Scanning issue, help!

[ageorge]

Below is a 100% crop of a TMAX100 negative straight out of an Epson 4990 scanned at 2400 DPI with VueScan. I have also included the a whole image, below as well for scale. As you can see, I am getting these awful dark edges on some of the rocks here. These are very difficult to deal with. Anyone know the cause of these. They are definitely not on the negative. Any help would be greatly appreciated.

Thanks in advance,
-alan

100% Crop
http://alan-george.com/tmp/crop-bad_edges.jpg

The whole image
http://alan-george.com/tmp/scan0121_raw.jpg

[Michael Mutmansky]

Alan,

Are you using multiscanning and/or dust removal of any kind?

It looks like the negative has moved in the middle of the scan to me. There is clear streaking in the grain in the sky areas, and it looks like it has the same axis of movement as the dark bands you are asking about.

---Michael

[Kirk Gittings]

Are you trying to run Digital Ice with a b&w?

[ageorge]

Are you trying to run Digital Ice with a b&w?

Not Digial Ice but I do have the "dust removal" set to medium in VueScan. I'll turn it off and see if that makes a difference.

Thanks,
-alan

[ageorge]

Alan,

Are you using multiscanning and/or dust removal of any kind?

It looks like the negative has moved in the middle of the scan to me. There is clear streaking in the grain in the sky areas, and it looks like it has the same axis of movement as the dark bands you are asking about.

---Michael

I do have the the "dust removal" set to medium in VueScan and I also have multi-scan turned on as well. I do see this on multiple negatives so I am not sure it would be movement.

Thanks,
-alan

[Michael Mutmansky]

Alan,

Change one variable at a time, and you'll find the culprit.

If you look at the foggy sea portion of the blowup, there is a streaking in the grain/density that looks like it has it's primary axis at about 15 degrees from vertical. Look carefully at the blowup. That may also be an artifact of the dust removal, but it could be an artifact of multiscanning as well.


---Michael

[ageorge]

Alan,

Change one variable at a time, and you'll find the culprit.

If you look at the foggy sea portion of the blowup, there is a streaking in the grain/density that looks like it has it's primary axis at about 15 degrees from vertical. Look carefully at the blowup. That may also be an artifact of the dust removal, but it could be an artifact of multiscanning as well.


---Michael

Thanks, I'll start eliminating things. I was hoping that it wasn't "normal" for my setup.

As for the streaking, this was a 2 minute exposure and that fog does come from moving water so it is possible that that is what is being recorded.

Thanks again,
-alan

[ageorge]

What does a Newton Ring look like?

-alan

[Michael Mutmansky]

Alan,

Is this a color scan from a B&W negative? If so, newton rings would look like somewhat concentric circles that have a slight rainbow effect to them. They're much easier to see in a color scan due to the rainbow effect.

If this is a B&W scan, the newton rings will be light and dark bands in circles, and they can be pretty subtle.

Either way, I'm not seeing any newton rings in the blowup.


---Michael

[Michael Mutmansky]

Here's some newton ring samples.

[Ted Harris]

Also, why VueScan? Why not either the dedicated Epson software or Silverfast SE which I assume came with the scanner?

Nothing against VueScan, simply curious if you are having the same problem with other software and if this may be a software artifact.

[ageorge]

Also, why VueScan? Why not either the dedicated Epson software or Silverfast SE which I assume came with the scanner?

Nothing against VueScan, simply curious if you are having the same problem with other software and if this may be a software artifact.

I was using VueScan for reading the density values from the negatives. It also has some predefined options for TMAX100 at various contrast levels which seemed to be doing a good job of managing the highlights (or most dense parts of the negative). I can also select the RGB channel to scan from into grayscale thus saving the disk space and time of scanning in RGB and then converting. I am still new to scanning, so I am still not yet set on a workflow. I'll probably give the Espon software another go. The Silverfast seems very unstable. It crashed several times on my G5 and th UI seems less than intuitive. Is there truly a unique feature to Silverfast compared to the Espon package? Considering I just need 4x5 B&W scans?

Thanks,
-alan

[Ted Harris]

Before you do anything else try what Michael suggests (BTW you might be interested in the scanning workshops he and I run PM me if so). There are many differences in the fullversion of Silverfast (Ai), not so many with the packaged SE. The current Epson software is excellent, IMO better than VueScan.

[Jim Jones]

Alan -- have you disabled sharpening when scanning?

[neil poulsen]

I know very little about scanner construction, but I notice there are three shades of gray along the top edges. It makes me wonder if that might correspond to the three colors, R, G, B.

Perhaps the rgb sensors in the scanner aren't properly aligned? That's kind of a wild guess. You might try scanning the image in color, even though it's a black and white negative, open the image in Photoshop, and see if you observe the same effect in each of the R, G, B channels.

Can you use a strong loop on the negatives and see if you see the effect there? That could rule out anything to do with the scanner.

Or, chromatic abberation??? (I doubt it.) But, something to do with the three colors.

[ageorge]

Alan -- have you disabled sharpening when scanning?

I didn't have sharpening turned on with this scan (or any scan for that matter).

-alan

[ageorge]

I know very little about scanner construction, but I notice there are three shades of gray along the top edges. It makes me wonder if that might correspond to the three colors, R, G, B.

Perhaps the rgb sensors in the scanner aren't properly aligned? That's kind of a wild guess. You might try scanning the image in color, even though it's a black and white negative, open the image in Photoshop, and see if you observe the same effect in each of the R, G, B channels.

Can you use a strong loop on the negatives and see if you see the effect there? That could rule out anything to do with the scanner.

Or, chromatic abberation??? (I doubt it.) But, something to do with the three colors.

No, I am scanning to 16-bit grayscale so there is no color info in the scan itself.

It's definitely not on the negative.

It's probably the "dust removal" or multi-scan options of VueScan. I will do some testing a report back to the thread tomorrow.

-alan

[Kirk Gittings]

George this looks to me like "Digital Ice on a black and white scan" artifacts. I know Vuescan doesn't technically have true Digital Ice software but theirs must be pretty close to DI to be able to use the DI hardware.

[ageorge]

Setting "Infrared Clean' from Medium to None in VueScan did the trick. It was also adding the directional blur to the water.

With out Infrared Clean 100% crop
http://alan-george.com/tmp/crop-good-edges.jpg

With Infrared Clean 100% crop
http://alan-george.com/tmp/crop-bad_edges.jpg

[Kirk Gittings]

Vuescan's Infrared Clean is their version of the Digital Ice software. I knew that would do the trick as I accidentally did the same thing two nights ago with Silverfast!