I have tried slabbing manually a few times with very good results, particularly for retouching, so I was intrigued to try SlabberJockey by Chris S.
http://www.photomacrography.net/forum/v ... hp?t=20071
The download is still available, but it does not include the PDF help file mentioned by Chris in his description.
I installed Access 2010 and SlabberJockey appears to run perfectly. It generates a script which I copied and pasted into a *.txt file using Notepad.
Zerene loaded the script and appeared to run it: Zerene loaded complete set of source slices [16-bit per channel TIFF files] - but only produced the first slab.
Zerene first ran through the whole set of slices then selected the first batch of slices and produced the first slab.
Zerene then goes through the entire set of slices again and appears to produce the second slab to the point where the progress bar appears to be saving it.
Zerene repeats this process: working through the complete set of slices each time thern appearing to generate the next slab.
But there are no more slabs after the first. Not in the folder I have created for the output, the source folder ornor the project folder. Only the first slab has been created and I suspect that Zerene is not supposed to go through the complete sequence more than once or a deep slab will take a very long time.
I have tried Zerene 1.2 and 1.4 [the non-professional versions] and run Windows 7 X64
Chris does refer to “a few steps before and after this” so my guess is that is what I am missing.
Can anyone tell me what I am doing incorrectly from my notes, please ?
Henry
SlabberJockey - Help Please
Moderators: rjlittlefield, ChrisR, Chris S., Pau
-
Greenfields
- Posts: 116
- Joined: Thu Nov 25, 2010 10:54 am
- Location: Nottinghamshire, England
SlabberJockey - Help Please
Feel free to edit my images.
-
rjlittlefield
- Site Admin
- Posts: 23621
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Henry,
I'm not intimate with Slabberjockey, but I do recognize some bits of Zerene's behavior in your description.
The part about going through the entire stack multiple times can be triggered by having each of the slabbing tasks point directly to a folder of source images, instead of having the tasks say "current project" and loading the folder first to create that project. There's also a recently discovered bug that might cause that problem; I'm not sure yet.
I don't recognize anything else, but I'm suspicious that the failure to save all outputs except one could also be associated with repeatedly loading a folder of source images.
The "unaligned and aligned JPEGS" that you see in the project folder are actually low resolution cached screen image versions. They're not used in batch mode, but rather they are generated by batch mode operations in anticipation of speeding up future interactive operations such as retouching.
If you send a copy of the batch script to support@zerenesystems.com, I'll be happy to take a look at it.
--Rik
I'm not intimate with Slabberjockey, but I do recognize some bits of Zerene's behavior in your description.
The part about going through the entire stack multiple times can be triggered by having each of the slabbing tasks point directly to a folder of source images, instead of having the tasks say "current project" and loading the folder first to create that project. There's also a recently discovered bug that might cause that problem; I'm not sure yet.
I don't recognize anything else, but I'm suspicious that the failure to save all outputs except one could also be associated with repeatedly loading a folder of source images.
The "unaligned and aligned JPEGS" that you see in the project folder are actually low resolution cached screen image versions. They're not used in batch mode, but rather they are generated by batch mode operations in anticipation of speeding up future interactive operations such as retouching.
If you send a copy of the batch script to support@zerenesystems.com, I'll be happy to take a look at it.
--Rik
-
Greenfields
- Posts: 116
- Joined: Thu Nov 25, 2010 10:54 am
- Location: Nottinghamshire, England
Henry,
Sorry you’re having trouble with SlabberJockey, and that it took me half a day to see your post. Do be aware that I'm quite happy to be contacted directly.
When you ran the SlabberJockey install file, it should have placed on your desktop (or wherever you installed to) a folder named “SlabberJockey” in addition to the SlabberJockey shortcut. This folder contains, among other things, the how-to pdf and a sample batch file produced by SlabberJockey. Are you certain that this folder is not on your computer?
Just in case (and for folks who prefer not to run install files) I’ve edited the SlabberJockey thread you referenced. It now includes links to individual elements of the SlabberJockey package. One of these is to the how-to document: How-to do slabs in Zerene Stacker using SlabberJockey--V1.0.pdf.
Also, once you have the how-to pdf, you'll find step-by-step instructions which have, so far, worked pretty consistently for nearly all users.
Would you please let us (or, at least, me) know how this works for you?
Best,
--Chris
Sorry you’re having trouble with SlabberJockey, and that it took me half a day to see your post. Do be aware that I'm quite happy to be contacted directly.
When you ran the SlabberJockey install file, it should have placed on your desktop (or wherever you installed to) a folder named “SlabberJockey” in addition to the SlabberJockey shortcut. This folder contains, among other things, the how-to pdf and a sample batch file produced by SlabberJockey. Are you certain that this folder is not on your computer?
Just in case (and for folks who prefer not to run install files) I’ve edited the SlabberJockey thread you referenced. It now includes links to individual elements of the SlabberJockey package. One of these is to the how-to document: How-to do slabs in Zerene Stacker using SlabberJockey--V1.0.pdf.
Rik has probably hit the nail on the head. I’ve seen one or two people have this problem. So please check this first. Here is a screen grab from SlabberJockey with the potential error highlighted in orange-red:rjlittlefield wrote:The part about going through the entire stack multiple times can be triggered by having each of the slabbing tasks point directly to a folder of source images, instead of having the tasks say "current project" and loading the folder first to create that project.
Also, once you have the how-to pdf, you'll find step-by-step instructions which have, so far, worked pretty consistently for nearly all users.
Would you please let us (or, at least, me) know how this works for you?
Best,
--Chris
Last edited by Chris S. on Wed Jul 23, 2014 8:53 pm, edited 1 time in total.
-
Greenfields
- Posts: 116
- Joined: Thu Nov 25, 2010 10:54 am
- Location: Nottinghamshire, England
Thank you very much for your help. I'll try again at the week end and report back.
A potential benefit of slabbing which I have not seen mentioned before is that it can reduce edge banding - provided that the edge banding is not due to misalignment of the optical axis with the direction of travel.
As Rik Littlefield has explained in detail:
http://www.photomacrography.net/forum/v ... 9878#79878
and:
http://www.photomacrography.net/forum/v ... hp?t=17926
edge banding can arise as a result of the accumulated adjustments [translation, rotation and re-scaling] the stacking software makes to re-map succesive slices as it work through a stack.
What seem to be happening is that the accumulated adjustments in a deep stack [say 200-400 slices] can move part of an image image out of the composition. The stacking software fills the gaps by copying pixels at the boundary of the image.
Because the adjustments are cumulative edge banding is more likely to occur with deep stacks.
I suspect - but have not tested this - that it's also more likely if the subject detail is limited to a patch of the image which shifts as different parts of the subject come into and out of focus.
An indicator of this happening may be sprawling hot pixel trails. I would expect misalignment to produce more or less straight trails.
Slabbing often reduces this effect and reveals details at the edges previously obliterated by banding. I assume that is because each sub-stack starts accumulating adjustments afreh and, in general they don't accumulate when the substack are themselves stacked.
All of these comments apply to microscope objectives and macro lenses in which the camera moves during a stack. Microscope objective are reported to be designed to be telecentric but its difficult to see this in a stack because of the re-mapping.
A potential benefit of slabbing which I have not seen mentioned before is that it can reduce edge banding - provided that the edge banding is not due to misalignment of the optical axis with the direction of travel.
As Rik Littlefield has explained in detail:
http://www.photomacrography.net/forum/v ... 9878#79878
and:
http://www.photomacrography.net/forum/v ... hp?t=17926
edge banding can arise as a result of the accumulated adjustments [translation, rotation and re-scaling] the stacking software makes to re-map succesive slices as it work through a stack.
What seem to be happening is that the accumulated adjustments in a deep stack [say 200-400 slices] can move part of an image image out of the composition. The stacking software fills the gaps by copying pixels at the boundary of the image.
Because the adjustments are cumulative edge banding is more likely to occur with deep stacks.
I suspect - but have not tested this - that it's also more likely if the subject detail is limited to a patch of the image which shifts as different parts of the subject come into and out of focus.
An indicator of this happening may be sprawling hot pixel trails. I would expect misalignment to produce more or less straight trails.
Slabbing often reduces this effect and reveals details at the edges previously obliterated by banding. I assume that is because each sub-stack starts accumulating adjustments afreh and, in general they don't accumulate when the substack are themselves stacked.
All of these comments apply to microscope objectives and macro lenses in which the camera moves during a stack. Microscope objective are reported to be designed to be telecentric but its difficult to see this in a stack because of the re-mapping.
Feel free to edit my images.
-
Greenfields
- Posts: 116
- Joined: Thu Nov 25, 2010 10:54 am
- Location: Nottinghamshire, England
-
rjlittlefield
- Site Admin
- Posts: 23621
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Slabbing via Stack Selected should not have this effect. That's because Stack Selected requires that all frames have already been aligned by the usual one big pass over all frames. If that hasn't already happened, then Stack Selected will do it first. So, no, each sub-stack does not start accumulating adjustments afresh. It would be possible to do slabbing in a different way involving multiple projects that would get aligned separately, but as far as I know, neither Slabberjockey nor any of the other slabbing tools will do that.Greenfields wrote:Slabbing often reduces this effect and reveals details at the edges previously obliterated by banding. I assume that is because each sub-stack starts accumulating adjustments afreh and, in general they don't accumulate when the substack are themselves stacked.
Please check again to be sure that the reduced-banding effect is reproducible. Assuming it is, then start a conversation with me on support@zerenesystems.com so that we can figure out what in your workflow is allowing this benefit to occur. Thanks!
--Rik
-
Greenfields
- Posts: 116
- Joined: Thu Nov 25, 2010 10:54 am
- Location: Nottinghamshire, England
Rik,
You are absolutely correct, of course.
When I wrote my original comment I had not managed to use SlabberJockey so did not understand that it aligned the entire stack before making the sub-stacks.
I had in mind manually prepared sub-stacks in which the slices are only aligned within the sub-stacks. You must disable automatic ordering to make sure that all the sub-stacks are built up in the same direction or you will get discontinuities when the slabs are stacked together.
As you have asked I will repeat the test carefully and record the workflow to see if I can verify the benefit.
In the example I had in front of me at the time a "straight" PMax stack with alignment on produced wide bands of edge streaking + long, winding pixel trails and what was left of the image was tilted and some of the high frequency detail was blurred. Turning off alignment eliminated the banding, included the whole of the original image, rendered high frequency detail sharp everywhere and increased contrast, so demonstrating that mis-alignment of successive slices is not the cause. Combining sub-stacks with alignment on produced a much better result than a "straight" stack but not as good a result as turning alignment off.
In this case turning alignment off was successful because the objective lens [Mitutoyo Plan Apo] is telecentric or close to telecentric so that out of focus features don't change in scale [not sure that this makes much difference at such a small depth of field] and was accurately aligned. That may mean that it was wrong of me to generalise from observations made with one lens.
Henry
You are absolutely correct, of course.
When I wrote my original comment I had not managed to use SlabberJockey so did not understand that it aligned the entire stack before making the sub-stacks.
I had in mind manually prepared sub-stacks in which the slices are only aligned within the sub-stacks. You must disable automatic ordering to make sure that all the sub-stacks are built up in the same direction or you will get discontinuities when the slabs are stacked together.
As you have asked I will repeat the test carefully and record the workflow to see if I can verify the benefit.
In the example I had in front of me at the time a "straight" PMax stack with alignment on produced wide bands of edge streaking + long, winding pixel trails and what was left of the image was tilted and some of the high frequency detail was blurred. Turning off alignment eliminated the banding, included the whole of the original image, rendered high frequency detail sharp everywhere and increased contrast, so demonstrating that mis-alignment of successive slices is not the cause. Combining sub-stacks with alignment on produced a much better result than a "straight" stack but not as good a result as turning alignment off.
In this case turning alignment off was successful because the objective lens [Mitutoyo Plan Apo] is telecentric or close to telecentric so that out of focus features don't change in scale [not sure that this makes much difference at such a small depth of field] and was accurately aligned. That may mean that it was wrong of me to generalise from observations made with one lens.
Henry
Last edited by Greenfields on Wed Jul 30, 2014 12:15 am, edited 3 times in total.
Feel free to edit my images.
-
rjlittlefield
- Site Admin
- Posts: 23621
- Joined: Tue Aug 01, 2006 8:34 am
- Location: Richland, Washington State, USA
- Contact:
Henry, thanks for the check and clarification. This now makes perfect sense.
You're correct that telecentricity doesn't matter much at high magnification. What is important about telecentricity is that it prevents the scale from changing visibly between front and back of the in-focus portion of each source image. At roughly 10X and above, this becomes true for most lenses regardless of whether they are nominally telecentric or not.
My usual recommendation is to turn off whatever alignment options you don't need. With a mechanical setup, almost always Rotation can be turned off. At high magnification, Scale can be turned off. If your focusing mechanism is smooth and stable, then X and Y Shift can be turned off. In all cases, the ultimate test is "try it and see", because sometimes there are unexpected issues that were masked by the computational alignment. (One surprisingly common example is when the camera is mounted on bellows and rotates at the attachment flange due to mirror/shutter shock combined with weight imbalance of the camera.)
One warning: there is a long standing but recently discovered bug that affects alignment with slabbing. If you have Brightness correction turned off but any other alignment correction turned on, then Stack Selected may erroneously conclude that it needs to re-align the whole stack even when it doesn't. This will be fixed in the next beta.
--Rik
You're correct that telecentricity doesn't matter much at high magnification. What is important about telecentricity is that it prevents the scale from changing visibly between front and back of the in-focus portion of each source image. At roughly 10X and above, this becomes true for most lenses regardless of whether they are nominally telecentric or not.
My usual recommendation is to turn off whatever alignment options you don't need. With a mechanical setup, almost always Rotation can be turned off. At high magnification, Scale can be turned off. If your focusing mechanism is smooth and stable, then X and Y Shift can be turned off. In all cases, the ultimate test is "try it and see", because sometimes there are unexpected issues that were masked by the computational alignment. (One surprisingly common example is when the camera is mounted on bellows and rotates at the attachment flange due to mirror/shutter shock combined with weight imbalance of the camera.)
One warning: there is a long standing but recently discovered bug that affects alignment with slabbing. If you have Brightness correction turned off but any other alignment correction turned on, then Stack Selected may erroneously conclude that it needs to re-align the whole stack even when it doesn't. This will be fixed in the next beta.
--Rik