Cushion & spherical optical aberation with DSLR on µscop

[yminter]

Here is my problem:

I have a Leica DMLM µscope which is currently equipped with an old Leica 780X852pixels digital camera.

I tried to adapt a Canon 550D on it using a commercial optical adapter that is to be screwed to tha C-mount adapter of the µscope and which offers the right adapter for the 550D.

The problem is that the result is not exactly as I thought it would be:

Here is what I get using the old camera:



Here is what I get using the adapter & DSLR:



(Both images were shot without moving the sample, just by replacing the camera by the DSLR + adapter & refocusing).

The provider of the adapter tells me that it must be due to the magnification factor of the Leica 0.63X C-mount adapter that is not high enough and that the outer part of the optics of the µscope is responsible for the aberrations. He tells me that a 1X mechanical adapter (hence certainly lens-less) would correct the problem.

The explanation looks a bit strange to me since I can see the same part of the sample through the oculars, without any aberration... The only difference in this case is that I don't look through the Leica 0.63X C-mount adapter.

The point that I find strange is that parts of the image that look crisp on the camera have beginnings of aberrations with the adapter & DSLR, thought the optical path is exactly identical up to the C-mount adapter...

Does anyone have already seen something like this or an idea of what kind of problem could cause it?

[enricosavazzi]

I cannot comment on the specific equipment, but here is what I can say - and ask.

Here is my problem:

I have a Leica DMLM µscope which is currently equipped with an old Leica 780X852pixels digital camera.

Does this webcam have built-in optics? I should expect it does. In this case, these optics are optimized for the sensor size of the webcam. Probably the webcam has a small sensor, but in the unlikely case that the sensor is large, in principle it would be possible to use it with an APS-C or Micro 4/3 camera body (ignoring for the moment all practical problems of connecting the equipment together). If the image quality (in the central area of the camera sensor equivalent to the surface of the webcam sensor) remains similar to the webcam image, then the fault most likely is with the new optical adapter.

You do not say whether the webcam is used together with a separate photoeyepiece - this may also be relevant.

I tried to adapt a Canon 550D on it using a commercial optical adapter that is to be screwed to tha C-mount adapter of the µscope and which offers the right adapter for the 550D.

The problem is that the result is not exactly as I thought it would be:

Here is what I get using the old camera:

Here is what I get using the adapter & DSLR:

(Both images were shot without moving the sample, just by replacing the camera by the DSLR + adapter & refocusing).

The provider of the adapter tells me that it must be due to the magnification factor of the Leica 0.63X C-mount adapter that is not high enough and that the outer part of the optics of the µscope is responsible for the aberrations. He tells me that a 1X mechanical adapter (hence certainly lens-less) would correct the problem.

It could be that a 1x adapter will work better, but it is far from certain it would be without optics. If the microscope objectives require eyepiece correction, then the photoeyepiece/adapter/relay lens (different people use different names) must provide this correction. If it doesn't, this could be one of the reasons for the aberrations you are observing.

The explanation looks a bit strange to me since I can see the same part of the sample through the oculars, without any aberration... The only difference in this case is that I don't look through the Leica 0.63X C-mount adapter.

Is the camera mounted on a trinocular head? Or do you mount the camera into one of the eyepiece tubes after extracting the eyepiece you are using for visual observation?

In this case, it is not strange at all. You are seeing one thing, and the camera sees another. You are observing the subject through matched sets of objectives and eyepieces, but the new adapter/eyepiece may not be correcting the objective aberrations.

The point that I find strange is that parts of the image that look crisp on the camera have beginnings of aberrations with the adapter & DSLR, thought the optical path is exactly identical up to the C-mount adapter...

Does anyone have already seen something like this or an idea of what kind of problem could cause it?

As mentioned above, we would need to know more about the equipment and setup. But so far at this point, I suspect that the optics in the new adapter are at least one of the causes of the aberrations.

The aberrations from the new adapter appear already in areas that are "good" in the webcam image. Are you sure that this adapter is intended for direct image projection onto a sensor/film plane without additional optics? And is the adapter designed to provide an image circle sufficient to cover your camera sensor? Many adapters are only designed to cover small videocam sensors. You mention above "Leica 0.63X C-mount adapter". C-mount adapters are typically used for small-sensor videocams.

[rjlittlefield]

Both images were shot without moving the sample, just by replacing the camera by the DSLR + adapter & refocusing

Part of your problem may be in that bit about "& refocusing". It looks to me like your second image has developed quite some curvature of field, compared to the first image. This can happen when an objective is focused very far from its design point.

What objective are you using, and about how far does the stage move when you switch focus between the two cases?

--Rik

[yminter]

Hi Enrico, thank you for this reply; I'll try to give you a little more details below.

Does this webcam have built-in optics? I should expect it does. In this case, these optics are optimized for the sensor size of the webcam. Probably the webcam has a small sensor, but in the unlikely case that the sensor is large, in principle it would be possible to use it with an APS-C or Micro 4/3 camera body (ignoring for the moment all practical problems of connecting the equipment together). If the image quality (in the central area of the camera sensor equivalent to the surface of the webcam sensor) remains similar to the webcam image, then the fault most likely is with the new optical adapter.

You do not say whether the webcam is used together with a separate photoeyepiece - this may also be relevant.

Is the camera mounted on a trinocular head? Or do you mount the camera into one of the eyepiece tubes after extracting the eyepiece you are using for visual observation?

In this case, it is not strange at all. You are seeing one thing, and the camera sees another. You are observing the subject through matched sets of objectives and eyepieces, but the new adapter/eyepiece may not be correcting the objective aberrations.

The camera, which uses a CCD sensor, does not have a any built-in optics, there is only an IR filter in front of the sensor. It is adapted on a trinocular head, through a 0.63X C-mount adapter.
What you say about the image quality in the area of the camera sensor is also what I supposed, but I wondered if I could not have missed some detail

It could be that a 1x adapter will work better, but it is far from certain it would be without optics. If the microscope objectives require eyepiece correction, then the photoeyepiece/adapter/relay lens (different people use different names) must provide this correction. If it doesn't, this could be one of the reasons for the aberrations you are observing.

I had searched what I could find about microscope build-up and this was one of the hypothesis that looked likely but there are 2 things that make me think it is not :
* I got the problem on both the DMLM µscope & MZ16 binocular (also with a trinocular head & 0.5X Leica C-mount adapter), but with different degrees of intensity (it looks worse on the µscope); though both have infinity optical system, which I suppose requires that no such correction is possible (I think that the correction is mandatory before the straight optical flux in the tube? I suppose an aberration would be distance dependent while an infinity system allows any tube length ... ?)
* Looking with my eye through the lens inside the C-mount adapter, I did not see any inverse aberration, that I would have expected if it had corrected something?

As mentioned above, we would need to know more about the equipment and setup. But so far at this point, I suspect that the optics in the new adapter are at least one of the causes of the aberrations.

That is also what looks most likely to me as of now but since the adapter maker is no beginner and seems rather sure that it is not, I fear that there is something that I don't understand

[yminter]

Both images were shot without moving the sample, just by replacing the camera by the DSLR + adapter & refocusing

Part of your problem may be in that bit about "& refocusing". It looks to me like your second image has developed quite some curvature of field, compared to the first image. This can happen when an objective is focused very far from its design point.

What objective are you using, and about how far does the stage move when you switch focus between the two cases?

--Rik

Hi Rik, I like your idea about it because it would also explain why I get this kind of effect with different degrees of magnitude on both the µscope & the binocular

After reading your question, I went and performed a few measurements on the microscope:

The objective used for the two photographs in the 1st post is a :

Leica N PLAN 5X/0.12 ∞/-/A BD ; its theoretical work distance is 13.2mm
The focus distance is ~160µm (measured using the micrometric focus wheel) lower when I use the optical adapter & DSLR instead of the usual DC180 camera.

I also performed the same operation with the next two objectives:

Leica N PLAN 10X/0.25 ∞/-/A BD theoretical work distance : 5.2mm
The focus distance is ~45µm lower using the optical adapter & DSLR

Leica N PLAN 20X/0.40 ∞/0/D BD theoretical work distance : 1.1mm
The focus distance is ~12µm lower using the optical adapter & DSLR

[Pau]

Some complemantary comments:

- You want the camera parfocal with the eyepieces, if your dSLR setup do not accomplish this it will be much less convenient to work. And as Rik said you're pushing the objectives far from its desing point.

- The amount of aberrations shown is IMO too high to be due just to the above problem, in special with low power objectives that are usually fairly tolerant to changes of the "tube lengh"

- Most if not all infinite corrected modern microscopes have the tube lens inside the trinocular head in the common optical path for eyepieces and photoport. If the Leica infinite objectives do need any kind of secondary correction it will be done in the tube lens, so this isn't likely the problem.

- With a APS-C dSLR the total relay magnification for an eyepiece FN of 18-20mm is about 1.5X-1.6X. If your scope is a largefield one the 1X adapter sugested by the retailer may be adequate to fill the sensor. Your Leica 0.63X C-mount adapter would not provide good coverage for your Canon. If it does, likely you're placing the camera sensor at a much different position of the relay lens than expected. This may be the cause of your parfacality and aberration issues. So perhaps he is right.
You can test it removing the 0.63X adapter and holding the camera with a tripod or repro support over the trinocular tube in a position that was parfocal, if it is posible.

- If you can post pictures of the microscope with both cameras and of the dSLR adapter it may be useful to understand the problem.

[Charles Krebs]

a commercial optical adapter that is to be screwed to tha C-mount adapter

Need to know more about this adapter and the optics it contains.

If you are going to a Canon APS-C sized DSLR sensor (22.2mm x 14.8mm) you should not have the 0.63X optics in there. This is intended to make the intermediate image fit on a sensor of about 4.8 x 6.4mm (known as a 1/2" sensor), which is likely close to the sensor size in your older camera.

While we need to know what the "commercial optical adapter" is, My suspicion is that the 0.63X is primarily what is messing thing up.

[rjlittlefield]

- The amount of aberrations shown is IMO too high to be due just to the above problem, in special with low power objectives that are usually fairly tolerant to changes of the "tube lengh"

I agree. I have seen similar effects with a 10X NA 0.25 objective, but in that case the difference in focus position must have been about 600 µm (as calculated now based on recorded magnification differences). This is much more than the 45µm reported by yminter for 10X.

--Rik

[yminter]

Here is a detail of the C-mount adapter; I had made a mistake between this one and the one of our binocular : this one is a 0.5X and not a 0.63X!
I don't understand the end of this marking (the ^= 16 ) if you know what this means ...


And another one of the whole setup.

I tried to place the adapter and DSLR directly over the tube => it works without image aberration but there is a large focus offset : to get it parfocal with the oculars it needs to be around 10cm over the upper end of the tube.

=> It looks like it's clearly the 0.5X C-mount adapter that creates the problem.

I tried the same with the binocular but did not succeed in getting an image but since the trinocular head has a tilt capability, it's rather difficult to place it in a stable manner.

[Charles Krebs]

Are there any optics in, or markings on the "black tube" adapter attached to the camera?

In any event, as I mentioned earlier, the 0.5X C mount adapter doesn't belong there when using a camera with an APS-C sized sensor. That adapter is made for cameras that use a 1/2" sensor (6.4x4.8mm, diagonal of 8mm).

I don't understand the end of this marking (the ^= 16 ) if you know what this means ...

This indicates that an attached camera with a 1/2" sensor will record a 16mm field number (16mm diagonal of the image produced by the objective. Most 10X viewing eyepieces "see" a diagonal of 18 to 20mm).

This PDF is pretty basic but may provide some useful information in regards to your problems:
http://www.krebsmicro.com/pdf/trinoc_a3.pdf

[yminter]

Are there any optics in, or markings on the "black tube" adapter attached to the camera?

Yes : The adapter has optics inside.

In any event, as I mentioned earlier, the 0.5X C mount adapter doesn't belong there when using a camera with an APS-C sized sensor. That adapter is made for cameras that use a 1/2" sensor (6.4x4.8mm, diagonal of 8mm).

I don't understand the end of this marking (the ^= 16 ) if you know what this means ...

This indicates that an attached camera with a 1/2" sensor will record a 16mm field number (16mm diagonal of the image produced by the objective. Most 10X viewing eyepieces "see" a diagonal of 18 to 20mm).

This PDF is pretty basic but may provide some useful information in regards to your problems:
http://www.krebsmicro.com/pdf/trinoc_a3.pdf

Thanks for this explanation; indeed this points once again the fact that this adapter is not the right one for an APS-C format sensor.

The sum of all the answers finally points out that the most likely source of my problems is the inadaptation of the C-mount adapter to the sensor.

I have ordered an 1X adapter that should be better suited for this; I'll tell you what the result is when it arrives.