Blue ''aura'' on video

[Cristian11]

Hello all,

I'm facing an issue with the recording quality on a mirror-less camera.

My setup is the following : GH5 mounted on a Carl Zeiss GF stand, with 160 plan Neo-Fluar objective.

I'm using an adapter to insert the camera through the ocular slot. There no intermediate optics in between.

The LED on microscope is a 3W 5000K

I do not know if there's something in the camera settings or some processing reason behind this. This is visible only when using the mirror-less camera.

I've also tried to mount the KPI 8x ocular directly on the camera sensor, and the blue aura gets worse.

What is interesting is that the way I see it, the blue aberation is only visible on one side.

Anyone has any suggestions or encountered something similar?

[imkap]

Hello all,

I'm facing an issue with the recording quality on a mirror-less camera.

My setup is the following : GH5 mounted on a Carl Zeiss GF stand, with 160 plan Neo-Fluar objective.

I'm using an adapter to insert the camera through the ocular slot. There no intermediate optics in between.

The LED on microscope is a 3W 5000K

I do not know if there's something in the camera settings or some processing reason behind this. This is visible only when using the mirror-less camera.

I've also tried to mount the KPI 8x ocular directly on the camera sensor, and the blue aura gets worse.

What is interesting is that the way I see it, the blue aberation is only visible on one side.

Anyone has any suggestions or encountered something similar?

I have a similar setup with Zeiss GFL and Sony a6000 (I use an eyepiece) and have CA on images too. More visible in darkfield than BF, still not sure what is the main culprit.

When observing through the binocular it is not that obvious...

We had some talk about it here:
viewtopic.php?f=25&t=15405&p=121771#p121771

[Scarodactyl]

That's chromatic aberration--it's not just blue, it's red on the opposite side. Zeiss eyepieces do a lot of corrections to the objective's image, more than any other major maker of that era. Going direct to camera will never work that well. On top of that the intermediate image is inside the eyepiece tube when everything is set up right, so thr optics habe to be pushed our of spec to get the image to move up where it is accessible to the camera sensor.

[Cristian11]

Cheers thank you.

Yeah I was suspecting that it wont work simply like that by projecting directly into the sensor.

I've noticed that it is okish on alive transparent specimens, but gets worse if something is more opaque. It will probably get worse on darkfield as well, as above post stated.

I once had for a brief while, a Motic B310E with a Fuji Xt3. I simply mounted the camera on trinocular and voilla, top quality.

Apparently its not that easy with old 160 scopes heh

[Cristian11]

Hello all,

I'm facing an issue with the recording quality on a mirror-less camera.

My setup is the following : GH5 mounted on a Carl Zeiss GF stand, with 160 plan Neo-Fluar objective.

I'm using an adapter to insert the camera through the ocular slot. There no intermediate optics in between.

The LED on microscope is a 3W 5000K

I do not know if there's something in the camera settings or some processing reason behind this. This is visible only when using the mirror-less camera.

I've also tried to mount the KPI 8x ocular directly on the camera sensor, and the blue aura gets worse.

What is interesting is that the way I see it, the blue aberation is only visible on one side.

Anyone has any suggestions or encountered something similar?

I have a similar setup with Zeiss GFL and Sony a6000 (I use an eyepiece) and have CA on images too. More visible in darkfield than BF, still not sure what is the main culprit.

When observing through the binocular it is not that obvious...

We had some talk about it here:
viewtopic.php?f=25&t=15405&p=121771#p121771

It's probably a combination between sensor and the fact that objectives are not infinity.
I've decided to give it another try with the eyepiece ocular directly in front of camera sensor, and it seems it does a better job, since the ocular is probably correcting the light( as Scarodactyl confirmed), but the CA is still visible, similar with your case.

My LED has low cri index. 74. I will buy a better one to see if anything changes. But I also suspect that a halogen light would be better for us in our case.

[Scarodactyl]

The LED isn't helping, since it will have a spike in blue emission which makes the blue being out of focus more obvious. Putting the bare sensor in front of the eyepiece is still using it out of spec though--the eyepiece wants to produce an image focused at infinity, not a focused image. There are ways to modify the eyepiece to focus an image, but they do come with wuality issues. You can also put a lens in front of your camera focused at infinity to capture the infinity-focused image from the eyepiece, but even then you'll find CA issues since this is a viewing eyepiece which isn't as fully corrected as a photo eyepiece (which are rarer and more expensive). I think you will be able to improve things quite a bit but at some point you will hit dimishining returns on this system even with good objectives like the plan neofluar.

[Cristian11]

Thanks for explaining me in a simple and comprehensive manner. Optics are damn complicated.

It looks like no matter what I do, I will have CA. I will try to settle for what ever option (simple and cheap) brings less CA.

I've noticed that somehow, the best one is using the mobile phone over the ocular, but I get that ''tunnel'' vision, which I really dont like, plus, the adapters are quite crap and hard to set.

Will try with the halogen as well to test how things could miraculously improve.

Cheers!

[imkap]

Thanks for explaining me in a simple and comprehensive manner. Optics are damn complicated.

It looks like no matter what I do, I will have CA. I will try to settle for what ever option (simple and cheap) brings less CA.

I've noticed that somehow, the best one is using the mobile phone over the ocular, but I get that ''tunnel'' vision, which I really dont like, plus, the adapters are quite crap and hard to set.

Will try with the halogen as well to test how things could miraculously improve.

Cheers!

You can get rid of some CA with software editing later...

[Cristian11]

Thanks for explaining me in a simple and comprehensive manner. Optics are damn complicated.

It looks like no matter what I do, I will have CA. I will try to settle for what ever option (simple and cheap) brings less CA.

I've noticed that somehow, the best one is using the mobile phone over the ocular, but I get that ''tunnel'' vision, which I really dont like, plus, the adapters are quite crap and hard to set.

Will try with the halogen as well to test how things could miraculously improve.

Cheers!

You can get rid of some CA with software editing later...

Yes I have Davice Resolve (free version). I'm still trying to spin my head around it.

[imkap]

Desaturating separate (mostly blue and purple) colors worked for me to some extent. I suppose there are other techniques too. Looks too artificial if you overdo it.

[Hobbyst46]

There are several acceptable methods to connect the camera to these old 160mm Zeiss optics.

1. Afocal: KPL eyepiece plus camera lens, focused at infinity (because actual focusing on the specimen should be done with the microscope, exclusively!). The camera lens should fit the sensor size and eyepiece magnification, for good coverage of the field of view (FOV). For example, if the sensor is an APS-C, and the eyepiece is the 8X KPL, the best camera lens (as calculated and recommended by others) should be a prime 50mm. (Manual lens, and not necessarily from the same make as the camera).
Others have suggested an equivalent combination for the APS-C sensor: a 10X eyepiece with an 40mm pancake lens.

Some CA will still be present in both cases above..

2. Afocal: with original an Zeiss relay lens (named "Mipro"). Supposed to work well. These lenses are rare and expensive. Never tried it and do not know the details.

3. Focal, or eyepiece projection: A relatively recent method, developed and published by Rolf Vossen (search "Vossen" on the forum. You need two eyepieces - a 8X KPL and a 10X KPL (none of them WF!!). You make a "hybrid" - replace the field lens of the 8X with that of the 10X (or some similar exchange, I can't remember for sure). You use the hybrid eyepiece and the camera WITHOUT a camera lens. According to the experience of several hobby microscopists (myself included) this method works well, with relatively mild aberrations. Vossen has done an extensive survey of the various fortuitous eyepiece lens combination - and the motivation for his experiments was exactly the inadequacy of methods 1 and 2 above (and various other methods).

[Hobbyst46]

By the way, to better judge the image quality and various aberrations, a flat, uniform pattern specimen would be better. I would suggest a stage micrometer, that is a "ruler".

[Scarodactyl]

I ended up settling with the hybrid eyepiece approach when I set up my own Zeiss universal. It seemed to give results at least as good as afocal but with a simpler and cheaper setup. That said the achromats it was locked into using for epi DIC put a cap on color correction anyway so my experience can be taken with a grain of salt.

[Cristian11]

There are several acceptable methods to connect the camera to these old 160mm Zeiss optics.

1. Afocal: KPL eyepiece plus camera lens, focused at infinity (because actual focusing on the specimen should be done with the microscope, exclusively!). The camera lens should fit the sensor size and eyepiece magnification, for good coverage of the field of view (FOV). For example, if the sensor is an APS-C, and the eyepiece is the 8X KPL, the best camera lens (as calculated and recommended by others) should be a prime 50mm. (Manual lens, and not necessarily from the same make as the camera).
Others have suggested an equivalent combination for the APS-C sensor: a 10X eyepiece with an 40mm pancake lens.

Some CA will still be present in both cases above..

2. Afocal: with original an Zeiss relay lens (named "Mipro"). Supposed to work well. These lenses are rare and expensive. Never tried it and do not know the details.

3. Focal, or eyepiece projection: A relatively recent method, developed and published by Rolf Vossen (search "Vossen" on the forum. You need two eyepieces - a 8X KPL and a 10X KPL (none of them WF!!). You make a "hybrid" - replace the field lens of the 8X with that of the 10X (or some similar exchange, I can't remember for sure). You use the hybrid eyepiece and the camera WITHOUT a camera lens. According to the experience of several hobby microscopists (myself included) this method works well, with relatively mild aberrations. Vossen has done an extensive survey of the various fortuitous eyepiece lens combination - and the motivation for his experiments was exactly the inadequacy of methods 1 and 2 above (and various other methods).

Thanks for sharing this.

Can you please confirm to me. The field lens of the ocular is the bottom one, correct?

[Hobbyst46]

...Can you please confirm to me. The field lens of the ocular is the bottom one, correct?

Yes.

[imkap]

I don't own the KPL 10x, so never tried the hybrid eyepieces. But I did try the combination with Olympus Splan 20 and Olympus NFK 6.7, which should be ok and it had the same or similar amount of CA. So my conclusion was that it must be something else causing this much CA, as it is quite pronounced. The light source seems like the most likely problem to me but I didn't try the different LED yet. I think I should try a 2700K led with a blue filter, rather than the 4000K I bought to use without a filter. I don't have it yet, so didn't try. Unfortunately when ordering I bought different powers, as I was concerned about the heating, didn't think about it from this standpoint. Currently I use 70CRI 4000K, I have one 4000K 90 CRI I'll try that one too before buying, but I suspect the difference won't be very big... Although hopefully I'm wrong

[Cristian11]

I don't own the KPL 10x, so never tried the hybrid eyepieces. But I did try the combination with Olympus Splan 20 and Olympus NFK 6.7, which should be ok and it had the same or similar amount of CA. So my conclusion was that it must be something else causing this much CA, as it is quite pronounced. The light source seems like the most likely problem to me but I didn't try the different LED. I think I should try a 2700K led with a blue filter, rather than the 4000K I bought to use without a filter. I don't have it yet, so didn't try. Unfortunately when ordering I bought different powers, as I was concerned about the heating, didn't think about it from this standpoint. Currently I use 70CRI 4000K, I have one 4000K 90 CRI I'll try that one too before buying, but I suspect the difference won't be very big... Although hopefully I'm wrong

As they said above, the fact that LED light has a blue spike, even at 4000k, contributes to CA, as part of multiple factors. Since I've never read anything about modern microscopes with LED that produce CA on their trinocular, it means that old optics setups might have it's limits when projecting into mirrorless cameras.

my next steps will be:
1. try another LED with warm light
2. try halogen bulb
3. try hybrid method (which is more expensive since these 10x KPL oculars are kinda hard to find at a decent price in Europe)

[Hobbyst46]

I think I should try a 2700K led with a blue filter, rather than the 4000K I bought to use without a filter...

The blue filter will intensify the (already significant) blue component of the warm-white LED. If the purpose is to suppress the blue, an amber filter would be better.

Currently I use 70CRI 4000K, I have one 4000K 90 CRI I'll try that one too before buying, but I suspect the difference won't be very big...

I believe that a 90 CRI will be definitely better than a 70 CRI. I now have a ~95 CRI 4000K LED (in a retrofit LED illuminator that I bought) and it gives very nice neutral white illumination.

[Cristian11]

I think I should try a 2700K led with a blue filter, rather than the 4000K I bought to use without a filter...

The blue filter will intensify the (already significant) blue component of the warm-white LED. If the purpose is to suppress the blue, an amber filter would be better.

Currently I use 70CRI 4000K, I have one 4000K 90 CRI I'll try that one too before buying, but I suspect the difference won't be very big...

I believe that a 90 CRI will be definitely better than a 70 CRI. I now have a ~95 CRI 4000K LED (in a retrofit LED illuminator that I bought) and it gives very nice neutral white illumination.

Is it possible to present an example of that LED?
I'm having a hard time finding something above CRI 84, which has decent lumens and neutral white.

[imkap]

There are all sorts of LEDs on Mouser, 90 CRI or more. Didn't find much elsewhere

[zzffnn]

Other members have made good points.

What was that particular sample in the opening post, by the way? If it was the only slide that produces lots of CA (as you said in the first post) and your regular samples are different and don’t produce as much CA, then why worry about that.

Naturally high contrast images always reveal more CA. Also samples such as diatoms and glass / sand pieces will cause more CA as well.

[Cristian11]

Other members have made good points.

What was that particular sample in the opening post, by the way? If it was the only slide that produces lots of CA (as you said in the first post) and your regular samples are different and don’t produce as much CA, then why worry about that.

Naturally high contrast images always reveal more CA. Also samples such as diatoms and glass / sand pieces will cause more CA as well.

I've checked afterwards on other samples as well, by recording in 4k and viewing on a monitor, including transparent specimens. The CA is quite bad on all and very difficult to look at.

I'm waiting for another LED to arrive and also to buy a 10x KPL to create a hybrid ocular, hopefully things will improve.

Will update with pic/video if any success.

[imkap]

I just put a 90 CRI LED of the same brand and same color temperature (last one was 70CRI) and the amount of CA isn't affected, it is the same. Whether I need to change the eyepiece or the condenser I don't know. Maybe I should get the hybrid eyepiece first , there is no, or very little CA (only on defocused dust) when observing visually. So the eyepiece might be the main problem.
If I'd get the Achr. Apl. 1.4 condenser, I suppose I'd lower the CA a bit, but I wouldn't expect dramatic difference, or should I?
Can a 1.4 N.A. condenser be used on a 25x/0.45 objective or less?

[Scarodactyl]

The condenser is not your problem, it's on the eyepiece end. Getting the eyepiece set up correctly will help a lot, but keep in mind your eyes will always be more forgiving than the camera. Your brain does some of its own postprocessing and through eyepieces you will also be focusing on the center of the image where ca is a nonissue.

[imkap]

but keep in mind your eyes will always be more forgiving than the camera.

And the thing about eyes is they can't zoom in, unlike the computer... I'll try to find an eyepiece and see what happens.

[Finelld]

The astronomy market has fringe killer filters that are basically a band reject filter for the CA. They can be screwed into the bottom of eyepieces and On cameras. They come in a variety of sizes.

[imkap]

The astronomy market has fringe killer filters that are basically a band reject filter for the CA. They can be screwed into the bottom of eyepieces and On cameras. They come in a variety of sizes.

It would be nice to try it out. Try out different places in the light path and see what happens...

[imkap]

Of course I have ordered a KPL 10 eyepiece. The CA is still here, although different than it was... The great thing is, that it was easy to make everything parfocal with the KPL hybrid eyepiece.

Here is an image. I didn't edit, just resized and sharpened a bit afterwards...
Maybe using planapo objectives might help. I have only one Planapo objective 63x/1.4 OIL. My condenser is 0.9, so resolution will be bad, but probably will see if there is a difference in CA...

DSC09832.JPG (79.02 KiB) Viewed 3556 times

[Sure Squintsalot]

Here is an image. I didn't edit, just resized and sharpened a bit afterwards...
Maybe using planapo objectives might help. I have only one Planapo objective 63x/1.4 OIL. My condenser is 0.9, so resolution will be bad, but probably will see if there is a difference in CA...

I use Photoshop Lightroom 6.0 (free) for photo processing of my NEX-5R images and removing chromatic aberration is dead easy. I don't even worry about it.

CA is also a huge problem with legacy telephotos imaging of rocks and snow, coupled with modern sensors. I'm pretty sure that extremely high fidelity of modern sensors brings out all of the inherent defects of ancient lens manufacturing.

[imkap]

Here is an image. I didn't edit, just resized and sharpened a bit afterwards...
Maybe using planapo objectives might help. I have only one Planapo objective 63x/1.4 OIL. My condenser is 0.9, so resolution will be bad, but probably will see if there is a difference in CA...

I use Photoshop Lightroom 6.0 (free) for photo processing of my NEX-5R images and removing chromatic aberration is dead easy. I don't even worry about it.

How do you remove it in Lightroom?