Dithering for dummies!

[mbalkham]

Hi all,

Can someone explain dithering to me from the basics (i.e. what it is and why I'd want to do it!) or point me in the direction of some reading.  I've seen it mentioned a few times and just want to know if it's something I need to understand better!

Thanks in advance.

Matt

[carastro]

Dithering is where your capture software and guiding software talk to each other and the guiding software moves the mount a few pixels between subs.  The advantage of this is that this helps to reduce noise and hot pixels etc by overlapping subs ever so slightly in the stacking process.

Carole 

[teleskopjo]

I believe some imaging software (like APT) can give their own dithering instructions to the mount (ie. without a guiding setup). I would say shoot at least 15 subs when dithering (and no darks).

[lensman57]

Hi all,

Can someone explain dithering to me from the basics (i.e. what it is and why I'd want to do it!) or point me in the direction of some reading.  I've seen it mentioned a few times and just want to know if it's something I need to understand better!

Thanks in advance.

Matt

If you are using DSLR for capture which I think that you are then it is essential that you get your capture software to dither. For a DSLR you set the dithering to as high as possible and take at least 20 subs. This combined with a Sigma Clipping rejection during stacking, DSS,Pixinsight, Nebulosity and all other stacking software do SC, will reduce noise and get rid of a lot of nasties

including satellite trails.

A.G

[Nathan_Pembs]

When you say set dither as high as possible, how high roughly? I have mine currently set to 5

[PhotoGav]

+1 for all of above.

As far as I understand it, the reason for dithering is that you avoid always imaging a particular point in your image with the same photosite of your chip. That way, if the photosite is slightly defective (and most of them are not perfect), by dithering, you use a variety of photosites for that particular point of the image, thereby giving you a spread of values for that point. That leads to a more accurate representation of the real value for that point. Multiply that across your whole image and you will have a better quality image, with fewer defects from the dodgy sensor. The biggest defect is 'noise', especially in a DSLR. Good news to eliminate the noise as much as possible!

[jgs001]

Dithering actually comes into play when you come to stack the subs... the dithering process, as has been said moves the positions of the stars in the image a few pixels between each sub. The result is, that when you come to stack the subs, and you use a statistical stacking technique (kappa-sigma in DSS), then once all the subs are aligned (based on the star positions), hot/cold pixels etc, as they are all in different locations (in each sub) relative to the stars, become statistically meaningless (appearing in only the one frame in that specific position), and the stacking algorithm promptly removes them. You need a minimum of 12 to 15 subs for the statistical processing to have enough data to work from, but it works extremely well. 

[lensman57]

When you say set dither as high as possible, how high roughly? I have mine currently set to 5

For a DSLR as high as the software allows it.

A.G

[mbalkham]

So now I'm sold on dithering, how do I go about it?!   What capture software do people suggest?  To date I've been capturing in camera.

Thanks in advance

[glowingturnip]

APT will do the job nicely - http://www.ideiki.com/astro/Default.aspx

[PhotoGav]

+1 for APT. Very easy to set dithering - just check the help file for exact details.

[lensman57]

So now I'm sold on dithering, how do I go about it?!   What capture software do people suggest?  To date I've been capturing in camera.

Thanks in advance

For dithering you first need to guide as APT or BYEOS do not guide but they do issue dithering commands. APT works well with PHD guiding which is a free guiding software and will accept dithering commands from APT and BYEOS. You also need to interface the mount and the guiding software through the laptop or PC. I use GPUSB but there are other options such as ST4 . You need to do a bit of reading and research before starting.

A.G

[gnomus]

APT in combination with PHD is working for me - and I barely know what I am doing. I get PHD guiding first, then launch APT. I don't know if this is the correct way or not, but, as I say, it seems to be working.

I wouldn't be without APT (even without dithering). And it is 'pennies' for what it does and the time it will save you.

[lensman57]

APT in combination with PHD is working for me - and I barely know what I am doing. I get PHD guiding first, then launch APT. I don't know if this is the correct way or not, but, as I say, it seems to be working.

I wouldn't be without APT (even without dithering). And it is 'pennies' for what it does and the time it will save you.

Yes that is the correct way, launch PHD guiding, go Tools, Enable server and then launch APT and under the Gear tab shift click  Guide and choose PHD as guiding program, set the parameters and you are done. APT will automatically connect to PhD in server mode and issue Dither commands once capture ends. You can monitor the axis being moved by looking at the PHD guiding graph.

A.G

[mbalkham]

So am I right in saying no darks required when dithering? Or do they help further reduce noise and/or increase signal to noise ratio?

Downloaded APT looks great.

[lensman57]

So am I right in saying no darks required when dithering? Or do they help further reduce noise and/or increase signal to noise ratio?

Downloaded APT looks great.

Whether you dither or not and unless you can control the temperature of the darks to within a couple of degrees C of the lights then there is no point in taking darks as they might do more damage than good. If Darks need to be taken then do them at the end of the session with the same exposure delay as the lights. It is said that around 20 darks are required but I have read that even 4 darks correctly taken can remove up to 80% of the thermal noise. I myself do not bother with darks with Cooled CCD. I use 200 Bias and a cosmetic correction routine to remove the hot pixels. Dithering will help to a great extent in combating noise but it must be combined with a sigma clipping rejection algorithm during stacking to remove the nasties. For this to be effective you need a lot of subs, minimum of 20, though in PI I can get away with 12 or 13 subs but DSS needs a lot more than 13 subs for sure to enable Kappa Sigma clipping.

A.G

[teleskopjo]

For dithering you first need to guide as APT or BYEOS do not guide but they do issue dithering commands. APT works well with PHD guiding which is a free guiding software and will accept dithering commands from APT and BYEOS. You also need to interface the mount and the guiding software through the laptop or PC. I use GPUSB but there are other options such as ST4 . You need to do a bit of reading and research before starting.

A.G

APT does issue its own dithering command without any guiding connection to PHD or other guiding software, you need the full version however.

From the website: "Built in Auto-Dithering, when there is no guiding program."

[lensman57]

APT does issue its own dithering command without any guiding connection to PHD or other guiding software, you need the full version however.

From the website: "Built in Auto-Dithering, when there is no guiding program."

Yes it will but for that you need to give control of the mount to APT, I presume using the Ascom Platform. I do not use this so naturally I can not comment on its performance. I know that APT and PHD guiding work well together.

Regards,

A.G