Purchased a QHY8 Pro from a fellow member. I need some help setting her up.
What Gain and Offset setting do you guys use as a 'rough' starting point? I am using Maxim DL5, and I know that settings changes across software (eg EZCap, Nebulosity, Maxim... etc).
When I take a 600s dark, cooling @ -25, camera covered, the shot is very grainy. What could be causing this? (Again, using MaximDL).
When I took test shots with Nebulosity 2, demo version, I could see lines across the image...
I must be doing something wrong with the G and O settings...
Try Gain = 0, offset = 120 as a starting point. do a bias frame and alter the offset until you get about 800-1200 counts in a histogram. Leave the gasin at zero...
Are you still having issues with the O and G on that cam Logan? If you want I can take a few screen pictures of where it is you need to look to utilize the guide that you have of mine. That should be a cover for all of the camera's as it uses fundamental theory that isn't specific to one type of camera.
Ill fiddle and get something to you later this evening.
Hi Logan, there is very little need to increase the gain on these cameras. I keep it to 1 not 0 because CCDStack doesn't like FITS files with a 0 gain. Set the offset so your skycount (ADU) in your subs so it's very close to the pedestal of your darks or bias frames. Mine is around 950 ADU so I set the offset to 60 that gives me that same value in my subs in the dark parts. That may vary from camera to camera. Just find the sweet spot for yours. From memory the higher the offset the higher the background ADU in your light subs. Don't remember really I haven't changed those settings in years.
Gary, good numbers in the 'ballpark' to start with. Appreciated.
Thx Brendo, read that PDF cover to cover, but I obtained the settings via a demo version of Neb2. They are not transferrable to Maxim (but I understand that the theory is). Will have another crack tonight.
Thx Mark. From memory, I was testing it at G5 / O123 in Maxim. Will bump her back down. You've taken some great shot's with the 8. I'll try HA down the track as well.
1. Take a Bias Frame: camera covered, cooled to the temp I plan on using, -25c, shortest possible exposure.
2. Check out the Maxim Histogram (Screen Stretch box): should be around 1000 count, ascertain by looking at the bottom numbers (min and max)? Increase / decrease offset to change.
My settings where transferable to maxim, hence they give you a % gain and a numerical value. unless the qhy8 pro ascom driver is vastly different in set up to the qhy9 ascom driver they should come straight up. The good thing about that document is that it is a scientific way to calculate your offset and gain not just a oh yeah somewhere about there which will inevitably give you a half arse effort, starting points are all good but yeah...
With maxim when you are in the Camera control panel over where the Single, Continuous, Auto-save under that is a black arrow which will bring a drop down tab, within that tab you should find Camera settings, when the qhy8 is connected you should be able to bring up the Ascom settings, in there should be the Gain (%) Offset (numerical). Be aware that you have to do your gain and offset for both bin 1x1 and 2x2. My settings which are no reference to your camera is 1x1 = G:17% O:116 2x2 G:0 % O 118. if not just do a screen dump and send me a photo of your Ascom control panel and I will try and figure out what you need to do.
The difference in the operation and performance of my qhy9 with the proper calibration and the oh yeah she will be right is massive! My stars don't oversaturate anywhere near as quick (qhy9's are notorious for small well depths) and the colours are brilliant now even with standard qhy filters! though the red standard filter gives fire truck red stars that require a de saturating tool but that is another matter!
My histogram in Maxim has a very small peak with a large non saturated curve to the white point with a lot more data landing non saturated for 10 min images. For a 10" 4.7 newt this brings a fair amount of light into the system very quickly. What this means in the end is my stars have far better colour because the majority of them are not saturating, and the colour that lands in nebulosity is vibrant. If you look at my image of the area of Corona Australis notice all the tiny red and blue stars and stars within the diffraction spikes even to the major double star at the top has most of is colour, no its not "added in" high res click on the image to zoom in at 80% full resolution
PS. be careful of maxims standard histogram.... its more for screen stretching to see what you have in the image! there is a more accurate version within maxim that ill have to tell you where to look for it!
Neb2 has that nice info box that lists the 'mean' value. What info box / feature do I use in maxim to get the same 'mean' number? I'll just follow the same steps - but in maxim this time...
At the end of the day you want to maximise your dynamic range. That's what's important. Increasing gain will clip one end of the histogram, the offset will the other if you're not careful. So start at minimum gain then try various offsets to spread the data. Different programs interpret Gain/Offset differently so what works in Neb might not in Maxim DL or others and vice versa.
the only difference i can see with the different programs is how they calculate the mean and the others values needed to get your off set and gain values. I cant see how the ascom driver will vary the off set and gain from program to program based on the the program used that defeats the whole premis of ascom. I could be compleatly wrong though as i am not an expert in it but if you can explain to me how its changing that would be good to expand my knowledge.
Brendo, came across many posts on Cloudynights etc where people couldn't understand why settings changed across the software platforms. I'm new to this, no idea why.
But some good news, here is a screen shot I took of my settings:
When I take long exposures dark, eg 10mins with camera covered, why does it still look grainy like this? I'm new to asto CCD's... if I coverd my SLR and did a similar exposure - she was dark...
Q: what were you using to take pictures with your DSLR?
Maxim and Nebulosity will do an Autostretch and try show you what is there ie white noise. Using something like Canon EOS Utility for a DSLR will show unstretched data ie Black.
you have the right things up in maxim Logan, and the offset is about right, but it all comes down to your gain aswell. If you have followed the guide to the letter, then you need to tweek your offset just a wee bit more but its a tweak. Ill do some screen dumps now of the histogram i was talking about.
No calibration Gain 0 Offset 118
Like i said you can notice the distinct difference in the shape of the histogram with almost 100% of the data Landing under saturated. Where as the non calibrated is topping out at just on 1000 pixels and a fair bit of the main neb landing in the saturated area.
Also to note that the calibrated image was done with the 254 and the non calibrated with the ED 80 which has a considerable downgrade in F ratio and light capturing capacity. Both images where Luminance, both images where 10 min.
Last edited by bmitchell82; 23-05-2011 at 09:28 PM.
the only difference i can see with the different programs is how they calculate the mean and the others values needed to get your off set and gain values. I cant see how the ascom driver will vary the off set and gain from program to program based on the the program used that defeats the whole premis of ascom. I could be compleatly wrong though as i am not an expert in it but if you can explain to me how its changing that would be good to expand my knowledge.
In the past I have used CCDcap, Nebulosity, Maxim to capture, before settling down for Neb2. What I mean is that if you set Gain X, Offset Y in any of those programs it might not match and give you the same pedestal / background ADU. Neb2 and CCD cap where out by a lot I recall. Maxim seems to increase the gain more than Neb2 when you move from 1 to 10 for instance, in a less linear fashion. I don't know if it's each software doing its thing or if it is the various different drivers needed to interface the camera with each software. That's why I'm saying if Gain1/Offset 60 works fine in Neb 2 and gives me 950ADU in skycount it won't in Maxim and vice versa. So the bottom line is Gain/Offset settings are not set in stone. They will vary depending on the capture software.
I can see what your saying marc, and yes thats most likely right with the black points and white points changing though i beg to differ with the research that i have done on it.
Each program has a slightly different method of calculating statistics, a good test would be to run the same image though nebulosity and say maxim if they both come up with the same statistics then its how the programs interperate the drivers and settings. But i come back to ascom, if ascom isn't the main interface between drivers, control programs mounts and what not. Other wise ascom would need a driver for maxim, one for nebulosity and every other program.
When you do the calibration its not just setting the black and white point or getting the ADU above a cirtain number, its about optimising the CCD so that the pixel wells are actually getting everything that lands in there. the Black point that you talk about from my understanding is to get rid of the back ground noise and pixels that don't quite register as signal.
The Offset and gain are also affected by temperature so changing the temp will automatically change your settings, this might be the reason why your settings differ between programs as the QHY8 doesn't have TEC its just flat out like a lizard running from a big old wedge tail eagle! This is also the reason why i run at -25 deg C when my TEC is sitting at 40% duty cycle, because i havn't done the settings for -30 or -35 deg C
I do understand that not all ccds are the same but at the end of the day they arnt that different to each other and proper calibration will maximise your sky time!
If you're using Nebulosity, untick the "auto" up there near the histogram to turn off auto stretching. It doesn't affect the actual image, just how it's displayed on screen.