15th Feb 2024: Astro Pixel Processor 2.0.0-beta29 released - macOS native File Chooser, macOS CMD-Q fixed, read-only Fits on network fixed and other bug fixes
7th December 2023: added payment option Alipay to purchase Astro Pixel Processor from China, Hong Kong, Macau, Taiwan, Korea, Japan and other countries where Alipay is used.
Hi,
I‘m hoping to be in the right place to pose my question here:
I recently got into shooting with a dedicated mono camera (namely the Altair 183M Pro fan-cooled). As some may be aware, the Sony chip as a pronounced star burst when exposing over a longer time. Altair‘s support has suggested me to use AstroPixelProcessor which seems to handle the star burst pretty well.
My question is now regarding the temperature of the dark frames. Since I cannot control the sensor temperature, I usually should match the temperature of the light and dark frames well. How well is well?
I don’t have a dark library yet and to assess my lights, I subtracted darks at higher temperature. In the first run I had lights at 10 deg C and darks at 20 deg C. The integrated image showed a lot of dark spots which I attribute to the clipping from subtracting the hot darks from the cold lights.
I am wondering if there is a way to handle strong temperature gradients within a session with either darks a bit too hot or too cold?
Last night, I started with 10 deg C lights and ended at 4 deg C. It would seem that it may become painful to have a library of darks for each degree Centigrade?
With a new set of darks at 10 deg C, the results are much better but still there are black spots, probably from the 4 deg C lights calibration.
I‘ve been playing with the rejection on the integration step lately and set a kappa low of 3.0 which improves a lot on the black spots on the final image.
I am wondering if there is a good combination between dark library temperature resolution and settings in APP to improve on that issue?
Any comments or suggestions are strongly appreciated. I guess, a temperature controlled/cooled camera might be a convenient remedy but I bought the camera with budget in mind and also the fact that this model doesn’t need additional 12V.
Finally, a few additional numbers: exp. time per light 120“ @ gain 200 @ 2x2binning. I have 104 lights in total, 30 darks, 30 bias and 30 flats (0,5 secs per flat)
Clear skies!
Björn
@barnold84 Björn, this is a challenge that all uncooled cameras struggle with. I don’t know the specs of the sensor well enough, particularly the amount of dark noise at various temperatures, but I would argue to use darks that have the same or lower temperature than the lights. Better not enough noise reduction that too much. You can use a fridge or other cool environment to cool the camera.
The effect of temperature on the sensor dictates how much warmer the lights can be. Two degrees of temperature difference produces a much larger difference in dark noise around +20° C than around 0° C so it is up to you what’s acceptable and what not.
@wvreeven Hello Wouter, thank you for your quick reply. For the first session with a few lights, I even just stacked the subs without any dark subtraction and the results wasn’t that bad expect for the strong star burst.
I‘ve seen a YouTube video where „my“ chip (Sony IMX183) was mentioned. It has a comparably low thermal noise (less than 0,25 electrons per second at approx 20 deg C) but since I also have very dark sky (fortunately), the thermal noise is no negligible compared to the background signal.
Maybe I should give it a try and use colder darks, hoping that the star burst is still removed. I‘ve a cooling box where I can achieve a temperature of about 2 to 3 deg C and quite constant. I‘ll try to create a set of darks at this temperature and see how the result looks.
Björn
@barnold84 Hi Björn, ZWO use the same sensor and they have published these diagrams for it:
That should give you a good idea of the behavior of the sensor at various gains. Unfortunately there is no temperature info for it.
@wvreeven Hello Wouter,
I actually found the dark current data for the chip from two other vendors:
IMHO, a quite low noise chip. However, it confuses me a bit that the QHY shows the expected exponential increase with temp, while the ASI seems to be very close to linear. Also the scale is very different: there seems to be a factor of 2 for a given temp between the two cameras. Something that shouldn't appear for the same sensor!?
In any case, the dark current is around 0,01 or 0,02 while my sky background should be something like 0,3 to 0,5 electrons per pixel per second (depending if I use my f/8 or f/6).
One could say there is an order of magnitude between thermal noise and sky background. Actually, I am just wondering if with these values, one would ever need dark subtraction!? Of course, the star burst ("amp glow") is the killer for this approach.
Björn
@barnold84 Hi Björn, good to see that you found those plots. The dark current not only depends on the sensor but also on the camera electronics. That explains the difference between the two plots.
The QHY plot cuts off at +15º C at which the dark current is expected to be 0.062 e/pix/s which is reached by the ASI camera at around +26º C. So there is quite a difference. And you are shooting with a mono camera so the sky background varies with the filter used as where the dark current is independent of that. So I'd still strive for cold darks though I'll admit that I don't know how much the amp glow will vary with temperature.
