Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

QHY268M Characterization - Deviation of Specifications

Astrophotography CMOS Imaging Equipment
  • Please log in to reply
12 replies to this topic

#1 BenKolt

BenKolt

    Vanguard

  • *****
  • topic starter
  • Posts: 2,229
  • Joined: 13 Mar 2013

Posted 10 April 2021 - 10:42 PM

Greetings!

 

I received my QHY268M a couple of weeks ago but have only just now received all the other new equipment that I needed in order to start imaging with it.  Hopefully, that can start tomorrow night.

 

In the meantime, I wished to characterize my camera by measuring the gain, read noise, full well capacity, dynamic range and dark current.  The QHY website shows these specifications in plots for the four readout modes of the camera vs. the camera's gain parameter setting.  I wanted to confirm whether or not my camera meets these specs.  It turns out that my measurements show a deviation, and I thought this might be of interest to other owners of this camera.  I invite others to test their cameras to see if there may be some variability from camera to camera in these specifications or if there may be a discrepancy on the QHY website.

 

I will be imaging in Mode 1 (High Gain Mode), at least at first, and so I restricted my testing to this mode.  I varied the gain parameter from 0 to 100 in steps of 5 but used steps of 1 between 50 and 60 in anticipation of the camera's transition through these values.  I cooled the camera to -10C, and I adjusted the offset as I changed the gain parameter in order not to cut off pixels.  For the following results I used (mostly) the equations for basic camera characterization from Berry and Burnell, The Handbook of Astronomical Image Processing, Section 8.2.  The results compared well with PixInsight's BasciCCDParameter script with the exception of the dark current, where the script version I used to check my results is known to be in error.

 

For each gain parameter value I took two bias frames, a dark frame with 60s exposure, another dark frame with 600s exposure, and two flat frames with peak illumination set at about 30,000 [ADU].

 

The first is a plot of measured gain in units of [e-/ADU] vs. the camera's gain setting.

 

GAIN.png

 

The first thing I noticed is that in this mode I am not getting a gain of about 1 [e-/ADU] at gain parameter 0 as QHY's plot suggests.  Instead, I am getting about 0.76 [e-/ADU] or so.  Next is my measured full well capacity:

 

FULL WELL.png

 

Due to my smaller gain, my peak full well capacity at gain parameter 0 is about 50 ke-, not the 62 ke- or so from the QHY plot.  I note, however, that in the specification table, QHY lists two full well values, one at 51 ke-, the other at 75 ke- for extended full well mode.  My value agrees quite well with QHY's lower full well from the table, which does not correspond with any of the peak values from the plots for any of the modes.  I am unable to reconcile this difference on the QHY website.

 

Next is my read noise measure, which agrees quite well with the QHY plots:

 

READ NOISE.png

 

I only measured dark current at the -10C cooling temperature, although I did measure it at difference gain parameter settings.  For the most part over my range of gain parameter values, the dark current was mostly constant, although the value dropped for large gain parameter values.  I don't know what that means, but the more constant answer is about 0.0016 e-/px/s, which is just a little higher than QHY's published value of 0.0013 e-/px/s at this temperature.

 

Next is the dynamic range, which I plot three different ways: in steps, dB and stops.  The last plot in stops agrees well with QHY's values.

 

DYNAMIC RANGE STEPS.png

 

DYNAMIC RANGE dB.png

 

DYNAMIC RANGE STOPS.png

 

It is this last plot in particular that shows the interest many have expressed in operating the camera in Mode 1 with gain parameter setting of 0 or 56 in order to maximize the dynamic range.  I look forward to trying this out starting tomorrow night if all goes well.

 

Again, I invite you to make some measurements of your QHY268M cameras to see if your specifications agree with mine or perhaps establish how much variability there is in the specs.  Thank you!

 

Best Regards,

Ben


Edited by BenKolt, 10 April 2021 - 10:45 PM.

  • BobT, SeymoreStars, munkacsymj and 2 others like this

#2 munkacsymj

munkacsymj

    Lift Off

  • -----
  • Posts: 11
  • Joined: 04 Jan 2016
  • Loc: Rhode Island, USA

Posted 12 April 2021 - 08:53 PM

Ben:
Thanks for a beautiful job characterizing this camera. Your numbers are almost exactly the same as mine (although I haven't covered the entire range as you have). I tested mode 1 gain 0 and mode 1 gain 56:

 

Mode 1 gain 0:

read noise = 3.6 e-/pixel

system gain = 0.78 e-/ADU

 

Mode 1 gain 56:

read noise = 1.69 e-/pixel

system gain = 0.35 e-/ADU

 

- Mark M



#3 BenKolt

BenKolt

    Vanguard

  • *****
  • topic starter
  • Posts: 2,229
  • Joined: 13 Mar 2013

Posted 12 April 2021 - 09:50 PM

Ben:
Thanks for a beautiful job characterizing this camera. Your numbers are almost exactly the same as mine (although I haven't covered the entire range as you have). I tested mode 1 gain 0 and mode 1 gain 56:

 

Mode 1 gain 0:

read noise = 3.6 e-/pixel

system gain = 0.78 e-/ADU

 

Mode 1 gain 56:

read noise = 1.69 e-/pixel

system gain = 0.35 e-/ADU

 

- Mark M

Thanks for the remarks and for adding your results, Mark.  Very interesting that you get the same system gains as I.  I wonder if the QHY plots may be of an earlier test model and that the product set point is now for a smaller gain as we are measuring.  I'm certainly using my numbers for planning and operation.

 

Ben



#4 2ghouls

2ghouls

    Apollo

  • *****
  • Posts: 1,227
  • Joined: 19 Sep 2016

Posted 13 April 2021 - 06:52 AM

Here's mine:

Mode 1 gain 0:
read noise = 3.465 e-/pixel
system gain = 0.780 e-/ADU

Mode 1 gain 56:
read noise = 1.562 e-/pixel
system gain = 0.334 e-/ADU

-Nico

Sent from my VOG-L29 using Tapatalk

#5 BenKolt

BenKolt

    Vanguard

  • *****
  • topic starter
  • Posts: 2,229
  • Joined: 13 Mar 2013

Posted 13 April 2021 - 09:52 AM

Here's mine:

Mode 1 gain 0:
read noise = 3.465 e-/pixel
system gain = 0.780 e-/ADU

Mode 1 gain 56:
read noise = 1.562 e-/pixel
system gain = 0.334 e-/ADU

-Nico

Sent from my VOG-L29 using Tapatalk

Thanks for posting your numbers, Nico!  That's three of us now that are getting very similar results that deviate from QHY's posted plots on their website.  I'm wondering if they performed testing on an earlier model and that there were some adjustments made to the final released model that we all have.  These differences are in no way an issue because the dynamic range I measure is very similar to what they advertise.  What differs may be choices and expectations on camera performance and operation for a given set of parameters.  Nothing beats characterizing your own camera to finalize your understanding!

 

I reached out to QHY last week to let them see my results and to ask if they can reconcile the differences.  I'll post something if they respond.  I'm sure they're quite busy and may not be able to get back to me about such detailed testing.

 

Best Regards,

Ben


  • rockstarbill likes this

#6 pcalexander

pcalexander

    Lift Off

  • -----
  • Posts: 16
  • Joined: 08 Dec 2019
  • Loc: South Carolina

Posted 18 April 2021 - 11:02 AM

I ran the SharpCap analysis two days ago and then once again last night.

 

First run - Mode 1, Gain 0 at -10C:

Read noise (e): 4.03

e/ADU: 0.79

Full Well (e): 52035.91

Dynamic Stops: 13.66

 

Second run - Mode 1, Gain 0 at -10C:

Read noise (e): 3.49

e/ADU: 0.88

Full Well (e): 57939.56

Dynamic Stops: 14.02

 

 

These runs weren't done at exactly the same flat panel brightness, but it was following the SharpCap prompts. I haven't run this test on other cameras that I've owned, so I am not sure what variances I am introducing myself, and I'm not sure if the variant results are typical when running the sensor analysis tool.

 

Phillip

 



#7 pcalexander

pcalexander

    Lift Off

  • -----
  • Posts: 16
  • Joined: 08 Dec 2019
  • Loc: South Carolina

Posted 18 April 2021 - 11:11 AM

I ran the SharpCap analysis two days ago and then once again last night.

 

First run - Mode 1, Gain 0 at -10C:

Read noise (e): 4.03

e/ADU: 0.79

Full Well (e): 52035.91

Dynamic Stops: 13.66

 

Second run - Mode 1, Gain 0 at -10C:

Read noise (e): 3.49

e/ADU: 0.88

Full Well (e): 57939.56

Dynamic Stops: 14.02

 

 

These runs weren't done at exactly the same flat panel brightness, but it was following the SharpCap prompts. I haven't run this test on other cameras that I've owned, so I am not sure what variances I am introducing myself, and I'm not sure if the variant results are typical when running the sensor analysis tool.

 

Phillip

Full sensor analysis results attached.

Attached Thumbnails

  • run1.PNG
  • run2.PNG

  • Miguelo and SilverLitz like this

#8 BenKolt

BenKolt

    Vanguard

  • *****
  • topic starter
  • Posts: 2,229
  • Joined: 13 Mar 2013

Posted 25 April 2021 - 04:16 PM

pcalexander:

 

Thank you for posting these results.  It's interesting to me how these two runs with SharpCap differ by so much.  Do you know if SharpCap performs an analysis over the whole sensor or just over a region of it?  For my calculations I picked one region of the sensor's field and stuck with it.  I had intended on doing at least one more pass through all the calculations on a different region in order to evaluate the variability.  I imagine there is wisdom in keeping away from the edges.

 

Ben



#9 pcalexander

pcalexander

    Lift Off

  • -----
  • Posts: 16
  • Joined: 08 Dec 2019
  • Loc: South Carolina

Posted 25 April 2021 - 05:06 PM

That is a good point. I wasn’t even thinking about that. The sharpcap tool uses a region that you select yourself. You can change the size and a region of the test. I know that I did not have the same region selected across the two tests, but both were towards the center of the sensor.

#10 carballada

carballada

    Mariner 2

  • -----
  • Posts: 232
  • Joined: 13 Apr 2016

Posted 30 April 2021 - 05:02 AM

My results

 

mode1 gain 56 offset 25 at -5º
gain 0.331 e/ADU
readout noise 1.725 e / 5.213 ADU
fullwell 21688 e
dynamic range 12571 steps

 

 

mode1 gain 0 offset 25 at -5º
gain 0.764 e/ADU
readout noise 3.728 e / 4.877 ADU
fullwell 50097 e
dynamic range 13439 steps

 

Updated with more gain values close to 56

 

qhd268m-gain-values.jpg


Edited by carballada, 30 April 2021 - 07:52 AM.


#11 BenKolt

BenKolt

    Vanguard

  • *****
  • topic starter
  • Posts: 2,229
  • Joined: 13 Mar 2013

Posted 30 April 2021 - 11:23 AM

My results

 

mode1 gain 56 offset 25 at -5º
gain 0.331 e/ADU
readout noise 1.725 e / 5.213 ADU
fullwell 21688 e
dynamic range 12571 steps

 

 

mode1 gain 0 offset 25 at -5º
gain 0.764 e/ADU
readout noise 3.728 e / 4.877 ADU
fullwell 50097 e
dynamic range 13439 steps

 

Updated with more gain values close to 56

 

qhd268m-gain-values.jpg

Thanks for posting your results, carballada!  Your values are close to the other sets listed in this thread.

 

Ben


  • carballada likes this

#12 carballada

carballada

    Mariner 2

  • -----
  • Posts: 232
  • Joined: 13 Apr 2016

Posted 01 May 2021 - 09:26 AM

Thanks for posting your results, carballada!  Your values are close to the other sets listed in this thread.

 

Ben

yes, and confirms that I will use 56-25 for narrow band and 0-25 to rgb in mode 1 :D



#13 jeffcrilly2

jeffcrilly2

    Explorer 1

  • -----
  • Posts: 60
  • Joined: 25 Jun 2011

Posted 06 May 2021 - 03:13 AM

Hi.. I just came across this thread last night and it looks like there's lots of good info here.

 

I got a QHY268M about a month ago and have been imaging (LRGB and some Ha) with :

 

Gain = 0

Offset = 30

ReadMode = PhotoGraphic DSO

 

After reading this thread, I ran SharpCap -> Sensor Analysis , and attached the results to this post.

 

I'm still trying to wrap my brain around how to interpret these results, which gain/offset values to use, and exposure times (tho my understand is exposure depends on object brightness).

 

If anyone has some pointers given these results, and values to use for DSO vs , such info would be appreciated.

I sorta need a "CMOS settings for the impatient imager" cheatsheet.

 

thx

 

-Jeff

 

QHY268M-minus-10-16-bit.jpg




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics





Also tagged with one or more of these keywords: Astrophotography, CMOS, Imaging, Equipment



Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics