Chromatic aberrations filtering

[bkt]

What I am describing, is changing the alignment of the color channels, shifting them around, in order to correct for certain types of CA, in a way that does not drop any channels. - snip - changing the alignment via the method used in registax and ser player.

Ser Player? Seems to be a video player, dedicated to astro... "The color channel alignment feature ensures that the colors in your recordings are appropriately aligned with the RGB color space": https://appsitory.com/windows/ser-player-64/ https://github.com/cgarry/ser-player/bl ... /ChangeLog "mp_frame_image->set_colour_align(red_align_x, red_align_y, blue_align_x, blue_align_y);"

Will check now what RegisStax does.

The "synthetic blue channel method" seems to be described here, basically "B_synthetic = 2G - R"; on page 2, one shows a software allowing to tweak the 3 channels: https://www.cloudynights.com/topic/9059 ... el-method/

[bkt]

your chromatic aberrations are all over the place and as before completely unsymmetrical as well
...

This is the kind of distortion and aberration one would expect of a 100.00 student scope.
I think perhaps, address the optics first and maybe the software won't seem such a necessity.

The scope build itself (and the options) doesn't look that bad to me. I checked, the 4x costs 50€, the 10x 58€... If I expect more, I could have to swap them out. A 10x pl apo costs approx 1000€

[apochronaut]

You are maybe missing the point that any objective whether cheap or expensive should be symmetrical . The fact that each objective has a serious skew to it's presentation of ca wavelengths across the field to the point of flipping the polarity midsection or should I say maintaining the polarity rather than flipping it means there is likely an optical fault somewhere that is common to them all. Any two hemispheres of your field should present mirror images of ca when there is no particular refractive differential between them. I would be looking at the telan lens in the head or the alignment of the head itself.
There is no ca at the central axis of any lens and it develops slowly off axis but radially and symetrically.
One thing for the future is that I suspect your micrometer is a surface coat type, not a sandwich. That is suitable for epi objectives but yours are cover slip corrected. With the 40X and up pop a cover slip over it to reduce unwanted extra ca but more importantly sa.

[Topcode]

It seems now blurry (in a crop), this is what I'd like to check using a variable length photo port tube:

MGKgUV0.processed.jpg

well, any CA will always leave you with some blur, because RGB are broadband, from one end of your camera picking up red(maybe, 700nm), to the other end(maybe, 600nm), there will be some chromatic aberrations, meaning its blurred a bit, no matter what you do.

What I am describing, is changing the alignment of the color channels, shifting them around, in order to correct for certain types of CA, in a way that does not drop any channels. - snip - changing the alignment via the method used in registax and ser player.

Ser Player? Seems to be a video player, dedicated to astro... "The color channel alignment feature ensures that the colors in your recordings are appropriately aligned with the RGB color space": https://appsitory.com/windows/ser-player-64/ https://github.com/cgarry/ser-player/bl ... /ChangeLog "mp_frame_image->set_colour_align(red_align_x, red_align_y, blue_align_x, blue_align_y);"

Will check now what RegisStax does.

The "synthetic blue channel method" seems to be described here, basically "B_synthetic = 2G - R"; on page 2, one shows a software allowing to tweak the 3 channels: https://www.cloudynights.com/topic/9059 ... el-method/

Ser player and registax are both designed for astro, but i have found the work very well for microscopy aswell. registax offers alot more in depth processing, ser player only allows for a bit of basic stuff, but it can handle any videos you may have. Synthetic blues work pretty well, ive used them in the past.
but you do have to remember that they completely get rid of one color channel, all that data disappears, so your colors will look different. Oranges and yellows will become completely different colors under many circumstances.

overall i do have to agree with apochronauts assertion that something is impacting your optics, that you should probably try to fix.

[bkt]

You are maybe missing the point that any objective whether cheap or expensive should be symmetrical . The fact that each objective has a serious skew ....

I would be looking at the telan lens in the head or the alignment of the head itself.
There is no ca at the central axis of any lens and it develops slowly off axis but radially and symetrically.

I didn't care about this skew, which is maybe the "bug" I quickly noticed and shared with the seller, he didn't react. The head is too loose, tilts a little bit, would need a spacer to prevent this... I need to fix that.

Not sure this impacts the images? It is maybe more perceptible with this flat micrometer?

One thing for the future is that I suspect your micrometer is a surface coat type, not a sandwich. That is suitable for epi objectives but yours are cover slip corrected. With the 40X and up pop a cover slip over it to reduce unwanted extra ca but more importantly sa.

Yes, surface coated. I cleaned it, added a drop of pond water, plus a cover glass. I still get heavy blue/green aberrations, but I assume this is due to the high contrast of the dark grid in bright field. On "real slides", I do not get those blue artifacts, only the other one; seems like the later are due to the objectives, and pop in out of focus areas....

I tried to better align the head for this. But it isn't use less, it depends of what I'm watching at. The below is a prepared slide.

[bkt]

Ser player and registax are both designed for astro, but i have found the work very well for microscopy aswell. registax offers alot more in depth processing, ser player only allows for a bit of basic stuff, but it can handle any videos you may have. Synthetic blues work pretty well, ive used them in the past.
but you do have to remember that they completely get rid of one color channel, all that data disappears, so your colors will look different. Oranges and yellows will become completely different colors under many circumstances.

overall i do have to agree with apochronauts assertion that something is impacting your optics, that you should probably try to fix.

I learned a lot since now with this topic, and with what was answered to, thanks all for this.

My major concern being my objectives, which are cheap ones. At some point, I may swap them. Meanwhile, I'll use software, which takes some processing time, especially for 2 or 4k videos, but it is free... Others having "scholar" scopes and objectives could also be interested, I believe.

I don't think the software I found swaps the channels on the full frame. But only where it detects artifacts, based on parameters and comparisons.

I started to test RegiStax V6, which seems not so bad to me, with an initial attempt.
By the way, digging, I found an alternate solution, G'MIC.
An I believe GIMP version 3 will include a solution for chroma aberrations also. That shall soon be available (beta exists, I didn't test it).

For testing RegiStax, I used a known poor image: https://fr.m.wikibooks.org/wiki/Fichier ... inging.jpg



RegiStax doesn't degrade it, seems it mainly rebuilds data in the altered areas... Like with other solutions, there are still some remaining artifacts, but the result quickly looks better: https://www.astronomie.be/registax/

registax.JPG (105.33 KiB) Viewed 2020 times

I didn't test G'MIC, produced by image processing experts also. Seems very interesting: https://linuxfr.org/news/sortie-de-g-mi ... vos-images

gmic.jpg (116.35 KiB) Viewed 2020 times

[bkt]

The above solutions can fix up to nicely purple fringes, very common in photography, in high contrast areas.

What I'm getting annoyed by being a bit different, I locate it in another area of the color spectrum. I may be wrong... I believe it is related to the flat rate achro objectives, build and coating, not by any other hardware or alignment issue. With artifacts popping in according to slides and magnifications, and according to what is in focus or not. I have some slides of a prepared set I cannot use at all due to this effect... I skip those.

4x, nice:


10x, nice


40x...


40x...

[apochronaut]

I go back to my comment that you cannot make a silk purse out of a sow's ear. Microscope imaging is based on resolution which can be theoretically ,mathematically determined based on the angle of incidence of the light. An N.A.is determined. Higher N.A.'s provide more information. More information is more point sources of light creating the image. Removing chroma is removing point sources of light and despite the fact that ca obscures resolution, by wiping the ca off the screen you are not adding back resolution but in fact reducing it. This is all before the hazy blur that the process leaves behind, plus the fact that the wavelengths and distribution of the ca in the various objectives is uneven. Any program applied to solve the ca would not apply to the microscope as a whole but would have to be specific to each objective. I'm pretty sure that somehow you may be opening up a can of worms regarding other distortions too. The microscope optics can be repaired.
As far as ca is concerned, modern achromats from various manufacturers have fairly low and consistent levels of ca across the range and since your system is an Olympus format 180mm , keeping an eye peeled for some used Olympus achromats might be in order. It is a bit of an eye opener that those Motic objectives are so diffuse in correction levels. Stuff made by reputable companies in the 1920's was better, aside from planarity.

There is a range of inexpensive Chinese made plan fluor objectives. The 4X is less than 100.00. A used one or a set might show up. They have been around long enough.
Another option that would be cheap is to pick up a set of Reichert U.S.A. planachros. A lot of those were sold and the markings are off many of them from wear, so many go up for sale inexpensively. As planachros go, they are very good, quite ca free with residual chroma leaning towards those a fluorite displays. They are probably in the advanced achromat category which have some low dispersion glass in them, possibly lanthanum. I have seen entire sets of those selling for 100.00, even less. Since they are for a 182mm tube, they will fit right in except for the telan lens corrections. You would likely have to switch eyepieces in order to ether correct or compensate a bit. I haven't worked that through but that would be a cheap option and a real fix with resolution attached. I could do some testing and see where to go with the eyepieces. The corrective mis-matches will be a small amount of lateral ca. Just about any defect in correction can be optimized with eyepieces.

[JWW]

I am just curious why something like a Chromacor hasn't been developed for microscopes. I have one for my 8" f/15 D&G achromat and am happy with the results. It's designed to intercept the light cone and correct for chromatic aberration. Of course, I have no idea if it would work on a microscope; I am just wondering.

-JW:

[apochronaut]

I can think of numerous answers to that, the most obvious being that the design as an add on is quite expensive and that you would likely need a different one for each objective due to the correction tuning required.. Modern objective families are more harmonious but then they are the least to need correction or compensation. The least complex microscope objective is more complex than the most complex telescope objective and in microscopes you keep the eyepiece and change the objectives, the opposite of telescope practice.


That complexity makes me think that in certain cases, some less expensive objectives touting superior colour correction but not quite apochromatic, possibly have a Chromacor like group maybe in the rear to fix the corrections some. Older apochromat/compens eyepiece sets functioned a little in that way with the rather complex eyepiece compensating for overcorrection.

[JWW]

Thanks. As I said, I was curious. In the telescope world, they are quite rare and expensive.

[Chas]

bkt, out of curiosity is your microscope halogen or LED?
I noticed that in the (old?) specifications I came across it said it was a halogen system but it could take LED 'bulbs' too, if required.

[bkt]

@apochronaut still thinking about what you added at #38, checking also what I may get from Chinese sellers... will answer you later.

bkt, out of curiosity is your microscope halogen or LED?
I noticed that in the (old?) specifications I came across it said it was a halogen system but it could take LED 'bulbs' too, if required.

Some were offered with LED. I got an halogen set but I added a 4500k LED for much less heat, stable color temp and longevity.

I forgot illumination, always set at 100%. I rarely look into eyepieces, using my PC screen instead. My camera adapting exposure time and running so mostly close to 30fps (4k) or to 60ps (2k).

As explained under another topic (viewtopic.php?f=27&t=18839&p=142491&sid ... 07#p142491), I'm using high frame rates for videos, and for quick/easy stacking also... @apochronaut just adding some pictures. According to needs or expectations, results could be acceptable. This being a prepared slide, a "bee mouth", using the 4x:

[bkt]

Hi, @apochronaut, and others,

> I go back to my comment that you cannot make a silk purse out of a sow's ear.

I know this. I've a full frame since the 80's. The cheapest and worst optics I have are:
- a Samyang fishey, without any CA correction
- a vintage 80mm russian tilt/shift, without any correction, and bit blurry
- a pinhole, producing high diffraction, but I'm getting nice pictures with it also.

> Microscope imaging is based on resolution which can be theoretically

I'm not interested by high resolution, but by budget reasonable CA correction. What I get in #37 at 40x (and very often popping in from 4x to 60x also) being annoying to me.

> Removing chroma is removing point sources of light and despite
> the fact that ca obscures resolution, by wiping the ca
> off the screen you are not adding back resolution but in fact reducing it.

But for some pictures, it may help to improve the quality. The rush of this is full of CA, software fixed it:



> Any program applied to solve the ca would not apply to the microscope
> as a whole but would have to be specific to each objective.

From my short experience using software, it needs to be tweaked based on images or videos, according to the amount of CA that popped in...

> I'm pretty sure that somehow you may be opening up a can of worms regarding
> other distortions too. The microscope optics can be repaired.

Sorry for this

I believe you made a typo, optics can't be repaired, shall be swapped out...

I talked again to my seller, cancelled a Motic EC 20x order. I would surely be disappointed or limited by it also.

He suggested me to upgrade my camera, I won't do that, or to find a much more expensive scope base. Plus high end objecives. I had choosen this BA310e after having watched lot of videos and reviews over several months. Assuming I would so have an acceptable mechanical base, plus the possibility to swap the objectives later. Seems other do this, swapping the lenses for apos, or any alternates, but never telling exactly by which ones...

Seems the BA310e and its stock objectives are sold for routine, were it is adequate, and not to people having much more expectations. I shall have bought a Zeiss or an Olympus, but those bases - without objectives - are up to four time and more this price...

> As far as ca is concerned, modern achromats from various manufacturers have
> fairly low and consistent levels of ca across the range and since your system is
> an Olympus format 180mm , keeping an eye peeled for some used Olympus
> achromats might be in order. It is a bit of an eye opener that those Motic
> objectives are so diffuse in correction levels. Stuff made by reputable companies
> in the 1920's was better, aside from planarity.

If I swap the objectives, according to my seller, I could be limited by the tube or telan lense...

> There is a range of inexpensive Chinese made plan fluor objectives. The 4X
> is less than 100.00. A used one or a set might show up. They have been
> around long enough.

I had a quick look, found some fluar and semi-apo offers. With the same or similar NA than mine. Without any details of their CA correction... I would spend money without knowing what I might get.

Did it once, as a newbie, bought a $10 2x objective. To see. Saw nothing at all, but dark through it

> Another option that would be cheap is to pick up a set of Reichert U.S.A. planachros.
> ...
> in the advanced achromat category which have some low dispersion glass in them
> ..
> Since they are for a 182mm tube, they will fit right in except for the telan lens corrections
> ...
> You would likely have to switch eyepieces in order to ether correct or compensate a bit

That is a no go for me. I'm inexperienced in optics, and I don't believe I could succeed in making so much changes. I'd prefer to keep the head and the eyepieces...

I've noticed the Olympus UPLFLN range, available from 4x to 60x (without immersion). Their NA are almost similar than mine. Could this be an option, with my current head/telan lense/eypieces? Maybe, according to this only video I found (the second part, vs the first, with more CA):

[apochronaut]

Let me guess. You're favourite band is Blur.

[bkt]

Let me guess. You're favourite band is Blur.

I have some Blur albums, yes, but not only. I've also some sharp full frame lenses, one being of the 80's, a 60mm f2.8D macro, pretty well adapted to modern Nikon DSLRs. Sharpness/apo is out of budget for me, especially after the prices I saw of Olympus objectives

[bkt]

@apochronaut

I called and sent a mail to a Nikon/Zeiss/Olympus seller. Waiting for their answer and proposal...

On the web some share such kind of information, my main concern, before resolution, being chroma aberrations:



Moreover, they add the higher the NA, the lower will be depth of field... If plan fluor could be an intermediate option, for price also, I'll probably go that way.

What you may have considered as "a way too blurry" was at 40x, taken and foreseen to test some sofwares. This is the rush, with a lot of CA:



This was the result after image processing (since, I tweaked this software for less de-saturation):



If I want something blurry and close to what a vintage russian tilt/shit or any bottle bottom can produce, I know exactly how to proceed. I just have to plug one of my Bresser achro objectives in this Motic microscope to appreciate the result (with no CA added by the setup, I had nothing to fix by software for that picture)

240208024528.jpg (85.69 KiB) Viewed 1846 times

[bkt]

There is a range of inexpensive Chinese made plan fluor objectives.

I enjoy shopping no brands in China. Any idea of what "Plan Flour" is?

PlanFlour.JPG (106.28 KiB) Viewed 1803 times

[Chas]

I tested also different methods to filter or attenuate colors, with yellow and amber filters, wasn't satisfying or magic either.

Try a "Kodak F1 Pale yellow" filter; it seems to chop off the 'stand alone' blue band pretty cleanly.
I cant find a crossover (the Lee pale-yellow gelatine is not the same ) The one I have seems to have been an accessory for Kodak Retina cameras, though I guess the filter colour might pre-date these:
https://retinarescue.com/retina_filters.html

[MichaelG.]

This page may be of interest:

https://www.kodak.com/en/motion/page/w ... 2-filters/

lots of transmission curves for download.

MichaelG.

[bkt]

I tested also different methods to filter or attenuate colors, with yellow and amber filters, wasn't satisfying or magic either.

Try a "Kodak F1 Pale yellow" filter; it seems to chop off the 'stand alone' blue band pretty cleanly.
I cant find a crossover (the Lee pale-yellow gelatine is not the same ) The one I have seems to have been an accessory for Kodak Retina cameras, though I guess the filter colour might pre-date these:
https://retinarescue.com/retina_filters.html

I thought about using such hardware piece of the early photo film age, used for daylight/interior color/white balance compensation. Somewhere, I should have a deep amber filter, that would then have its place in the illumination path. If I remember well, They were were declined in 85A or B5B, amber, or 85c, yellow. I do not think they chop off, they only more or less attenuate part of the spectrum. Nowadays, those filters seem rare and costly.

I did different tests with photo gelatines, but that wasn't very satisfying. So I didn't look where my amber filter is.

An alternate modern and cheap solution could consist to tweak the color channels, or to filter images using software, to exclude part of the spectrum. But as apochraunote added, that would exclude part of the image information also.

The less destructive would consist in using software to deal with aberrations... that way, there only, images would get altered or more or less rebuilt.

[Hobbyst46]

An amber KR12 filter works for me. It removes much of the 450nm blue tint of a 6500K LED. The filter was not costly then (in 2018/2019).
The 85A did part of the job, not enough IMO.
Searching "KR12" on the forum will lead to some 4-5 years old posts about LED light filtration.
Amber filters were useful for special B&W film photography...

[bkt]

It removes much of the 450nm blue tint of a 6500K LED

My "cosmetic" problem not being the blue tint, which would be easy to fix. But aberrations popping into the blue/green spectrum, like here: