scan of the same film to compare sharpness? Just one more possibility to eliminate.
  16. Here is one more V500 image. I scanned this one at 6400ppi just for fun to see what the scan would look like. The scan was then downsampled to 4000 ppi if I remember.
  17. Hello!
    I am also interested in this problem .It seems that these flatbeds simply don't do a good scan from 35mm:
    Look at this review of the V700 , which is superior to the v600 : briefly ,they don't do practically more than 6-7 MP, at best : , so, Robert is right.
  18. By the way, 135 format film well scanned and mildly post-processed can give technically high quality images. The following test frame was shot on Kodak Gold 100 then scanned on a Nikon 5000. The scanned image is about 22MP at 16bits deep per channel; I'd hazard to estimate that this level of detail is reflected pretty closely in the actual digitized image:
    Here's a crop:
    Here's another:
  19. Don't be so quick to throw the baby out with the bath water, folks! Yes, in an ideal world, a dedicated anything is probably better than a general tool (flat bed). I have both. I bought a Plustek 7200i for 35 and a 4490 for 120.
    The 7200i gives me real 3200 dpi and is sharp. Adequate for my needs.
    The 4490, out of the box, shows poor thought and engineering in film holders. It's like they want to promote an after market. The real test of how good a holder/scanner combination is to scan and look for grain. If it is in focus, you are good to go.
    Before you spend any money to fix your situation, put a negative face down on the bed (center) of the scanner. Get a sheet of glass from a picture frame on top of it, to hold it in place and to flatten it. You should get a sharper image. Yes you will have newton rings, but this is just to prove something. If your image is sharper, your scanner is OK and up to the task. If not, don't waste your time. has replacement holders and ANG glass for 120. They don't work that well for excessive curl (LuckySHD 100), but have adjustable legs for focus. If the curl is too great, the film cannot rest on the edges of the form. Nice try, but it was no magic bullet. has ANG glass for 35. They rely on the holders that Epson gave you. Film with curl has the same problems from above. The focus is not adjustable with this solution. I do not know of any replacement holders for 35.
    As far as I know, there are no commercial wet mount jigs available for the 4490/v500/v600. This is too bad, considering how many of them are out there (and how badly the engineering of Epson is for this product). I decided to make my own. Here it is......
    1 - 1 8x10 clip frame glass ($1-$4 depending where you get it).
    2 - Cardboard from a cardboard box of some sort. Thickness to equal .5 to 1 mm, as required.
    3 - Cellophane tape.
    4 - 1 sheet of acetate (or mylar scanning sheet). Mine cost me $3 at an art store.
    5 - Scanning fluid ( $25.
    6 - 1 microfiber cloth ($1 at a dollar store)
    Take the glass out of the clip frame. Cut the cardboard into 1/2 in squares. Tape them onto the 4 corners to form legs. Layer to get optimal height (experiment and adjust). If you use the fluid I suggested, you can use clear acetate, otherwise a sheet of clear mylar will have to be used. Cut a strip big enough to cover a strip of 120, in the usual length. Oversize it enough as it will for a flap covering the film. One edge (length wise) to the sheet glass on the underside (side with the legs) in the center. This has to be just where the stock holders would place the 120 film. You now have a holder.
    Flip the acetate back and spray some fluid on the glass. Just enough to have the film stick to the glass. This is not the optical side, so its not that critical. Place the film (emulsion side) onto the glass as square as possible. Spray a healthy amount of fluid onto the film. Experience will dictate how much. Lower the acetate down, going from taped side to end. Take the microfiber cloth and squeeze the air out, going from taped end to the other, while holding the glass from the bottom, just under the film. This gives a gentle pressure from under the glass. You are only concerned with the film itself. You can see some bubbles in the shot above.
    To scan, place the holder onto the scanner resting on the cardboard legs. The film never touches the scanner bed. Make sure the holder does not go up to the top edge of the scanner as banding from mis-calibration may result. You will know when you see it. The alignment of the holder is less important and making sure the strip is correctly oriented. It is easier to correct here than in Photoshop.
    To remove the film, use a xacto knife to pry up the edge of the acetate and then the film. You do not need to re-spay the glass, just the top side of the film as you change strips.
    The microfiber cloth is important, because it is soft. The scan surface is the acetate and can be easily scratched. You should expect to replace the acetate from time to time. The glass plate is only required for form and to let the light through. It should be a keeper.
    The Lumina fluid is gentle on plastic parts and evaporates off the negs easily. Just expose them to air. The tray, as I have experianced, is not prone to making a mess. I just let it dry on it's own.
    Sadly, that's all there is to it. The holder should be generic to any scanner. I can even scan my 110!
    Here is an example . Choose all sizes to see the full detail.
    I did all of this to get my MF scanned properly. As a bonus it does 35 (and 110) much better. The scanner does about 1800dpi natively, but the quality has vastly improved. I've been meaning to post this somewhere and I hope this is of use to someone.
  20. One more from the v500 at 1500ppi APX100 rodinal. Not a tack sharp neg hand held around 1/30th 35mm series E f2.5 wide open. It would produce a 10 inch wide print at 200ppi. There is not much point in trying to produce huge scans with these flat bed scanners. I find it better to shoot TriX in HC110 and scan at around 2000ppi that way the files don't get too bloated. At the moment I have APX100 and Rodinal to hand so I am using that. Otherwise I would just shoot TriX.
  21. My V500 is so fuzzy its disgusting. Bad example I guess.
    Thats after owning the 4870 which did a good bit better and the v750 that did a whole lot better.
    I have owned a lot of other film scanners and also own a drum scanner.
    IMO there must be a wide variation in quality between V500 V600 machines, as I have seen some nice V500 scans, but mine like yours is very fuzzy. If your looks this fuzzy I would send it back.
    This is how my mighty V500 holds up against a Primefilm 7250 pro3 and a Howtek 4500 drum scanner.
  22. I can resolve film grain in 1950's B&W large format negatives with a V750. See my example in this thread. The Better Scanning film holder is key.
  23. Les, I don't have that strip anymore. It was from a test roll. However I have some frames from the same roll. Different pigeon, but same scene.
    Go to all sizes to download and inspect. No grain reduction. In my opinion, the 7200i is marginally better. These are both 2400 scans. The 7200i can go up to 3200 real. and surpass the Epson, but my $10 mount is cheaper than a new scanner.
    For me, it was all about getting a good MF scan. This is just a bonus as it can be a pain switching back and forth. Before, I would not even consider the 4490 usable with 35mm.
  24. The V750 is a totally different animal and resolves a lot more detail than a v500 or probably V600 by the looks of it.
    At least that is what I have seen.
    Comparing my old V750 to my current V500 is like apples and oranges.
    If I were to guess the V500 is resolving about half of what my V750 did.
  25. Two more. I have not figured out how to embed 2 images in a post so here goes.
    This is some really sharp bluefire film.
    This scan is from a PI Primefilm 7250 pro3 at 3600 dpi with no sharpening. Only histogram.
  26. Oops lost the 1st image. Here is the PI 7250 image.
  27. Here is the same piece of film scanned with my V500. I have tried, reaising the film holder, glass, on the glass, wet mount etc etc. None of it does any good. I think I just have a lemon. This is 4800 dpi rediced to 3600 dpi to match the 7250 pro 3 scan.
  28. Troy your V500 seems to be quite blured. There must be quite a bit of variation between different V500.
  29. Jamie, the V500 is somewhere in the middle. Way less than a Coolscan V or 5000, but reasonably useful, especially for the price.
    I tested the V500 with an Edmund Scientific USAF glass test target. As expected the actual achieved resolution is about 2000 ppi, way less than the quoted "optical" resolution. But, surprisingly this is only achieved in resolving lines in one direction (lines parallel to the scan track); for resolving lines running across the scan track, it's only 1270 ppi. See this thread:
    Practically speaking, that means you'll get 2000 x 2000 resolved pixels from a 35mm negative. Less than your DSLR. For 120 film, the story is better.
    I get very usable scans of 120 film (6x9) from which I make nice 12x18 inch prints. I doubt with this scanner that you'll be able to scan 35mm to get results better than a DSLR. In other threads, people have said that the V500 gives scans that print OK at 4x the linear dimension of the negative/slide. That sounds about right to me.
    Peter, I'm intrigued by your wet-mount. I might give that a try.
  30. Troy I played around with your image a bit. It is still not great but it is somehow improved. It is still not as good as the plustek.
  31. I thought about buying a Fisher holder for the V500 but didn't after running a quick test.
    Use a stack of the yellow 3M sticky pads to shim up the stock Epson holder to various heights and compare the subsequent scans. If film height actually mattered for the V500, then this (binary) search process should quickly locate the best height offset. Unfortunately, for my V500 at least any height variation within 2mm or so made absolutely no sharpness difference.
    The premise of the third party carrier is that it it flattens film curl and keeps the entire scanned film surface at this optimal height. Unlike the V700/V750 (and dedicated film scanners to an even greater extent) the V500 optics have a great deal of DOF. Unfortunately, the resolution isn't particularly high within this range.'
    I'm sure the Fisher holder is a good addition for the high end Epsons, but is not particularly useful on the V500.
  32. Here is another v500 scan first is the full frame image.
  33. Now if I print the 2400ppi scan at 10 inches wide the face would be some what smaller than this next section I have uploaded. The 8x10 print would not look bad at all.
  34. update: After some setting changes I can get acceptable scans from running straight off the glass holding the negative flat with another piece of glass. I can see grain now!
    Perhaps id11 and neopan 1600 weren't such a good idea ;)
  35. Jamie,
    ANother important thing to consider is the multipass and multiexposure. The stepper motors in these units are not all the accurate. With multipass, it is rare to get the system to line things up correctly.
  36. I second the idea that you can't compare the sharpness of a recent DSLR and a flatbed scanner, no contest the flatbed will be a lot worse. You may be able to get some images to share on the web or print less than 8x10, but only after spending a lot of time and effort on an image. The difference in time and effort compared to a dedicated scanner is like night and day, I could get better prints from a dedicated scanner with no post scan manipulation than most efforts with a flatbed.
    IMO I could not suggest that you get another epson, or even say to fix your current one. Stick with your DSLR and learn how to use it, you will be more productive and get better image quality in your prints. Note that the DSLR is a different tool than shooting film, you need to expose the image differently. Your DSLR is more like shooting slide film, you want to get your highlights as bright as possible without blowing them out. IMO there is a lot of craft in DSLR shooting, just as much as scanning film. The difference is that the DSLR will let you focus on image composition rather than spending your time trying to correct flaws in a flatbed.
  37. Stuart,
    I know anything can be improved with sharpening and that does look better but still that does not improve a machines ability to render detail.
    Also I look for a 2 pixel soft edge max in whatever I scan/shoot etc due to the fact that when
    you go over that amount sharpening just becomes harder and creates more artifacts.
    Also over 2 pixels and photos just never look crisp.
    BTW I have a Pacific Image 7250 pro 3 not the plustek.
    That's the new 7200 dpi scanner that will scan an entire roll of film.
  38. Sorry Troy I read Plustek somewhere here and confused it with your Pacific Image 7250. That Pacific Image scanner looks to be a good deal with the ability to scan a whole role and the scans from it look much better than the V500. I would like one of those myself shame I did not know about it before I bought the V500 however for now I am stuck with the V500.
  39. "IMO I could not suggest that you get another epson, or even say to fix your current one. Stick with your DSLR and learn how to use it, you will be more productive and get better image quality in your prints. Note that the DSLR is a different tool than shooting film, you need to expose the image differently. Your DSLR is more like shooting slide film, you want to get your highlights as bright as possible without blowing them out. IMO there is a lot of craft in DSLR shooting, just as much as scanning film. The difference is that the DSLR will let you focus on image composition rather than spending your time trying to correct flaws in a flatbed."
    I don't think the OP is asking about whether his DSLR is better than film. He used the DSLR with a macro lens to digitize the negative and compared that to a V600 scan of the same negative. His full film shot may be a masterpiece but he's just showing how crappy the flatbed scan looks.
  40. Stuart,
    No problem, I just wanted people to realize that image was scanned with a 7250 pro3.
    The 7250 pro3 is decent and my copy is certainly sharp enough at 3600 dpi.
    That said it does have some issues. Its difficult to batch scan the emulsion side of the film and with this scanner it is significantly sharper scanning the emulsion side.
    The software is very simple and nice but a bit flaky. Also WYS is not always what you get. They must use some other color space or its a bug.
    Silverfast works better than Cyberview or Vuescan but thats just more $.
    If you can find the right combination of film profile and ROC, Cyberview is faster for automated scanning of a roll of 36. Thats a long piece of film.
    Its noisy in very dark area, but I am still working on that. Forget severly under exposed slides.
    Very thin B+W negatives scan well.
    Mine color streaks on one side with slide film. That does not show up with B+W
    Still its good enough to pull a 18 mp file from film like this.
    I bought it mostly to preview an entire roll before drum scanning but if you dont need more than a sharp 3600dpi/18mp then it will work.
    Not bad for $420.
  41. I have tested filmholder height on my V500... I found no difference in the ranges I tested, from zero "lift" up to a couple of mm. This matches Robert's report (above) and similar comments by others.
    I hear that on the V750 film height can make a significant difference in resolution.
  42. By the way, for those looking for ANR glass to use on flatbeds, this seller on EBay has been pretty good. Keep in mind that to really remove film curl, you'll need two pieces of glass to sandwich the film (but only one has to be ANR.)
    I bought the ANR glass, but ended up holding my nose and got the Nikon glass carrier as well. The Nikon was just more convenient to use, but so damned outrageously expensive.
    Lastly, it is possible to use the glassless carrier and simply focus stack multiple scans of the same frame. This is worthwhile for that single special exposure. The fewer glass/air/mylar gives slightly sharper scans (for a tremendous amount of more work.)
  43. The Epson V500 is really not up to the task of scanning negatives. I own a V500 (one of the best copies it seems) and a Coolscan 9000. This is a comparison scan of TMAX 4000 6x7 at 4000dpi. Also shot the same shot with the Canon 40D and upsampled to match.
    (The negative was very banged up when I scanned it - This is at no fault of the scanner).
  44. This is even digging deeper with a crappy microscope to show that there is even more there than what the Coolscan can fetch at 4000dpi. (the Coolscan is upsampled to match the microscope).
  45. The V750 is in an entirely different league and it does benefit from tweaking.
    Still the lens is just not good enough for high end film scanning and between that and the glass mine created a significant amount of CA. Also no AF, but it was built for a price point.
    If they were to equip it with an AF lens that cost as much as the scanner itself, better glass, better dmax, better software then they would really have something, but probably more like $2000 price tag.
    Its still a good scanner and miles ahead of the v500/600.
    To me the v750 is about 2x as good as the V500/600, and a real film scanner like a Nikon 5000/9000 is about 2x as good as V750.
  46. Maruo, I really appreciate the real-data examples, here and in other threads. Thanks!
    How do we interpret your results, quantitatively? Looks like your Coolscan 9000 is delivering a 10, the V500 and the 40D are about 4.5. What does that mean in terms of PPI?
    I do note the V500 resolution is less than half of the Coolscan 9000. That sounds about right. (Coolscan 9000 at 4000 ppi, V500 at < 2000 ppi).
    But, before we throw our V500's into the trash bin in disgust, let's consider what one can do with the results they deliver. I'll post an example in a minute.
  47. So, what can we do with a V500? I think I can scan 6x9 color negative film to produce a good 12x18 inch print. Here's an example... It's a favorite test scene I use, the famous Cheers Pub in Boston. Here's the overall scene, shot with a Mamiya Universal, Mamiya Sekor 100 f/2.8 lens, f/16, probably about 1/60th, on a solid tripod. Scanned on V500, Epson film holder, at 2400 ppi. Processed in Photoshop (more below).
    Epson V500 Scan, 6x9 Color Negative, Overall Scene
  48. Here is a crop 600x600px, actual pixels, of the final processed file. This will be a 2" x 2" square in the 12x18" print. I think this is sharp enough, and it has accurate enough color, though it took some steps to get there. I forgot to mention above, this is Ektar 100. Regulars here know that I have posted examples from this scene in the past; this is a new exposure Feb 2010.
    And, I know the OP is asking about B&W, but I think this color negative shows the resolution and contrast that will be relevant in using this scanner for B&W.
    2x2" crop from 12x18" image, 600x600 actual pixels, Final Processed file.
  49. Here is a second crop from the same scene. Same comments as above.
    Another 2x2" crop... Same comments as above.
  50. You can produce some nice prints with the V500. I may not be the best with fine grain film and detailed lanscapes but it works great with more grainy TriX or Agfa APX. This one is from APX100
  51. What did it take to get there? Here's the layer stack:
    And, a brief listing:
    • V500 Scan to 16bit tiff at 2400 ppi (I read here that I could get a little more resolution by scanning at 4800 ppi and then down-sampling). Color Space sRGB, with profile.
    • Noise reduction and capture sharpening. I find this essential when scanning.
    • Hue/Sat for minor color adjustments... (reds and yellows shifted a bit to orange)
    • Vibrance to bring up the saturation and punch
    • The top layer Curves adjustment for the major color correction
    These are shown in the layer stack above. Then (not shown)
    • Resize to 3600x5400 to make a 12x18" print
    • Flatten Image
    • Sharpen for output (per Bruce Fraser's nice book)
    • Mode = 8 bit. Save as JPG, high quality.
    In case it might be helpful to anyone else using a V500 on Ektar 100, here is a link to a fully layered PSD file (5MB) of the 600x600 pixel first crop from above. The bottom layer is the original 16bit tiff scan from the V500.
    I would be interested in any comments about the processing above, or the scanning.
    I'm going to keep my V500... I think I can make nice 12x18 prints from my 6x9 negatives and chromes. For 35mm, I would only try for a 4x6 or a web image with the V500. I hope this is helpful to the OP and to anyone else reading this far.
  52. Finally, in case anyone would like to look at the final file, ready to print at 12x18", here is a link to the JPG, 3600x5400 px, 5MB.
  53. Here are 100 percent crops and 50 percent crops. They both appear in the monitor much larger than they would in a 10inch wide print the details in the rope would hardly be visible and the crop area would be around 1.5 cm high.
  54. Richard the file you uploaded is only 3600x2400 pixels. It does look nice though.
  55. Oops.. So it is. My error... In everything above, I meant 8x12" print, not 12x18" print. Sorry for getting this confused today.
    Yes, I'm happy with 8x12" prints from my V500. (I don't know about 12x18")
    And, Stuart, your boat image looks good too. How large did you print it?
  56. Richard, every marker on the test chart corresponds to aprox 730 lines per picture height.
    In this particular test the results were:

    *** 40D:
    marker = 3
    lpph = 2160 lines per picture height
    effective total real resolution: 2160 * (1+1.5) = 5.5 megapixels [as opposed to the 10MP nominally advertised]
    *** TMAX+RZ67 scanned with a flatbed
    best marker: 6 (it is on 4 in the cross direction)
    lpph = 4320 lines per picture height (about 2000 dpi - about 1300dpi in the other direction)
    effective total real resolution: 4320 * 4320 * 4/6 * 69mm/56mm = 15.3 megapixels
    *** TMAX+RZ67 scanned with Coolscan:
    marker: 11.5
    lpph= 8400 lines per picture height (about 3810 dpi - it is only 3650 dpi in the other direction)
    effective total real resolution: 8400 * 8400 * 3650/3810 * 69mm/56mm = 83.3 megapixels
    *** TMAX+RZ67 visually inspected on the microscope:
    marker: 15.5
    lpph= 11310 lines per picture height (about 5130 dpi - 200+ lines per millimiter)
    effective total real resolution: 11310 * 11310 * 69mm/56mm = 158 megapixels
  57. As you can see, the flatbed captured only 1/10 of the information on the film and the Coolscan only 1/2.
  58. I am always amazed that some folks are "discovering" a flatbed is not as good as a dedicated film scanner; whether today; 5; 10 or 15 years ago.
    A pro flatbed 15 years ago was about 1200 dpi and cost about 3000 bucks; today a Modern consumer one is about 1.5 to 2 times the ability to pull out details; and considering inflation only costs 1/10 to 1/20 as much.
    Before 1200 dpi pro flatbeds came out; we had 600 and 800 dpi models that cost more than a new Hassleblad kit; ie body; 80mm Planar and back too.
  59. Mauro, thank you. That exactly matches my measurements for the V500, 2000 ppi in one direction, 1300 ppi in the other.
  60. The test above uses 6x7 film. Using 35mm film the film+scanner effective real resolution would have been 4.5 times smaller:
    Canon 40D = 5.5 megapixels
    TMAX 35mm + flatbed = 3.3 megapixels
    TMAX 35mm + coolscan= 18.5 megapixels
    TMAX 35mm film= 35 megapixels
  61. I think in the end if you have enough resolution to render the subject as you would expect to see it then prints will usually look pretty good at reasonable sizes. I will get it printed around 10inches wide and I think it will be fine. I don't make big prints. I used to like making 8x8 inch darkroom prints from 2 1/4 square negs.
  62. You are welcome Richard.
  63. Having tripped up in my own confusion about print size, I just prepared a file for printing 12x18" at 300 dpi from the same V500 scan. Since I originally scanned at 2400 ppi, I had plenty of pixels.
    Looking at the file, I think it would make a nice sharp print at 12x18". Here is the ready to print JPG file at 12x18" (8MB) in case anyone is interested.
    I think the V500 has some usefulness for MF, especially at the modest price. For modest sized prints, web images, and files to view on the iPhone, the V500 lets me enjoy the results of my MF work.
  64. Mauro wrote:
    [40 D] effective total real resolution: 2160 * (1+1.5) = 5.5 megapixels [as opposed to the 10MP nominally advertised]​
    I agree with your results, but I think the math for the 40D is: 2160 * 2160 * 1.5 = 6.9 Megapixels
  65. I think you will get a 12x18 from that file no problem at all. I would maybe sharpen it a touch more for printing. It gets really crisp when I applied a bit of unsharp mask. ( 0.7 Radius ), ( 140 percent ), ( 0 threshold ).
  66. Richard, you are correct. I'm sloppy today.
  67. When I was scanning 135 with a V750 I was getting better results with regular old film than from a good Canon P&S. Oftentimes much better. MF + the V750 was at least as good as a DSLR, oftentimes much better. Now obviously DSLRs and P&Ss are good enough for almost everyone (in fact P&S are wildly more popular than DSLRs) and therefore a good flatbed can be better than what most people are accustomed to from digital. That's all I wanted to say. I'm sorry, I can't help with your particular scanner. Obviously something's wrong with it.
  68. You are better off investing in a Nikon coolscan V ed or Nikon Coolscan 5000 Ed. My images scanning use Nikon if that should tell you something with at least 4000 DPI.
  69. Thanks, Stuart. The proof will be on paper, and I'm going to print it to see. That would mean that from a V500 scan we can get good prints 6x the linear dimension of the film.
  70. Here I got one of these *FREE* when I bought some paper for my business.
    It seems to test about 2700 dpi compared to a high end scan; great for a flatbed. The film holders are abit clunky; but are rugged. Jim helped me set up the drivers; the scanner bundle even came with a test slide; some type of dorky red vegetables. The scans seem normal; ie about as quick as other scanners as far as time goes.
    When you do a scan one does get a popup; an tiny advert for Sabre's paper division; I am not sure yet how to turn it off yet. I tried it on Win2000 and Vista Home; but not the iMac duo yet.

    The scans are decent; and tech support was great; the scanner supports USB 1/2/3. It scans abit faster on a USB 2 port; but with 3.0 the scanner is the bottleneck as far as time goes.
    OK, first let's go over some parameters; it is really NOT a 4200 dpi scanner; "more like a 2600 dpi one with a high contrast original; probably about 2000 with low contrast pictorial stuff." That's what she said; the lady in tech support.
    TOTALLY COOL; a scanner company's employee mentioning PRACTICAL numbers in an honest dialog!!!. I did not even have to pay for shipping; Darryl in shipping bubble wrapped the scanner and had it on top of the paper order. I was worried that the scanner would get ruined by a fork truck!
    The scan software is some what strange; Sabre uses dmi instead of dpi..
    Sabre ExelScan i4200
  71. "That's what she said; the lady in tech support"
    The salient point in Kelly's last post is the one immediately to the right of his (her?) name, following the 'film canisters' icon.
  72. OK, jokes aside...
    We can keep our V500's! I have in my hands a 12x18" print from my V500 scan of 6x9 Ektar color negative. The print is plenty sharp. Here's another link to the ready-to-print file if you want to look for yourself. Based on this, I'm ready to say the V500 is OK for prints up to 6x the linear size of the film you are scanning. This makes it a cheap and helpful tool for my MF shooting. Yes, there's more on the film than is realized in a V500 scan; I'll use my V500 for convenience and then buy better scans of the keepers.
    Stuart, I'm still digesting your suggestion for more sharpening.
  73. I found that just a touch more sharpening made the image pop a bit more and it became really crisp. The brick work became more defined. Did you try sharpening it a touch. It may just be a taste thing more than anything. It is not unsharp as it it and the scan looks very nice. You could print it at 16x24 at 225ppi. The scan looks good even when viewing at 100% so I don't thnk it would be a problem. I spent around six years working as a wide format printer and doing photgraphy work I had to make many large prints from much much worse scans than this one.
  74. You must have one of the good ones. Mine wouldn't even come close to that.
  75. Hi Richard, this image could stand a lot more sharpening to bring out detail. I just did a quick and dirty edit using unsharp mask and also increased the overall contrast a bit (using curves/autocolor). See how this compares to your other print.

Share This Page