47min first light Qhy268m

[assouptro]

Hi guys!

I’ve had a real rollercoaster of a ride since buying a second hand QHY268m

Firstly, I had to repair the QHY 2” filter wheel that I got with the camera, then find some 2” filters! Up till now I am used to using astrodon 1.25” filters that are faultless, they never let me down, but to buy the 2” equivalent I would be looking at taking out a mortgage!! So I took a punt on Antlia 4.5nm edge filters that cost more than the camera! Then I had to set up a new laptop that had the required usb3 inputs! Then I had to learn how to use new software, namely Nina so I could make it all work.

I finally got all the software working and had a short spell of clear sky

I took some 2 min subs of the Cygnus region in narrow band with my Samyang 135mm lens, some of which were spoiled by high thin cloud that affected the Oiii filter more than the other 2, I’ve yet to confirm halo free oiii with the Antlia filters but the ones taken without cloud look promising 

I have had a nightmare trying to achieve flat frames, so the results are calibrated without them 

Anyway….rant over…

Here is my poorly processed first light 

47 min in total SHO jpg 

Thanks for looking and sorry for the long description, it can be a lonely hobby, family and friends don’t always get the frustrations and obsession over capturing photons of ancient light from objects that are not only so far away but no longer probably look anything like the image captured 

I hope to learn how to tame the new equipment 

Bryan 😊

[AMcD]

I feel your pain, Brian. 

I purchased an QHY268M a month or so ago, together with a filter 2" wheel.  I love the new equipment but have concluded that making any significant change to a setup (in my case, from 20 years using OSC cameras) means having to solve problem after problem before you get back to where you were. 

Once the 268M was attached, it took me a while to figure out back focus as I had mis-remembered the backspacing for my TS152 (solved with the help of @Richard_ and @teoria_del_big_bang and others). I then had to re-cable two or three times due snagging and balance issues (solved with the help of @ollypenrice and others).  I then had a period where plate solving would not work in SGP as I had set my mount safety limits to account for the increased size of the camera and filter wheel but with the result that I excluded an area of the zenith from being accessible to the mount's go-to (solved with the help of Brian Valente on the Gemini IO User Group). After that, it transpired that whilst my Crayford focuser could handle the QHY8 without difficulty, as the nights got colder, I started suffering focuser slippage with the 268M plus filter wheel and had to re-condition and adjust the Crayford (and investigate rack and pinion focusers in case the fix does not hold).  Finally, I have had to order a new laptop for processing as my aging PC cannot be upgraded beyond 8GB of RAM which simply does not cut it for Photoshop given the size of the subs produced by the QHY268M (I can watch a movie in the time it takes for PI to execute Starnet2 on my current PC). 

However, your really excellent image shows that all the effort to upgrade in this crazy, and sometimes lonely, hobby is ultimately worth it!  😊

[assouptro]

1 hour ago, AMcD said:

I feel your pain, Brian. 

I purchased an QHY268M a month or so ago, together with a filter 2" wheel.  I love the new equipment but have concluded that making any significant change to a setup (in my case, from 20 years using OSC cameras) means having to solve problem after problem before you get back to where you were. 

Once the 268M was attached, it took me a while to figure out back focus as I had mis-remembered the backspacing for my TS152 (solved with the help of @Richard_ and @teoria_del_big_bang and others). I then had to re-cable two or three times due snagging and balance issues (solved with the help of @ollypenrice and others).  I then had a period where plate solving would not work in SGP as I had set my mount safety limits to account for the increased size of the camera and filter wheel but with the result that I excluded an area of the zenith from being accessible to the mount's go-to (solved with the help of Brian Valente on the Gemini IO User Group). After that, it transpired that whilst my Crayford focuser could handle the QHY8 without difficulty, as the nights got colder, I started suffering focuser slippage with the 268M plus filter wheel and had to re-condition and adjust the Crayford (and investigate rack and pinion focusers in case the fix does not hold).  Finally, I have had to order a new laptop for processing as my aging PC cannot be upgraded beyond 8GB of RAM which simply does not cut it for Photoshop given the size of the subs produced by the QHY268M (I can watch a movie in the time it takes for PI to execute Starnet2 on my current PC). 

However, your really excellent image shows that all the effort to upgrade in this crazy, and sometimes lonely, hobby is ultimately worth it!  😊

Glad to know I’m not alone!!

my laptop has 16g of ram but still takes about 20min or more  to remove stars, I’m considering building a processing desktop 

I haven’t tried plate solving or installed the auto focuser yet! That’s all to come🙄

can I ask you a quick question? 
I was using mode 0 gain 29(I think) offset 25 

what settings do you recommend to play around for narrowband? 

Clear skies are rare here atm and I can’t afford the luxury of testing different modes and gains right now 

Thanks for the input 

Bryan 

[AMcD]

53 minutes ago, assouptro said:

what settings do you recommend to play around for narrowband? 

For narrowband I have been using High Gain Mode, with gain at 0 and offset at 40, but am going to experiment with the High Gain Mode and settings of gain 56 and offset 40 as I think this may be better with narrowband filters given the reduced number of photons.  For LRGB I use High Gain Mode with gain 0 and offset 30.  I cannot claim there is any 'science' behind these numbers.  They are what I arrived at having trawled through SGL, CN and Astrobin 😂

With the focusing, I have found auto-focusing the Ha filter a bit of a challenge.  The solution appears to be to bin the camera 2x2 for focusing runs with an exposure time of 60-75s.  It results in a longer auto-focusing routine but captures enough stars to be accurate.

I have ordered a 32GB laptop as there was a (seemingly) good deal on a Lenovo 32GB with RX Vega Graphics on Amazon.

Apologies for misspelling your name!

Edited October 25, 2022 by AMcD

[Richard_]

Brilliant image Bryan and a great result from your first light 🙂 I've got the same camera and filters (36mm unmounted) and I'm really impressed with them too. I haven't tried any reflection/halo tests on bright stars with the OIII filter yet. I completely agree with your comment regarding "lonely hobby", as I also appreciate that friends and families aren't as impressed with the frustration of getting things to work as much as they are impressed with the final image!

1 hour ago, assouptro said:

Glad to know I’m not alone!!

my laptop has 16g of ram but still takes about 20min or more  to remove stars, I’m considering building a processing desktop 

I haven’t tried plate solving or installed the auto focuser yet! That’s all to come🙄

can I ask you a quick question? 
I was using mode 0 gain 29(I think) offset 25 

what settings do you recommend to play around for narrowband? 

Clear skies are rare here atm and I can’t afford the luxury of testing different modes and gains right now 

Thanks for the input 

Bryan 

For narrowband, I am using Readout Mode #1 with Gain 56 and Offset 25 which is kinda the "default mode" that you'll find people using for narrowband. At Gain 56, the camera switches from low to high gain mode which is results in a significant reduction in read noise. I've linked one of my posts in the QHY268 thread explaining how to run a sensor analysis in Sharpcap for this camera. By default, Sharpcap tests gain in steps of 10 (eg 0, 10, 20, 30) but I included some instructions saying how to include discrete intervals (eg 55, 56 and 57). When you execute the analysis, you'll see the sharp drop off at Gain 56.

For broadband (ie LRGB) I've recently been imaging using Readout Mode #1, Gain 0 and Offset 25 to give me a larger full well capacity. At narrowband settings (Gain 56, Offset 25) I found stars were saturating as early as 1-2 minutes. Rather than fill up my hard drive and slowing Pixinsight down with tons of 1 minute subs, I opted for a lower Gain and to keep my subs at 5 minutes to match narrowband sub length. 

I haven't tried testing the other read modes or gain/offset values not stated in this post 🙂

 

Regarding star removal, are you using Starnet++, Starnet 2 or StarXterminator? Also, do you have an NVIDIA graphics card in your computer/laptop? If you do, there's a way to utilise your graphics card instead of the CPU when running Starnet which significantly reduces the time it takes! Running Starnet 2 via pixinsight on my gaming laptop with an RTX 2060 Max-Q takes less than 1 minute on images taken with my QHY268M 😀 I've shared a link below which explains the process. 

https://www.williamliphotos.com/starnet-cuda

Edited October 25, 2022 by Richard_
Corrected spelling

[assouptro]

5 minutes ago, Richard_ said:

Brilliant image Bryan and a great result from your first light 🙂 I've got the same camera and filters (36mm unmounted) and I'm really impressed with them too. I haven't tried any reflection/halo tests on bright stars with the OIII filter yet. I completely agree with your comment regarding "lonely hobby", as I also appreciate that friends and families aren't as impressed with the frustration of getting things to work as much as they are impressed with the final image!

For narrowband, I am using Readout Mode #1 with Gain 56 and Offset 25 which is kinda the "default mode" that you'll find people using for narrowband. At Gain 56, the camera switches from low to high gain mode which is results in a significant reduction in read noise. I've linked one of my posts in the QHY268 thread explaining how to run a sensor analysis in Sharpcap for this camera. By default, Sharpcap tests gain in steps of 10 (eg 0, 10, 20, 30) but I included some instructions saying how to include discrete intervals (eg 55, 56 and 57). When you execute the analysis, you'll see the sharp drop off at Gain 56.

For broadband (ie LRGB) I've recently been imaging using Readout Mode #1, Gain 0 and Offset 25 to give me a larger full well capacity. At narrowband settings (Gain 56, Offset 25) I found stars were saturating as early as 1-2 minutes. Rather than fill up my hard drive and slowing Pixinsight down with tons of 1 minute subs, I opted for a lower Gain and to keep my subs at 5 minutes to match narrowband sub length. 

I haven't tried testing the other read modes or gain/offset values not stated in this post 🙂

 

Regarding star removal, are you using Starnet++, Starnet 2 or StarXterminator? Also, do you have an NVIDIA graphics card in your computer/laptop? If you do, there's a way to utilise your graphics card instead of the CPU when running Starnet which significantly reduces the time it takes! Running Starnet 2 via pixinsight on my gaming laptop with an RTX 2060 Max-Q takes less than 1 minute on images taken with my QHY268M 😀 I've shared a link below which explains the process. 

https://www.williamliphotos.com/starnet-cuda

Thanks for all the info Richard! 
Greatly appreciated! 
 

I am using starxterminator I do have an Nvidia graphics card and have tried the “tenserflow, cuda hack” with starnet/starnet2++ but I failed, I think it’s down to the age of the card and or not having an original driver, I downloaded the last driver released for the card installed and upgraded the laptops firmware to the last known release, I wasn’t using Pixinsight just stand alone, so I even bought Pixinsight to try the hack but I just cannot seem to get it to work! 
I would love it if Pixinsight would run through Nvidia but it seems only a few installed Programs have the ability to activate it! Maybe it’s something I’m missing? I’m not a tech wizard but normally manage to fettle these things in the end, just not with this one 🙄

Thanks again, I’ll take another look 

Bryan 😊

[assouptro]

31 minutes ago, AMcD said:

For narrowband I have been using High Gain Mode, with gain at 0 and offset at 40, but am going to experiment with the High Gain Mode and settings of gain 56 and offset 40 as I think this may be better with narrowband filters given the reduced number of photons.  For LRGB I use High Gain Mode with gain 0 and offset 30.  I cannot claim there is any 'science' behind these numbers.  They are what I arrived at having trawled through SGL, CN and Astrobin 😂

With the focusing, I have found auto-focusing the Ha filter a bit of a challenge.  The solution appears to be to bin the camera 2x2 for focusing runs with an exposure time of 60-75s.  It results in a longer auto-focusing routine but captures enough stars to be accurate.

I have ordered a 32GB laptop as there was a (seemingly) good deal on a Lenovo 32GB with RX Vega Graphics on Amazon.

Apologies for misspelling your name!

Great info! 
I’m looking forward to finally getting plate solving and the focuser set up, it will be great for my health as I’ll be able to sleep instead of staying up to re-align and re-focus! 

I have noticed that the ha filter is the same focus as the red filter with my astrodons, I haven’t tried comparing with my Antlia filters yet but if this was the case I was considering seeing if I could set up the ha focus with the red filter? and see if any of the other broadband filters align with sii and oiii 

You have given me some really useful info there, though and don’t worry about my name, it happens all the time, I hardly notice these days!! 😊

Cheers! 

[ONIKKINEN]

Looks good for a very short integration with narrowband.

Re: processing desktop. Yep, definitely worth it to build one. My aging i7 6700K equipped pc runs through an M31 mosaic with 20k+ stars in a couple of minutes at worst with StarXterminator. Not sure if StarX uses GPU acceleration but for that the also aging GTX1080 works well, most modern kit will make processing a breeze.

[Richard_]

 

1 hour ago, assouptro said:

Thanks for all the info Richard! 
Greatly appreciated! 
 

I am using starxterminator I do have an Nvidia graphics card and have tried the “tenserflow, cuda hack” with starnet/starnet2++ but I failed, I think it’s down to the age of the card and or not having an original driver, I downloaded the last driver released for the card installed and upgraded the laptops firmware to the last known release, I wasn’t using Pixinsight just stand alone, so I even bought Pixinsight to try the hack but I just cannot seem to get it to work! 
I would love it if Pixinsight would run through Nvidia but it seems only a few installed Programs have the ability to activate it! Maybe it’s something I’m missing? I’m not a tech wizard but normally manage to fettle these things in the end, just not with this one 🙄

Thanks again, I’ll take another look 

Bryan 😊

I recently updated Pixinsight and had to reinstall Starnet 2 and redo the CUDA process. The first time failed as I forgot to unzip the directories I downloaded (tensor flow and CUDA) when I moved the files into each location. I repeated the excerise but unzipped the folders first and everything worked properly. Odd, because I didn't have an issue with this before. Maybe you did the same as me? 

1 hour ago, assouptro said:

Great info! 
I’m looking forward to finally getting plate solving and the focuser set up, it will be great for my health as I’ll be able to sleep instead of staying up to re-align and re-focus! 

I have noticed that the ha filter is the same focus as the red filter with my astrodons, I haven’t tried comparing with my Antlia filters yet but if this was the case I was considering seeing if I could set up the ha focus with the red filter? and see if any of the other broadband filters align with sii and oiii 

You have given me some really useful info there, though and don’t worry about my name, it happens all the time, I hardly notice these days!! 😊

Cheers! 

Oh you'll definitely love having auto focus set up and being able to sleep. 

What image acquisition software are you using? In NINA, you can record the focuser step position for each filter and save them as a "filter offset". When you move to each filter, NINA moves the focuser to the correct position. This is useful in two ways:

1. If you're imaging in narrowband, instead of running long auto focus routines and plate solving with longer exposure times, you can do all of this with the luminance filter to save time

2. When shooting SHO image sequences, you can change filters without the need to repeat an auto focus routine

Before you go down this route, it's worth checking that your focuser has no slipping at all. For a long time, I was fighting with my FLT120 and Sesto Senso 2 auto focuser because the auto focuser was rotating but the telescope focuser wasn't. The rack and pinion focuser was slipping between the coarse and fine focus gears (the Sesto Senso attaches to fine focus shaft). After much fettling and eventually swapping to a Pegasus Focus Cube v2 (which connects to coarse focus shaft) my focuser is rock solid with no slipping even when pointing to zenith! Using filter offsets now work really well and images are sharp. 

With regards to the Antlia filters, my offsets were as follows:

• S = 5,757
• H = 5,760
• O = 5,740

Your values will be different to mine, but as you can see they are pretty close to being parfocal, especially S and H! NINA has an addon which calculates filter offsets and records them (essentially, it runs multiple auto focus routines for each filter and records the optimum focus point for each cycle. It then calculates the average across the runs for each filter and saves them). 

[tomato]

Great first light, your perseverance has paid off.

A mono version of the IMX571 camera plus all of the associated filter hardware will be my next major purchase, if we ever get out of the COL crisis.

[Stuart1971]

17 hours ago, assouptro said:

Hi guys!

I’ve had a real rollercoaster of a ride since buying a second hand QHY268m

Firstly, I had to repair the QHY 2” filter wheel that I got with the camera, then find some 2” filters! Up till now I am used to using astrodon 1.25” filters that are faultless, they never let me down, but to buy the 2” equivalent I would be looking at taking out a mortgage!! So I took a punt on Antlia 4.5nm edge filters that cost more than the camera! Then I had to set up a new laptop that had the required usb3 inputs! Then I had to learn how to use new software, namely Nina so I could make it all work.

I finally got all the software working and had a short spell of clear sky

I took some 2 min subs of the Cygnus region in narrow band with my Samyang 135mm lens, some of which were spoiled by high thin cloud that affected the Oiii filter more than the other 2, I’ve yet to confirm halo free oiii with the Antlia filters but the ones taken without cloud look promising 

I have had a nightmare trying to achieve flat frames, so the results are calibrated without them 

Anyway….rant over…

Here is my poorly processed first light 

47 min in total SHO jpg 

Thanks for looking and sorry for the long description, it can be a lonely hobby, family and friends don’t always get the frustrations and obsession over capturing photons of ancient light from objects that are not only so far away but no longer probably look anything like the image captured 

I hope to learn how to tame the new equipment 

Bryan 😊

A beautiful image, and a great first light… but some real odd star shapes in all 4 corners, especially top 2 corners, looks like COMA on the left top, not sure about the others….🤔🤔

[assouptro]

19 minutes ago, Stuart1971 said:

A beautiful image, and a great first light… but some real odd star shapes in all 4 corners, especially top 2 corners, looks like COMA on the left top, not sure about the others….🤔🤔

Thanks Stuart 

I agree! I think it’s a combination of tilt (due to the qhy canon lens adapter which is quite “springy”) and possibly distance to sensor 

I have an adapter for the back of the lens to change it to an m48 thread now and will double check sensor distance when I get out next 

I hope it isn’t just a bad lens as the quality can be hit and miss with Samyang (apparently) 

I do pixel peep, and can be over critical with my images but I was just eager to try the camera that I didn’t spend a great deal of time getting everything exactly as it should be 😊

[assouptro]

48 minutes ago, Richard_ said:

 

I recently updated Pixinsight and had to reinstall Starnet 2 and redo the CUDA process. The first time failed as I forgot to unzip the directories I downloaded (tensor flow and CUDA) when I moved the files into each location. I repeated the excerise but unzipped the folders first and everything worked properly. Odd, because I didn't have an issue with this before. Maybe you did the same as me? 

Oh you'll definitely love having auto focus set up and being able to sleep. 

What image acquisition software are you using? In NINA, you can record the focuser step position for each filter and save them as a "filter offset". When you move to each filter, NINA moves the focuser to the correct position. This is useful in two ways:

1. If you're imaging in narrowband, instead of running long auto focus routines and plate solving with longer exposure times, you can do all of this with the luminance filter to save time

2. When shooting SHO image sequences, you can change filters without the need to repeat an auto focus routine

Before you go down this route, it's worth checking that your focuser has no slipping at all. For a long time, I was fighting with my FLT120 and Sesto Senso 2 auto focuser because the auto focuser was rotating but the telescope focuser wasn't. The rack and pinion focuser was slipping between the coarse and fine focus gears (the Sesto Senso attaches to fine focus shaft). After much fettling and eventually swapping to a Pegasus Focus Cube v2 (which connects to coarse focus shaft) my focuser is rock solid with no slipping even when pointing to zenith! Using filter offsets now work really well and images are sharp. 

With regards to the Antlia filters, my offsets were as follows:

• S = 5,757
• H = 5,760
• O = 5,740

Your values will be different to mine, but as you can see they are pretty close to being parfocal, especially S and H! NINA has an addon which calculates filter offsets and records them (essentially, it runs multiple auto focus routines for each filter and records the optimum focus point for each cycle. It then calculates the average across the runs for each filter and saves them). 

Thank you very much! 
This is great info! 
I am using Nina (or trying to!) 

I too have sesto senso and Pegasus, and know what you mean by slipping! I’ve created more problems with my focuser by fiddling with the fine focus “pin” when it came out with the knob!! 
I like the idea of sesto senso but in practice the Pegasus is my favourite! 
I will save your info for when I get around the focuser!

Thanks again 

Bryan