Dithering Not Working


#1

I’m beginning to have trouble with dithering. Last night, I set up with my mount and camera using a 200mm lens. My guiding was rock solid and I had it set to “Very High” dithering settings.

I began noticing odd behaviour in PHD2. Every time SGP finished a frame, it would state that it was sending a dithering command. This caused my PHD graph to jump off the charts, usually in both RA and Dec. Most of the time, it would recover after 30-45 seconds, but sometimes it would fail and go into recovery mode. Furthermore, it appears that no dithering is actually happening, as I checked five images and found the stars in exactly the same spot as before. It appears that PHD recovered, but put the star back in its exact location, not the dithered location.

Any ideas?


#2

Please help us help you… have a read through this and see what other information you can provide:

http://191.234.49.38/t/how-to-ask-for-help/26


#3

My guess is that your max motion is set too low. I recently did this to myself too and it was taking too long before SGP would fail out because it wasn’t setting. You can either a.) set the max motion RA/Dec higher or you can set your ‘settling’ time to a higher value.

Dithering is very tiny to begin with. Even on Very High, I only end up with like 5-10 pixels on either side of a stacked image. And, you’ll only see the movement if you use a ‘blink’ program or stack the image.


#4

I have had the exact same problem. I never used to have it with earlier SGP versions and PHD 1. The dither happens, PHD 2 jumps off the charts as though the dither command were just tracking error, and then it brings it back with guiding and the stars end up on the same pixels after a bit of time settling (my max motion is set high enough that SGP doesn’t go into recovery mode). The effect is no dither between frames since PHD 2 is correcting the dither. Even worse, if your settling time isn’t long enough and your DEC axis has some backlash then frequently the first few seconds of your next exposure will have the guiding overshoot and try to come back due to the huge dither movement making every frame worse than it would be otherwise. I had to disable dither because of this bug which is not good since that means I can’t use drizzle when I integrate my frames which is important to me since I use a Hyperstar setup for imaging and want to recover some resolution via drizzle.

Next time this happens I will get logs from SGP and PHD to attach. I lost my logs from last session in a computer mishap.


#5

KGoodwin, sounds like dithering is working exactly as expected. Do a blink through your pictures, I think you’ll see that they’re off by a few pixels on each side of the image. It’s not very much. Just enough to keep the background noise from being exactly the same.


#6

I can guarantee you that is not happening. Dither moves the lock position in PHD[2]. It’s up to you to decide how far you want it to move (dither settings in SGP and dither scale factor in brain in phd2).

That’s right. But if you have that problem you need to increase your settle time! (Or reduce the mechanical backlash in your mount.)

If you want a more detailed analysis of what happened in SGP and PHD2, we will look at your logs when you have them available.

Andy


#7

It has never happened to me before on previous versions of SGP/PHD, so it seems odd to me that I suddenly need to increase my settling time and I have tried the blink test and with the same dither settings and same focal length and same camera pixel size I am seeing much less dither offset (0-1 pixels) now than I did before (3-5 pixels). I think that the difference must be between my old PHD settings and my PHD2 settings, perhaps the dither scale factor. The settling thing is still a mystery to me since I’ve measured the dec axis backlash and it is the same as it has always been on my mount.


#8

I suppose the PHD settings could also affect the settle time since if the max motion is less or something like that settling after dither will take longer. It’s not more backlash it’s just taking it up less quickly. I’ll try changing these things next time out.


#9

KGoodwin,

I think you are probably on to something there. If your dither scale factor and your max RA/Dec settings are lower now in PHD2 than they were in PHD1, I think you would see exactly the results you are seeing.

The actual base dither amounts (before multiplying by the dither scale factor) are:

Small +/- 0.5 px
Medium +/- 1.0 px
High +/- 2.0 px
Very High +/- 3.0 px
Extreme +/- 5.0 px

These are maximum dither amounts. The actual amount for any specific dither will be a random value up to the max on each axis.

Andy


www.mainsequencesoftware.com