Cooled astro-cameras based on Nikon Z6 Sensor

[MarkS]

The 24Mpixel Sony IMX410 full-frame colour sensor used in both the Sony A7iii mirrorless and my Nikon Z6 mirrorless cameras will shortly appear in two cooled astro-cameras:  ZWO ASI2400MC PRO and QHY410C.

They could both be phenomenal one-shot-colour cameras!

Mark

[RobertM]

Very interesting - the ASI2600MC is also 24Mpix(ish) but 16bit which means the 2400 will have files 25% the size.  The photosites will also capture 2.5 times more photons and a massive field of view !  Drawback - they are 14bit and getting on for twice the price !

Ah well ...

[NoelC]

Very interesting and a lot of money.
Does cooling the chip make that much difference?  The quoted dark current is about 1/100th of the read noise at 20 degrees.  And zero amp glow; a distinct advantage, but is it hardware or an algorithm?
Noel

[MarkS]

Very interesting - the ASI2600MC is also 24Mpix(ish) but 16bit which means the 2400 will have files 25% the size. 

Err, no.  The size difference would be a factor of 14/16 or for typical file formats (e.g. FITS/TIFF) none at all.

Does cooling the chip make that much difference?  The quoted dark current is about 1/100th of the read noise at 20 degrees.  And zero amp glow; a distinct advantage, but is it hardware or an algorithm?

Very interesting questions.  Most amp glow effects are caused by internal IR radiation from the circuitry of the sensor.  The solution to this is in the hardware design of the sensor and which parts of the sensor are switched on during a long exposure, which might be controlled by the camera firmware.  I've noticed no significant amp glow on this sensor and in any case it would calibrate out.

The question of cooling is very interesting and I'll give you my considered opinion on this because there is so much "mythology" surrounding the subject in typical forum discussions (OAS excepted!).  My own experience, from taking careful measurements and performing careful analysis of sources of image noise, is that well built cameras such as the Sony A7S or Nikon Z6 dissipate internal heat very well and consequently have very little camera self-heating during a long imaging session.  Night-time temperatures here in the UK never become hot enough for thermal noise to become a problem when doing full colour imaging under my typical sky conditions.  This is because the noise from the background sky signal (light pollution) swamps other sources of noise such as read noise and thermal noise.  This is the reason I've never seen the need to buy a cooled camera.  It's quite simply because it won't make any visible difference to the results of my imaging. 

However the situation is different for those imaging in very warm places or for those doing narrowband imaging.  Narrowband filters make a huge reduction in background sky signal and so thermal noise can become more dominant. That being said, you are unlikely to buy a colour sensor for narrowband imaging.  On the other hand, the new crop of dual/tri/quad narrowband filters could change the argument here and make cooling more important for colour sensors.  I don't have such a filter but I keep meaning to perform some further analysis on this question.

Mark

[Roberto]

Drawback - they are 14bit and getting on for twice the price !

Indeed; this is still the issue.  Lower stops and more expensive.  The pixel size makes them very attractive for a wider category of instruments though.

I have been looking at the ASI6200 to eventually replace my KAF-16803 based CCD.  The pixels of the ASI are too small even with a 0.75x reducer for the area I want to cover.

I'm sure there'll be another CMOS chip out in the next year or so that will fit the bill though!  They are coming out thick and fast.

Roberto

[MarkS]

Drawback - they are 14bit and getting on for twice the price !

Indeed; this is still the issue. 

The difference between a bit depth of 14 and 16 is a red herring.  Given a full well depth of 120k electrons, at the lowest gain setting the 14 bit representation leads to a quantization of approx. 7e / digital unit (incredibly the read noise is less than 1 digital unit).  Under typical deep sky imaging situations the sky-fog swamps this and you will actually see no difference in your final stacked image between 14bit and 16bit acquisition.   It's the sky-fog that sets the practical limit to the dynamic range of the image, not the capabilities of the camera.

As a matter of interest and ignoring sky-fog, the per-pixel dynamic range (i.e. full_well/read_noise) of the 16bit ASI2600MC Pro is lower than the 14bit ASI2400MC PRO !  I say per-pixel dynamic range because there are valid alternative metrics for dynamic range.

Mark

[RobertM]

Yes you're right Mark I was thinking 16bit integer values which is indeed one quarter the numeric value but not the saved file size which also depends on the storage unit and compression (if used).

With the quantization and sky fog, I do know that but for some reason it did not register.  I actually think the 2400 could end up being the ideal colour camera but I'm not going to be on the bleeding edge of that one especially as it is getting on for twice the price.  The ASI2600 hits a sweet spot for me at the moment but I'll keep an eye on the new camera to see how it goes.  One thing I do notice already is that with the increased well depth I get far fewer blown stars.

As for multi-band spectral filters, this is precisely the reason I bought the 2600.  I also bought the Optolong duo band filter with the idea being that I can switch filters from RGB to NB depending on subject or conditions.  Much time was wasted trying to get all the different filter captures that it was getting difficult to complete images - at least with a OSC the integration time for all channels will build up linearly.  Time will tell if this was the right choice given my sky conditions.

Robert



Robert

[MarkS]

As for multi-band spectral filters, this is precisely the reason I bought the 2600.  I also bought the Optolong duo band filter with the idea being that I can switch filters from RGB to NB depending on subject or conditions.  Much time was wasted trying to get all the different filter captures that it was getting difficult to complete images - at least with a OSC the integration time for all channels will build up linearly.  Time will tell if this was the right choice given my sky conditions.

Aha - that makes sense.  I'm looking forward to your multi-band results.

Mark