ASIAIR PRO and Guide Camera Binning: Page 2

ASIAIR PRO and Guide Camera Binning

w7ay

JeffMorgan Chen that is an interesting observation on the IR cut filter that may have broader application.

Atmospheric turbulence is wavelength dependent (inverse relationship). So, 700nm should have appreciably better centroids than green light (say, 520nm).

That said, multiple stars also should have reduced atmospheric turbulence ("seeing") -- the variance of the measurement error fall by two (RMS of the measurement falls by sqrt(2)) for each doubling the number of stars, if the stars all have the same signal to noise ratio. ASIAIR uses SNR weighted centroid for multi-star centroids, so in general, the improvement will be less than sqrt(N).

Still, I was pleasantly surprised that near-IR guiding actually helped (a little) even for multi-star case. But you need to carefully consider the loss of SNR (when you cut off most of visible spectrum) vs any improvement in atmospheric turbulence. It is not a slam dunk monkey-see-monkey-do case.

The 290 has pretty small pixels when it comes to star size. The first zero of the Airy ring is about 2.5 pixels (i.e., main lobe of Airy ring is a full 5 pixels in diameter) with an 85mm aperture and a 900mm focal length. Typical "seeing" could further bloat the diameter star to around 8 pixels (HFD of about 4 pixels) since the "seeing" convolves with the native point spread function of the optics.

Additionally, with an OAG, you will be looking at the edge of the image circle (where aberrations are not well controlled), so you end up also with not just dimmer stars, but also stars with hard to quantify centroids -- that is why I personally avoid OAGs as much as I can.

Chen

w7ay

Just a quick check since it was clear outside. This is with a 7 kg or so payload (OTA + dual saddle plate + Borg 55FL as guide scope) on a RainbowAstro RST135 mount. Guide scope focal length is 251 mm with 685nm IR cut filter and ASI290MM (USB 3 version) at bin2. Mount and a William Optics Mortar tri-pier has been left outdoors for 24/7 for a over a year now (two summers and one winter) under dry bags. The electronics (ASIAIR/eero router/USB hub) are left outdoors too inside a waterproof enclosure at the base of the tri-pier.

The last time polar alignment was performed was in early September (so it is probably off by a couple of dozen arc seconds).

When calibrating at 0.5 second exposure, ASIAIR never went above 1 FPS. However once it started guiding, I got 2 FPS with 12 guide stars (a couple of guide stars are just visible behind the guide graph)

Gusty outside to around 15mph, so I backed off the feedback gain (aggressiveness) to 35%. It might guide better if not for this.

Guiding around the region of NGC7000 (my latitude is +45.45º, so NGC7000 transits close to Zenith), with North America about 2 hours west of the Meridian. I had earlier calibrated near 0º declination with pier west, and did not recalibrate, and simply asked ASIAIR to flip the guide pulses.

This (0.5 sec exposure, 2 FPS) seems better than my average before v1.8 (exposure 0.5 sec, 1 FPS); but this is just one night, so take it with a grain of salt.

BTW, the Guide Graph won't update faster than 1 FPS either. So, you more often than not, see multiple guide pulses for each guide graph refresh. Notice lots of guide corrections, but the RMS is satisfactorily low.

So, no problem with guide resolution in my case even with 2x2 binning with a 2.9µm sensor.

Next possible improvement is for guiding in ASIAIR to support dark frame subtraction.

Chen

w7ay

Looks like 2x binning is still in flux with ASIAIR v1.8 Beta TestFlight releases.

In the latest beta (Build 6, Firmware 8.79), the Bin2 switch has been moved to "Experimental Features" in the "About" window. It is no longer in the Guide Window.

Weird, since 2x2 binning was working perfectly for me (see recent posting above) when I tried using it a couple of nights ago. Arguably the most consistent guiding I had gotten with the RainbowAstro RST-135 on ASIAIR.

The next chance I get, I will be trying to use the ASI462MC with 2x2 binning. Rains and clouds for at least the next 10 days, though.

If the binning is implemented as mono-binning, the ASI462 could be much better for near IR guiding. In the past, with 1x1 binning, the ASI462 had some inconsistency, possibly because of the underlying Bayer filter (non-round star shape, so inconsistent centroid with field rotation). With 2x2 mono binning, that should be gone, and since it has the same pixel size as the ASI290, I should also have sufficient resolution. With higher sensitivity in the near IR region, and I may be able to even push the filter to a 720nm IR pass.

BTW, I have since also experimented with a Mini (USB 2) version of the ASI290MM and I was also been able to get 2 FPS with 0.5 second exposure.

Unfortunately, an ASI485 only managed to achieve 1.6-1.7 FPS when using 2x2 binning. The largest camera that I have that will still achieve 2 FPS with 2x2 binning is the ASI178 (a slightly larger FOV than the ASI290/ASI462). The ASI178 has much larger read noise, however, which makes it less useful for short exposures. A shame.

Chen

astrosatch

w7ay In the latest beta (Build 6, Firmware 8.79), the Bin2 switch has been moved to "Experimental Features" in the "About" window. It is no longer in the Guide Window.

There was another plate solve routine in experimental features some time ago. I wonder what happened with it. I hope bin feature won't disappear too.

w7ay

astrosatch I hope bin feature won't disappear too.

I suspect it will stick around, but who knows; some rank beginner may be more confused by what it does. As it is, so many people are having problem just following instructions on how to autoguide, in spite of the instructions always appearing each time you even open the autoguide window -- albeit, the icon for recalibration could probably be clearer (like a ruler icon) instead of an "i" in a circle.

Fortunately, ASIAIR is not the only system that offers autoguiding.

Chen

PMTeam@ZWO

w7ay albeit, the icon for recalibration could probably be clearer (like a ruler icon) instead of an "i" in a circle

It will be updated in the next update.

SpaceGoat

w7ay Any recommendations on a IOS app for autoguiding? Thx!

w7ay

SpaceGoat Any recommendations on a IOS app for autoguiding?

The iOS app does not do autoguiding. The tablet simply provides the display and parameter setup functions for ASIAIR's multistar centroid image capture and PHD2 backend. The autoguiding is done as an ad-hoc discrete feedback system in the ASIAIR processor.

Read here to see the GUI for ASIAIR's multi-star guiding:

https://www.yuque.com/zwopkb/asiair

See here to see what some of the parameters in PHD2 do:

https://openphdguiding.org/man-dev/Advanced_settings.htm

The Open PHD Guiding website above has many more documentation.

The Open PHD documentations describe what to do to calibrate the autoguiding. The guiding parameters are not dependent on iOS, but rather on your guide scope/camera's plate scale, the amount and precision of the guide pulses passed to your mount, the amount of torque and backlash from your mount (balance is only a problem if the mount's motor does not supply enough torque), and atmospheric turbulence ("seeing"). Although that last item is much mitigated by using measurements of the centroids of multiple stars ("multi-star guiding").

And finally some understanding of what a German mount's meridian flip does to the guide pulses before and after the meridian flip. There is plenty of documentation on the web on what a meridian flip does, all the way back to Fraunhofer's description of his original mount.

So, you cannot simply copy someone else's parameters in a monkey-see-monkey-do fashion. Even nightly variation in "seeing" will affect how well you can guide (and the need to back off the aggressiveness when seeing is poor from night to night, for example) and even the camera gain will depend on the astronomical transparency of the sky that changes with time and even direction.

Without autoguiding, tracking depends on the precise rate of your RA motor (the Perioidic Error of gears, for example), and the precision of the polar alignment. With autoguiding, the rate of the RA motor is no longer a factor (i.e., no translation term problem); however, image rotation will still occur with very long exposures, commensurate with the polar alignment accuracy.

Chen

Aizqpor

Hi, since the new version 1.8 my autoguide doesn´t work fine, it´s horrible. What happens with this?

DavidGom1

Aizqpor The same thing happens to me and to everyone I know, they have noticed a change for much worse.
I have a friend who bought an asiair plus and is going to sell it because he does not get the same guidance as before with nina.
Feel frustrated

Ekato

With 2x binning I am getting the best guiding ever. Very stable. Great job Zwo team!

Eriktim82

Ekato I agree my guiding has been better since 2x2bin. The only that bugs me though is that fact I have re-enable 2x2bin mode Everytime it boots up. Same thing about having to re-enable USB storage

astrosatch

Did anyone notice that rms numbers with bin2x feature are not right? I checked phd2 logs with ph2 viewer and it shows approximately 1.5-1.8 times higher values than asiair vhen bin2x is checked. I was amazed when tried bin2x for the first time, but now I'm not sure anymore how to interpret the numbers. There is obviously some calculation error in how asiair makes those numbers.

kvastronomer

There is an opinion on CN forum that AAP 2xbin guiding numbers are misleading. It almost feels like ZWO forgot to multiply guiding accuracy by 2.
https://www.cloudynights.com/topic/798243-new-asiair-bin-2-guide-feature/

w7ay

kvastronomer

2x binning will not improve guiding accuracy per se. In fact, for some guide plate scales (short focal length/large sensor) it can make guiding much worse. (I.e., if the star mass is mostly inside the 2x2 super pixel, then there will be a hysteresys in the centroid computing - a star can move and the centroid won't change if the whole star mass is still inside the superpixel).

If SNR is good to start with, 2x binning will at best be only as good as 1x binning.

There are really only three reasons I can see why you use 2x binning.

(1) The first is 2x binning can give you better SNR to get you more stars when using multi-centroid guiding. But only if the guiding plate scale allows it. Star sizes still needs to be larger than the 2x2 superpixel. Otherwise, the guiding will be worse, not better! If you can already guide with 12 stars, additional SNR will not help that much (a little since the multi-star centroid that ZWO uses is SNR weighted).

(2) if you have a tiny pixel size relative to the focal length of the guide scope. This is probably true for most OAG users whose focal length are longer than 600mm. Most OAG with 2.4 µm sensors are severely oversampled -- the pixel is so small that it sees only a small part of the photons from a star and thus decreases the SNR. So, a 2x binning will help in those cases since the superpixel will catch more photons.

(3) sampling rate -- and this really applies only to the ASIAIR. The ASIAIR is severely crippled -- at 1x binning, you cannot get a guiding frame rate that is faster than 1 FPS even when you drop the exposure time to 0.5 seconds (believe me, I had been bugging ZWO's head bottle washer for more than a year now about the poor frame rate). ASIAIR must be doing in-camera binning, since I now see a frame rate of 2 FPS with 0.5 second exposure by using 2x binning. Even so, don't expect miracles, I am only seeing a 10%-20% improvement in guiding with 2FPS even with a harmonic drive mount (the large periodic errors of harmonic drives mounts require as fast a feedback loop sampling rate as possible).

In general, this last class of improvement will not be seen with most mounts, so most of the gains with 2x binning will be for people using an OAG with long focal length and small pixels. RainbowAstro and Hobym mounts (and perhaps the Avalon M-series) are probably the only mounts that benefit from 2 FPS feedback rate. For mounts with backlash, the faster sampling rate can even hurt.

Chen

« Previous Page