Stackshot Issues & Partial Solution
Moderators: rjlittlefield, ChrisR, Chris S., Pau
Stackshot Issues & Partial Solution
I have been attempting to use Stackshot at 2 micron step size and having issues with image alignment and non-uniform steps. I know this is asking a lot from this rail, and will be moving to a higher precision rail.
However in the meantime I needed to get a 3 image stitch chip image at 20X with 2 micron steps with about 300 steps per image. Having Alignment ON in Zerene causes the rendering image to somewhat wander as it is stepped across, yielding an OK result for a single image but not too good for trying to stitch with other "wandering" images. So I usually try to stitch with Alignment OFF in Zerene, this however introduces issues with the actual image alignment for image to image because of uneven step sizes and some slight image shifting.
My system uses a small incline (20 degrees) from subject to camera/lens, to allow some thread loading in the rail due to gravity. I decided to load Stackshot focus rail with a array of rubber bands stretched to 6" and attached to the optical base at about 3" down and 3" to the right from the rail mid-point. This loads the rail, pulling with gravity, but also adding an off axis load to pull the rail thread surfaces to one side and down. My thinking was to try and force the rail to mesh the threads on the right side and bottom to get better uniform travel and not see any staggered movement. This really worked well for the stack I just completed at 2 microns, I may try 1 micron to see how well it works out later.
Anyway, folks that have the Stackshot rail and experiencing slight shifts during high resolution sessions should try this. 20 rubber bands and a 1" long 1/4 by 20 bolt are pretty cheap.
Best,
Mike
However in the meantime I needed to get a 3 image stitch chip image at 20X with 2 micron steps with about 300 steps per image. Having Alignment ON in Zerene causes the rendering image to somewhat wander as it is stepped across, yielding an OK result for a single image but not too good for trying to stitch with other "wandering" images. So I usually try to stitch with Alignment OFF in Zerene, this however introduces issues with the actual image alignment for image to image because of uneven step sizes and some slight image shifting.
My system uses a small incline (20 degrees) from subject to camera/lens, to allow some thread loading in the rail due to gravity. I decided to load Stackshot focus rail with a array of rubber bands stretched to 6" and attached to the optical base at about 3" down and 3" to the right from the rail mid-point. This loads the rail, pulling with gravity, but also adding an off axis load to pull the rail thread surfaces to one side and down. My thinking was to try and force the rail to mesh the threads on the right side and bottom to get better uniform travel and not see any staggered movement. This really worked well for the stack I just completed at 2 microns, I may try 1 micron to see how well it works out later.
Anyway, folks that have the Stackshot rail and experiencing slight shifts during high resolution sessions should try this. 20 rubber bands and a 1" long 1/4 by 20 bolt are pretty cheap.
Best,
Mike
Lou,Lou Jost wrote:That's interesting. Can you please send a picture of the set-up?
See below. Note the angle and placement of the rubber bands and bolt: front, down and away to the side. The attachment point for the rail is the knob for the ARCA clamp. For the particular Stackshot rail I have this worked remarkably well.
Best,
Mike
Lou,
Thanks. My thought was to load the thread surfaces and side arm slides such that they would not move much as the rail moved, but not so much as to introduce any type of binding.
After looking at the rail sides and threads I decided the location of the bolt would probably be the best based upon the angle of tension, and decided the tension of the rubber bands would produce what I thought would work. This is what is shown, I have not tried another location, nor number of bands, nor changed the tension. This worked so well I have not tried to change anything!!
Best,
Mike
Thanks. My thought was to load the thread surfaces and side arm slides such that they would not move much as the rail moved, but not so much as to introduce any type of binding.
After looking at the rail sides and threads I decided the location of the bolt would probably be the best based upon the angle of tension, and decided the tension of the rubber bands would produce what I thought would work. This is what is shown, I have not tried another location, nor number of bands, nor changed the tension. This worked so well I have not tried to change anything!!
Best,
Mike
- Charles Krebs
- Posts: 5865
- Joined: Tue Aug 01, 2006 8:02 pm
- Location: Issaquah, WA USA
- Contact:
Mike,
I understand what you are doing, and the improvement is considerable.
So in the example shots all alignment settings in the software are off. I realize for your stitching purposes this is desired... but I am very curious what the result would look like with all alignment options "on", as might be the case with a single image and no stitching intent..
Would it be possible to try that?
I understand what you are doing, and the improvement is considerable.
So in the example shots all alignment settings in the software are off. I realize for your stitching purposes this is desired... but I am very curious what the result would look like with all alignment options "on", as might be the case with a single image and no stitching intent..
Would it be possible to try that?
Last edited by Charles Krebs on Fri Nov 04, 2016 3:23 pm, edited 1 time in total.
- Charles Krebs
- Posts: 5865
- Joined: Tue Aug 01, 2006 8:02 pm
- Location: Issaquah, WA USA
- Contact:
BTW I had surmised that some of this was due to screw wobble. Earlier images had more periodic effects than the first image shown. I moved my image capture point closer the the end bearing in the rail, could not move very close to the motor end as interference between the rail end and my subject positioning fixture would have been an issue.
This improved the result to what is in the first image shown, and lead me to the off axis tension concept with the rubber bands. I had used rubber bands before to hold a lens end to a fixture to reduce lens movement due to vibration, and also to pull along the optical axis to reduce backlash.
Mike
This improved the result to what is in the first image shown, and lead me to the off axis tension concept with the rubber bands. I had used rubber bands before to hold a lens end to a fixture to reduce lens movement due to vibration, and also to pull along the optical axis to reduce backlash.
Mike
Yes, I have run images with alignment ON in Zerene and it does a pretty good job but can cause the image to wander off at the later stages of the stack I found. I did a 3 stitch session with the same chip and alignment ON and had a bear of a time stitching the three together to make a presentable image. (We chip designers are sticklers for precision, and fudging an image is not an option!!).
However with alignment OFF I got the result you see here.
I want to get a more precise rail, but needed to get this particular chip done and couldn't wait until I got another rail configured and setup.
Mike
However with alignment OFF I got the result you see here.
I want to get a more precise rail, but needed to get this particular chip done and couldn't wait until I got another rail configured and setup.
Mike
-
- Posts: 1966
- Joined: Sat Oct 07, 2006 10:16 am
- Location: Bigfork, Montana
- Contact:
Mike,
Since you're from the semi-conductor world why not something industrial like this?
http://www.ebay.com/itm/THK-Used-KR2602 ... SwZ8ZXAzvZ
-JW:
Since you're from the semi-conductor world why not something industrial like this?
http://www.ebay.com/itm/THK-Used-KR2602 ... SwZ8ZXAzvZ
-JW:
Smokedaddy,
Thanks, I have been looking at something similar. The one you identified has a 2mm thread pitch, whereas the Stackshot rail has a 1.59mm pitch, and the Wemacro has a 1mm pitch.
However now I am not so convinced these industrial grade rails are that much better than the ones that Stackshot or maybe Wemacro offer, especially with the off axis tension concept mentioned. They will handle a larger load, and have more travel, but may not be much better in short movement uniformity with little image shift. I don't require precise location, nor repeatability but do require very little image shift.
Johan at Extreme Macro evaluated the Wemacro system, and Beatsy has posted some very nice images at 1 micron steps. These confirm that 1 micron steps are possible with these rails, but I may require tighter control of image shifts that they do because of my chip subjects.
I am shooting at 20X with 2 micron steps as shown here and will be going to 50X at 1 micron steps soon, so any image shift will just get amplified more.
If I had the $, or my company would pay, I would just get a really precise expensive rail...but that is not the case.
Anyway, thanks for the pointer, if you find a 1mm pitch rail please let me know.
Best,
Mike
Thanks, I have been looking at something similar. The one you identified has a 2mm thread pitch, whereas the Stackshot rail has a 1.59mm pitch, and the Wemacro has a 1mm pitch.
However now I am not so convinced these industrial grade rails are that much better than the ones that Stackshot or maybe Wemacro offer, especially with the off axis tension concept mentioned. They will handle a larger load, and have more travel, but may not be much better in short movement uniformity with little image shift. I don't require precise location, nor repeatability but do require very little image shift.
Johan at Extreme Macro evaluated the Wemacro system, and Beatsy has posted some very nice images at 1 micron steps. These confirm that 1 micron steps are possible with these rails, but I may require tighter control of image shifts that they do because of my chip subjects.
I am shooting at 20X with 2 micron steps as shown here and will be going to 50X at 1 micron steps soon, so any image shift will just get amplified more.
If I had the $, or my company would pay, I would just get a really precise expensive rail...but that is not the case.
Anyway, thanks for the pointer, if you find a 1mm pitch rail please let me know.
Best,
Mike