4990 Glass plate Negatives

[simon_fuller]

Can anyone confirm that they have had success using this scanner with Glass plate Negatives? We was going to purchase the predecessor but this one coul be the better option

[DannoB25]

I'm interested in the same application. The only other option is the ArtixScan 1800.

Note the review on this forum titled: Epson 4990 new owner appraisal for film

If I read his conclusion correctly, the 4990 delivers approx 2400dpi real resolution, all above that is due to 'sharpening.'

If you are doing 7x5 or 8.5X6.5 full plates at 48 bit tiff, that will deliver a big file.

If that is good enough for your negs, then the question is speed, bit depth, and noise in dense parts of the negatives. If the negs were developed to be contact printed for azo, platinum or albumen, they can be very contrasty.

Daniel

[Kevgermany]

Lots of guys used the predecessor 4870 for glass plates as well as xrays. Should work fine...

[honza]

We used Epson 4880 to scan roughly 5000 glass plate negatives (gellatine and collodion, ranging from 6x6cm to 13x18cm in size) and now switched 4990 (as this one scans 18x24cm too) to scan roughly 2000 more. We use vuescan (registered so you get 16bit). Both scanners works very fine producing just somewhat out of focus scans at 2400DPI (sharpening with radius around 2.7 is needed for our scans) allowing pretty gigantic reproductions. On 4880 after rougly 2000 scans joints holding cover broke and service refused to replace them within warranty. That eventually lead to breaking wires going to the cover that brought light off that was finally repaired within warranty together with the joints. 4990 seems to be slighly better in this respect as the joints are no longer clicking when being openned and is also noticeably faster.

We use cartoon frame to avoid scratches of glass plate that holds negative emulsion down in roughly same distance as film holders does. It seems superrior to 9x12 BW film scans as the matherial is not getting deformated during scan and image is sharper. Dynamic range is fine for both models I would say except for real extreme negatives that are grainy anyway. (Grain is bit problem as with any black&white matherial perhaps because of grain aliasing, but there is probably nothing to do about that. I use noise ninja when sharpening for print.) Only case where I run into density issues was scanning old agfacolor slides. Infrared cleaning won't work, but you probably know that. When the glass plate has ugly yellowish spots caused by acetic papers, one can scan the infrared channel and those spots are usually invisible in it (registered vuescan allows saving infrared channel into separate raw tiff)

The negatives usually needs hand tweaking of curves in photoshop to come out right as the density differs from scanner software expecations about BW films.
You might see our archive at http://sechtl-vosecek.ucw.cz . In the digital archive the previews are automatically generated without adjustments, other images are mostly hand edited.

We also experimented with Cannon and Microtek scanners and Epsons seemed to be most suitable (Cannon had mostly problems with software, but vuescan is available now I think), Microtek (Arixscan 1800) just seemed slower, not sharper nor less grainy or deriving any benefits from better density with significantly higher price. It seems to work somewhat better on color slides however and perhaps it won't fall apart after 2000 scans

Honza

[bookmom]

We use cartoon frame to avoid scratches of glass plate that holds negative emulsion down in roughly same distance as film holders does.

What do you mean by "cartoon frame"

Thanks,

Angela

[Doug Fisher]

They probably mean the thin cardboard sandwich type frames used to hold individual animation cells. Basically imagine a huge 35 mm slide mount/frame

Doug

[Kevgermany]

Doug, I'm guessing they mean something thinner - like thin card at about 2-300gsm. Much like we've discussed before. This would tie in with the german Karton, which may well be used in the Czech republic (it's next door to us...). Would also work better as it's thinner.

[bookmom]

Thanks - I was thinking of making a frame using matte board and another from card stock and see which resulted in a better scanned image.

Angela

[honza]

Hi,
We really use "karton" as in germany, so it really is cardboard. In general I don't see much difference in thickness of matherial the holder is made from. Our one is about as thick as CD medias. This is about the distance used by film holders, but it looks like the depth of focus is not too problematic here (ie I originally experimented with few different distances and the results was about the same. Major point is keeping the glass clean and without scratches - the glass glass negs are made from seems to be hard enought to scratch the glass plate in scanner when you are unlucky and it falls down somehow on corner).

Honza

[honza]

Since I complained about the joints in 4870, I shall also mention that we are now after another 2000 scans on 4990 and it don't seems to be decomposing . Since the joints are different, I would say that Epson fixed this problem (rather than that we was unlucky with previous unit/used it incorrectly).

We however got some fogging of glass mentioned elsewhere I don't see on 4870. On glass it is visible only with the light in scanner on when you look from angle and it is also barely visible when you scan empty frame. I would say that it is not really issue (yet) and I will try to get unit replaced if it becomes apparent in real scans. Overall due to high contrast of old negs, scans are much more tolerant to dirt than with modern matherials (I am scanning negs from 1960s on Imacon too now so got some experience with cleaning them) or slides.

Honza

[Jaycey]

Does anyone know whether this scanner will scan glass negs measuring 4" x 5"? (If so, would they fit in the holder, or just scanned as they are?)

[Kevgermany]

yes, take a look at Honza's last post here http://www.photo-i.co.uk/BB/viewtopic.p ... highlight=

[honza]

Yes it does fit. You can take advantage of the wider scanning area compared to 4870 or v700 and place them horisontally so they scan faster (more pixels are scanned at one shot).
Vuescan also allow scanning two of them at once if you are patient enough to play with the multiple crop features (I find it easier to just can one by one).

Honza

[europanorama]

new hp g4050 is cheapest- has autofocus and can enlarge at least 20x. razorsharp- look at the flatbed-scanners.

[honza]

This ineed sounds like really interesting scanner. Before ordering it, I would ensure that the autofocus logic won't tend to pick the wrong side of glass plate (this was problem on some pre-press scanners I tried. They apparently used some weird trick for focusing instead of simply maximizing contrast). I wonder if it can be competetive with film scanners (Nikon/Imacon/drum) for medium/large format film...
Honza