APT/Guiding and dithering

[Carole]

What settings are you guys using for guiding in APT and how much gap are you giving between subs?

I don't quite understand What the instructions mean:

Dithering Settle Time – defines how many seconds to wait after Dithering Stability distance is reached. It gives time guiding to settle down. See the explanation below. (implemented for PHD only)
I think I understand that, I have chosen about 10 secs.

Dithering Timeout - defines the maximum time for dithering. After that time APT will continue with the next exposure no matter what is the distance to the guiding start. 0s means that there is no time limit to complete the dithering.I don't understand this, does it mean if it hasn't dithered by the number of pixels I have chosen within the dithering time I have chosen - 15 secs (and I have chosen 2 pixels), then it will stop dithering.

I don't know how long it takes to dither.
I normally have about 30 secs between exposures, but could extend this if necessary. 

Any advice would be appreciated.  
Thanks
Carole

[The Thing]

I have a 30 second pause after a 10 second anti-vibration time.

I use the defaults.
Stability = 1.5
Settle Time = 5
Timeout = 10

I don't use the auto cancel function.

[MarkS]

I don't know how long it takes to dither.

You should display the PhD graph - you can then see the jump in position caused by the dither and then get an idea of how long it takes to settle afterwards. 

In my experience, it normally settles in 10-15 seconds but it does tend to vary from imaging session to imaging session and vary according to the size of the jump (I always use extreme dither for my DSLR).  I would give it 30 seconds to give you a safety buffer for when it takes longer.

[Carole]

Thanks Mark and Duncan. 

I have actually got it running outside, and seems to be working even before I read your responses.  But I wasn;t sure if I had allowed enough gap between exposures for the dithering.  It did say dithering had finished by the time the 2nd sub started.

I think I'll change to yur settings though Duncan, as it obviously works for you and as you say Mark, watching the graph is a good idea, I'll nip out and do that too.

Thanks

Carole

[Carole]

I changed to your settings Duncan but dithering timed out, so I've increased the length of the gap between subs and see if that makes any difference.

I also have used the framing mask facility tonight, except I could not get it to load the mask I had already saved, but it would let me load a fresh one and then switch from the "real image" to live view. 

Carole

[Carole]

I got APT/PHD and dithering working OK, but it's only dithering about 1 in 3 times, the rest of the time it "timed out", is this normal?  

Carole

[RobertM]

Sounds like you're really getting yourself sorted now Carole and that stability is showing in you're images.

There are a lot of factors involved in the guider settle time for dithering which mean that can vary quite a lot.  You'll sometimes find on bad nights that maybe even 30s isn't enough !  Each setup is different and I'm sure you'll get to know what you can get away with.  Like Mark an extreme dither can take 15s to settle to within 0.1 pixel (@176mm f/l) which is good enough to image at 1000mm.  For small dithers that can reduce to 5s or so.

Try it and see what works for you.

You reposted while I was typing this -- 30s should be enough but suck it an see...

Robert

[Carole]

Thanks Robert, just need to sort out a few more things though and as you say I will be sorted. 

Although I put 45 sec interval time last night, APT only in fact waited about 15 secs before starting another sub, which is confusing.  I have put a post on APT forum to see what IVO says.

Carole

[MarkS]

Carole,

I guess you mean you set a dithering timeout of 45 seconds?

If it started the next sub after only 15 seconds, it probably means that the guiding had stabilised back to the required tolerance after the dithering jump.  It should have been obvious from the PHD graph.  You were looking at the PHD graph, weren't you?

There's obviously no point in it waiting another 30 seconds for no benefit.

Mark

[Carole]

Hi Mark,

No what I meant was I set 5min images at 45 sec intervals.
Then in the APT/PHD settings I had a timeout of 10 secs (which duncan uses).  However despite settings an imaging interval of 45 secs, APT started the next sub after only about 15 secs.

Yes I was watching the graph but even on the occasions when dithering worked (which was about 30% of the time), I could not see any change to the graph at the time that dithering was going on. 
Yet APT said that dithering had finished, whereas at other times it said it had timed out.

Carole

[MarkS]

I'm not yet a user of APT but it sounds to me that if you have dithering switched on then the dithering parameters will take precedence over the interval parameter of 45 seconds.  The timeout of 10 seconds is far too short in my experience but it does depend on the magnitude of the dither.  A big jump will take longer to settle than a small jump.

APT should be able to tell PHD the maximum size of the dither "jump" and then PHD chooses random jump sizes up to this maximum - the most extreme maximum jump size accepted by PHD is 5 pixels.  The randomness does mean that sometimes the dither jump is very obvious on the graph and at other times the dither jump is not noticeable.  Hoever, the randomness is an important element of the dithering process - improving the final image quality.

I generally use the extreme dither setting because it means that the influence of any hot/warm/cool/cold/dead pixels is spread over a larger area and becomes unnoticeable in the final image.  However it does mean that you lose a small border around your stacked image, necessitating a crop.  With a DSLR sized CCD this is hardly an issue!

Mark

[Carole]

Hi Mark,

Yes those were the conclusions that I came to.  I have the dither set to up to 2 or 3 pixels, I can't quite remember which but i didn't think it should go beyond that.  So looks like I should set my timeout a bit longer then as it doesn't give the camera much rest between subs, not that it seemed to affect the temperature that much.  (I've been through and checked the termperature of all the subs on Kuso Exif as I am trying to match darks more closely).

Carole

[The Thing]

Hi Mark,

Yes those were the conclusions that I came to.  I have the dither set to up to 2 or 3 pixels, I can't quite remember which but i didn't think it should go beyond that.  So looks like I should set my timeout a bit longer then as it doesn't give the camera much rest between subs, not that it seemed to affect the temperature that much.  (I've been through and checked the termperature of all the subs on Kuso Exif as I am trying to match darks more closely).

Carole

Carole, the value you need to set is the PAUSE in the plan settings, nothing to do with dithering. For a 300s exposure I set this to 30s so the camera 'rests' between exposures and hopefully loses some heat.

[MarkS]

Rest periods   

Personally I don't think there is anything to be gained by letting the camera "rest" between subs - I've never seen any evidence that it helps.  Although my camera doesn't have a built in temperature sensor, the noise on the dark frame makes a perfect substitute because, after all, reducing the thermal noise is supposed to be the point of it.  I actually did some experiments once, taking a long series of dark frames at constant room temperature with and without rest periods and was not convinced that the rest periods would make any noticeable difference to the final image quality.  All it does is to reduce the number of subs you can shoot in a given time.

Look at it another way: at a constant ambient temperature, it takes 60-90 minutes for the CCD temperature to gradually climb to its plateau at thermal equilibrium.  If you have a 30sec pause between 300sec exposures then, at best, you are reducing the heat going into the CCD by 10%.  How much difference is that really going to make?

Mark

[The Thing]

Interesting. I will remove the pause next time I'm out imaging. I get a lot of amp glow anyway so it's obviously not helping.