What on earth is Dithering?

[pne_rob]

hi all, despite me being a member of this forum for a good few years i still class myself a beginner and resort to many youtube videos for answers. My question is.... Is it possible to dither images with just a DSLR and a normal tripod.. i have scrolled through page after page and not seen anything about this hence the question being asked.

Now i understand the very simple premise of Dithering that the sensor or camera is moved a few pixels over so colour mottling is reduced but from what i have been reading everywhere is that they are using mounts on their cameras and software programs to do auto dithering (if thats even a thing).

Now this question will most definitely show my inexperience and naivety but between exposures could you not just manually move the tripod ever so slightly in either direction. The reason i think this could happen is that the program for stacking i would use is Registar and according to Tony Hallas the program stacks all the photos together with precision even though the stars are not in exactly the same position in each exposure. So i thought you could manually move the camera slightly and import the photos into Registar and get them stacked or am i as previously said naive ?

 

thanks.     

[ollypenrice]

If by an 'ordinary tripod' you mean a fixed tripod then this is attached to a moving earth and you will get all the dither you need by not touching it at all! It is with autoguided mounts, which are tracking the stars to a fraction of a pixel in accuracy, that it becomes a good idea to move the sensor a little relative to the sky so that the same pixels don't always sample the same bit of sky.

Until the Earth stops rotating your fixed tripod is dithering...

Olly

[pne_rob]

Just now, ollypenrice said:

If by an 'ordinary tripod' you mean a fixed tripod then this is attached to a moving earth and you will get all the dither you need by not touching it at all! It is with autoguided mounts, which are tracking the stars to a fraction of a pixel in accuracy, that it becomes a good idea to move the sensor a little relative to the sky so that the same pixels don't always sample the same bit of sky.

Until the Earth stops rotating your fixed tripod is dithering...

Olly

so because the cameras location is fixed and due to the natural movement of the sky, the pixels will be in a different location anyway so dithering would be natural and automatic then?

 

if that is so then that is great to hear.

 

cheers olly 

[Filroden]

7 minutes ago, ollypenrice said:

Until the Earth stops rotating your fixed tripod is dithering...

So does dithering in one direction work? I'd always assumed it would need to be in two, or at least more random than drift.

[pne_rob]

4 minutes ago, Filroden said:

So does dithering in one direction work? I'd always assumed it would need to be in two, or at least more random than drift.

thats a fair point, i was made aware when moving it needs to be move left right up down and diagonally to randomise the pixel pattern so would still need some manual movement instead of leaving it for natural drift yes. Thats my assumption anyway

[ollypenrice]

There are, potentially, two reasons for dithering. The first is to move the chip relative to the sky so that pixel defects land on different parts of the sky and are averaged out when multiple exposures are stacked. This is done to reduce noise in the final stack. It does not matter which way you move the chip relative to the sky in terms of improving signal to noise. How far you should move between subs depends on the camera. Tony Hallas, a DSLR imaging expert, suggests a 12 pixel dither. He is not, here, trying to combat just individual bad pixels but is trying to reduce the 'colour mottle' background sky effect associated with DSLR images. I would be very surprised if the earth's rotation did not suffice easily for this purpose. It is best to use a Sigma Clip stacking routine because this will identify rogue (outlying) pixels in each sub and give them the average value of the rest of the stack, powerfully attacking this kind of  noise.

Dither can have another more complex justification, though. It can be used to support drizzle stacking, a Hubble Team technique for enhancing resolution. It can allow images to be resampled upwards to reveal new levels of resolution provided each sub has sampled a slightly different part of the sky. You can follow this up on the net but I don't think it applies in this thread. I'm not sure how far and in which direction the dither should be done in order to optimize the results but it does work well on long focal length 'lucky imaging' with fast frame cameras.

By the way, a camera on a fixed tripod at mid latitudes collects a dither which is certainly not in only one direction. Don't forget field rotation.

Olly