Loss of contrast after stacking (Zerene)

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

Matthias
Posts: 33
Joined: Wed Jun 08, 2011 6:38 am
Location: Germany
Contact:

Loss of contrast after stacking (Zerene)

Post by Matthias »

I need some help in Zerene stacking.

This is the result after stacking 175 images with Zerene PMax (no retouch). Unfortunately there are very large areas of low contrast. Although PMax tends to increase contrast? I tried DMap (default preferences) too. The result was almost the same.


Image


For comparison a single shot (no retouch), which looks much better to me:

Image


Any advice for preservation of contrast in this case?

lauriek
Posts: 2402
Joined: Sun Nov 25, 2007 6:57 am
Location: South East UK
Contact:

Post by lauriek »

That's quite a deep stack, I suspect optical 'tainting' from oof bits of subject - have you tried running short sub-stacks of the problem areas and then retouching those over the full stack?

DrLazer
Posts: 293
Joined: Wed Feb 16, 2011 4:45 am

Post by DrLazer »

This is an interesting thread. Laurie, are short sub-stacks part of your regular workflow? How do you decide how many images to include in a substack?

DrLazer
Posts: 293
Joined: Wed Feb 16, 2011 4:45 am

Post by DrLazer »

Also, would it be possible to make multiple sub-stacks. Save the output and then use those images as the source for a stack of stacks? or do you need to retouch to achieve best results? I am just a little worried that I will brush some soft focus back into the final image due to the shape of the slice. Which in this case is wibbly wobbly.

lauriek
Posts: 2402
Joined: Sun Nov 25, 2007 6:57 am
Location: South East UK
Contact:

Post by lauriek »

Yes I'm always using sub-stacks. It's quite possible to export output images and then import them as input images and stack them that way but I tend to do it with retouching... I always stack the full stack, then I may stack a shorter 'full stack' if I decide I don't want the rearmost detail in focus. Then I run sub stacks and retouch the bits from those I want over my 'master' stack.

Regarding how to decide what to sub-stack, afraid I'm not sure I can describe the process really any further than - look for problem areas in the main stack, and then work out what input images are required to sub-stack just that area.

Matthias
Posts: 33
Joined: Wed Jun 08, 2011 6:38 am
Location: Germany
Contact:

Post by Matthias »

have you tried running short sub-stacks of the problem areas and then retouching those over the full stack?
Well, I just have tried it. I splitted the 175 images in 5 sub stacks. Then I run these 5 sub stacks to a full stack. But the result was only slightly better. So I've increased the contrast of the 2 "problem" stacks from the low-contrast centre in photoshop and imported them back as input images into Zerene. Then i used the retouch-tool.

I'm not particularly happy with the outcome and the whole procedure ... But of course there's a lot room for improvement after retouch with photoshop ...

Image

lauriek
Posts: 2402
Joined: Sun Nov 25, 2007 6:57 am
Location: South East UK
Contact:

Post by lauriek »

Can you post an image from each end of the stack?

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

This is an unusual problem. It took me a while to put the pieces together.

I suspect you're getting hit by a side effect of brightness adjustment. Go to Options > Preferences > Alignment and remove the checkmark on Brightness. Then try your DMap again.

The reason for my confusion is that, in the end, DMap works by just selecting pixel values from individual frames, or sometimes a weighted average of two adjacent frames. In either case, the output image will be faithful to the pixel values that got stacked. Also DMap cannot construct pixel values outside the normal range of black to white, so there's no opportunity for "Retain full dynamic range" to work any mischief when the images are saved.

On the other hand, brightness adjustment works by normalizing the mean and variance of pixel values in each image, so if the master image for the stack happens to have low contrast because it's substantially out of focus, then all other frames will have their contrast cut to match. That seems like a good fit to what you're seeing.

--Rik

Matthias
Posts: 33
Joined: Wed Jun 08, 2011 6:38 am
Location: Germany
Contact:

Post by Matthias »

Can you post an image from each end of the stack?
Yes Laurie, I could, but after your post Rik gave some good advice and perhaps it's no longer necessary ...

Go to Options > Preferences > Alignment and remove the checkmark on Brightness. Then try your DMap again.
Rik, that's what I did. I guess your suspicion about the brightness adjustment was absolutely correct! I made a complete stack (DMap) with all the 175 images and I can see a big improvement ... much more contrast now =D>

Image



After some cropping and retouching in Photoshop I am reasonably satisfied with this final result (Adscita statices):

Image



But now the question is: In which cases is it worthwhile to change the preferences that way? Perhaps I would (should?) have used this once before ...

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Matthias wrote:I made a complete stack (DMap) with all the 175 images and I can see a big improvement ... much more contrast now =D>
Agreed -- much better.
But now the question is: In which cases is it worthwhile to change the preferences that way?
Good question, and interestingly, one that I don't recall coming up before. Let me think through this...

You definitely do need brightness adjustment if you have uneven exposures and you're using DMap. Otherwise you'll get dark/light contour bands.

You definitely do not need brightness adjustment if your exposures are uniform and you're using just DMap.

You also do not need brightness adjustment if you're using just PMax and your exposures stay away from the ends of the histogram.

The ambiguous case is for PMax with exposures that push the ends of the histogram. PMax is intrinsically resistant to uneven exposures, but especially with deep stacks it's vulnerable to contrast increases that can cause out-of-range pixels to be computed. Those get pushed back into range using a sort of HDR method (high dynamic range), but still there can be some loss of detail in extreme regions. Because deep stacks also tend to have low contrast ends, "brightness" correction can sort of accidentally help in this case by reducing the contrast of middle frames that would otherwise cause out-of-range values.

Brightness correction is selected by default because that's the "safe" setting. It will fix what would otherwise be serious problems, and whatever problems it introduces in otherwise good stacks are generally mild and not obvious. I suspect that the current thread came about because the loss of contrast is unusually much with this stack and the viewer was more attentive than most.

Thanks for the question!

BTW, I would be interested to see the frame that was first in the list after ZS was done stacking. That's the one that was used as a master for brightness/contrast adjustment.

--Rik

Matthias
Posts: 33
Joined: Wed Jun 08, 2011 6:38 am
Location: Germany
Contact:

Post by Matthias »

Rik, thank you for the clear explanations. Now I'm more comfortable about using brightness adjustment.

Here's a screenshot from the moment where Zerene starts stacking:

Image


But don't forget, I also did 5 sub-stacks with "better" first frames and the result was just as bad as the first full stack.


I've also tried to run a DMap-stack with "the other end" of the 175 images, which is the following ... it made no difference at all.

Image

(it's a pity that my English is limited ...)

rjlittlefield
Site Admin
Posts: 23564
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Matthias wrote:But don't forget, I also did 5 sub-stacks with "better" first frames and the result was just as bad as the first full stack.
Assuming you used Stack Selected, this is exactly what is supposed to happen.

The purpose of Stack Selected is to generate "slabs" suitable for use in retouching.

So, each frame that goes into the Stack Selected process receives exactly the same shift, rotation, scale, and brightness/contrast adjustments that it receives when you stack all the frames.

In fact if you do a Stack Selected without having stacked all the frames first, you will have to wait some time while ZS spins through the whole stack determining what all the adjustments will be.

--Rik

DrLazer
Posts: 293
Joined: Wed Feb 16, 2011 4:45 am

Post by DrLazer »

Matthias ... would it be possible to see a full resolution version of these images?

Matthias
Posts: 33
Joined: Wed Jun 08, 2011 6:38 am
Location: Germany
Contact:

Post by Matthias »

... would it be possible to see a full resolution version of these images?
Hi Craig, of course. Of which pictures you want to see the full resolution version? Maybe I can send them to you via e-mail. The final image you can find here: http://www.flickr.com/photos/63933727@N ... hotostream

DrLazer
Posts: 293
Joined: Wed Feb 16, 2011 4:45 am

Post by DrLazer »

Thanks!

I just wanted to compare your original low contrast image, to the one straight out of Zerene, to the final thing. I am trying to work out similar things atm. My personal email is crgtlr@gmail.com.

Post Reply Previous topicNext topic