Guidecam exposure, I'm confused

[wuthton]

I've seen several comments lately in discussions about higher end mounts, Avalon, Mesu etc. These have been along the lines of "This mount only needs a 1 second exposure". Isn't that the wrong way round?

As is my understanding the longer your guidecam exposure the more accurately PHD et al can calculate the star's centroid though the atmospheric turbulence... but this needs to be balanced with the accuracy of your mount. For my mount about 3 seconds seems to be the sweet spot, longer on poor nights.

Surely on more accurate and stable mounts less corrections and a more precise calculation of the centroid would be preferable or have I missed something?

[ollypenrice]

I've seen several comments lately in discussions about higher end mounts, Avalon, Mesu etc. These have been along the lines of "This mount only needs a 1 second exposure". Isn't that the wrong way round?

As is my understanding the longer your guidecam exposure the more accurately PHD et al can calculate the star's centroid though the atmospheric turbulence... but this needs to be balanced with the accuracy of your mount. For my mount about 3 seconds seems to be the sweet spot, longer on poor nights.

Surely on more accurate and stable mounts less corrections and a more precise calculation of the centroid would be preferable or have I missed something?

No, I think you're quite right. The very accurate Mesu thrives on seeing-stable 4 second subs or longer. I tend to find our other mounts, including the Avalon (and contrary to Avalon's published advice) like shorter subs.

But here's a tricky thing: if you go for very short subs you'll get a very nice nice graph because the graph is based on the assumption the the guide star is where the guide sub says it is. However, it may not be there at all because the seeing is bouncing it around all over the place. Good graph but, in reality, not so good guiding. How do we know where to go for the compromise? I'm blowed if I know... I suppose that, if you could be bothered, you could do some 20 second tests at different guide sub lengths and see which gave the lowest FWHM. You'd need to average the results of quite a few test subs though.

Olly

[fwm891]

Mmmm...

I have an Avalon Linear Fast Reverse and use 2 or 3 seconds binned 2x2 normally. The guide cam is a Lodestar slotted into an OAG on my RC8, controlled through Maxim. Very rarely get any guide errors, even when blocked by passing cloud it stays on track until the cloud disappears.

[swag72]

I used to use 3s exposures and then read somewhere that Avalon recommended short 1s exposures.I have no idea why and neither do I remember where I read it

I must go back to longer exposures and see if I get a difference in anything.

[ChrisLX200]

The good thing about the Avalon (and Mesu) is that there is very little backlash in either RA or DEC, so you can make these very small corrections rapidly and with the confidence those requested movements will be accurately applied by the mount. Most other mounts (pure wormgear-driven) won't do this reliably. With my G11 then sometimes, depending where in the sky it was pointing (and a host of other things I never figured out after 13 years using it!), it would respond well to even 0.5s corrections, but at most other times this would go pear-shaped and I needed 2-3sec which even then still produced a wavy PHD graph. I don't think it was seeing-related, just the mount being 'special'.

ChrisH

[MartinB]

A mount with quite modest performance might have a peak to peak PE of 30 arc secs. Give it a typical worm cycle of 7mins. It will go from peak to peak in 3.5mins That averages an arc second every 7 seconds. However, it isn't even across that time being a wave form. My maths isn't up to calculating the maximum PE movement but my sailing is and we use the rule of 12ths. Over the fastest 3rd of the 3.5 mins it will move half of the 30 arc seconds. This roughly equates to 15 arc seconds in 70 seconds = 1 arc second every 4.66 seconds. This doesn't look at random errors but these tend to guide out very badly regardless of guide exposure times

So, if you aim for a guide accuracy of half your sampling rate even with a such a mediocre periodic error you should easily be able to manage 4-5 seconds at short focal lengths especially given that you are less likely to end up chasing the seeing and can use more aggressive rates of correction.

Another factor for good guiding is having a good signal. A bright star will have less noise and will give a more accurate centroid. Sometimes, especially when off axis guiding you need to lengthen the exposure time to ensure your star is good enough.

[Psychobilly]

I  used to aim  for a star  where a  3-4s exposure was   70-80% of max and whose profile gave a distinct centroid...

Peter...

[fwm891]

I know some people like to display the guiding graph while capturing - I hate it because the temptation is always there to fiddle with settings and you end up chasing settings as well as the seeing. If at the end of a sub I have tight stars I leave well alone, if they're not its usually because I'm working with a fairly faint guide star and need to lengthen the capture time a little.