Is this good guiding ?

[alan4908]

In an attempt to improve my guiding and hopefully the quality of my imaging, I recently spend 4 hours adjusting various MAXIM DL parameters from the default values and finally ended up with the guider tracking graph shown below. Please note that I haven't yet implemented PEC.  

With respect to the graph, the vertical axis is in guide camera pixels where 1 pixel = 4.42 arc seconds, the horizontal axis is in seconds.  My main camera has 2.39 arc seconds/pixel.  I was also dithering at the time by 1.5 guide camera pixels and the subframe length was 300s.  I also found that I needed a guider settle delay of 120s since the DEC axis (Y on graph) was taking a long time to recover from the dither.  In a further attempt to quantify the situation, I took several images and got MAXIM DL to measure the quality on several calibrated sub frames, the FWHM was coming out as about 2.8 and the average error in star roundness was about 8%, this is through an air mass of 1.06.  As you can see from the graph, MAXIM is recording an RMS error on the RA axis of 0.233 pixels and an RMS error on the DEC axis of  0.227 pixels.

For those that might be interested I ended up with the following MAXIM DL parameter set:

X Aggressiveness = 10 - RA axis (I was a bit concerned at the value of this parameter but it didn't seem to introduce any instabilities).

Y Aggressiveness = 7 - DEC axis

Min move = 50ms

Guider Settletime between subframes = 120s

Dither in X and Y = 1.5

Single star guiding

Guider exposure = 3s

In an attempt to get some extra value from this session (I'm not totally sure about the accurately of the methodology so please comment):

PEC: I turned off the dither and correction on the RA axis and set the  X Aggressiveness  = 3 and watched the guider error graph on the RA axis - this seemed to suggest that my mounts period error is about 12 arc seconds peak to peak and is occurring at a rate of about 1.7 arc seconds/s.

Polar misalignment error: To find out my polar misalignment error, I switched on the RA corrections and switched off the DEC corrections and watched the guider error graph on the DEC axis - this seemed to suggest a polar misalignment of about 0.007 arc seconds/s. 

Alan

[PhotoGav]

Are the stars pin points of sharpness? If so, that's good guiding. If not, that's rubbish guiding.....!

[Freddie]

Looks good to me as you are sub pixel on the main cam most of the time. I would just want to understand what happened at that first peak as it was DEC and RA pretty much at the same time. A gust of wind? Did you nudge the mount?

[alan4908]

Freddie and Gav - thanks for the comments.

Freddie - Not sure which peak you are referring to.... if its the one between 2340s and 2520s - this is the result of a dither and during that time I'd not be imaging since the guider is still settling.  Are saying that if the guide error in arc seconds is less than the main camera pixel in arc seconds then you have good guiding ?

Alan

[MartinB]

Freddie - Not sure which peak you are referring to.... if its the one between 2340s and 2520s - this is the result of a dither and during that time I'd not be imaging since the guider is still settling.  Are saying that if the guide error in arc seconds is less than the main camera pixel in arc seconds then you have good guiding ?

 

 

Alan

That would be a good benchmark. It looks ok although the RA is a bit spikey. I'm not sure that an aggressiveness of 10 is a great idea or necessary. Given that some star movement will be due seeing changes then you don't want to make a full correction in any one go otherwise you can compound the problem. 7 should be plenty. What you will find is that the RMS figures vary from night to night depending on how good the seeing is.

On a different note, it's taking 120seconds for the dec to settle down after the dither. I don't think that's so unusual and is caused by backlash. Each sub is taking 5 mins but you are only collecting signal for 3 minutes. You are losing 40% of your imaging time. That is downgrading your final s/n ratio way more than any benefit you might gain from dithering.

[alan4908]

Martin

Thanks for the response.  

You've made an excellent point regarding the my settling time (120s) eg doing "nothing time" versus imaging time (300s) eg a wasted time of 40% ! but given that I've found dithering good for improving the quality of imaging - I think I need to increase my subframe duration and/or maybe I shouldn't dither in the DEC direction.....Given that I have a OSC camera - what sort of subframe time do you think I should be aiming for ?  

I'll also take your advice and reduce the RA Aggressiveness down to 7. 

Alan

[Phillyo]

As Gav has said, I would look at the final image and check your stars. If they're round then you're good, if not then...you're not. I used to worry about guiding a lot, now I've learnt that it's the final image that matters

Phil

[MartinB]

Alan, the shortest sub time you are aiming for is one that makes the read noise insignificant and it depends on a number of factors such as the read noise figure for the camera, dark current and sky glow. It is therefore camera dependent. CCDware have a free online calculator here http://www.ccdware.com/resources/

This gives you the minimum suitable exposure however, you can go beyond this time if you like. Because the download time with my camera is quite long I tend to go beyond the minimum. Other factors then come into play - you need at least 10 subs before you can use stacking methods which use statistical rejection methods. Also the longer your exposures the greater the chance of them being messed up by cloud, wind, contrails etc.

[Freddie]

As Gav has said, I would look at the final image and check your stars. If they're round then you're good, if not then...you're not. I used to worry about guiding a lot, now I've learnt that it's the final image that matters

Phil

It's not just about being round. You can have guiding all over the place and if out equally in all directions it will still produce round stars. As Gav pointed out, it's also important to get tight guiding to produce small stars as well as being round.

[alan4908]

Hi All - thanks for all of that, my understanding is slowly improving  

Phil and Freddie: Yes, on the final image point: I appreciate that... I was attempting to quantify this by getting Maxim to provide me with the average error in roundness and the average FWHM on the calibrated but unstacked frames.

Martin - thanks for this - I shall investigate this calculator.

Alan