ZWO ASI290MM mini noise characteristics

[astrosatch]

Hi!
I'm waiting for asi290mm mini to arive to replace my asi120mm mini and wondering how low noise of this camera affects sensitivity. Both cameras has approximately same QE and 290mm has smaller pixels than 120mm. This would in theory make latter better choice except for higher noise it has. Other users experience speak for 290mm to be a better camera for guiding making it quite more sensitive.
Does lower noise really make that a difference?
And is there any mathematical way to prove this?

Andrej

[vlaiv]

Depends what you define as sensitive camera.

For guiding, I think QE is much more important than read noise.

Read noise is added once per exposure. All other noise types grow with exposure time. You can always control amount of read noise in comparison to other noise sources by controlling exposure length.

Only when you can't use longer exposures - level of read noise becomes important. For example in lucky type planetary imaging. Here you need to lower your exposure length to freeze the seeing. Exposures are often in range of few milliseconds. Having low read noise camera becomes really beneficial in such cases.

Noise math is not that complex at all. Read noise is modeled by Gaussian distribution, while other noise types (target shot noise, thermal noise, LP noise) are Poisson type distributions. Value of read noise is fixed per exposure, while other noise sources are equal in value to square root of associated signal intensity.

Noise adds like linearly independent vectors (square root of sum of squares).

What will you use this camera for and what type of sensitivity are you after?

[astrosatch]

16 hours ago, vlaiv said:

Depends what you define as sensitive camera.

For guiding, I think QE is much more important than read noise.

Read noise is added once per exposure. All other noise types grow with exposure time. You can always control amount of read noise in comparison to other noise sources by controlling exposure length.

Only when you can't use longer exposures - level of read noise becomes important. For example in lucky type planetary imaging. Here you need to lower your exposure length to freeze the seeing. Exposures are often in range of few milliseconds. Having low read noise camera becomes really beneficial in such cases.

Noise math is not that complex at all. Read noise is modeled by Gaussian distribution, while other noise types (target shot noise, thermal noise, LP noise) are Poisson type distributions. Value of read noise is fixed per exposure, while other noise sources are equal in value to square root of associated signal intensity.

Noise adds like linearly independent vectors (square root of sum of squares).

What will you use this camera for and what type of sensitivity are you after?

Thanks Vlaiv for reply.

I would be using it with oag 800mm fl F4 with asiair pro to implement new feature of multistar guiding. With asi120mm had occasional star lost and found again situation, but in general ok guiding. However dimmer stars were rarely good for guiding. Maybe broader dinamic range of asi290 would improve things.

Andrej

[vlaiv]

I think that you can do couple of things to improve guiding that will have larger impact than swapping the guide camera (although getting new guide camera is not bad thing - do use that new one).

1. Place OAG as close to sensor as you can.

What is your imaging camera and what OAG are you using. DSLR, for example, is rather poor choice to be guided with OAG at F/4. This is because there is 44mm of distance between T2 adapter and sensor.

Let's suppose that you have 8mm prism. At F/4 - it needs to be placed closer than 32mm in order for prism not to stop down the beam used for guiding. If you have DSLR - best you can hope is is probably around F/6 instead of F/4 - and that is if you place pick off prism right at T2 adapter.

2. Place prism as close to optical axis as you can.

I'm guessing you are using fast newtonian since you have 800mm F/4 scope. Fast newtonians really need large secondary mirrors in order to illuminate large diameter - and often they don't have it. If you place prism too far from optical axis - two things happen:

- you are in vignetted part of the field - so further light loss happens. This combined with prism acting to stop down the beam can result in guider getting (only F/10-12 or so of light)

- you are in coma affected part of the field (even with coma corrector - far from axis stars are not going to be sharp). This spreads light from a star and reduces signal that each pixel receives (here having low read noise helps). Keeping prism closer to optical axis makes guide stars tighter - they have better SNR

There are of course two additional things that you can do:

3. Increase guide exposure.

Most people use 1s guide exposure as it is sort of a default value, but in reality, if your mount is mechanically sound - you should be able to use 3-4s guide exposures. You can examine your mount behavior when not guided to find out maximum drift rate. You can then use maximum error that you want your guiding to have and divide with max drift rate and that will give you how long your guide exposure can be.

If you can - do PEC on your mount - that reduces max drift rate in RA. Good polar alignment often reduces DEC drift rate to very low levels - as far as that one is concerned - you can have 10-20s guide exposure no problem.

Very crude methods for estimating drift rate would be to use polar alignment error calculator here:

http://celestialwonders.com/tools/polarErrorCalc.html

and for RA to measure P2P of periodic error and divide with worm period (then multiply that with 2). Say you have Heq5 and have 35" of P2P. Worm period is 638s. This means that in 319s mount drifts 35" (other 319 it drifts back) - and if that motion was uniform speed would be 0.11"/s. However, most often it is not uniform and in fast parts it is about x2 this value so 0.22"/s. I personally would keep such mount under 3s guide exposure as in 3s it can drift 0.66" in worst part of worm period. However, smaller P2P Ra error will enable you to use 4s for guide exposure.

Another benefit of using longer exposure besides better SNR and more guide stars is that you give seeing enough time to average out - and you won't be chasing the seeing as much (that is why you want to use multi star guiding after all, right?).

4. Bin your pixels. That also increases sensitivity.

You have plenty of focal length with OAG and you can easily bin your pixels. General rule is that you take your target RMS, divide it with 3 and multiply that value with 16. Your guide resolution should be that number or higher (lower number in "/px but higher resolution).

Say you want to precisely measure your RMS down to 0.3" (which is great value by the way). Divide that with 3 gives you 0.1 - now multiply that with 16 and you have 1.6"/px. 2.9µm pixel size of ASI290 at 800mm will give you 0.75"/px. You can bin that x2 without any issues for 1.5"/px and you'll still be able to reliably measure centroid position for 0.3" RMS guiding.

In fact - depending on the mount you are using and your target resolution - maybe even bin x3 your guide pixels. That will give you 2.24"/px and that is still good for 0.5" RMS guiding.

[astrosatch]

Thanks Vlaiv. A lot of good information here.😀

About distance from sensor. I have asi1600mm. There's 11mm extension tube between FW and OAG. If oag is moved next to FW, would benefit be  noticeable? In that case I must buy another 10mm M48 extender to connect it from oag to CC/FF to retain correct distance from camera.

[vlaiv]

That is already quite close. If I'm not mistaken, ASI1600 has 6.5mm of optical path. FW is another 20mm and 11mm extension would make it total of 37.5mm.

Yes, there is probably some stopping down of guide aperture by prism if prism is 8mm in size (there are also large models), but I would not bother with it if it means purchasing and waiting for additional accessories (like that 10mm M48 extension).

I think that you should first try other things - like new camera, increased guide exposure and pixel binning to see how much improvement you'll get.

 

[astrosatch]

Will try that once asi290 arrives.

Clear sky,

Andrej