Budget Friendly BH2 Video?

[ethicalhacker]

My sister and I have both recently obtained BH2s with trinocs. Hers is setup for transmitted light, and mine is setup for reflected light. She's just starting residency, and I still have a mortgage. We both can't afford expensive camera setups right now. We are interested in a budget friendly video camera setup, and I was curious if someone here could give us some advice. It seems as if most people go the full-frame, APS-C, or 4/3 route. Ideally, those with tighter budgets among us could use a cheaper Chinese microscope camera (typically a 1/2" sensor), a NFK photo eyepiece (not the 1.67x as its unobtanium), and some adapters to get the same field-of-view that a full-frame image sensor could. The one thing I am willing to sacrifice on is looks. I don't care if its weird (e.g. PVC). My priorities are listed in order below. Just so we're clear, I'm aware of alanwood.net and krebsmicro.com. They are great resources, but they don't seem to address this use-case. I think this is going to require some out-of-the-box thinking. If I missed something though, please point it out.

1. Budget friendly
2. "Live" (no frustrating lag if focusing via camera video feed)
3. 1920x1080 resolution @ 29.97fps minimum
4. Computer connectable (for manual slide scanning, focus stacking, live lecturing, etc.)
5. Color accurate (has the color correction from the NFK photo eyepiece or equivalent...we both have SPlans, and my sister has FL/Apo objectives...this is where our money went...we can upgrade the camera after image sensor technology changes in a few years and we become rich and famous or whatever)
6. Reasonably similar field of view to the eyepieces (not just a tiny spec out of the center..."per low/high power field" counts matter)
7. Parfocal with eye pieces

[micro]

How much are you willing to spend? 1080p will be a few hundred $ at least.

[ethicalhacker]

That won't be a problem for us, but I don't want to squash the creative process by giving a particular number. Someone else reading this thread will have a number lower than ours.

[micro]

If you want 1080 and no lag while live viewing you need a camera $400+, Elgato Cam Link for connecting to monitor with hdmi cord for live viewing $100+, adapter camera attachment prices vary $50-100? Additional optic possibly required depending on camera $50-100
You can get live viewing with just a $250 or less usb camera but for HD no lag you need to go the dslr or 4/3 route. There are expensive usb cameras that claim they have HD footage and all that and cost $500+ but I wouldn't risk it with the mixed results people get with those.

[ethicalhacker]

...adapter camera attachment prices vary $50-100? Additional optic possibly required depending on camera $50-100...

I think the devil's in the details here. As per Alan Wood, the APS-C and 4/3 sensors are best matched to the NFK 1.67x photo eyepiece, which can go for as much as $1K USD. For that price, one is probably better off just going with a used full-frame sensor and a NFK 2.5x photo eyepiece plus adapters, but that fails the budget friendly test. Hence this post. Am I missing something here?

[micro]

With my setup I don't use additional optics. I have my 4/3 camera attached to the trinoc port with just a tube that has no optics.
These are screen shots from videos I took with my lumix g85
viewtopic.php?f=6&t=10810&p=90632#p90632

and here is a video

[micro]

If you got a panasonic g7 body only $450, high speed memory card $100, elgato cam link $110, camera adapter attachment $50 it would cost about $700-800 and you would basically have my setup.

[MichaelG.]

With my setup I don't use additional optics. I have my 4/3 camera attached to the trinoc port with just a tube that has no optics.

.

So you are effectively using the microscope objective as a macro lens
... do I understand correctly ?

MichaelG.

[micro]

With my setup I don't use additional optics. I have my 4/3 camera attached to the trinoc port with just a tube that has no optics.

.

So you are effectively using the microscope objective as a macro lens
... do I understand correctly ?

MichaelG.

Oh I don't really know actually.

[jfiresto]

With my setup I don't use additional optics. I have my 4/3 camera attached to the trinoc port with just a tube that has no optics.

So you are effectively using the microscope objective as a macro lens
... do I understand correctly ?...

Indeed, to the camera, the microscope looks like an exceptionally steady, manual focus lens – so turn off in body image stabilization!

[Tom Jones]

Easily done.

Panasonic GH4 (~$500-600 used), M4/3 to C-mount adapter(~$10, Fotodiox, Fotasy, etc.), NFK 3.3x (~$50-75), U-PMTVC 0.3x C-mount to microscope adapter (~$100-125). The combination of the 3.3x and 0.3x is close to 1x, so you have essentially direct projection, with the color correction necessary from the NFK 3.3x projection lens. Lots of the 3.3x around, pretty cheap. The U-PMTVC is focusable so parfocality is easy. Quite a few of them around, too. It's actually a BX series component, and equivalent to the BH-2 series MTV-3, but focusable and essentially the same price.

That gives you nice 4K, or less if you want it . HDMI out to an external monitor (HD or 4K), with or without the camera screen information. It will feed HD out while recording 4K if you don't have a 4K monitor (auto-switches, actually). The image circle matches the sensor, and the eyepiece FOV very well. Plays well with my SPlans. The NFK 1.67x isn't necessary, and actually doesn't match the M4/3 sensor very well. I have one and don't use it. It's designed for the APS-C frame size which is significantly larger than the M4/3 sensor.

Batteries will last more than an hour, and a 64gb card will last over an hour as well if you are recording in 4K. An external 120v adapter is available too. It doesn't over heat either. I've gone well over an hour recording continuously with no problems. Very nice color. Better than C-mount cameras. Files are relatively easy to edit. Easy setup, and it will work very well for lectures and presentations. I've used it for both. You don't have to record. It will feed your monitor/projector as long as the battery lasts.

That is exactly the setup (including the 4K) I used with my BHS for this. Watch in 4K if you can, it makes a difference: https://www.nikonsmallworld.com/galleri ... er-feeding

Here's the full version if you're curious. The Nikon Small World submission was limited to 1 minute with no titles or music: https://www.youtube.com/watch?v=mqZns4wo0AQ

This used the same setup on an Olympus SZH stereo scope: https://www.youtube.com/watch?v=OMNHuab9hKE

I use the direct projection method (without the projection lens or the adapter lens) using a Diagnostic Instruments DBX adapter to the C-mount for my BX-50s.

You're going to love 4K. Much better than HD for very little more money. It's not that HD isn't adequate, it is, but the resolution doesn't match what you can see through the eyepieces very well. 4K is SO much better! If you want to get an idea how much (ignoring YouTube compression issues) watch the above videos in HD first, then switch to 4k and watch them again. It's actually better than that.

Oh yeah, toss in a camera lens or two and you have a nice regular stills and video camera, too, not limited to use on your microscopes . It's not some limited-run Chinese multi-branded camera with little to no long-term support.

Tom

[ethicalhacker]

I have my 4/3 camera attached to the trinoc port with just a tube that has no optics.

Nice phase objective, micro! I can see your solution being viable for others too. What adapters/tube do you have in between your camera and your trinoc?

Panasonic GH4

Tom, amazing scope and footage! What is the reason you recommended this particular camera body? Are there other competitive options that would have the correct mount and distance, etc. too? (I'm camera illiterate.)

NFK 3.3x

Great! I happen to have one of these.

U-PMTVC

I was aware of these, but you've tempted me further.

The image circle matches the sensor, and the eyepiece FOV very well.

Would you be so kind as to elaborate? Is there vignetting?

An external 120v adapter is available too.

Thanks for mentioning this. I'm one of the few that prefers wires over batteries.

You're going to love 4K. Much better than HD for very little more money.

When I was told to work entirely from home due to the pandemic, I upgraded to a 43" 4K monitor, and I do like it. I didn't list 4K as a requirement to reduce cost.

Olympus SZH

I might have to get me one of these some day.

It seems like ~$800 is the budgetary estimate I'm hearing from both micro and Tom. That's probably the upper limit of what I'd consider budget friendly, but I think it's reasonable.

[Tom Jones]

I recommended the GH4 because I've had one since they were new, and I know it works well in this application. It's also pretty cheap (used) right now. Origianally it was recommended to me by a microscopy/photography professor friend. The GH5 is newer, same sensor and mount, but probably ~$1500. I know they work and am not usually in the habit of recommending something I haven't tried. I used a full HD C-mount Tucsen camera for a while but sold it after I went with the GH4.

No vignetting in UHD 4K. I don't think there is in DCI 4K, but I don't think I've tried it. I have not tried the lower resolution modes. Since I bought it for 4K, I shoot in 4K. What you see in the videos is straight off the sensor, no cropping.

Since you say you are camera illiterate, you are probably not aware that digital sensor size is not the only variable in matching projection image size with the sensor. Each make and model camera have different sensors, pixel sizes, and a lot of variable resolution options, particularly in video. If the manufacturer uses an actual pixel-based resolution, which is the most common method, size needs change at least a little with each resolution on the same physical sensor.

If they line-skip or pixel bin, the sizes stay the same since the full size of the sensor is always used. For example, if a camera like the Canon R5 uses full-frame for 4K (binned probably) or 8K, but in 4K 120 frames per second uses actual pixels in a Super 35/APS-C mode, the projection requirements change. An NFK 2.5X for full frame, then the NFK 1.67x for the 4K 120fps mode unless you're willing to accept a ~1.5x crop.

All of that makes for an interesting mess if you want to add accurate scale bars to your video or stills!

Do a quick Google search for each of the video terms I've used, as well as sensor sizes and resolutions. After wandering around the rabbit holes for awhile you'll get a better explanation than I could give you.

Tom

[jfiresto]

... I use the direct projection method (without the projection lens or the adapter lens) using a Diagnostic Instruments DBX adapter to the C-mount for my BX-50s.

I found a DMC-GH3, which is too old for 4K, gave minor vignetting:



Its live HDMI-out decimation could also be better. I ordered a newer model in April, I thought would be much better, but can not get to it because of the pandemic. From your reports, it should be the business. Thanks!

[ethicalhacker]

Since you say you are camera illiterate, you are probably not aware that digital sensor size is not the only variable in matching projection image size with the sensor.

Nailed it! The fact that a camera with a full-frame sensor might only use the center most APS-C surface area of the sensor while in certain modes was news to me. I'm sure some folks have had to figure that out the expensive way. Whew!

All of that makes for an interesting mess if you want to add accurate scale bars to your video or stills!

Would using a calibration slide be an easier way around this problem?

[micro]

I forgot you can plug the camera directly into a monitor without the elgato cam link. I use the cam link because then I can use the computer at the same time without it just being hooked up to the camera. I also use it for streaming the microscope footage. So that makes the setup $110 cheaper. The tube attachment I have is from some third party company I'll have to look it up in my internet history.

[micro]

https://www.ebay.com/itm/Olympus-Micro- ... 4914938000

This adapter isn't parfocal with the eyepieces but the images I get look good so I never bothered trying anything else. And Tom is right about 4k looking better. I recently started experimenting with 4k more and it seems to look far better than my old 1080p footage. I'm kind of mad I wasn't always shooting in 4k. If you buy a used camera find out how many pictures and videos it has taken. Cameras do get old and die.

[Tom Jones]

I think using a calibration slide is the only way to get accurate dimensional information for scale bars.

Basically,

Photograph the calibration slide with the objectives you intend to use (or all of them to be complete) with the same camera and projection setup you will use with the specimens. If you use a mag changer, be sure to include photos for each objective and each position of the changer. Make sure the camera resolution is the same as you intend to use, so the sensor area is the same. If you want to use multiple resolutions, you will need to do this for each resolution, with each objective, and each mag changer position. And yes, it's a pain...

Take each photo in your editing software, and measure the number of pixels between any two (or more) of the calibration lines on the calibration slide. I'd do it multiple times and take the average.

Those numbers provide the pixel dimensions to calculate whatever size scale bar you want to include.

In your photo or video editing software, draw a line the appropriate number of pixels for the scale bar. If 50µm takes 521 pixels, and you want a 100µm scale bar, draw a line 1042 pixels in length. Make sure you label it somewhere in the image. Make sure you either add it before any resizing is done to the image, or rescale the bar exactly the same as the image after resizing.

Not too tough until you try to figure out how to count actual pixels in an image, and draw lines a specific number of pixels long. Or have multiple microscopes with mag changers and multiple objectives, and want to use different resolutions say for regular and slow motion. Then decide to get a new camera and have to start again... And obviously the accuracy is up to you. For most of this stuff, close counts. Hopefully this make sense.

Tom

[ethicalhacker]

Tom, you mentioned a magnification changer. I've considered getting one because of the built-in bertrand lens. (I also happen to have a phase condenser laying around.) My question now is if you find it necessary with the camera setup you recommended above? Does a magnification changer help solve some additional problem that I haven't thought of yet?

[Tom Jones]

A mag changer can help with photographic framing. You can actually do the same thing with different projection eyepiece magnifications too.

I originally got one to help with framing specimens for outreach demonstrations, so the kids didn't have to search so much for what I was trying to show them.

And no, it's not necessary. Some say it's not even really useful as it results in empty magnification.

A better solution is a full range of objectives!

Tom

[75RR]

Panasonic GH4 (~$500-600 used),

Slightly off topic question - are you happy with the Electronic Shutter (Silent Mode) of the GH4?

[Tom Jones]

I only use the GH4 for video on the microscopes, so I have no experience with it's electronic shutter in the photography mode.

Tom

[75RR]

.
Thanks for your reply, 4k video really is an amazing step up from 1080.

Might risk it anyway if I run into an economical one, can't kick more than my Olympus E-P2 ;)

[bphuber]

I conduct microscopy classes for my local park district and I purchased a number of scopes as parts on ebay and made 14 working scopes out of the 20 or so total that I bought.
All have trinoc heads, and most are various vintages of Olympus scopes, most being BH(A, B or C). I purchased a few KHC and KHS (B&L Galen III) scopes as well.
The Olympus Photomicro Adapter L has been talked about as the necessary accessory to get parfocality with an Olympus scope. That isn't necessarily the case- different C-mount rings stacked together ending in a C-42mm adapter to the camera mount of your choice works well. A relay eyepiece isn't always needed- this depends upon the sensor in the camera.
I happened to be an old Olympus OM-1 camera user from the 70's and migrated to M43 bodies a few years ago and that's what I use on my scopes.

One of the other useful accessories I've found is the PM-10. This was designed back in the film camera days to attach a specific Olympus film body. Removal of the electronics, throwing away the Olympus film body and purchasing a 67>42mm adapter (Amazon) gives a nice solid mount for a camera with V-C-CV viewing selector in the body of the PM-10. There are a number of PM-10 versions, I have purchased variations of them on ebay and most cost ~$30. There is also a PM-6, which is a manual version of the PM-10. It works just as well as the PM-10... Add a camera adapter to the 42mm adapter and you're in business for ~$50.

I've also found that variable length extension tubes (ebay) work nicely to tweak up any needed length adjustments for parfocality... I have modified ~9 scopes like this, it is pretty easily done.

[ethicalhacker]

Removal of the electronics, throwing away the Olympus film body and purchasing a 67>42mm adapter (Amazon) gives a nice solid mount for a camera with V-C-CV viewing selector in the body of the PM-10.

I happen to have an unused PM-10AD in a box within arms reach. Would you be so kind as to explain how this could be modified and list all adapter(s) and camera(s) you'd use with it? A few pictures might be helpful too since this sounds like surgery.

[Scarodactyl]

I wrote up a brief guide here
viewtopic.php?f=9&t=9093
It is nothing too difficult.
The relay eyepiece does some essential corrections, particularly in the periphery, so not using one might give some issues unless you capture only the very center of the field of view.

[ethicalhacker]

I wrote up a brief guide here

Thanks. Your write-up mentions a SLR and a [Canon] "EOS" adapter (presumably an EF mount). Are you using a full-frame or APS-C sensor, and what photo eyepiece are you using? I'm assuming that going with this approach discourages the use of a micro 4/3s camera because of the reduced field-of-view. I also assume that the ideal setup with a hacked PM-10 would include an NFK 2.5x photo eyepiece and a full frame DSLR camera to maximize field-of-view, right?

relay eyepiece

Are you referring to the photo eyepiece - not the thing sticking out the front of the PM-10? (view finder?)

[Scarodactyl]

I wrote up a brief guide here

Thanks. Your write-up mentions a SLR and a [Canon] "EOS" adapter (presumably an EF mount). Are you using a full-frame or APS-C sensor, and what photo eyepiece are you using? I'm assuming that going with this approach discourages the use of a micro 4/3s camera because of the reduced field-of-view. I also assume that the ideal setup with a hacked PM-10 would include an NFK 2.5x photo eyepiece and a full frame DSLR camera to maximize field-of-view, right?

Correct, I used it with a 2.5x, on aps-c though it's meant for full frame.
For your usecase the most ideal combo is likely a 3.3x nfk eyepiece with an mtv3 reduction lens, but that is just what I have heard secondhand--I have never used an m4/3 camera. It is also kind of expensive.

relay eyepiece

Are you referring to the photo eyepiece - not the thing sticking out the front of the PM-10? (view finder?)

Correct, the photo eyepiece. The one on the side is a viewfinder (which is inoperable after the beamsplitter comes out and is thankfully obsolete. I can't imagine what the film era was like for photomicrography).
You could also use a viewing eyepiece in the trinoc port with a lens on your camera to focus the image onto your camera sensor (the 'afocal' method). The viewing eyepieces are not as well corrected as the nfk photo eyepieces, and a 10x will not sit low enough in the trinoc port--a 15x will though. I have not tried this on an olympus or on a m4/3 camera but it might be the best bang per buck.

[bphuber]

I have a BHA scope with the PM-10 sitting on the trinocular tube. I've removed only the electronics and retained the slider for the beam splitter. (I wanted to keep the eyepiece on my scope to verify parfocality easily, and yes my M43 camera is right in focus). The eyepiece I use is nothing. I tried .5X, .3X and decided I didn't need anything.
I tossed all the electronics after looking a bit at the cool shutter, but couldn't justify keeping something I didn't want to tinker with...
The beam splitter is redundant with the trinocular head, but I figured I'd keep it anyway.
I've attached some photos- they are of various PM-10 versions. Two photos show inside of a PM-10 (one with beam closed, the other open. The last photo shows a removed shutter from a PM-10ADS.

[ethicalhacker]

I only use the GH4 for video on the microscopes, so I have no experience with it's electronic shutter in the photography mode.

The Panasonic Lumix GH4's sensor resolution is 4620 x 3474 and has a maximum image resolution of very close to that (4608 x 3456). The actual image sensor is 17.3mm x 13mm with a 21.64mm diagonal. However, when shooting in either UHD 4K (3840×2160) or DCI 4K (4096×2160), there is a crop factor because the pixels are taken from the center of the image sensor area. For UHD 4K, that would be the centermost 13.38mm x 8.08mm with a 15.63mm diagonal, and for DCI 4K, that would be the centermost 15.33mm x 8.08mm with a 17.33mm diagonal. If one takes the maximum image horizontal resolution and divides by a horizontal crop factor of 1.2 for UHD 4K or 1.125 for DCI 4K, one ends up with the exact horizontal video resolution. If one takes the maximum image diagonal (21.56mm) and divides by a diagonal crop factor of 1.38 for UHD 4K or 1.24 for DCI 4K, one ends up with a number close to the video diagonal dimension. HD will be different. For example, I've read that the "Ex. Tele Conv." menu option has a significant effect on the dimensions of the sensor area used.

The Panasonic Lumix GH5's sensor resolution is 5196 x 3907 and has a maximum image resolution of very close to that (5184 x 3888). The actual image sensor is the same size as the GH4; however, unlike the GH4, when shooting in either UHD 4K or DCI 4K, the GH5 has effectively no horizontal crop factor because the maximum image sensor surface area is used. The camera employs subsampling to achieve the appropriate resolution. In fact, a later firmware update allowed video capture of the maximum sensor resolution (4:3 aspect ratio). (It sounds like the HDMI output is disabled in this particular mode though.)

I don't know what a NFK 3.3x LB photo eyepiece, a U-PMTVC, a C-Mount to micro four thirds adapter, and a GH5 in 4K should produce because I don't know what the diagonal of the light hitting the image sensor actually is. I don't know how to mathematically relate that to the various NFK LB photo eyepiece magnifications, the U-PMTVC's 0.3x reduction lens, and how any of that compares to what one sees through the WHK 10×/20 L eyepieces. Could someone show me the numbers and explain the math? The GH5 is probably outside of my price range, but I'll likely have to go through this exercise with another camera eventually anyway.

As a bonus question, it's common for people to use "speed boosters" with micro four thirds cameras in combination with full frame lenses. I am thinking "speed boosters" are a reduction lens and spacer in one package. Could one use this to combine a micro four thirds camera with a BH2 setup for a traditional full frame camera? Given the current prices, I doubt this would be more cost effective, but maybe that will change in the future.