Atik camera to max cool?

[Astroblagger]

Hello

Iv been using my Atik 460ex set to -10 degrees, so it will use less power out in the field (I use battery's) will it make a huge difference in noise taking it down to -20? at -10 the subs look fairly clean?

[newbie alert]

The Sony chip is fairly clean, -10 is about right  and don't cool it at its full speed, not that I've had a chance to use mine yet but it's set to cool at -10 over 10 mins..which in my head is pretty fast

[ollypenrice]

One discovery I made is that using a master bias as a dark works fine at -20C but does not work fine at the -8C which is all I could manage on our warm summer nights. At -8C I got visibly cleaner stacks using proper darks. So the temperature does matter, even with the Sony chip. I think -10 should be OK but I would use darks with it.

Olly

[Magnum]

Im a great fan of  dithering and SDmask stacking, and have found is all I need with Sony chips, ive never taken darks or bias with my 314 or my new 460. Any hot and warm  pixels stack out perfectly.

Lee

[ollypenrice]

4 minutes ago, Magnum said:

Im a great fan of  dithering and SDmask stacking, and have found is all I need with Sony chips, ive never taken darks or bias with my 314 or my new 460. Any hot and warm  pixels stack out perfectly.

Lee

What I find is that my 460 camera produces a noisy background sky without calibration. The noise takes the form of a lot of overly dark pixels so, after stretching, some background pixels will be at 23 but others will be far below that.

Olly

[Magnum]

19 minutes ago, ollypenrice said:

What I find is that my 460 camera produces a noisy background sky without calibration. The noise takes the form of a lot of overly dark pixels so, after stretching, some background pixels will be at 23 but others will be far below that.

Olly

Maybe I don't see that as my skies are much brighter than yours Olly? sky glow is probably swamping any read noise in mine. though I haven't noticed that with my narrowband subs either. I just looked at a 15 min Ha sub with mostly empty sky and the darkest pixels are 350 ADU with the average around 400ADu. so thats massively brighter than your background.

Dithering should smooth out any noise in the image though, wether its read noise or shot noise or dark current, as all of those things will be effectively moved against the true signal, so the algorithm should through away everything else . I find I need at least 12 subs to make SDmask work at its best with at least 5 dithers, though I just dither every sub so I know it will work reliably.

Lee

Edited November 4, 2019 by Magnum

[vlaiv]

Why do people choose not to use darks with set point cooling is beyond me .

Not much to be lost if one uses cloudy night to build master dark.

As for original question, 460ex has about 0.0004 e/px/s at -10C per original specs - and that is really low dark current. It means that in 10 minute sub, average dark current per pixel at -10C will be 0.24e and associated noise will be less than 0.5e - much lower than read noise of said camera which is 5e.

Sensors have doubling temperature of about 6C, so going to -16C will slightly improve things - 10 minute sub will be about 0.12e of dark current and associated noise will be ~0.35, so not much improvement over 0.5e. It is same for going warmer - at -4C dark current will be about 0.48e and associated noise ~0.7e - again not anything to be worried about.

[Magnum]

26 minutes ago, vlaiv said:

Why do people choose not to use darks with set point cooling is beyond me .

Not much to be lost if one uses cloudy night to build master dark.

 

Its precisely for the reason that the dark current is so low, that I dont use darks. As both Atik and Starlight Express state, using dark with these Sony chips can actually introduce noise into the image. I think Craig Stark also came to this conclusion when he first tested the 314: quote "The dark current is so low, however, that you may-well be better off simply taking a large stack of bias frames once to use for bias correction and use a simple bad pixel map for hot pixel removal". 

I think this is the method Olly choses to implement, Ive tried that then tried dithering and both gave me similar results except dithering also has other advantages, I guess I could do both but im probably the laziest imager there is. 🙂

Sorry if I went off the original topic slightly, my answer to that is I find -10C fine with Sony sensors, though he mentions running off batteries which I wouldn't recommend, you want to be above 12V as low voltage can increase noise, though thats more of an issue with the Kodak chips. With my new Atik 383 it is noticeably cleaner at 13V than at 12V.

Lee

Edited November 4, 2019 by Magnum

[vlaiv]

12 minutes ago, Magnum said:

Its precisely for the reason that the dark current is so low, that I dont use darks. As both Atik and Starlight Express state, using dark with these Sony chips can actually introduce noise into the image. I think Craig Stark also came to this conclusion when he first tested the 314: quote "The dark current is so low, however, that you may-well be better off simply taking a large stack of bias frames once to use for bias correction and use a simple bad pixel map for hot pixel removal". 

Yes, I'm well aware that you introduce additional noise when doing calibration. You can however control how much noise you introduce.  You add both dark current noise and read noise back in when you use master dark. If you for example dither and use 64 dark subs to create master dark you are in fact raising both dark current noise and read noise by 0.8%. Too much? Use more dark subs to create master dark.

Is there actual reason to use dark calibration if dark current is low and uniform? Yes there is - without removing dark signal (even really small like less than 0.5e per sub) you will have wrong flat calibration. Flat calibration should operate on light signal alone. Not removing dark current makes it operate on both - you will be "correcting" what is uniform offset of dark current thus creating "imprint" of master flat on your image.

[andrew s]

I found this Sony chip had a temperature dependent snow of warm pixels. See here for a discussion http://www.spectro-aras.com/forum/viewtopic.php?f=8&t=1919

Regards Andrew

[Magnum]

49 minutes ago, andrew s said:

I found this Sony chip had a temperature dependent snow of warm pixels. See here for a discussion http://www.spectro-aras.com/forum/viewtopic.php?f=8&t=1919

Regards Andrew

Not sure how they can call it a snow storm, the Sony sensors ive used show maybe a dozen, Kodak chips on the other hand could be described as a snow storm.

[andrew s]

1 hour ago, Magnum said:

Not sure how they can call it a snow storm, the Sony sensors ive used show maybe a dozen, Kodak chips on the other hand could be described as a snow storm.

These are not hot pixels  but warm. They do seem to calibrate out. The term was mine I would call 10,000 a storm 😀

Regards Andrew 

[Astroblagger]

10 hours ago, Magnum said:

Sorry if I went off the original topic slightly, my answer to that is I find -10C fine with Sony sensors, though he mentions running off batteries which I wouldn't recommend, you want to be above 12V as low voltage can increase noise,

Sounds like i'll stick to -10 subs then, getting a abit technical there, but all good advice, Unfortunately I have to run off batteries, if I give up batteries I have to give up astrophotography! Its hard to acquire images from an apartment 3 floors up.  Although I do use a separate 36A/h battery purely for the camera.

[Magnum]

3 minutes ago, Astroblagger said:

Sounds like i'll stick to -10 subs then, getting a abit technical there, but all good advice, Unfortunately I have to run off batteries, if I give up batteries I have to give up astrophotography! Its hard to acquire images from an apartment 3 floors up.  Although I do use a separate 36A/h battery purely for the camera.

If its a known good battery then that should be putting out around 12.5V anyway and be fine, just keep an eye on the voltage every few months.