What can you expect from new cooled CMOS based cameras in DS astrophotography?


In the upcoming weeks we will see new cooled ZWO and QHY cameras with Sony CMOS sensors. Those cameras can do Solar System as well as DS imaging and will be somewhat cheaper than most DS cameras. So do those cameras are as universal as they suggest?

Noise sources in CMOS sensors

Soon we will be able to buy a cooled ASI224 variant (pixel 3,75, digonal 1/3") as well as ZWO/QHY cooled IMX174 variant (pixel 5,86, diagonal 1/1.2"). The color IMX224 has extremely low read noise and is already widely used in planetary astrophotography. Mono IMX174 offers excelent performance for Lunar or Solar imaging as well as quite big pixel for a new sensor. But what happens when we use them on longer exposures?

When using a CMOS sensors we will notice Fixed pattern noise (FPN) as a darker/brighter vertical columns. As the noise is fixed it can be removed by dark or bias frame. Second noise source are random horizontal lines that can't be removed by dark/bias frames. Stacking multiple frames can blur it out, but that's not always possible or won't work. There are two solution for this. One is to use special algorithm during processing (latest Autostakkert2 has such option) or flood the lines noise in uniform noise floor by using higher gain (possible in miniCAM5). Then we get uniform noise among the whole sensor.

Aside of noise there seems to be a quite apparent amp glow in Sony IMX sensors - in the bottom right corner. As it's fixed noise it will be removed by a dark frame. ZWO and QHY had to spend some time on the drivers to limit it, but still amp glow will be visible on longer exposures.

Newer Sony IMX sensors have much less FPN than older Micron sensors (miniCAM5, ASI120MM, QHY5L-II), however without processing they still don't give as uniform noise as most Sony CCDs in typical DS cameras.

Example pictures

On cloudynights James posted a raw Rosetta nebula frame where you can see the amp glow as well as some banding. On the final image they are gone. Note that the exposure time was only 32 seconds. Second James picture showcases M42 nebula.

On Facebook Dominique is testing cooled ASI224. As you can see cooling limits the amount of warm/hot pixels present on the dark frame. Amp glow and banding remains even on cooled dark frames. However quick 10 x 60 sec Pleiades picture can reveal the reflection nebulosity present in that area.

Also Emil has a lot of excelent examples made with ASI224, PGR Blackfly IMX249 (slower IMX174 variant) and other cameras. The NGC 7331 galaxy looks nicely on one second exposures, while M51 was done on 4 second exposures.

New DS imaging technique

As you could notice most DS images made with those cameras uses short exposures. Some were uncooled which could force short exposures but also planetary photographers are quite good at getting most from images taken on short exposures. Aside of some very high end DS imaging your DS camera will have only fraction of histogram filled on 5-10 min exposures. In such cases high well depth isn't used, but low read noise allowing stretching the signal in processing. The same idea is used with new CMOS based cameras. Very low read noise allows reducing exposure time and capturing many frames for stacking. Few second exposures may make guiding unnecessary or also allow high resolution DS imaging without seeing blurring effect.

So if you want a universal camera then the new cooled cameras may be a good pick. Just remember that for DS imaging they are best used differently than typical DS cameras and that they won't replace high end big DS cameras for really high end DS imaging.

Comment article