|
At GRAS, the satisfaction of our subscribers is one of our highest priorities. Consistent with that idea we were delighted when Bob Denny, the author of ACP, offered to participate in the implementation of ACP at our GRAS facilities.
We choose to install ACP due to the proven reliability and robustness of this fantastic application. One of the other reasons was the unique manner in which Bob supports his software. We have found him to be extremely cooperative, supportive and available. Purchasing ACP has turned out to be like subscribing to a service as opposed to simply buying a software package.
One just has to read Bob’s following detailed and informative response to John Ebersole’s original note on the preceding page to fully understand the depth of his commitment to user satisfaction. I think you will quickly agree the additional information provided and the more complete explanations contributes to a greater understanding into the logic and the mechanisms that make ACP such a fine product.
On behalf of the whole GRAS Team, I would like to publicly offer our thanks and appreciation to Bob for his proactive support of ACP.
Arnie
The GRAS Team
John (et al.) -
I understand your frustrations and want to help you and the rest of Arnie's users to have the best possible experience with the new ACP-based systems. I read through your message, got Arnie to send along the log files from your runs, and asked him to pass along the guiding settings in use at AREO-3. After reviewing all of this I have some observations:
1. The focusing between filters is not necessary. ACP has a facility for using relative focus offsets when switching filters, eliminating the need to refocus on a filter change. This will save lots of slewing, plate solving, recentering, and focus time (probably 2 minutes per filter change). At present, AREO-3 does not have this configuration set up, so that's why you are seeing all of that motion and focusing. Arnie's going to get this set up asap.
Thanks to Bob for pointing out this oversight on my part, this has already been accomplished in both G1 and G3.
2. ACP is designed to manage guide star centering and acquisition automatically, and thus it has no controls for manual centering. It's simply not needed. If the coordinates and position angle (rotation against the sky) YOU give for a target are correct for a guide star to be on the chip, it WILL be on the chip. That was not the problem.
3. ACP doesn't require a brighter guide star than BA. Actually it's MaxIm DL that does the guiding, and its requirements for guide star "brightness" are on a par with CCDSoft. ACP does automatically choose the guiding exposure interval based on a test exposure, picking an isolated guide star (with several constraints), measuring the signal to noise ratio (SNR) of the chosen star, and calculating the guide exposure to yield a target SNR. In looking at your log file, I see that ACP had indeed acquired a usable guide star:
04:02:14 (long exp(s), no orbital tracking, trying to autoguide)
04:02:34 **No stars found in guider image
04:02:34 No usable stars in guider image; trying again at max guider exposure
04:03:40 (guide star SNR=3.0 X=313 Y=241)
04:06:02 **Script Error (Tracking has been stopped)**
04:06:02 Source: ACP
04:06:02 Message: The script was aborted.
The settings at AREO3 are 5 seconds for the initial test exposure, and 30 seconds for the maximum guiding exposure. The above log shows that the initial test exposure yielded no guide star. But the next line says it's going to try again at the maximum guide exposure (30 seconds). The next line shows that it DID acquire a guide star at X=313 Y=241 with an SNR of 3.0 (the target SNR that they have set for AREA3).
Once the star is acquired, ACP switches MaxIm to tracking mode and waits for up to 20 guiding cycles for the guiding error in BOTH X and Y to stabilize below a pre-set limit before starting to image. I forgot to ask Arnie what that limit is set to. I'm guessing it is a bit tight. Therefore, the guider was running, tracking a usable guide star, but the xy errors both didn't get below the threshold before you killed the run. I'm guessing that it required something like 20+ seconds guiding exposure for that Ha filter and the star you chose to get to SNR 3.0. You killed the run after 160 seconds (120 sec on another run) which would be 8 guide cycles or so.
It would be benificial to do a bit of engineering work on AREO3 to find out the faintest usable guide star magnitude for Ha. Once known, you guys could avoid the situation where the chosen guide star is too faint for Ha under any circumstances. While planning, you'd just know you need to find one of sucnandsuch mag or brighter period. It would also be useful to look at the max guider error on startup to see if it could be relaxed a bit.
But the bottom line is that imaging via Ha is going to require either a "bright" guide star (so the guide cycles can be shorter) or a longer time (longer guide cycles) for faint guide stars. There's no way around that tradeoff, and I claim that ACP's calculated exposure interval based on SNR gets you to the shortest possible guiding cycle time more consistently than you could manually. The guiding servo's tracking loop performance is dependent on the guide star SNR, so by controlling that tightly ACP provides the most consistency.
4. The thumbnail and 512x512 preview images are highly compressed JPEG, and yes they are lossy and compressed. The system is designed for use over slow (satellite and dial-up) links. But the idea is that ACP is a "set and forget" system. The previews are there to detect gross problems like tracking loss, light leakage, etc.
I could double or triple the size of the big image and it would have better resolution, though it would download slower. Once you get used to ACP, if this continues to be a problem for GRAS folks, I'll consider an option for that.
The system is designed to really truly just run and do what you ask of it without manually jiggering things. Target positioning will be virtually exactly what you tell it in your observing plan or via the web forms. The micro-slewing accuracy of the Paramount is outstanding. Your images should (and will!) be well within 0.5 arc min of where you specify.
Imaging via narrowband filters (Ha, O3, etc and an internal guider is something that a bunch of my customers routinely do. The trick is to determine the faintest guide star mag that will work for each filter on that particular observatory's scope, camera, and average skies. With the Paramount, which doesn't need to be pounded to death during guiding, it's eminently practical.
The "black box" features of ACP do work. I only ask you to have a little patience while Arnie and Brad get some experience with the system anmd get things dialed in.
-- Bob
|
