rudibarani Did you get to test the 294MM in Bin1 mode, too?
Hi Phillip,
The only 294 that I own is a color one coupled with a Canon-mount wide angle as an all-sky camera, so I didn't have one to spare when measuring the guiding frame rates.
I had for a long time used the 290MM and 290MM mini. Then went to the 462MC to get better near-IR guiding. Neither one allowed faster than 1 FPS sampling rate even with 0.5 second exposures, which I really need with my harmonic drive mount.
The ASIAIR finally managed with Bin2 to get to 2 FPS with 0.5 seconds with some of their cameras; thus the study that I made to find a better guide combination with the ASIAIR. Bin2 is far from ideal, because the stars are now undersampled. The 178MM is not too bad a compromise, since its pixels are smaller compared to the 290 and 462, while having a larger FOV (essential to get enough stars for multi-centroid guiding). I had to forgo near-IR guiding (miserable near-IR sensitivity from the IMX178), but near-IR was only giving me an extra 10% improvement anyhow.
I cannot guide with 1 second exposures since at the peak time-derivative of the periodic error curve of my mount, a 1 second exposure would no longer produce a round star to achieve an accurate centroid.
I would not be surprised that ZWO's own mount with large peridioc error (and on top of that, irregular non-sinusoidal bumps) will need to guide at faster than 1 FPS too. So it behooves ZWO to further improve the guiding frame rate on the ASIAIR. Afterall, the ASIAIR is just a gateway drug to sell their other products.
With my experiments with using INDIGO Sky (also Rasperry Pi 4/Rasbian based), I could get almost 9 FPS full-frame capture rate from an ASI178MM, even through an Ethernet bus from the Raspberry Pi to my Mac. So, the camera is not the bottleneck; the current code in ASIAIR is the bottleneck.
After I upgraded to an Mac Studio for my desktop, I have since converted the Mac Mini M1 that the Studio replaced to an 12V system, and I will be using that as my telescope computer, running the full INDIGO -- with that, I should have no problem getting really high feedback rates to control the mount within the Mini itself, without sending the data back to the faster machine in the house.
What I do not get is, if the displayed fps are "correct"? Even if it displays a rate of 0.2 fps, the image refreshes every 2 sec if I set the exposure time to 2 sec. It would be good to know, if it just displays the "old" image again - or if it is a new image and the displayed fps is wrong?
I was at one time (perhaps two years ago when I pleaded with their chief for faster guide frame rates) told by one of their develops to not trust the frame rate reported. Now, that was also before they went from asynchronous captures to synchronus captures (setting the camera to "video mode") for guiding.
With synchronous captures (took some of us who are outside the loop by surpise, and exposed a lot of bugs) the camera is always running at a constant frame rate. When you send a pulse to the mount, you need to wait at least a full cycle (more, if the mount responds slowly, or the pulse is longer than the exposure period) before accepting a frame again from the synchronous stream. I suspected from the behavior that ZWO was not tossing the frames that were captured while the mount was being pulsed. After a few exchanges with them, and sending them diagrams on what they need to do to fix the problem with their synchronous mode model, the problem appeared to be finally fixed.
I think you will notice that if the ASIAIR sends a lot of consecutive pulses to the mount, the reported "frame rate" slows down, unless the max pulse length is significantly shorter than the exposure time. This is because the synchronous frames that coincide with the pulses (and settling time) have to be tossed.
The problem is also worse when both RA and declination require pulses from the same frame. I notice that ASIAIR does not appear to send a pulse to both axes at the same time. It sends a RA pulse and a declination pulse (short slews) one after another -- so that adds to more frames that need to be tossed :-).
Chen
Edit: the ASI678MC is promising -- I have already asked my favorite Washington State dealer to put one aside for me. The 2 µm pixel, when binned, gives a smaller super pixel than a 178 that is also binned. Only problem is there is a small error left in the centroid even with mono binnng, and because of the smaller pixels, the FOV is also smaller than the 178. But the lower noise should be able to pull more guidable stars. And, I may be able to do near-IR guiding with it.
