iSRD doesn't work properly

[John3:16]

I'm using Silverfast SE Plus 6.6 with CanoScan 9950F. When I use iSRD it properly identifies the dust particles, but the applied fix is offset such that it has no effect. I first noticed that iSRD had no effect. Checking a little closer I noticed that after the preview scan and auto button I clicked the button to show the corrections and I can see the red marks where the corrections are occuring and they are just next to the actual dust specs, but not covering them. I'm guessing that the infrared scan is not lining up with the other scan(s). Is there a fix for this problem?
Thanks.....

[LSI_Luebker]

Dear Customer,

we just updated the Canon version.
Please donload the new SilverFast version 6.6.0r4d this should solve this issue.

best regrads,

Thomas Luebker,
LaserSoft Imaging AG

[John3:16]

Hello,
Thanks for the prompt reply. I just downloaded and installed the update then did a test scan. Sadly, the problem seems even worse, the corrections appear to be offset even further than before. The higher the resolution of the scan the worse the problem. Was this specifically addressed in this latest release? Am I possibly doing something wrong?
Thanks,
John

[LSI_Luebker]

Dear John,

ok so you downloaded SilverFast 6.6.0r4d and the iSRD results where worse than before ?
Can you please tell me the Scan Type you are using ?
I have just tested it again with a Canoscan 8800 and here iSRD works fine.
Maybe you can upload a sample picture of your problem here ?

best regards,

Thomas Luebker,
LaserSoft Imaging AG

[John3:16]

Thomas,
Thank you for the prompt reply.
I am using the CanoScan 9950F and Silverfast SE Plus 6.6.0r4d. I have done a screen capture of the problem and I will try to insert it here. I assume I use the "Img" button above.

The captured screen image is from the iSRD screen. You will see the white particles of dust, and offset below them you will see the red marking where Silverfast is placing the corrections. Of course, in this position they have no effect. If I continue the scan all the dust is still visible. The image is of an old Kodacolor negative using NegaFix in Linear mode. It is set for 48 -> 24 bit color and 1200 DPI resolution. If I adjust the resolution lower then the offset distance narrows. Around 150 DPI they are close enough that they begin to have some affect.

Okay, here is the screen capture image (if it works?)
[img]C:\Documents%20and%20Settings\John%20Sr\My%20Documents\My%20Downloads\SilverFast\iSRD%20Capture.jpg[/img]

Thank you again,
John

[LSI_Heidorn]

Dear John,
this looks like a misalignement of IR & RGB Scan, you can see that it marks at a shifted position of the scratches if you look closely.
@Thomas: maybe that helps in reenacting this and further improving the Alignment !

best regards,

Nils Heidorn

[John3:16]

Yes, that is exactly what appears to be happening. Any idea what is causing this? Is there anything I can do or is it a software glitch?
Thanks...
John

[LSI_Heidorn]

Dear John,
you can stabilize the situation if you fix the films / slides on the flatbed ( or use the holders ).

We also did some more work to make the alignement between IR & RGB Scan safer, thise changes will be there in 660r5 !

It will be more likely to fail for images that do not have a lot of high contast details, can you verify that ?

As i said, try to fix the films and give the 660r5 a shot as soon as its out !

best regards,

Nils Heidorn

[John3:16]

Nils,

I posted last week, but it doesn't seem to have made it online. On your two points from your previous post:

* I always us the negative holders.
* Contrast doesn't seem to be a factor, but perhaps I just haven't noticed it with lower contrast negatives.

I have installed the latest update 660r4e. Attached is a capture of another screen capture image from a scan using the latest update. This image is from the lower left corner of the negative. The shift is very obvious in this image as you will see the distinct line, up and right, from the edge of the negative where the defects are marked.

This scan is from a 35mm color negative, still using the CanoScan 9950F, scanning at 1200 dpi, 48->24 bit. The negative holders are used and the scanner is on a solid foundation.

Am I the only one having this result? You stated that release 660r5 will have a fix, when will it be out? This feature was one of the main reasons I chose to pay to update from my previous release. The other was the multi-exposure.

If it is not possible to get the infrared and RGB scans to align programmatically, how about giving the user the ability to fine tune the alignment? This is a great feature and I can't wait to see it working correctly.

Also, is the use of iSRD limited only to color negatives? I get some bizarre results when applied to B&W negatives. Is it supposed to work with reflective scans?

Thanks again for your help.
John

[LSI_Luebker]

Dear John,

I jsut talked to Nils about your Problem and as he allready told there will be improvements for Canon iSRD.
We plan to release the r5 later this month but I can not give you an exact date yet.

best regards,

Thomas Luebker,
LaserSoft Imaging AG