Vuescan testing - multipass, long exposure pass, grain reduction, and sharpening

[kaiyen]

All,
I have finished some testing of the various settings in Vuescan. Specifically:

-single vs. multiple passes
-grain reduction settings
-long exposure pass
-impact on sharpness from above

Now, many of you may find issues with this testing - not rigorous enough, not scientific, needs larger sample size, etc - but this is what I got so far. I hope it is useful to some.

I will eventually build this out into a web page. For now...

All of these scans are from a single 35mm frame from a roll of Ilford Delta 3200 exposed at EI 3200, developed in F76+ 1+9 for 10:00. Agitation was continuous for the first 30 seconds, then 10 seconds every minute.

The frame was scanned with color settings to Auto Levels with the TMAX 400 profile, with CI =.55 in D76. The scanner is a Nikon CoolScan IV (so "only" 2900 dpi, which is important to note), and I used the SA-21 strip loader.

First: the original straight scan

I'll followup with the different tests.

allan

 

[kaiyen]

single vs. multi-pass

single vs. multi-pass

So the first look was at single vs. multipass.

Links:
1 pass
3 pass
8 pass

I did not include 100% crops of this because I found there to be not much difference in terms of shadow detail. I was quite surprised by this. I was not able to examine noise because there really wasn't any to begin with in any of these scans. Very interesting. I am hoping to do some tests on slides and c41 color next to see if there is a practical benefit.

For what it's worth, I feel like I must have messed something up since I am not seeing a difference.

 

[kaiyen]

Grain reduction

Grain reduction

My next test was re: grain reduction settings in Vuescan. Several folks have indicated that they combat the grain-enhancing qualities of the highly collimated light in dedicated film scanners by using grain reduction in the software. I wanted to see whether this helped much and how much it impacted sharpness.

Please note my choice of a very, very grainy film. Ideally, I should then do similar tests with a lot of other combinations, but that would take too long.

This post deals with just the unsharpened scans.

Links to 100% crops:
1 pass, no reduction
1 pass, light grain reduction
1 pass, medium grain reduction

I did not upload examples with multipass because they exhibited roughly the same changes.

My opinion is that the grain reduction doesn't do all that much for the grain, but it impacts sharpness quite a bit. I am not a fan. I think I'd rather change my developer to reduce grain than use the software solution.

next up: impacts of sharpening on the usefulness of grain reduction

allan

 

[kaiyen]

Photokit Sharpener vs. Grain reduction fuzziness

Photokit Sharpener vs. Grain reduction fuzziness

So, just for the heck of it, I thought I'd try a state-of-the-art automated "smart" sharpening tool against grain-reduced scans to see if I could bring them up to the straight scan, without grain reduction. I used Photokit Sharpener for this, applying the Narrow Edge capture sharpening. I did not use the output sharpener at all.

Reference link:
reference: 1 pass, no grain reduction

Just for the heck of it link:
1 pass, no grain reduction, PK sharpener (this is invariably going to be even sharper than the other scans, but still useful as a reference)

Comparison links:
1 pass, light grain reduction, PK sharpener
1 pass, medium grain reduction, PK sharpener

As you can see, while PKS has done a nice job, it still isn't the same as without grain reduction at all. I think I might be able to get closer if I used multiple sharpneing techniques - wide edge then narrow edge, for instance - but that's a lot more effort than simply turning off grain reduction in the first place.

allan

 

[kaiyen]

Long exposure pass

Long exposure pass

My final test was for the long exposure pass. This is often cited as a useful tool with slides, for punching through the shadows. Now, I'm scanning negatives so I am going to compare any changes in the highlights (so, again, dealing with highest density).

None of these images have any grain reduction turned on.

Reference link:
1 pass only. So this is a single pass, WITHOUT a long exposure pass. This is what the highlights look like.

Comparison links:
1 pass, long exposure pass
8 pass, long exposure pass

I didn't bother with a 3 pass one. It seems to me that adding the long exposure pass made the grain a bit software and smoother in the highlights. Additional highlight detail is negligible at best. The 8 pass version looks much less sharp, which is confusing. The SA-21 should not have allowed the film to move so that shouldn't be the reason. And the earlier examples of multi-pass didn't indicate a reduction in sharpness, either. I will have to test that again.

allan

 

[aad]

Thanks for all the work. It looks to me, based on the monitor, that it isn't worth a whole lot on this film type. I tried a multi-pass on a slide, and it seemed a bit better, but that was a flatbed.

Maybe all this stuff is kind of fluff? Post process in PS seems more effective.