Thank you that's great feedback. We've done up to 10 000 fps in the past (another lab, another system) we were indeed fine using heavy duty fiber optic white light lighting, it works fine and doesn't heat up the sample no, especially since we're in a continuous flow and the channel has a cooling system anyway (the bubble move and vibrate due to ultrasound, and the ultrasound piezo is cooled to 20°C with water circulation + big ##### Peltier). I was just hoping not having to buy one, they're pretty expensive and voluminous.rjlittlefield wrote: ↑Thu Jan 30, 2025 1:49 pmI disagree. I just now tried a simple experiment, placing one of THESE flashlight bulb LED's about 1/2 inch from a piece of copy paper with toner dots that was being imaged at 10X NA 0.28 onto a Canon R7 camera (APS-C sensor). Auto exposure produced a nice image at ISO 800 and 1/5000 second. A fingertip held in the same position is not even mildly uncomfortable. That bulb runs the LED at about 1 watt, the whole package is only 1.8 watts. So, I'm not seeing any big problem with light and heat.
The optics also seem straightforward: infinity objective with tube lens appropriately sized to fit the camera sensor. I'd put an adjustable iris behind the objective, to allow stopping down to trade off DOF against resolution.
Vibration might or might not be an issue, depending on the setup and environment. I live in a quiet residential neighborhood; vibration is no problem as long as I sit quiet. But it can get arbitrarily worse in other environments. I suspect the usual trick of a massive platform on soft supports would work fine. A platform loaded with bricks and placed on tire tubes inflated to low pressure would be a cheap test.
Touching briefly on specific questions:
1) Since it's C-mount, what kind of setup with tubes etc... would I need for, say, a x10 Olympus or something ?
Assuming the camera has no lens of its own, then the chain would be objective, iris, tube lens, empty tubes, C-mount adapter.
You can buy all that stuff at Thorlabs, or for a lot less money but a lot more fiddling you can find all the pieces at eBay and AliExpress.
2) What kind of lighting would you use, for 5000 FPS I'm guessing something pretty bright...
See experiment reported above.
3) When you film something, do you have a setup connecting your subject to your camera, like both standing on the same supports to avoid vibrations ? We're thinking translation stages for the camera, it will be X Y Z I guess, doesn't have to be motorized
That all sounds fine. Definitely put them on the same platform. You want the objective coupled to the in-focus subject frame as closely as you reasonably can, but you don't have to get crazy about it. My most common setup is similar to what's shown at https://www.photomacrography.net/forum/ ... 31#p266431 . It's regularly used up to 50X and has close to a meter of mechanical supports between the objective and the subject. That one meter measurement is essentially the circumference of a sort of big "G", with objective at the end of the top loop and subject on the little bar. In between there's a long Arca-Swiss plate with optics and camera mounted on it, that plate mounted on a harvested microscope focus block, the block mounted on an optical column, which sticks up from an optical table, which connects through some other adapters to a positioning stack that contains two linear stages, a goniometer platform, and a rotary table. All this is sitting on top of what is essentially an ordinary folding worktable. If I walk around the room then focus and framing at 10X shift around like crazy; if I sit still or leave the room then everything is fine. Your mileage will vary.
4) With those fixed tubes and adaptors I've seen, how do you focus ? Is it done with the translation stage ? Or do you have the option of a focusing ring like on any normal zoom lens ?
Most commonly it's translation stage, moving either the subject or the camera+lenses. You can get some amount of fine focus adjustment by mounting the objective on an internal focusing tube lens, but that has quite limited range.
Dunno if people here have experience with really fast cameras...
I have almost no experience with fast frame rate, only long stacks at high mag.
--Rik
Stroboscopy couldn't work either, as it would only work for periodic motions. The bubbles already move from the bottom to the top with the flow, and with ultrasound they interact in all sorts of way, so it's not periodic at all. But I'm not worried about the lighting part since we've done it before. It's really attaching a microscope objective to a C-mount camera directly, with accessories, and no microscope. That part I've never done.
From what you're saying, it's pretty straight forward, as long as I order the right gear. I'll have a look at Thorlabs stuff, thank you. I'll also take inspiration from the setup you've shown, this is great stuff.