Ghosting

[jorgenm]

Greetings,

I've had this annoying problem for a while, and it's really getting me frustrated.
*sometimes*, when doing more than one scan for a single picture (either multiscan or multiexposure), I get ghost images. It seems like the scanner doesn't start scanning from the same position.
I've had some luck with the "delete current frame" option but sometimes not even that'll do.

I'm running:
Plustek OpticFilm 7300 with the newest 3.1.0.1 driver (w/ Silverfast SE 6.6.0r1); 3600dpi; Kodak Pro Image 100; stand alone mode.

This on:
Athlon 64 3200+; 1GB RAM; WinXP

There are other quirks, like the whole system freezing when scanning or having a white strip on the upper part of the image (although the film is correctly placed in the film holder), but my biggest problem is the ghosting part.

A sample of a typical output: (multiscan x4)

Any help would be greatly appreciated.

[LSI_Muenier]

Dear jorgenm,

I am not aware of such a problem.
Did you make sure that your original is fixed in the scanner and not movable?
What about your USB connection (thinking of your freezes) -- is it short enough?

Best regards
Martin

[jorgenm]

Did you make sure that your original is fixed in the scanner and not movable?
What about your USB connection (thinking of your freezes) -- is it short enough?

At first I also thought this might be the problem but I don't think this the reason since neither the scanner nor the film is touched after hitting the scan button. Besides, I've noticed that when this starts to happen, each time I do a prescan (without touching the film holder), it gives me slightly larger or shorter images (in the y axis). I will try to post pictures the next time I get this error.

Regarding the USB connection, I've thought about it. The connector hasn't given me trouble with memory sticks that I'm aware of, but I can't rule out a flimsy front USB connector. I now have the scanner connected to a back port, just in case. I'm using the cable supplied with the scanner, by the way.

Cheers.

[jorgenm]

Since my last post, I used the scanner only a couple of times -scanning around 18 negatives total- with good results. However, today, on my third photo, I got the "ghosting" problem again:
http://www.petaimg.com/uploads/1216231951.png
The process was
- the second image was correctly scanned. Then I "Delete the current frame"
- Choose a DPI of 3600
- Do a prescan
- Adjust the frame
- Click on "Image Auto-Adjust"
- Select the darkest and brightest place with the "Highlight-Shadow Tool"
- Select a Multiscan 4x
- Scan

As you can see,
a) It scanned above the selected frame
b) Each of the four scans started on a different place (as can be noticed by the dark lines)
c) There is an empty white space above the image (even though the negative fills the window in the holder almost completely)

Any thoughts regarding this issue will be very appreciated.

[LSI_Muenier]

Dear jorgenm,

when you observe this phenomenon, and you immediately scan again -- does the same happen again, or is the scan okay then?
Did you observe this at different resolutions or at the highest one of 3600dpi only?

Best regards
Martin

[jorgenm]

I scanned another negative after the one I posted, and it came with the same problem, and it's happened also before (after hitting this problem, subsequent images turn bad).

I haven't tried with lower resolutions except, I think, 150, which I kind of remember having been ok. I once noticed the problem with 5000.

I will try more resolutions when I get it and post results.

Cheers.

[jorgenm]

Yesterday, scanned 5 negatives which were ok (including the last one from the day before yesterday). Then the 6th came out wrong (the first one giving me trouble from the day before yesterday). Then I did several tests (after each "delete current frame" action, I did a prescan and framed the image, except with the first one):
- As it is, moved the slide to 2400dpi -> came out blue-ish and with slight "ghosting"
- "delete current frame", Multiscan 4x, 2400dpi -> still blue, ghosting is worse
- delete the frame, 4x, 1800dpi -> perfect
- again, delete frame, 4x, 1800dpi -> looks slightly blurred against the last one, but ghosting is not obvious
- delete frame, 4x, 1440dpi -> blue, but doesn't look blurred
- delete frame, 4x, 3600dpi -> looks good.

Today, same negative as yesterday and the day before
- delete frame, 4x, 3600dpi, Highlight-Shadow Tool -> yellow, but no ghosting
- delete frame, 4x, 3600dpi, Highlight-Shadow Tool -> perfect
- Changed to another negative (of the same roll). Delete frame, 4x, 3600dpi auto -> one of the four scans is way off
- delete frame, 8x, 900dpi, Highlight-Shadow Tool -> at least one scan is off. Also, as in the image I posted before, parts above the frame I selected can be seen. http://www.petaimg.com/uploads/1216435714.png
- delete frame, 4x, 720dpi -> looks good, but the upper part of the image shows lines of repeated pixels: http://www.petaimg.com/uploads/1216492436.png
- delete frame, 4x, 1200dpi -> slight ghosting
- delete frame, 4x, 3600dpi, Highlight-Shadow Tool -> at least one of the scans is off

[LSI_Muenier]

Dear jorgenm,

if I understand your experiments right, then the same slide comes out differently with no obvious connection to the settings (resolution etc.) in SilverFast, but it possibly seems to be a scanner hardware problem.

Could you check this again by simply scanning one and the same "ciritcal" slide several times after another with the same settings (resolutions etc.)? This should show if identical parametrization leads to the same or different results.

Best regards
Martin

[LSI_Muenier]

Dear jorgenm,

I once again tested Multi-Sampling of negatives with this scanner under WinXP, but could not reproduce your ghosting results.
All partial scans were perfectly aligned to each other, so that the final image was okay, too.

Best regards
Martin

[jorgenm]

Well, I just scanned several times the same negative and, indeed, every scan looks diferent. Some look yellow, some blue, most of them have the "ghosting" stuff...

I'll contact the people at plustek now.

Cheers

[LSI_Luebker]

Dear jorgenm,

I would also agree with Martins opinion about the problem, it really appears to be a Hardware Problem.

best regards,

Thomas Lübker
- LSI Software Support Specialist -