Thread: More sparkles
View Single Post
  #12  
Old 20-10-2020, 07:27 PM
Peter Ward's Avatar
Peter Ward
Galaxy hitchhiking guide

Peter Ward is offline
 
Join Date: Dec 2007
Location: The Shire
Posts: 8,088
Quote:
Originally Posted by rustigsmed View Post
excellent shot Peter, interesting regarding the background noise levels though, i thought they were suppose to be super clean above gain 30 - perhaps that was just readout noise and no amp glow... although it might be worth punching out shorter subs as you're dealing with a cmos sensor now..?

cheers

Russell
I'm still working this out as I go, but I suspect read noise and pixel to pixel uniformity are two different things. sCMOS looks to have a novel notion of how to best define read noise, e.g. median or RMS...but with many sensors this can still be a remarkably low number and makes you wonder, despite the disparate numbers, why CCD's still look so smooth.

Here's my take: CMOS have a transistor for every pixel. So each pixel has its own "weight scale".
Which in effect deliver an albeit very repeatable signal (i.e. low read noise) but the overall sensor has a "film grain" pattern to it. (as all the" scales" read a little differently)

Yet a CCD is read-up effectively by one transistor for the entire array.
Just one "scale" that each pixel has to jump on to be weighed. It takes longer, but there is no variation in individual "scale" readings to contend with.
The total signal might drift high or low after a readout and shows as a higher total read noise rather than an embedded "grain" in the sensor.

I'd be more than happy to hear a decidedly more erudite take on the above, from those better versed in electronic engineering than myself

Last edited by Peter Ward; 21-10-2020 at 09:53 AM.
Reply With Quote