First 120 film scanned - underwhelmed (Epson V700)

Discussion in 'Digital Darkroom' started by raymondc, Jun 24, 2016.

  1. Yes, it is a flatbed scanner. Guess the question is what do you guys consider the best solution for the average person? Medium format film. Ok, so it may not be a drum scanner or a hybrid scanner like a Imacon and fluid mounting can be messy so that's out too. Just look on a light table with a loupe? Darkroom printing with b/w film instead? For the few shots request drum scanning from a commercial place?
    Cheers ...
     
  2. SCL

    SCL

    Specifically what underwhelms you with the first roll scanned? What post processing software are you using, and are you reasonably competent with it? It took me a while and a switch in scanning software before I became comfortable scanning it. I rarely print anymore, so most of my stuff is viewed on a computer screen. Unless you're doing mural sized printing or selling to a house, most flatbed scanners can do a credible job with medium format images.
     
  3. It is the snappy sharpness that is not there. I have a Nikon D600 and the first thing I notice was if my Nikon could do that then what's the point shooting a 6x7 (RB67). It was shot on a tripod with mirror lock up.
    Not so concerned with the color or contrast yet etc yet as those can be adjusted and shadows can be lifted where required (more so than slides).
    Using Epson Scan software and the original film holders. With 35mm film I had more detail with a Nikon D2h (4MP). Both on tripod etc. Maybe for the important ones use a darkroom and have it printed there with my camera club old timer buddies or have the odd one drum scanned? Simply go the full hog than just get better film holders, better scanning software, better sharpening software? Go on a trip with a digital camera and take the medium format on specific important outings only?
    I bought a Coolscan 4000 some yrs ago off a pro maybe, but after a year or two it broke. And .. they are no longer sold / serviced. Should have got the Coolscan V new before but that doesn't address 120 film.
    If I print, its generally 11x17 inches or A3 / Super A3 and I would like it to be printed A2 (16x24) if the others enjoy like if I visit a place that was friends/family's hometown / country. But for now it is the sharpeness that is absent viewing the image inside my Lightroom catalogue. Seems like I could do the job easier and better with the D600 less the filmi tonality / and the film enjoyment.
     
  4. Ray, I'm no expert at scanning, but scans from a flatbed require quite a bit more sharpening than those from a deicated film scanner such as the Nikon Coolscan. I use a V700, but not the crappy film holders supplied with the scanner. As I'm using slide film, I scan with the slide mounted in the GePe mount (I still project slides!) which holds the film pretty flat.
    I've had a couple of trannys scanned on a Coolscan and I've scanned the same trannys on my V700, and with careful sharpening I've managed to produce a print which is all but indistinguishable from a print scanned on the Cool Scan. Even examining the prints with a loupe it's difficult to see any difference.
     
  5. SCL

    SCL

    Ray - I was generally unhappy with my Epson Scan software (several years old by now). I eventually switched to Vuescan Professional, took abouta year to really feel comfortable with it, and now use it exclusively on my flatbed for 120 film. I have an older scanner and was a little unhappy with the film holders, so experimented around and finally settled on some homecrafted ones for my work. I think as you learn the ins and outs of your scanner, its software, as well as finding the ideal exposure and development for your films you'll become much happier. Of course if $ is no object, you can have the important shots professionally scanned and printed. I've done that on several of mine and over the years and thrilled with the results.
     
  6. Scans, like raw files, need sharpening. Period. And that's left as an exercise for the scanner operator, as the correct means and amount of sharpening are going to vary based on scanner, settings, film stock, personal preference... Welcome to the jungle. Give yourself some time to nail down your process. ;)
    And, lest it go unremembered, Lightroom is automatically sharpening your D600 files. Lightroom, on the other hand, does not automatically sharpen non-raw images. So, yes, your D600 raws are naturally going to be sharper—because they've been sharpened. (And also because there are some frequently-tangible differences in the way Lightroom renders raw and raster images at reduced viewing sizes.)
     
  7. Ray,
    Please post a 700 pixel wide crop from a scan which demonstrates your concern. If the grain is sharp, that's as good as it gets.
    In general, the effective resolution of a flatbed scanner is only about half the "optical" resolution. This is due to a combination of poor optical resolution, overlap between sensor cells, and marketing puffery of the specifications.
    The depth of field is often very thin, and the plane of focus is not necessarily at the glass surface, or at the height the holders maintain the film. A precise way to check is to put a machinist's scale on the glass and prop one end with a pencil. Looking at the image, you can see which scale markings are sharp. The actual height can be calculated from the sine of the angle of the scale.
    Sharpening improves the micro contrast of the image, which is often low out of abundant caution.
    Focusing a MF camera is critical, as is holding the camera still. If you want medium format sharpness, lock the mirror up (if applicable), use a tripod and a cable release.
    Finally, film isn't as sharp as most people like to think. High resolution is usually obtained with extremely high contrast (1000:1) targets. The edge contrast in real life subjects rarely exceeds 6:1, which cuts the resolution in half (or more). I have a 16 MP Hasselblad back (4080x4080 pixels), which is noticeably sharper at the pixel level than scans from my Nikon LS8000 (8500x8500 pixels).
     
  8. Les - no idea don't have a proper loupe.
    Edward and others - att. Film scan resetted sharpening.

    Ilford Delta 100 Pro @ Ilford ID-11 @ 20degrees Celsius @ 11 minutes as by the Ilford Data Sheet to the dot.
    00e1EH-563920584.jpg
     
  9. Also now the D600 crop, at F16 also shot in RAW downloaded into LR with no adjustments at all. Ok, sharpening by default. But you can sharpen the film scan ... I just wanted to leave it as is so easier to assess (without subjective input by myself).
    00e1EJ-563920684.jpg
     
  10. A 50mm lens for 35mm used in reverse makes a good loupe substitute. No reason a DSLR 50mm lens will not do the same. Epson scan likes the unsharp mask to render sharp scans, its enabled as default and can be varied in Epsonscan 5.1. Good scanning takes about 12 to 24 hours with a known good negative/slide starting at default settings then changing one variable at a time 10% and comparing the result to the default scan. Make the test scans at 1200 to 2400 dpi and view at 100% to evaluate.
     
  11. Charles yes but I turned off
    sharpening in Epson Scan. Ticked
    off the box. Yes the 50mm lens
    look sharp though. I rather
    maybe prefer drum scanning than
    spend on additional holders and
    software. The older digital backs has some brownie points lol.
     
  12. In Epsonscan Professional mode click on the Configuration tab, bottom center of the scan plane, click on the Color tab, at the bottom select the check dot No Color Correction, click OK to close the configuration box. All adjustments/color correction/autoexposure are now turned off.
    Now put a known sharp negative in the holder insuring it is flat as possible and make a 2400 dpi scan (higher if you wish), this is your reference scan. Place strips of paper between the film holder tabs that it sits on and the scan glass then repeat the scan. #24 pound printer paper is .004 inch thick, card stock is .013 inch thick. If the scan is sharper then add a second layer of paper and scan again. The holder may need to be raised .002 to .006 inch to get the optimal focus. Permanently attach a shim of the proper thickness to the holder when the proper height is achieved.
    I have a V700 and shortly after getting it I purchased a Betterscanning holder. I was disappointed with it as one has to cut the thick plastic blank mask material to the negative size, tape the negative to the antinuton ring glass, then adjust set screws changing holder height .001-.0015 inch at a time to get the optimal sharpness of the scanner. I have yet to scan enough with my V700 to get its optimal scan. Production tolerances dictate that each scanner will be slightly different. You will have a similar learning curve with a drum scanner and I wish I could afford a drum scanner.
     
  13. I have adjusted the factory holder tabs to the + sign. Reviews say that is better. What you mentioned is an awful amount of work to get one one scan done.
    I was thinking about sending the odd film out to be scanned, not buying my own one. But, on the local auction here in New Zealand I saw a used one for around $700US (a Howtek) but it required another $700US to replace the power supply.
    The macro lens approach sounds easier too ....
     
  14. What you mentioned is an awful amount of work to get one one scan done.​
    But once you find the optimal height it will be the optimal height for all negatives scanned with that holder. You only have to do it once every year or two. The bed glass is attached by thin double sided tape which will give out after a few years. You will know when its time to recheck as the sharpness will start to shift. Another member on this site used aluminum foil for his shim. Aluminum foil, like paper, comes in different thicknesses.
     
  15. Thanks I will give it a go.
    I know you said double sided tape onto the tab. How did you stick paper on paper? Ie - 2 layers of paper. Glue?
     
  16. No, the scanner bed glass is attached to the scanner bed frame with double sided foam tape. I've seen it give out and let the glass sag after 5+ years.
    I would start by taping strips of paper to the holder with the tape not on any of the support tabs. I believe there are 3 rows of supports, one left, one center, one right. For a double thickness I would fold a wider strip of paper in half. Once the proper height is determined then I would look for a more durable material and possibly glue it to the holder or tape it to the scanner frame without blocking the calibration area or negative path. Some glues expand as they dry and glue or tape will add thickness if used on the holder at the support tabs.
     
  17. How much of copy paper equiv are you using?
    Here is my progress so far. Left is the double sided tape with 7 sheets of copy paper. Right is just the (1) sheet.
    00e1H4-563929384.jpg
     
  18. Epson added clips to hold the glass better for longer in addition to the double sided tape for some of their newer models. When I first designed my holders and took my 3200 apart, I was surprised to see the double-sided tape. Even when it was in good shape, the tape still added variance.
    Since you are using the OEM Epson holder, when you snap their frame shut, look at your film from an angle to make sure the frame hasn't bunched up the film and added a curve or ripple to the film. If it has, you need to try insertion again. The film has to be at the right height for your scanner's optics AND it has to be flat.
    Doug
     
  19. Ray,a few pieces of advice from my experience with my V700.
    The focus adjust needs to be done just once. Edward Ingold's suggestion of using a slanting machinist scale looks good. Alternate methods I've used: maximize jpeg file size when repeatedly scanning same small crop area; use slant edge MTF.
    Once the focus is right, you need sharpening to get the most out of your V700. That is not "cheating", there is information present at spatial freqs where the MTF has started to drop. Specifically, I found that scanning at 3200dpi, USM parameters radius 1.0, 100% are close to optimum (and easy to remember!). 3200 dpi is probably overkill for MF; scan @ 3200, USM, then downsample to 1600 (or 2400, or whatever). Scanning directly at 1600dpi carries the risk of grain aliasing.
    I bought the betterscanning holder, do not regret that purchase. Bought at framing store some lightly frosted glass (a.k.a. anti-glare glass), cut to size to use in the betterscanning holder as AN glass. I scan with emulsion down, apply mirror in scanning soft.
    Good luck
     
  20. I have not optimized my V700 yet.
     
  21. Thanks.
    The film is flat as possible no real curl, it did curl but I corrected it. Not as good as ANR I suppose. Thanks also for the sharpening. I've had an extra pair of eyes and it seems optimal at double sided tape and 7 layers of copy paper.
    I suppose it is ok when viewed by others. After seeing what a Coolscan can do I just know there is something lacking in the scan.
     
  22. Does the scanner change it's focal point between scanning film where it is raised with the holder vs. a print where it sits on the glass? If so, how would a machinist scale help as described above to determine optimum focal point?
    Is it different between models as I have a V600 not a V700?
     
  23. There is absolutely nothing wrong with the sharpness attained from the V700. There is issues with their software and holders. If anyone wants to take the V700 serriously, start wet mounting and learn VUScan. The results are incredible.
     
  24. Does the scanner change it's focal point between scanning film where it is raised with the holder vs. a print where it sits on the glass?​
    The V700 has different focus points (different lenses, actually) when scanning the full format of the glass plate (against the glass) or smaller formats (4x5", 120, 135) in a holder, a few mm above the glass, as determined by the holder. That is for the V700. Other models I don't own, won't say.
    If so, how would a machinist scale help as described above to determine optimum focal point?​
    Imagine the "0" of the machinist scale is resting on the glass and a 5-mm block lifts the end of the scale at the 200mm mark. Scan this objet, making sure the scanner (if a V700/800) is in the "film holder" configuration. See which scale marking is sharpest, or center of sharp zone. Apply proportion: e.g. if the middle at "100mm" is sharpest, optimum focus is at (100/200)*5=2.5mm from glass.
     
  25. Does the V600 have different focal points between
    scanning a print and film?
     
  26. No. http://www.epson.com/cgi-bin/Store/jsp/ProductCategory.do?oid=-16223&UseCookie=yes
     
  27. One thing to remember scanning medium format with a flatbed scanner is that at 3200 ppi you get a really big print size at the standard 300ppi, like over 25 inches. You can then downsize to a smaller size and get a much sharper print by the downsizing. As others have mentioned, medium format scans on a flatbed always require sharpening, and often, the film isn't as sharp as you would think when viewed at 100% anyway. I have had great results with my older Epson 2450 flatbed film scanner and Vuescan pro. If you look at the dust on the neg often it will be quite sharp even though the image is not, indicating the actual image is not as sharp as you thought, though it rarely matters once downsized and sharpened.
     
  28. +1
    Another thing I used to do is scan at 6400 (full res) and downsize the save in Vuescan. Letting the scanner actually scan at full res gives clearer images when downsized to match the size of a lower res image. Only when scaning 4x5 and above does this make no sense.
     
  29. @ peter carter: not sure that you gain from scanning at 6400dpi; the V700's optical resolution is about 2500dpi, and scanning at 3200dpi extracts all the available information; that is based on my personal slant-edge MTF tests. Furterhmore, scanning at 6400 increases scan time and file size considerably. Especially if saving in raw tiff in order to profile outside the scanner software.
     
  30. Says who? I've always gotton much better results than that.
    Here is a scan of a reversal frame I took a while ago. It was scanned on a PrimeFilm XA (which scandig thinks the light shines out of it's ass). Click on the picture and it will go up in res. Further down in the comments I show the original scan I did (wet mounted, V700). Click on it and you can blow up the image in the same way. Aside from levels and curves, they are identical.
    https://www.flickr.com/photos/peterbcarter/27236081853/in/dateposted/
    ScanDig says the XA is about 4000dpi and the V700 is about the same as you stated.... I think they have ther head up their ass's about one or both. I think both. They certainly love to slam flatbeds.
    The reality is if you don't have a quality neg/pos to scan, you are going to get a soft image. If you don't scan to get the most info out of your film, you are not going to get all the resolution. If you don't think what I have suggested works, you are blind.
     
  31. The reality is if you don't have a quality neg/pos to scan, you are going to get a soft image. If you don't scan to get the most info out of your film, you are not going to get all the resolution. If you don't think what I have suggested works, you are blind.​
    Peter apparently you are not aware that one's eyes are better than any optical instrument when it comes to resolution of a scanner. :D
     
  32. :) I am. Aperently some have not seen what they are talking about. :)
     
  33. This is a 50% screen-shot of a scan from a V700. Film was 4x5 Polaroid Type 55, scan from negative.
    00e1ff-563997584.JPG
     
  34. Says who?
    If you don't think what I have suggested works, you are blind.​
    You could have made better use of the bandwidth by stating what were for the pictures (and yes they both look sharp) you link to:
    • negative size
    • dpi (at negative) of files as shown
    Just to be clear, I also feel the V700 is too often dismissed just because it is a flatbed. But I prefer to argue on a factual basis than to throw around implications re: the disabilities of other party.
     
  35. It was a 35mm reversal. The Primefilm XA scanned at 5000 dpi with vuescan (max) and the V700 used 6400 (max). The Primefilm XA image was uploaded @ 4698 x 6917 and the V700 was 3180 x 5052.
    I normally (as a rule) let Vuescan reduce by half on save (one third on MF) for pratical storage and 13x19 printing. If I have something special to do, I rescan and produce a full size image. The XA image was actually saved to a jpg at %100.
    Out of the box, both these scanners were out of focus. This gives the ilusion of low DPI on both units. I took the time to figure it out and got results. The reviewers obviously have their bias and everyone seems to have been convinced that flatbeds are evil. I go a little nuts when people start reciting these errored conclusions as gospel, with no experiance with the units in question.
    If you look at this.....
    [​IMG]
    These are scans of the same USAF targets that ScanDig uses to figure out their version of DPI. On the right, is scanned as directed by the manufacture. ScanDig rates this scanner as the second coming and directly compares it to the Nikons. On the left is me scanning the neg upside down (and flipping it digitally). More DPI? Nope, just focus. You can't really trust these reviewers. 10 mins with a scanner does not give them any expertise on the unit. It took a month (back and forth) with the manufacturer (they always insist there is nothing wrong) to actually figure out what was wrong with the unit. When I found out what was actually wrong and knew a workaround, I kept the unit. It's a known evil which is always better than an unknown one.
    Soo, the V700 is a fine unit and I would still be using one if mine hadn't had a catastrofic physical falure. I just didn't have the $$ to replace it. In time I will. I am stuck with 35mm in the meantime. Yes, this is an endorcement of the V700.
     
  36. Years ago, View Camera Magazine did a very detailed test of resolution, dynamic range, etc with the V700. No matter
    how focus was adjusted, they never got more than about 2600ppi. They did recommend scanning at 6400ppi when
    possible as they did find minor benefits...but based on what I've done, I agree with the 2600ppi figure. For 110, 35mm,
    and MF, I scan at 6400ppi. I then immediatly down sample the MF and 35mm to 3200 ppi and then process.
     

Share This Page