Guiding with Dither vs Unguided - Confused

[Adreneline]

I've been trying to get guiding working with my ASIair + CEM25-EC + EvoGuide with 120MM-mini - it's proving difficult.

Whatever last night I managed to get 10 x 120s guided and dithered and 10 x 120s unguided (no dithering) of M101 - the target was not relevant.

The result was not what I expected. The guided and dithered clearly has a lower background noise level but what's with the hot pixels?! Hot pixels are not present in the unguided version at all - and there was me thinking dithering was one way to get rid of hot pixels.

I've done my best to pre and post process both images exactly the same in PI.

Am I doing something wrong? Any help or insights would be appreciated.

Thanks.

Adrian

[vlaiv]

Dithering works for hot removal only if you sigma reject in your integration step.

10 subs seems a bit low to do reliable sigma reject.

[Adreneline]

46 minutes ago, vlaiv said:

Dithering works for hot removal only if you sigma reject in your integration step.

10 subs seems a bit low to do reliable sigma reject.

Thanks @vlaiv - I thought I had it set but I’ll check tomorrow.

Adrian

[Davey-T]

Dithering works well with colour sensors to help mitigate the colour mottle background.

Hot pixels can be removed using hot pixel map in most processing software.

As Vlaiv said for sigma reject it needs quite a few subs to compare.

Dave

[discardedastro]

Are you doing CosmeticCorrection before integration? Best to remove hot pixels before you do integration in PI, at the very least the obvious ones. With a master dark plus/or a 3-6 sigma automatic removal you'll clear up almost all the hot pixels before you hit the integration stage.

Dithering doesn't actually get rid of hot pixels in subframes without some logic. The reason being that you're moving the observing field. The camera stays static to "itself", if that makes sense, so the logic has to be that if you see the same "hot pixel" in different positions in the sky it's probably wrong so the statistical rejection says it's not actual signal.

However, if you reject the hot pixels before you even get to ImageIntegration then there's no signal to act upon and the error is rejected entirely.

[Adreneline]

9 hours ago, Davey-T said:

Hot pixels can be removed using hot pixel map in most processing software.

 

9 hours ago, discardedastro said:

Are you doing CosmeticCorrection before integration?

Thanks for the feedback. I used a master flat, dark-flat, dark and a bad-pixel-map in APP to perform the calibration and subsequent integration so I was surprised the BPM didn't sort the hot pixels. They were also all new calibration frames taken yesterday as I had decided to use -10 in stead of -20 degress to take the lights.

I'll have another go at calibrating, etc. in PI today and see what I get.

Thanks again for the help and advice.

Adrian

[vlaiv]

52 minutes ago, Adreneline said:

Thanks for the feedback. I used a master flat, dark-flat, dark and a bad-pixel-map in APP to perform the calibration and subsequent integration so I was surprised the BPM didn't sort the hot pixels. They were also all new calibration frames taken yesterday as I had decided to use -10 in stead of -20 degress to take the lights.

Could be that you accidentally mixed calibration subs?. At -10C it is more likely that you'll end up with some hot pixels versus -20C and if you've shot your lights at -10C and calibrated with -20C by mistake - those hot pixels could be left in image.

[Adreneline]

54 minutes ago, vlaiv said:

Could be that you accidentally mixed calibration subs?. At -10C it is more likely that you'll end up with some hot pixels versus -20C and if you've shot your lights at -10C and calibrated with -20C by mistake - those hot pixels could be left in image.

Hi Vlad,

I was careful not to mix things up but I agree that the -10C trial may not have been a good idea. I normally image at -20C but decided to give -10C a go whilst I was trying to sort the guiding. I repeated the calibration, etc. in PI and got essentially the same result which at least confirms that there is little to choose between APP and PI when used by me with my level of knowledge.

As for the guided/unguided trial aspect there was virtually no difference between the two as far as star shapes were concerned; the CEM25+Samyang135+ASI294 does not seem to benefit from guiding other than some slight reduction in background noise. That said the RA guiding was not as good as I hoped; DEC guiding was much better. Try as I may I could not get the same/similar RMS in RA as in DEC when using the ASIair integrated PHD software. It was also frustrating how long it took to recover from a dither perturbation - far longer than I remember my NEQ6 + PHD2 taking using the exact same guide scope setup.

Adrian

[vlaiv]

I've heard about guiding "fighting" with encoders. Maybe if you use longer guide exposure and let encoders do their job between guide exposure?

What was your guide exposure like? I would set it to somewhere between 5-8s to start with.