QHY268M & QHY268C Users Thread (Please share your knowledge and tips)

[Stuart1971]

Well just tried mine again, and got the dark patch in the middle again, this time I cooled to -10mover a 5 min period, and when at -10 the humidity was 64% which does not seem that high TBH, and it’s been that beofre and never had this issue…this is a flat frame below, as soon as I warmed up and it was at 2 degrees the flat frame was perfect again..

 

Edited January 3 by Stuart1971

[Rob_Jn]

If you’re still using the pale pink silica gel desiccant supplied with the camera then ditch it. As mentioned in one of the messages  use molecular sieve 4A. After a couple of days my RH dropped to 24% a bit extreme but it would ensure your chamber was thoroughly dried out. After a week or two the RH increased to 60% which says the chamber is not sealed effectively. I run it at about 65% at -10C and never had the dark spot problem.

[Starflyer]

What desiccant are you guys using? If it's the originally supplied QHY stuff it may well be saturated already from sitting there for a year or two.

This is what I'm using, 100g will last a lifetime, just store it in a small glass jar instead of the poly bag it arrives in.

[Stuart1971]

48 minutes ago, Rob_Jn said:

If you’re still using the pale pink silica gel desiccant supplied with the camera then ditch it. As mentioned in one of the messages  use molecular sieve 4A. After a couple of days my RH dropped to 24% a bit extreme but it would ensure your chamber was thoroughly dried out. After a week or two the RH increased to 60% which says the chamber is not sealed effectively. I run it at about 65% at -10C and never had the dark spot problem.

Mine that came with the camera is orange and came in a sealed bottle. Was supplied by Bern at modern astronomy when I bought the camera..

I am also at 65%  when at -10, and have figured out it’s the air in the filter wheel condensing on the cover glass when cooled quickly to -10, when I cool over a longer period, say 10 mins, it’s all fine…

Edited January 3 by Stuart1971

[Stuart1971]

The issue with that molecular desiccant is that you get dust with it that can get in the chamber, which is not good…I will stick with the stuff I have as it seems to work..

[assouptro]

Here in the North West Lancashire it has been really really wet for what seems like a ridiculous length of time! 
I have my rig permanently outside on a pier under a telegizmos 365 cover (has been for 15 years or so) 

I have a dehumidifier under the cover and up till now that has been enough, just recently, I’ve added another dehumidifier and a seed tray warming mat to attempt to alleviate the excessive moisture in the air and the length of time between letting the scope breathe! (Which I do as often as I can, weather and other commitments permitting) 

Basically, for my location, it’s the wettest it’s been for a long time. 
 

I’m not surprised there are a few soggy cameras out there! 

[Rob_Jn]

Yeah, you have to make sure you have a decent plug of cotton wool in the tube first.

[assouptro]

Apparently the short back focus of the qhy268m can cause issues with the front glass not warming quick enough to match the cooling chip 

slower cooling helps 

also the orange desiccant provided has the added benefit of changing colour to a green/blue when saturated 

cheers 

Bryan 

[tomato]

I used the orange coloured beads that came in a sealed bottle, my understanding that is usually the colour of 'active' silica gel beads, they go blue or green when saturated with water.

This offset number.. I'm sure I had it set it at around 50 previously and somebody on SGL  informed me that the figure was too low, I thought the the idea is you need a little bit of null space to the left of the histogram peak. I run the OSC camera at 155, it's something to play around with while it stays cloudy.

[Stuart1971]

12 minutes ago, Rob_Jn said:

Yeah, you have to make sure you have a decent plug of cotton wool in the tube first.

Wouldnt the cotton wool get saturated before the desiccant, which would then get back in the chamber….??

[Stuart1971]

8 minutes ago, tomato said:

I used the orange coloured beads that came in a sealed bottle, my understanding that is usually the colour of 'active' silica gel beads, they go blue or green when saturated with water.

This offset number.. I'm sure I had it set it at around 50 previously and somebody on SGL  informed me that the figure was too low, I thought the the idea is you need a little bit of null space to the left of the histogram peak. I run the OSC camera at 155, it's something to play around with while it stays cloudy.

Yes it is to Create a null space to the left, but you need that gap as small as possible, or you really reducing the dynamic range, 25-30 gives a nice small gap, this is the figure you should use all the time, 255 is way way too high, and also use the mode 1 with gain 56 it gives the best all round performance…

if you read back in this thread you will see what I mean, about offset, trust me, yours is way too high…

Edited January 3 by Stuart1971

[Stuart1971]

28 minutes ago, tomato said:

I used the orange coloured beads that came in a sealed bottle, my understanding that is usually the colour of 'active' silica gel beads, they go blue or green when saturated with water.

This offset number.. I'm sure I had it set it at around 50 previously and somebody on SGL  informed me that the figure was too low, I thought the the idea is you need a little bit of null space to the left of the histogram peak. I run the OSC camera at 155, it's something to play around with while it stays cloudy.

Oh, and if using NINA to check the offset by taking a bias frame, you need the graph full screen to see the gap on the left of the histogram, dont use it in a small window as the scale is far too small, you are looking for a very low reading on the MIN setting..

To simplify you should have no pixels with a value of zero on the min side, but try to keep down below 1000

and as few as possible above zero on the max side…

you will find with your setting of 255 that you have a hight number of pixels clipped on the min side of the graph, which is not good..

you also will need to adjust your gain too, as on the mode you use a gain of 26 is about right..

this is from QHY

 

Edited January 3 by Stuart1971

[Rob_Jn]

40 minutes ago, Stuart1971 said:

Wouldnt the cotton wool get saturated before the desiccant, which would then get back in the chamber….??

The cotton wool is just stop any dust getting into the chamber from the mol sieve. The sieve I use came from a lab I used to work in and is a high quality grade, I can’t vouch for the dustiness of what’s available now but I think the dust risk with a cotton wool plug is low. I’ve used mol sieve on all my cameras for the past few years, lasts longer than silica gel and just as easy to regenerate.

[Tomatobro]

If the range is 16 bit (0 to 65535) and the black level is set to 255 then the available range is 0 to 65280.

I use the DSO mode in Sharpcap and set the black level high and APP accepts the data with no issues.

I set the black level high because when I started with the camera I had some issues with processing and fixed it with the high black level (brightness in Sharpcap). I think my brother had similar problems and took his advice on the black level setting.

Interesting comments though and will experiment further 

[Stuart1971]

2 minutes ago, Tomatobro said:

If the range is 16 bit (0 to 65535) and the black level is set to 255 then the available range is 0 to 65280.

I use the DSO mode in Sharpcap and set the black level high and APP accepts the data with no issues.

I set the black level high because when I started with the camera I had some issues with processing and fixed it with the high black level (brightness in Sharpcap). I think my brother had similar problems and took his advice on the black level setting.

Interesting comments though and will experiment further 

Well, if you read through this thread especially the early part we do talk about all this and also QHY recommendations..

I have to say you are the first I have ever seen using this setting, and when I got my camera I read hundreds of forum posts and watched a ton of videos to get the best out of my camera….but each to there own I guess, but I dont use APP or Sharpcap I use PI and NINA, not sure why that would make a difference though…

Also I do think the 0-255 setting works like that. I think 0-255 if a relative scale of the whole 16 bit range but I may be wrong…

[Tomatobro]

Agree. The problem I find now is that imaging time is so precious that experimentation eats into the session so tend to stick with what has worked before.

I wonder what vliav thinks about this

[Stuart1971]

13 minutes ago, Tomatobro said:

Agree. The problem I find now is that imaging time is so precious that experimentation eats into the session so tend to stick with what has worked before.

I wonder what vliav thinks about this

@vlaiv

[tomato]

I read a lot reviews when the QHY268c first came out, the consensus then was to use DSO mode with gain 30. I was happy with the results so left it at that. However I dare say there have been developments since then and I will try the other settings. May I ask how much impact will these changes make to the final image, given all of the other contributing factors that make up the process?

In an ideal world we would image from the top of a mountain in the Atacama desert and get dozens of hrs of integration to improve SNR, so optimum camera settings would make a measurable difference but in practice we have Light Pollution removal tools, Blur and Noise Xterminator to help us out, which I’d venture make orders of magnitude difference to the final result.

I guess @vlaiv can make the science based case, and if someone cares to loan me another QHY 268 I could run a side by side trial on the dual rig. I tried a comparison with a QHY268c and an RC571c but they are not the same camera so running them with different gain and off set settings didn’t make any sense.

[vlaiv]

I see that I've been mentioned, but can anyone point out to the actual piece of discussion that you want my opinion on?

 

[Stuart1971]

3 hours ago, vlaiv said:

I see that I've been mentioned, but can anyone point out to the actual piece of discussion that you want my opinion on?

 

We were discussing the ideal gain and more importantly the offset settings of the QHY268 cameras…it seems that we have some people here that use 255 as the offset and others including QHY themselves,  just use 30 for the same mode 0 setting, and wondered what effect this might have, and if you have an opinion…

[tomato]

I had a wider question in that that although we should of course operate the cameras at the optimum settings for deep sky imaging, do these changes make a visible difference to the final result, given the magnitude of the changes made to the raw data by modern processing techniques (e.g LP gradient removal, BXT, NXT etc.

[vlaiv]

6 minutes ago, Stuart1971 said:

We were discussing the ideal gain and more importantly the offset settings of the QHY268 cameras…it seems that we have some people here that use 255 as the offset and others including QHY themselves,  just use 30 for the same mode 0 setting, and wondered what effect this might have, and if you have an opinion…

Ah, ok.

If everything is ok with camera hardware (so no funny business on some gain / offset settings) - here is how it behaves and what you should use:

Gain - there are only few things that change with gain - that is read noise and full well capacity.

Neither is very important when stacking.

If you can't make longer exposures for some reason - then it pays off to reduce read noise (increase gain), but you should always look to optimize your sub duration depending on read noise - swamp read noise with some other noise source like LP noise or thermal noise.

People love larger full well capacity - because less stars will saturate in the image, but simply put - you will always have stars that are bright enough that will saturate and this is handled in a different way. If you have star or parts of very bright target that saturates in your exposure (regardless of full well capacity) - you simply make few short exposures that won't saturate those parts - and then use signal from those images in places where you have saturation in regular stack. It is sort of HDR composing of the data.

I personally like unity gain with most cameras as it has good balance of read noise and FWC (and removes some minor quantization issues), but some camera models - that is not best option as unit gain is in high read noise part of graph - so slightly higher value is then recommended - like with ASI533 where it is not very clear where low read noise kicks in and people often use 101 instead of 100 that is marked as unity gain.

As far as offset goes - too high is fine, too low is not.

Too high will simply just rob you of some FWC - but like we've seen above - FWC is not that important and main drawback of FWC - which is saturation of pixels is dealt with differently (by using few shorter subs).

However, too low can really mess up your data as it creates different kind of clipping - not at high signal but at low signal.

There is simple method to determine if you have your offset set correctly:

- shoot some number of bias subs (say 16 or so)

- stack those bias subs with minimum method

- do statistics on resulting stack and see what is the minimum pixel value

- if minimum pixel value is 0 or generally smallest number that camera reports (maybe some cameras will report some other number as lowest number - I don't think I've seen that yet, but in principle it is possible) - raise offset a bit and repeat procedure.

If you don't want to do this measurement - just raise it enough to be safe (like people do when the push it to the max).

[vlaiv]

1 minute ago, tomato said:

I had a wider question in that that although we should of course operate the cameras at the optimum settings for deep sky imaging, do these changes make a visible difference to the final result, given the magnitude of the changes made to the raw data by modern processing techniques (e.g LP gradient removal, BXT, NXT etc.

I think I covered that as well, but just to make sure:

1. too much read noise can be seen in raw data, especially when you go for very faint stuff that is on border of detection.

Say you image with very cool camera and very slow system in very dark skies. Then you need to expose for very long and difference between one minute subs vs something like half an hour subs - can be significant, particularly if you have camera with say 3-4e of read noise on lowest gain setting and you use that gain setting versus something more sensible + longer subs.

In general case - difference will be maybe too small to be noticed by eye but can certainly be measured. 10-20% higher total noise is barely noticeable by naked eye.

2. If you set your offset too low - that you will see as artifacts in stacking and very strange things at lower end of intensities close to noise floor.

Nothing can help there. In point 1 - you can simply image for longer to achieve better SNR and compensate for increase in noise from using sub optimal settings, but with this - you can't fix the data, so you better make sure you don't clip on the low side.

[tomato]

I have just noticed this morning that along with the ambient air pressure, the sensor pressure reading has risen also. I'm aware this will go up and down in a sealed system in line with temperature changes, but it does appear to be tracking the outside air pressure, more evidence that the chamber is not sealed.

I know @Tomatobro has a barometer/RH station in his dome, he can maybe make a more precise observation on this when he sees then post.

Edited January 5 by tomato

[vlaiv]

2 minutes ago, tomato said:

. I'm aware this will go up and down in a sealed system in line with temperature changes, but it does appear to be tracking the outside air pressure, more evidence that the chamber is not sealed.

Does not have to be the case.

There is glass window and glass is not 100% rigid (nor is aluminum casing for that matter, but it is far more rigid than glass window). It deforms under pressure - enough to be able to equalize pressure on the other side if difference is small (and volume vs surface area of glass is such that small bend in glass can cause enough change in volume).