ASI462MM sensor analysis report

[tonystar]

Hey Robin,

How do I read this sensor report for the ASI462MM? Anything worth noticing? I cant compare to any other ASI cameras since I sold all of them!
I just followed direction as best as I could....hope it was acquired under the right settings.
Thanks,
Tony

ASI462MM sensor analysis.jpg (129.01 KiB) Viewed 7503 times

[tonystar]

Addendum: someone just mentioned tha ZWO's website suggests read noise as low 0.47e at 350 gain vs my value of about 0.88e

[admin]

Hi Tony,

the measurements seem a little off - firstly the e/ADU for minimum gain has come out at nearly 3, whereas ZWO have about 2.75 (see https://astronomy-imaging-camera.com/pr ... 462mccolor) - this is about 7% high, and because of the way the analysis works, this error affects all the other e/ADU and read noise values, making them all 7% higher than they should be. This also makes the full well calculations high by the same amount.

Another thing to note is that the 'Sensor linear to' reading is only 39% - this almost certainly means that the light level was varying somewhat during the stages of the analysis when the sensor was illuminated - this will affect the figures too.

The read noise of 0.5 on the ZWO is interesting - the errors above would not account for the size of difference between the 0.86 or so that you got and the ~0.5 claimed by ZWO (maybe about half of it). I'm not quite sure what is going on there - I don't think I have a 462 sensor here to test myself, must see if I can get hold of one to see if the analysis measurements work at such low noise levels.

I would be tempted to try another couple of measurements to see if you can get steadier illumination and a better value for the e/ADU for minimum gain.

cheers,

Robin

[tonystar]

Hey Robin, thanks for prompt feedback!
I repeated the measurements, this time being more careful not to change the light levels around the camera (although the test is quite fast since I am only selecting maybe 300x300 pixels area for the test).
I get a much better linearity and e/ADU level close to 3 @G=0. Is this linearity typical for a CMOS sensor?

Also, when I compare to the data reported by Player One (https://player-one-astronomy.com/produc ... ra-imx462/) for their camera with the same sensor, (done with Sharpcap) the readout noise lever @G=350 is very similar, which shows that the amplification circuitry for both manufacturers performs very similarly.

ASI462MM sensor analysis_repeat.jpg (124.06 KiB) Viewed 7491 times

[admin]

Hi,

normally I would expect linearity to run to about 95% or more with a CMOS camera - that will require optimal conditions though - I would still suspect some sort of variation in the illumination levels giving you the lower reading of 62%.

I just realised I was talking rubbish about not having an IMX462 camera - I have a PlayerOne Mars-C (it was actually sitting on my desk). I will run an analysis now and see how the results compare.

cheers,

Robin

[admin]

Here we go - PlayerOne Mars-C IMX462 results

Capture.JPG (77.87 KiB) Viewed 7460 times

I should have really added a few extra measurement points around gain 60 to 100 to capture the change from LCG to HCG mode, but the raw figures are very similar to yours - ie the e/ADU closer to 3 (a bit over 3 in my measurement) and the read noise only going down to about 0.75. I did get linearity to 98%, which is what I would expect typically. (Linearity measurement done under artificial light, making sure no light from the screen contributes to the illumination).

Interesting that ZWO claim 0.5 read noise - there are two techniques to measure the read noise and SharpCap uses the quicker and simpler one (the other would require a series of exposures at each gain value with different lengths and then to calculate the slope and intercept of the best fit line). Maybe that other method gives different readings in some cases? In practice, I'm not sure that it is very important - for deep sky you would typically not use such high gains because of the reduced full well depth (and reduced dynamic range of each frame). For lunar/solar/planetary, a read noise of 0.5 or 0.75 is still tiny compared to the shot noise in well exposed frames, so will contribute practically nothing to total noise.

cheers,

Robin