PDA

View Full Version : artixscan 1800f



phil sweeney
26-Jul-2004, 16:58
I am interested in any experience folks have with this scanner, especially for scanning 8 x 10 negatives. The scanner is glassless for all but 8 x 10, and supposedly a microtek rep said the anti-newton ring glass for the 8 x 10s takes care of 99% of newton rings.

Curious how often the newton rings show up?

Does the occurance of the newton rings increase at higher resolutions?

Paul Butzi
26-Jul-2004, 18:38
I have an artixscan 1800f, which I bought this past spring. I use it primarily to scan 4x5 negatives, using the scans to make inkjet prints on an Epson 9600.

I'm very pleased with the scanner overall. The glassless carrier for 4x5 is vastly better than the cheesy carrier that came with my Epson 1640su Photo. The scans are vastly better. I've compared scans from 4x5 negatives head to head against scans from the Imacon Flextight 848. The imacon scans are better, but not by a whole lot. A pretty minimal amount, actually, once you consider that the 848 costs $17000 and the Microtek 1800f cost $1100.

I've scanned a few odd format negatives on the 8x10 glass carrier. I don't know why the Microtek rep said the glass was AN glass - it sure as hell isn't on mine. I know that Clyde Butcher has replaced the glass in his 8x10 carrier to avoid NR problems. The few scans I've made on the glass definitely had NR problems. If I were planning on scanning 8x10's as a regular thing, I'd probably take the 8x10 glass holder apart, measure the glass, then call the guys at Focal Point to get some AN glass.

I don't know what it is about scanners - I don't think I've ever made a scan on glass and not had NR problems.

George Stewart
26-Jul-2004, 19:07
I had some on-glass 4x5 scans done on the 1800f (for comparison) and was not impressed. I too wanted the scanner for 8x10 so I didn't bother with glassless scanning. I found that the scans were good for up to a 3x enlargement and then turned to mush. The comparison scans were done on a Creo flatbed scanner - they were much better. I shoot 8x10 for quality reasons, and for me the 1800f didn't shine.

fishfish
27-Jul-2004, 07:47
I have a Microtek Scanmaker 8700 which uses the same approach to scanning - holders for smaller formats, and a glass tray for 8x10. I have been nothing but pleased with this scanner, and never experianced newton rings, except when scanning entire rolls of 120 (4 strips) which I sandwitched between glass...and I knew that was asking for problems.

As opposed to using the Microtek software, I use Vuescan, and this does a great job.

Ken Lee
27-Jul-2004, 10:23
As many have pointed out, scanner manufacturers publish the sampling frequency, rather than the actual results. The scanner may indeed sample at 1800 ppi, but deliver closer to half of that ideal number, in terms of actual resolution.

If the scanner acutally delivers 900 or 1000 ppi, then given an output resolution of around 300 dpi, you should be able to make a fairly nice enlargement (as Prof. Evens calls it, "critically sharp") up to around 3x, but no more. If you intend to send 360 dpi to the printer, as many do when printing on inkjet printers, then enlargment will need to be correspondingly lower.

Donal Taylor
27-Jul-2004, 10:45
"As many have pointed out, scanner manufacturers publish the sampling frequency, rather than the actual results. The scanner may indeed sample at 1800 ppi, but deliver closer to half of that ideal number, in terms of actual resolution.

If the scanner acutally delivers 900 or 1000 ppi,"

That isn't the case with the 1800f - it's actual resolution is somehwere between 1650 and 1800 in testing

Ken Lee
27-Jul-2004, 18:15
Mark - I'm happy to hear of a scanner which comes so close to the advertized reslolution. Could you post a few sample images, and discuss how you measured the resolution ? Perhaps my Microtek 2500f does better than I thought !

Ken Lee
27-Jul-2004, 18:20
According to an earlier post by Leonard Evens, "According to digital sampling theory, the maximum resolution in line pairs per unit length obtainable from a scanner is half the scanning resolution. ....No scanner will actually achieve the theoretical maximum, but a high quality scanner like an Imacon will come close.



Perhaps I have misunderstood, but that's why I am asking how you measured the resolution you mention.

Nathaniel Paust
27-Jul-2004, 20:46
I was looking at buying an 1800f before I realized that my budget couldn't handle it and did a lot of research on the scanner. I seem to remember reading that the scanner actually has an 1800 dpi ccd line array. It should be able to resolve something like 900 line pair per inch.

Now I just need to win the lottery.

Donal Taylor
27-Jul-2004, 21:40
Ken, using a circular resolution chart, among other things, you can measure how many sensors per line you are getting

The Epson 3200, for example, gave us around 1.7 sensors per line which is nowhere near the stated resoltion of 3200 - something under 2000dpi (if I've still got my maths right....)

The figures for the 1800f came out at around 1650 to 1720dpi (I think it was maybe around 1.5 or 1.6 sensors per line)

This is all going from memory now and I no long have all the figures. And my partner is the math/engineering whiz... and he's away on an assignment right now.

Hope that makes some sense.

Our experience was that the lower end flatbeds like the Epson etc have the figures bumped way up to attract the photogorpaher market.

The mid range, graphics scanners Mictrotek 1800f, Epson 1680 (?) etc tend to give much closer to real sensor resolution figures (generally the same for the DRange tests we did as well)

Paul Butzi
28-Jul-2004, 10:35
There's a terrible tendency when comparing scanners to assume that all pixels are created equal.

This is definitely not the case. Having the scanner produce more pixels does not mean that it is extracting more information from the negative.

Beyond the obvious problem that some scanners have pixel densities that are much higher than the the density the scanner optics can produce (and I'm thinking of my Epson 1640su Photo, here), there's also little details like noise, maximum scannable density, etc.

As far as I can tell, there is very little information in the literature that scanner manufacturers put out. Lots of claims, but very little real information. Claims of dmax, for instance, are often apparently just number picked entirely out of thin air. If you could get scanner manufacturers to take bets, you could make a dandy living just taking your step wedge around and asking them to successfully resolve the difference between the 3.0 and 3.1 density steps. Since they all now claim a Dmax of something like 4.7, it should be a piece of cake. None of them can actually do it.

The only way to compare scanners is to actually scan negatives on them and compare the scans. But you have to be aware that even when you specify 'no sharpening', many scanners still do quite a bit of sharpening, etc. This makes it very difficult to actually do apples to apples comparisons.

D. Kevin Gibson
28-Jul-2004, 10:58
"The only way to compare scanners is to actually scan negatives on them and compare the scans. But you have to be aware that even when you specify 'no sharpening', many scanners still do quite a bit of sharpening, etc. This makes it very difficult to actually do apples to apples comparisons."

Use vuescan as your basic software for all your comparisons - you can then basically turn everythign off (and it's the best software for running most scanners half the time anyway)

Paul Butzi
28-Jul-2004, 13:09
It's not clear to me whether Vuescan supports the 1800f or not. On the 'supported scanners' page, the 1800f is listed as supported. Then lower, under 'unsupported' it specifies that "All Microtek USB models except X6USB, X12USL, 4700, 5600, 5700, 6700, 6800, 8700, FilmScan "are unsupported. Well, the 1800f is a USB scanner.

Go figure.

In any case, it apparently does NOT support the Imacon.

tim atherton
28-Jul-2004, 13:30
I'm pretty sure we used it with Vuescan at work when we had one?

"Added support for Microtek ArtixScan 1800f - version 7.6.50"

Check with Ed Hamrick - he's pretty helpful.

As for the Imacon - yes, it's a different kettle of fish altogether. A friend uses one for his 5x17cm and 5x12cm panos. Knocks the socks of anythign else. He's never felt the need for drumscans anymore (+ he got it "cheap" off ebay :-) ) Though, of course, it's no good for 8x10

Jeffrey Zweig
29-Jul-2004, 20:58
The 1800f is supported by Vuescan and Silverfast as alternatives to the ScanWizard Pro software. We have used it in our studio as it produces superior transparent scans compared to the i900. It uses a combination of firewire and USB ports so that maybe what the confusion on Vuescan's website is all about.

To solve the newton ring problem and to keep your image completely flat we use a piece of anti-newton ring glass also from Focal Point. The 8x10 piece of glass fits perfectly between the glass tray and the transparency, sandwhiching it between the two. This is helpful because it avoids having to use tape or a chemical spray to keep the transparency perfectly flat.

Jeffrey