[www.betterscanning.com] Film scanning mounts, any good? Your experience and advice??

[Brian Ellis]

. . . Given that these Epson scanners have limited resolution (approx. 2300 spi for the v700 and 1600 for the 4990), it's best to get the full resolution out of them - especially when using Medium Format or 4x5 film. . . .

Others haven't found this kind of difference between the 4990 and the 700/750. The late Ted Harris' tests found them to be essentially identical, both right at 2000 ppi. See his August 30, 2006 post in this forum, in which he said in relevant part:

"We have tested the 4870/4990/V700 (V700 scanning at lower than the 6400 setting) and the improvements all seem to be related to things other than resolution and DMax. The resolution is virtually the same and the DMax, which I recall improves marginally from the 4870 to the 4990 doesn't improve at all on the V700. What Epson is doing is stacking one chip on top of another so that they can claim a continuing higher resolution every time they add another chip. It is an honest theotretical claim but has no real world validity as we all now know. . . . Yes, you will get nearly 2000 spi when setting the 4990 or V Series to 3200. . . . Finally, given the low price of these scanners I a (sic) delighted by the quality of their performance rather than dismayed that they are not delivering more."

I don't really care one way or the other and since I haven't done my own tests I don't want to argue about it. But given the way relative offhand statements here sometimes become gospel, I just think it should be noted that not everyone has found a 1600 ppi maximum with the 4990 or a 700 ppi difference between the 4990 and the 700/750 scanners.

[Ken Lee]

"not everyone has found a 1600 ppi maximum with the 4990 or a 700 ppi difference between the 4990 and the 700/750 scanners."

True enough.

Other than Ted Harris, has anyone else demonstrated greater than 1600 effective spi with the 4990, or demonstrated that the 700/750 scanners deliver an effective resolution other than ~2300 spi ?

One could make the claim that both scanners deliver 9000 effective spi - but it would be an exceptional result - rather than the norm. We would have to consider it as an anomaly, rather than a vindication.

Since Ted's test images aren't available - and since he is no longer with us to elaborate on his methodology - we might want to consider his results, respectfully, as just that: an exception to what others have found and shared.

Here are links to my comparison tests: raw scans and sharpened. I have not shot a USAF target, but my tests on the 4990 accord with that of the others: resolution peaks somewhere between the 1200 and 2400 settings, and does not improve after that. In fact, the only compelling improvement among the scans, is seen when we go from 800 to 1200 spi. After that, you really have to squint.

[Brian Ellis]

"not everyone has found a 1600 ppi maximum with the 4990 or a 700 ppi difference between the 4990 and the 700/750 scanners."

True enough.

Other than Ted Harris, has anyone else demonstrated greater than 1600 effective spi with the 4990, or demonstrated that the 700/750 scanners deliver an effective resolution other than ~2300 spi ?

One could make the claim that both scanners deliver 9000 effective spi - but it would be an exceptional result - rather than the norm. We would have to consider it as an anomaly, rather than a vindication.

Since Ted's test images aren't available - and since he is no longer with us to elaborate on his methodology - we might want to consider his results, respectfully, as just that: an exception to what others have found and shared.

Here are links to my comparison tests: raw scans and sharpened. I have not shot a USAF target, but my tests on the 4990 accord with that of the others: resolution peaks somewhere between the 1200 and 2400 settings, and does not improve after that. In fact, the only compelling improvement among the scans, is seen when we go from 800 to 1200 spi. After that, you really have to squint.

I really don't know whether anyone other than Ted Harris has found greater than 1600 ppi with the 4990. And I don't care enough to research it. Ted was generally considered a pretty reliable source of scanner information and test data and I'm happy to go with his results.

As for his results being "an exception to what others have found and shared," the numbers thrown around here relating to ppi of the 4990/V series have been all over the place so I'm not sure who is an exception to who. Just in two recent threads I've seen one person claim 2300 for the 4990 and another who claimed something like 600 for the 4990 or the V series, I forget which.

My point wasn't to argue over who is right and wrong. My point was that just because you've finally settled in your own mind on these 1600/2300 numbers doesn't mean they're universally accepted truths.

[sanking]

People can claim whatever they like and indeed you are likely to see all kinds of reports on the web.

I have personally owned and used both the 4990 and the V700. I tested both with resolution targets. If you compare the 4990 to the V700 (using the lens that engages on the V700 when you choose film area guide) they give very similar results, about 1600 spi, maybe 1800 spi if you have optimistic eyes. If you compare the 4990 to the V700 (using the lens that engages when you choose film holder) the V700 is clearly superior in terms of resolution, with about 2300 spi. I also tested two other V700 scanners, with the same result. I have posted my results from the 4990 and the V700 on this forum.

Similar results to mine for the V700 can be seen at http://www.filmscanner.info/en/Films...tberichte.html. Several others, Nate Potter for example, have show similar or even better resolution from the V700.

I have never seen any real test results with a target, by anyone, that shows resolution from the 4990 over about 1600 spi - 1800 spi.

Bottom line for me is that the V700, if you use the better of the two lens, is a significant improvement over the 4990 in terms of resolution. And that is an opinion derived from a lot of experience with both scanners.

Sandy King

[Nathan Potter]

Sandy, good point above. I used the film holder, engaging the better lens, for the above plot and scanned at 2400SPI. The technique is fairly simple. I lay a glass resolution mask directly on the scanner glass platen with one end touching the platen and the other end suspended above by several mm. I just use a shim on the suspended end. Then I scan with no sharpening or other software functions. The image of each line pair cell is then expanded in PS (on screen) to about 100X and examined for a condition where there is a complete line of clear pixels between a complete line of black pixels. This is somewhat arbitrary in that one can change the individual pixel density using numerous PS functions. But I try to adjust the contrast to what I would use in printing the average image. I think this complication of interpreting the resolution is just a manifestation of the resolution/contrast (MTF) problem, and why MTF kind of data is so much more definitive.

I was initially concerned that the sloped glass plate technique would cause some inaccuracy in the results but a check using the same resolution plate parallel to the platen and at the sweet spot in height above the platen showed essentially identical results.

I think some of the difficulty here is the lack of a consistent methodology between different users when attempting to estimate scanner resolution. This especially in how one reads and interprets the data from the scanner. This is digital data that is read out using software and the conversion from scan to a viewable resolution image is somewhat of a black hole.

The scanner I used was new in June 2010 (V750Pro) so could there possibly be some recent hardware upgrades? I still would conclude it fair to say the actual max. resolution for my scanner is in the 2000 to 2500 SPI range.

BTW scanning at 3200SPI showed very slight difference (got to find the data) but < 10% improvement. At 6400SPI no improvement over 3200SPI. I'm satisfied using 2400 SPI for chromes yielding about a 260 MB file.

Nate Potter, Austin TX.

[Ken Lee]

On a related note, one way to improve the resolution on the 4990 - and I presume on the 700 series as well - is to scan in RGB, and use the Green channel only.

Of course, this doesn't help when scanning color media, but for b&w, there is a discernible improvement.

Here's a link to a short article with sample b&w images, scanned at 1200 spi in RGB. Each channel is shown separately, before and after moderate sharpening.

Dust and scratches, while normally removed, can be very helpful for testing purposes.

[Richard Mahoney]

btw Richard, I really like the photograph on your home page. I wonder if those chimneys are still standing?

My apologies for the delay in answering but I've only today gone up country to see how the place has been faring during the recent run of earthquakes. Probably built sometime in the 1860's the cottage has seen a bit. Sad to say that either the big shock on September 3rd or one of the after shocks took down the kitchen chimney (far right). It was held together with lime mortar -- not cement -- so shouldn't be too hard to rebuild. Prospects are probably better than many in Christchurch:

Chimneys and people come unstuck
http://www.stuff.co.nz/the-press/opi...e-come-unstuck

Fingers crossed that this wasn't the last shot ...




Kind regards,

Richard

[Lachlan 717]

Anyone tried replacing the plastic clip holders on the V700 with a piece of ANR glass that fits in the film space? Referring to the 120 holder...

Seems my biggest issue is film flatness with pano strips.

[wclavey]

Lachlan, I use a 4990 and I had some museum glass cut to replace the snap clips on the MF film holder. I use it with film that has a really strong curl, like Foma 100. I put the emulsion side up and put the non-reflective surface of the museum glass against it. It works well keeping the film flat - - much better than the clips do. But it makes no difference when I am scanning film that already dries really flat, like Tmax 400, so I don't use it then. The non-reflective surface has never given me rings, but why risk it if I don't need it?

[Lachlan 717]

Lachlan, I use a 4990 and I had some museum glass cut to replace the snap clips on the MF film holder. I use it with film that has a really strong curl, like Foma 100. I put the emulsion side up and put the non-reflective surface of the museum glass against it. It works well keeping the film flat - - much better than the clips do. But it makes no difference when I am scanning film that already dries really flat, like Tmax 400, so I don't use it then. The non-reflective surface has never given me rings, but why risk it if I don't need it?

Thanks for that.

The long strips of 120 in pano shots seem to curl both laterally and longitudinally. Buggers of things to deal with when not flat!!

Add to that the fact that I'm impatient, and I need a better solution.