resolution for archiving 35mm slides (Epson V700)

[gour_d.]

<p>@Kelly:</p>

<blockquote>What matters is the results off of your scanner; not my neighbors V700 thats good to 2400; or anothers results at a zillion dpi.</blockquote>

<p>Hmm, I do n't have much experience with serious scanning (besides doing simple reflective scanning for everyday's use, but I wonder if my neighbors V700 could be <strong>much</strong> different than mine?</p>

<p>Sincerely,<br>

Gour</p>

 

[kelly_flanigan1]

<p>Gour; just do some sample scans at different dpi scan setting and you be the judge, With any scanner of any age there is no reason to go beyond what either scanner does; or what the original really holds.</p>

[yoshio]

<p>Gour D<br>

to achieve optical resolutions of 6400dpi (dividing an inch into 64000 even segments) a scanner needs to be a precision optical instrument. Such things are usually in need of adjustment and calibration if they are to remain as such. Also manufacturing tolerance would need to be very high and also be hand tuned. Precision equipment often does not like to be moved, so after transport to your home calibration would need to be done in situ.</p>

<p>Due to manufacture tolerance it is quite possible that your model differs from your friend up the road.</p>

<p>The flatbed scanner shifts focus (internal parts) between document and film scanning mode, I doubt that they return to each location with such precision (in deed my observations are that they do not). Take a stack (say 5 or 6) glass microsope slices and write on one end of them with a OHP permanent marker. Stack so that overlap happens to show the bottom of each higher slide. Scan in one mode and observe, scan in the other mode and observe. Repeat a few times. I have not found that focus returns to precisely the same location each time.</p>

<p>does your room temperature vary? If so then the plastics will likely create a calibration difference too So one would need to recalibrate for tray height every time the scanner is used as the season changes room temperature.</p>

<p>experiments such as this cost you little and confirm much for you.</p>

<p>I would like very much my scanner to be just that, but I do not believe it is true.</p>

<p> </p>

[matthew_newton]

<p>You may want to consider scanning in 24bit RGB mode. 48 bit has a huge amount of tonality information that your eye just can preceive and it increases the file sizes and scan times. I've tried scanning 24 bit and 48 bit modes and I can't preceive any difference at all, so I stick to 24 bit RGB scanning.</p>

[robert lee]

<blockquote>

<p>The slides will be used for producing video slide-shows (using Ken Burns style) and that why I need some 'extra' dpi to be able to zoom details.</p>

</blockquote>

<p>The highest HD resolution is only about 2MP. Scan at around 2000dpi. This gives about a 6MP image, plenty of margin to pan and scale across the frame, for video production.</p>

<p>Scanning to digitally archive is another issue altogether. You do need a good scanner like the Nikon Coolscans or better to record the bulk of image information. Common, high resolution film like Acros or Velvia will carry around 20MP worth of detail; 400ISO and faster film has usably maybe around 6MP worth.</p>

[gour_d.]

<p>@Robert: Video-usage is 'by product', the 1st thing is to digitize collection.<br>

However, all the slides are done on consumer Kodak film, no high-resolution nor high-speed.</p>

<p>Sincerely,<br /> Gour</p>

<p> </p>

[gour_d.]

<p>@Matthew: That's true, although both scan time and file sizes at 48bits (I was testing with 64bit to include IR channel) are acceptable with 3200dpi.<br>

4800 & 6400dpi makes the huge difference.<br>

I was busy yesterday with other stuff and will resume testing today to find out about optimal height of holder's feet first before deciding on resolution.</p>

<p>Sincerely,<br>

Gour</p>

<p> </p>

[robert lee]

<blockquote>

<p>However, all the slides are done on consumer Kodak film ...</p>

</blockquote>

<p>Gold 100 or 200? You'll want to use a Nikon or better then. These film are actually quite high resolution if grainy. <a href="../film-and-processing-forum/00HEsT">This thread </a> has a few scans of Gold 100. Scroll down to the beach scene crops.</p>

<p>I use a 4490 and a V500 flatbed. I expect the V700 would be capable of real higher resolution scans, but probably not significantly more. For what it's worth, I did a comparison scan of the 4490 and the Nikon 5000 in <a href="00J1d2">this thread</a> .</p>

[gour_d.]

<p>I believe most of the stuff was Gold 100 with some Fuji 1oo as well.</p>

<p>Well, I own V700 and have to do the best job possible with it :-D<br>

Thanks for the links..</p>

<p>Sincerely,<br>

Gour</p>

 

[bryan_wolf]

<p>If you haven't finished yet, something to consider is that consumer flatbeds have a lot of ccd noise in their scans. One huge benefit of scanning at say 4800 and down-sampling to 2400 is that there will be significantly less scanner noise in dense parts of the slide/negative. Vuescan has a built in .tiff size reduction feature that while not as good as bicubic sharper, still works surprisingly well at cleaning up slides and effectively giving you more d-range.</p>

<p>To see the results take a scan at 2400spi into your photo editor and apply a very strong curve so that all the noise is visible in the shadows. Now take a scan at 4800 with 2x size reduction and apply the same curve. What you'll find is that the down-sampled image has far less noise and far better shadow/highlight detail. You may also want to compare this to the multi-sample feature at say 2 or 4x. Most scanners end up blurry when using the mutisample feature because they use 2,3,4,... passes and then try to line them up. Consumer flatbeds are too imprecise to do this effectively, so you end up with very soft scans. However, I've heard that the v700 supports single pass multi-sample which takes multiple samples without moving the lens/ccd, then averages the values, effectively eliminating noise. The results should essentially be the same as scanning at a higher resolution and down-sampling, but it may be faster to scan @2400 with 2x multi-sample than scanning at 4800->2400. You'll have to test it yourself.</p>

<p>I do have to agree with Yoshio with regards to the preciseness of consumer scanners. You may be able to coax 2400 out of a consumer scanner, but there will be no comparison to the same scan at the same dpi on a professional flatbed. Even one that is 8-10 YEARS OLD. There is a reason why a Creo Eversmart weighs 156lbs vs a v700 at 14lbs or so and cost 30x as much new. If you can deal with all the shortcomings of buying these older scanners(repair costs, legacy computers/operating systems, limited support, Size/weight, etc.) they are well worth their hugely discounted prices on the used market and offer great value if performance is your #1 priority. </p>

<p>This for example, should blow the doors off any Epson:<br>

http://cgi.ebay.com/EVERSMART-JAZZ-COMPLETE-EXCELLENT-CONDITION_W0QQitemZ230331635559QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item230331635559&_trksid=p3286.c0.m14&_trkparms=72%3A1205|66%3A2|65%3A12|39%3A1|240%3A1318|301%3A1|293%3A1|294%3A50</p>

<p>That being said the V700 can pump out some surprisingly good scans if you're willing to play with the files a bit. And since you've already got one, you might as well use it to it fullest. Good luck with all those images! I don't envy you. : )</p>

<p> </p>

[roy_montgomery]

<p>I am new to scanning, with some of the same questions as Gour D., so I am very glad to read this thread.<br>

My question...is there a no-charge (or relatively inexpensive) standard test pattern available that can be placed on the flatbed and used to determine scanning resolution?<br>

Roy D.</p>