For H-alpha Imaging: Canon 7D or Flea3 14S3/20S4?

[stu]

I was lucky enough to get my hands on a Lunt 60mm B1200 (blocking filter) 'scope that'll get here on Friday. I spoke on the phone with the main guy at OPT when I ordered and he said he'd been using his Canon 5DMkII camera, unmodified, perfectly fine with Lunt 'scopes for the past two years. So, my 7D he said should be fine, but what he really recommends is the modified webcams like the DMK 51AU02. Only he was out of stock on those.

My on-campus observatory has a spare DBK 41AU02 color that they said I could use, but speaking with the main guy there, he was saying that the Imaging Source cams are only 8-bit, which he personally doesn't like.

Fleas aren't. And, the Flea3 14S3M model, at least according to this FOV calculator, will see the entire sun due to its 1/2" chip. It's also 16-bit. Or, there's the 20S4M models that use the same chip as the DMK 51AU02 does (and would also see the whole solar disk).

Thoughts?

[RickJ]

Color cameras are poor for H alpha work because only 1/4th of the pixels are used. All the others are blind to H alpha light. While the sun is bright enough to get light through the IR filter it increases exposure time which can harm seeing as the longer exposure time give the atmosphere more time to do its dirty work. This, in turn, reduces frame rate which can be important. If you live in an area of far better seeing than I have it might work.

While 8 bit seems poor compared to 16 bit it has the advantage of speed. It takes twice as long to download each frame. To keep the frame rate up a lesser signal to noise ratio is usually the result. With read noise higher dynamic range of a 16 bit camera isn't much better than that of an 8 bit one. When making the decision be sure to compare noise figures as well as frame rate. The sun in h alpha changes very rapidly. You only have seconds to capture detail. The more frames you can get in those few seconds the more apt you are to get enough with seeing sufficient for stacking to get a very high signal to noise ratio and thus able to process strongly without creating artifacts. I like a 60 FPS camera for solar work. Helps for planets too. I've not tested the newest cameras however. But when I did a couple years ago the higher frame rate of 8 bit cameras beat out the 16 bit cameras. Don't know if that still applies or not.

For deep sky work 12 bit or more is necessary but I've not found it useful for planetary work. At least with 2 year old technology.

Rick

[stu]

Thanks Rick. What I find weird is that the DMK 51AU02 uses the exact same Sony ICX274 sensor as the Flea3 20S4 cameras. BUT, the DMK is 8 bit while the Flea is 16 bit. I guess it's a different A/D converter? Seems odd that one company would skimp on this and cripple the range of their product. The QE of this chip at H-alpha is around 20%ish based on the tech specs on the Point Grey's website.

I don't see any significant speed difference, though. The Point Grey 20S4M models report 15 fps, while the DMK 51AU02 reports 12 fps. Former is $1k, latter is $1k, both in mono.

For deep-sky stuff, I'd rather go the "real" CCD route and get an SBIG STF-8300M, but that's at least a year away. I'm trying to figure out what will give me something reasonable for Tuesday.

[RickJ]

Unless your seeing is well above mine 15 FPS is way too slow. You only get a couple seconds worth of frames to find a good one. 120 is far better than 30 in my seeing. But that's a price you pay for a large chip and reasonable noise levels. For a whole sun image you don't need resolution however. Limits future use. For low res whole sun images either is fine.

You have a lot of work ahead getting it all to work. Keep your old system on hand as backup.

Rick

[stu]

Seeing here is around 1-1.5 arcsec, though I've gotten down to 0.8ish. I see their GigE model on that chip can do up to 30fps for $1.3k, while they sell a USB3 version (which my laptop does not have a port for ...) for $800 with a different, higher Mpx chip that can do up to 60 fps. Problem with the faster chips in the USB3 is that the FOV is not enough to get the whole sun.

[RickJ]

Noise increases with speed so to keep noise low, as speed increases chip pixel count has to decrease to keep noise constant. That's the penalty for large chips. Technology continues to improve. 5 years ago only a very tiny chip could do better than 15 FPS. Eventually large chips will be able to handle the speed.

For whole sun work however resolution isn't all that important so slower frame rates are acceptable. Every camera is a compromise and always will be. Only you can decide which compromises you are willing to make.

BTW, PC slot cards for USB 3 are available for laptops. My laptop has two USB 3 ports so I've not tried the card solution. I see no reason that they shouldn't do the job. I find USB 2 external drives are far faster in the USB 3 ports actually reaching the speeds claimed for them. USB 3 drives are virtually as speedy as internal ones. If speed is important the USB 3 really delivers what it claims, at least on my machine.

Rick

[stu]

Yeah, I think what I'm going to do is see how my 7D works when the 'scope arrives tomorrow (morning? ... though UPS has consistently delivered in the AM for me every day of the week except Friday, when it's later afternoon ). Depending on that, and whether I think it's acceptable or not, I'll then decide on whether to go with something else, I think. I recognize that tech constantly improves and in a year stuff'll be 2x better (or some multiplier >1) and so I'd rather wait to get a cam that samples the sun at seeing (so, if I get 1" seeing, I'd want something 1800 px tall). Current tech for that is crazy-expensive. So I'm willing to use my 7D if it's reasonable results-wise.

BTW, PC slot cards for USB 3 are available for laptops.

Mac ...