PDA

View Full Version : Weird patterns on scans of long exposures



Mark McCarvill
23-Jun-2005, 17:09
Hi folks. I've started trying 5 sec+ exposures on TMX 100 4x5 and I'm getting some weird results when I scan them. On some I'm seeing a vertical wave pattern (when I'd expect to see only a horizontal one due to cloud and water movement) and on others I'm seeing a grid pattern. Easier to show than tell – here are two examples:

http://www.stepwise.ca/images/Grid_pattern.jpg

http://www.stepwise.ca/images/Vertical_pattern.jpg

Please ignore the Newton rings – I'm scanning with a cheap Microtek flatbed that I use just for proofing.

Scans of shorter exposures look fine.

I can't figure this one out. Any ideas?

Thanks!

Michael Veit
23-Jun-2005, 17:33
I'm scanning with a cheap Microtek flatbed

That could be your problem right there: only expensive scanners can tell how long the exposure was and adjust accordingly.

Seriously, there's no way length of exposure could have anything to do with your problem unless it's caused you to blow out your exposure. I know my scanner has problems with high density area of negatives.

Ted Harris
23-Jun-2005, 17:42
Thefirst one looks to me like it could be caused by dirt in the scanner control path......the skinny area either above or below the main scan area. Which model Microtek?

Mark McCarvill
23-Jun-2005, 18:24
Michael, I will scan the negative on a good scanner and see if the pattern persists. Still, I'm curious to know what’s causing this grid pattern.

Ted, it’s a Microtek Scanmaker 5900.

What about the vertical waves (they look sort of like flames) ... are they normal? I thought the cloud pattern should be more horizontal. Or should I just claim this was purely intentional?!

Ted Harris
23-Jun-2005, 18:35
Mark,

Hard to make any judgement at al without seeing another scan from a much better scanner. Consider that the Microtek i900,Epson 24503200/4870/4990 or Canon are usually considered the 'entry level' or 'consumer' scanners for scanning LF and you are using a scanner that sold new for less than 100 (25% of those mentioned) and goes for 20-30 on the used market. AIt likely just isn't up to the task but send us another scan before we pass any real judgement.

adrian tyler
24-Jun-2005, 00:06
why not make a "real" print at a "real" lab, i find those things indispesible references, even if i have already scanned or am going to scan, for one copy it shouldn't be that expensive and it'll help you work out your exposures.

...

domenico Foschi
24-Jun-2005, 01:43
It has nothing to do with the negative.
I have dealt with those patterns for sometime and i decided to send important scans to a good lab.
From what i understand , the grid patterns are a fairly common phenomena .
Almost all scans have them, the problems arises if they are visible at 200% zoom-in or less.
I can see grids in drum scans done from A&I here in LA, a very reputable lab, at about 340%, and are not visible whatsoever in an hard print.
Struan Grey, a member of this forum , explained me once that the patterns are caused by electrical interference
by sources in proximity of the scanner .
One way to reduce this problem is scanning to a multiple of the initial scan, don't ask me way.
Good luck.

Mark McCarvill
24-Jun-2005, 08:05
Thanks for the feedback, folks. I will rescan on a good scanner and post the results for the curious.

Domenico – interesting point about grid lines being in most scans. Perhaps that’s the answer, and the reason the lines only appear in these long exposures (and only in the sky and water portions of these images) – because they have large areas which are the smooth and show little contrast, thus providing an ideal background for the lines to show through.

Ed O'Grady
24-Jun-2005, 10:00
Mark, the second looks to be exibiting the results of lack of agitation during processing. That mottled appearance is pretty typical from lack of agitation or exhausted chemistry.

Mark McCarvill
24-Jun-2005, 10:54
Thanks, Ed. I called my lab and they say they do test strips every morning but maybe my film was processed in the afternoon when the chemistry was getting exhausted. I will try a one-shot chemistry lab next time.

Edwin B.
24-Jun-2005, 12:52
I've seen a grid pattern in drum scans from a professional lab and was told that it's normal and that I was getting the best quality possible. I attributed it to the fact that the lab scanned at 4800 ppi with a scanner that had an optical resolution of 4000 ppi; in other words, an interpolation problem. It could have been an electrical interference problem but that isn't really a very good excuse for a "high quality" lab. With my own drum scanner (same model) I have never had this problem although I only use optical resolution and have it plugged in through a voltage regulator and RF filter.

Struan Gray
3-Aug-2005, 05:29
My guess is that in this case the problems are inherent in the scanner. Grids and stripes from external noise or poorly-shielded power supplies tend to wander about a bit like sand ripples. The highly regular grid in Mark's first example is almost certainly an interpolation artifact from a crude interpolation routine. The solution is to scan at simple multiples or ratios of the scanner's native resolution and do the interpolation yourself in your photo editor.

I have seen the second sort of blotchy effect in my own flatbed scans of nighttime or twilight exposures. My best guess is that it is caused by slight variations in the background signal reaching the CCD, either caused by electrical noise, or more likely, by crud on the underside of the scanner glass. Try scanning a colour filter or a simple piece of coloured cellophane and use your photo editor to remove the 'colour cast'. If you see similar patterns, it's not your film.

If you do the last trick, you may well also see hot pixels in the form of dotty lines of one or other primary colour (with my Epson 3200 it's always red). These too show up in my scans of low-contrast originals. In theory you should be able to remove them with a calibration scan or a single 'dark frame', but the scan-to-scan repeatability of these units makes that a futile exercise.

My solution has been to accept that my 3200 is an excellent proofing tool, and capable of making great prints from normal negatives and slides, but that I will have to send out for 'real' scans for the best resolution and tonal reproduction. I take a lot of my LF photos at twilight or after sundown, so it's a real frustration. In B+W one can increase contrast with N+ development, but in colour a change of contrast means a change of film, and thus a change in feel. Sometimes cheap tools just don't work.