Helicon Soft Forum

Hi,

A feature I would really like to see in HF would be a per-channel stacking option. It would be interesting for people using HF for microphotography. Let me explain :

Microscope objective have two main defects : they can be non planar and have chromatic aberrations.

Imagine (to make things simple) that I take a picture of a planar object. if my objective is not planar, by staking a few images with HF, I can correct this defect, and have a sharpen image on the whole field. This is GREAT.

BUT, if my objective has chromatic aberrations, this means that each channel (R,G and B) will NOT be focused at the same time. At the moment, as far as I can see, HF stack images in RGB space. If there was an option if HF to run the stacking algorithm 3 times (treating each channel separatly), and registering each plane created, this would be awesome, because HF, with a stack of image taken with a cheap objective, could recreate the image a plan apochromatic objective would give...

Do not hesitate to ask me for more info and/or sample files.

g0b

Hi,

What about correcting source images before stacking? In Helicon Filter on the Aberration tab we have a tool to compensate chromatic aberrations cased by optics. We can embed such a filter to be applied to source images before stacking. Please have a look and let me know.

Hi,

My explaination was not clear i'm afraid.
Il is not possible to corret each image individually.
The issue with non APO objectives is plan of focus is not the same for each wavelength.
Remember i'm talking about microscope objective, not camera objective (where CA is far less visible)

Real example (in real world things are a bit more complicated, but the idea is here) :
I take a stack of 3 images (each has a different focusing distance).
Image 1 : Red channel is focused Blue and green are not
Image 2 : Red and blue channels are out of focus, green is focused.
Image 3 : Red and green are out of focus, Blue is fine.

I cannot correct this aberration for each image individually under Helicon Filter (or maybe I missed something and Helicon Filter allows that kind of correction on a stack of images...)

g0b

I understand you now. Each plane should be treated individually and focused ares may not be the same for for R,G anb B channels.Current solution is to split all source images into 3 channels, process 3 different stacks and then merge them together. In principle, this task can be automated. The question is how typical it is. I would appreciate if you upload sample stack to our ftp server (ftp://heliconsoft.com/ftp.html) so we can perform some tests.

Hi,

Thanks for your quick answer !
I've uploaded the stack and results of my test on your ftp (CA directory).
It's a picture of a DLP, I thinks the regularity of the patterns make analysis easier.

HeliconFocus_RGB is the result of stacking with color images
HeliconFocus_R, _G and _B are the result of stacking each channel independantly.
HeliconFocusMerged is the 3 above images merged into a color image.

Splitting channels to treat them independantly gives clearly better results (especially on the blue channel). But as I expected, on the corner , R G and B channels are not aligned anymore (unlike in the center) and even lateral CA correction in Helicon filter is not enough to correct it.

g0b