The adventure with the QHY247C camera has begun due to my portable imaging setup doubts. After several summer holidays with my traveling setup, I decided to alter it a little. Well, not a little, because only one element left 🙂 For most of the time, I have been using Samyang 135 telephoto lens with QHY163M camera with filter wheel. All that stuff sits on the iOptron SmartEQ mount. All worked just fine, was not quite heavy, but was also far from simplicity. iOptron was replaced by SW Adventurer, and the mono camera was replaced by astro modified Canon 550D. But that not lasted long, and I decided to make another attempt at the dedicated OSC camera. I used to have QHY163C for some time, but I was not impressed. This time I wanted something bigger (APSC or more), with low read noise (for live stacking) and with Linux support (for my next AstroLink project). Plus it should not ruin my budget. And I have chosen QHY274C – 24Mpx APCS sensor camera. Not BSI, so a little bit less sensitive, but with low read noise due to rolling (not global) shutter, with relatively large sensor and high resolution.

QHY247C camera with Samyang 135 lens

There is a very good reason to use a DSLR camera for astrophotography – the price (unless DSLR is more expensive than a dedicated camera, then it makes no sense). I used to picture the sky using DSLR cameras quite often and for quite a long time. I have modded several Canon cameras, EOS 20D was the first one. One may achieve excellent results in astroimaging using a general-purpose camera. However, there are substantial differences between consumer market cameras (DSLR or mirrorless) and dedicated cameras. Let’s have a look.

First of all, dedicated cameras usually are cooled. Cooling decreases the noise level present in the long exposure image. Modern cameras thermal noise is not as much of an issue, as it was some time ago, however it is still better to have a cooled camera, than not have. Especially if you use any kind of filtering – light pollution, narrow or duo narrowband.

Another important aspect of a dedicated camera is that they provide the true raw image. Some time ago raw images from DSLR also were true raw, but modern cameras modify even the image saved as so-called raw, and that is not acceptable in astrophotography, because you need to have raw data and full control during postprocessing.

Probably you already heard about astro-modifications of cameras, so they are more sensitive in red. Some modern cameras are already a little bit more sensitive in red, so they capture more hydrogen alpha band (at 656nm) that is crucial for imaging emission nebulae. And here is the trick, that actually does not allow for a camera to work well for astrophotography and regular photography at the same time. Take a look at the plot below. There are three spectral response curves: of the human eye, of the regular camera, and of the astro-modified (or dedicated) camera.

QE human eye
QE ASI1600

It is clearly visible, that sensitivity of the human eye (the first plot) decreases in the red starting from about 600nm. At hydrogen alpha band (656nm) this sensitivity is about 15% of the maximum sensitivity. Now the thing is, that consumer market cameras must follow this human eye sensitivity curve to be able to reproduce colors correctly. And that is true – you can see at the typical color camera quantum response (second plot) that is very similar to human eye response. With all consequences, also with low sensitivity in the hydrogen alpha band. This is not really desired for astrophotography, because as you know for sure, the Universe is filled with hydrogen, and hydrogen light carries much information. That is why we mod DSLR cameras, so they are not limited in this spectral range, and are more sensitive in far-red, as you may see in the third plot that shows dedicated camera spectral response. And once you mod the DSLR for astroimaging you will not be able to do 100% correct regular photography, because captured image already misses the information required for correct color reproduction. Color balance may be adjusted later on, but will never be 100% accurate (unless there is no red in the image 🙂 )

Yet another aspect of a dedicated astroimaging camera is an anti-aliasing filter. Consumer cameras usually have one, that works stronger or lighter and make images less distorted, but also softer. If you do astrophotography, then you do not want it. For the same reason as before – you must have full control over the image during postprocessing. It is easy to simulate an anti-aliasing filter during processing, but it is not easy at all to remove the effects of the already applied anti-aliasing filter, without losing data.


As you can see in the image above, the QHY247C camera is quite large. It will work with Samyang 135 lens on the SW Adventurer mount. To operate the DSLR it is enough to have a programmable shutter release cable. The dedicated camera needs digital control, and in my travel setup, it will work with the Raspberry Pi 4 device under Astroberry software. All connections worked fine during the first desktop test, and the 48MB raw file is transferred to RPi4 within 3-4 seconds. Not blazing fast, not really suitable for VAA or Live Stacking, unless I have misconfigured something.

Ekos panel in Astroberry
QHY247C sample image

QHY247C camera is quite bulky, its outer diameter is 90mm. The sensor is hidden behind the heated AR glass cover, so we do not need to worry about moisture condensation. There is an M54x0.75 female thread around the sensor window. AR glass does not provide any IR filtering, so we need to add a filter to the optical train – IR cut, UV/IR block, UHC, LP, dual-band, or else.

The camera driver contains usual adjustments – gain and offset. Gain can be set in a range of 0-4065 and offset between 0 and 255. Unfortunately, the gain regulation is extremely non-linear – it is proportional to the reciprocal of the real gain. Assuming the pixel capacity is 32000e, then for selected gain values we will have the following camera parameters:

e- e- EV
0 1 2,45 32000 13,7
500 1,14 2,36 28000 13,5
1000 1,33 2,21 21000 13,2
2000 1,96 2,13 16000 12,9
3000 3,75 1,99 8500 12,1
4000 43 1,59 745 8,9
4050 100 1,38 321 7,9

So the gain rises very slowly, and at the very end of the slider, it becomes significantly larger fast. I estimate for the long exposure imaging with broadband filters we should keep the gain from 0 to 3800. For dual-band or narrowband filters, I would set the gain above 3000 and for short exposure imaging or VAA above 3500, or even above 4000. This is how it looks in the current driver version, but it may change in the future, as we know that QHY played in this area a few times.

I also measured the camera current consumption:

  • 0.35A when the cooling is off
  • 2.9A at 100% cooler power
  • when the temperature is set 20*C below the ambient, the cooler works at 25% and the current is 0.65A
  • when the temperature is set 35*C below the ambient, the cooler works at 55% and the current is 1.40A

And that is all for this moment, as I am still waiting for the clear sky to do some night tests 🙂