QHYCCD

QHY10 strange behavior

QHY10 strange behavior
« on: October 11, 2019, 06:07:49 pm »
Hello,

I have a QHY10 that has not seen much use since I acquired it a few months ago. The times I've used it, I've not obtained reasonable results. I typically use gain=0, offset=123. These were set early on by following QHY's recommended method of setting these two parameters. What I've observed are two problems:

1. Color gradient visible across images. This gradient is not removed by calibration frames.
2. Exposure time setting in EZCAP and through ASCOM (APT) does not seem to change the captured image much - certainly not as much as one would expect.

To root-out the problem, I put the 10 on the bench and attached a Canon 28mm lens. I ran a series of light and dark frames over several exposures (1us, 1ms, 100ms, 1s, 10s), taking three frames at each exposure time. The camera was cooled to -15C.

In order to verify that the problem is not USB port/cable, computer, or DC-201 power supply, I ran the exact same sequence of exposures on an 8L camera. The only difference in the configuration was the offset, which was set to 115 instead of the 123 that was used on the 10 camera.

Attached are two frames, one each from the two cameras. I chose to share only the 1 millisecond exposure images. These were debayered (RGGB for the 10, and GBRG for the 8L), then auto-stretched using STF in PixInsight, then saved as highly compressed jpegs. Ignore the focus, as the Canon lens has an electronic focus that is not functional when I connect to the QHY cameras. The important thing I was looking at was increase in brightness as exposure increased, as well as decent color balance, and finally proper progression of under to over exposure. With the 8L this happens as expected, but with the 10 it almost appears to not respond to the configuration settings.

Any idea what might be the problem? In this controlled experiment, the only difference was the cameras (and only a small change in offset setting between the two).

Al

Re: QHY10 strange behavior
« Reply #1 on: October 11, 2019, 08:33:28 pm »
Can you send me the FIT format of this QHY10-C picture? This helps me make a judgment.

Re: QHY10 strange behavior
« Reply #2 on: October 12, 2019, 11:50:25 am »
Can you send me the FIT format of this QHY10-C picture? This helps me make a judgment.

Hi llsong,

I've sent you a PM with a link to access the images from my testing.

Al

QiuHY

  • *****
  • 4957
    • View Profile
    • Email
Re: QHY10 strange behavior
« Reply #3 on: October 14, 2019, 10:14:25 am »
Hello,
        Is the camera moved during the exposure or not?  It looks like the camera moved during the the even/odd field exposure .

Best regards,
Qiu Hongyun
Qiu Hongyun

Re: QHY10 strange behavior
« Reply #4 on: October 14, 2019, 10:17:55 am »
Hello,
        Is the camera moved during the exposure or not?  It looks like the camera moved during the the even/odd field exposure .

Best regards,
Qiu Hongyun

The camera was stationary. I mentioned that focus was not spot on because I used an electronic focus Canon lens, so I used the closest spacers to get enough focus to do the comparison. Focus is not the issue I am concerned about. It is the lack of substantial change in the image as exposure is changed, and also the strange color shift that you see. The 8L images behave as expected.

Al

Re: QHY10 strange behavior
« Reply #5 on: October 17, 2019, 12:02:14 am »
Hello I did not receive the picture, can you send it to my mailbox? The email address is cha@qhyccd.com. Is this the first one, is the picture behind it all?

Re: QHY10 strange behavior
« Reply #6 on: October 18, 2019, 12:16:52 pm »
Hello I did not receive the picture, can you send it to my mailbox? The email address is cha@qhyccd.com. Is this the first one, is the picture behind it all?

Hello llsong,

I have sent you a direct email. Please go to the URL I provided and download whichever pictures you need to review.

Al

Re: QHY10 strange behavior
« Reply #7 on: October 19, 2019, 04:14:52 am »
I have received the picture you sent, the picture is almost saturated, can you use QHY10-C to take some backlit pictures? Or turn OFFSET down. Try to make the average value in the image less than 10000. It is difficult to display more information in the image when the average value in the image reaches 50,000 or more.

Re: QHY10 strange behavior
« Reply #8 on: January 16, 2020, 11:03:02 am »
Hello,

I am following up on this topic. In response to your question, I took a new series of images using both an 8L and 10 once again. This time, I used my WO Redcat 51 lens and simply swapped out the cameras.

For the 8L, I set gain=0 and offset=115. For the 10, I set gain=0 and offset=127.

I captured several images, ranging from 10 microsecond to 1 second exposures, increasing the exposure by a factor of 10 each time. In other words, the set was: 10us, 100us, 1ms, 10ms, 100ms, and 1s.

I captured both light and dark exposures.

I will not comment at this time, except to say that the 8L behaves as one would expect over such range of exposures. The 10 behaves strangely.

Please review the complete set of images and provide feedback.

You will find the images in the same shared folder location as before. You confirmed that you were able to find the previous ones. I placed the new data in a sub-folder named: New_Data

Thank you,

Al

I have received the picture you sent, the picture is almost saturated, can you use QHY10-C to take some backlit pictures? Or turn OFFSET down. Try to make the average value in the image less than 10000. It is difficult to display more information in the image when the average value in the image reaches 50,000 or more.

Re: QHY10 strange behavior
« Reply #9 on: January 16, 2020, 11:11:05 pm »
For the benefit of other QHY 10 users, and to clearly illustrate the issue I am trying to document, I am attaching a few jpeg images. There are two pairs. One pair is taken with the QHY 8L. The other is taken with the QHY 10. The same lens (WO Redcat 51) was used, and the subject and lighting were the same. It was mid-afternoon on a sunny winter day.

The files are:

Filename                                                                       Camera       Gain      Offset        Exposure
g0o115_8L_Bin1x1_0_001s__20C_RGB_Bilinear.jpg         8L              0           115           1/1000 sec
g0o115_8L_Bin1x1_0_01s__20C_RGB_Bilinear.jpg           8L              0           115           1/100 sec
g0o127_10_Bin1x1_0_001s__19C_RGB_Bilinear.jpg         10              0           127           1/1000 sec
g0o127_10_Bin1x1_0_01s__19C_RGB_Bilinear.jpg           10              0           127           1/100 sec

Since these sensors are essentially what was available in DSLR cameras some years ago, we can assume that exposures with a factor of 10 difference would result in dramatic differences in an image -- independent of equivalent ISO or aperture.

With the 8L, the results are as expected. The fast shutter results in a more-or-less realistic depiction of the subject (perhaps a bit under-exposed). The slow shutter results in an over-exposed image. No surprise here.

With the 10, both shutter settings result in virtually the same depiction of the subject. Both are over-exposed.

So I believe there is something wrong with the 10 camera. I may be wrong and it may be the drivers, but I am using drivers as presently available on QHY web site for these cameras, which are described as latest stable versions. I tried both APT (ASCOM driver) and EZCAP (native windows), with no difference in results.

8L drivers are: QHY8LASCOM-StarSenseSci-V33, QHY8LDrv64LatestEdition
10 drivers are: QHY10ASCOM-StarSenseSci-V30, QHY10Drv64V10-11-8

I was reading a few other threads in this forum about the 10. What really made me worried is that QHY does not provide any easy way of determining the history (change log) for their drivers. In a couple of threads driver versions were directly provided to the topic originator, but those versions do not appear on the product URL page; and they are most certainly not described as "latest, stable". So as a customer who is trying to figure out how to get the best performance and stability out of a camera, it is extremely confusing.

I have two ZWO ASI cameras as well. ZWO provides one universal driver for all of their cameras. They regularly update the common driver set and it is easy to find versions and see the progression of version numbers. Why does QHY not follow a similar approach?

Al

Hello,

I am following up on this topic. In response to your question, I took a new series of images using both an 8L and 10 once again. This time, I used my WO Redcat 51 lens and simply swapped out the cameras.

For the 8L, I set gain=0 and offset=115. For the 10, I set gain=0 and offset=127.

I captured several images, ranging from 10 microsecond to 1 second exposures, increasing the exposure by a factor of 10 each time. In other words, the set was: 10us, 100us, 1ms, 10ms, 100ms, and 1s.

I captured both light and dark exposures.

I will not comment at this time, except to say that the 8L behaves as one would expect over such range of exposures. The 10 behaves strangely.

Please review the complete set of images and provide feedback.

You will find the images in the same shared folder location as before. You confirmed that you were able to find the previous ones. I placed the new data in a sub-folder named: New_Data

Thank you,

Al

I have received the picture you sent, the picture is almost saturated, can you use QHY10-C to take some backlit pictures? Or turn OFFSET down. Try to make the average value in the image less than 10000. It is difficult to display more information in the image when the average value in the image reaches 50,000 or more.

Re: QHY10 strange behavior
« Reply #10 on: January 19, 2020, 12:50:42 am »
What software do you use to shoot? Because the old CCD chip has different settings in the short-term exposure switch, it is set in the SDK. If you are using ASCOM software, please use QHY8L and QHY10 separate ASCOM plug-ins. If you are using QHYCCD software, use EZCAP instead of EZCAP-QT.
In addition, QHY10 has reached saturation in the shortest exposure time. I think you need to compare the shooting situation when you find the environment is dim. Because of how they read
The main difference between the two cameras is how the sensors read. The QHY8L uses two readings. After exposure, half of the red and green pixels are read first. Then read the other half of the blue and green pixels. On the other hand, QHY10 gradually scans the entire CCD at the end of the exposure. For long exposures, there is no difference between the two. However, to capture fast-moving objects, it is more advantageous to use the QHY10 progressive scan sensor.

Re: QHY10 strange behavior
« Reply #11 on: January 20, 2020, 03:23:15 pm »
I used APT and EZCAP and obtained nearly identical results with each. APT uses ACOM driver, while EZCAP uses native Windows driver.

For each camera, I used its specific ASCOM driver (see previous post where I describe the exact FW). I do not use EZCAP-QT.

I understand exactly how each camera works, as I had read the QHY information about each camera. I am not trying to capture fast moving objects.

What I am trying to do is understand if there is something wrong with the QHY 10. No other camera I have used, CCD or CMOS, has ever shown virtually no change in the image as the exposure time is changed so dramatically.

In particular, I am now wondering how this affects calibration frame capture. For example, a bias frame would need to have IR blocked (using metal cap) in order to guarantee that no IR is on sensor since even at 1ms exposure there would be a lot of possible ADU due to stray IR.

I am also wondering the same about darks and flats. Darks will need insure absolute darkness over all frequencies that the sensor has sensitivity to. For flats, which are typically done with relatively low exposure time using uniform light source, how then can one achieve average ADU in the 20,000-30,000 range reliably if the exposure time does not truly change the behavior of the sensor?

Al


What software do you use to shoot? Because the old CCD chip has different settings in the short-term exposure switch, it is set in the SDK. If you are using ASCOM software, please use QHY8L and QHY10 separate ASCOM plug-ins. If you are using QHYCCD software, use EZCAP instead of EZCAP-QT.
In addition, QHY10 has reached saturation in the shortest exposure time. I think you need to compare the shooting situation when you find the environment is dim. Because of how they read
The main difference between the two cameras is how the sensors read. The QHY8L uses two readings. After exposure, half of the red and green pixels are read first. Then read the other half of the blue and green pixels. On the other hand, QHY10 gradually scans the entire CCD at the end of the exposure. For long exposures, there is no difference between the two. However, to capture fast-moving objects, it is more advantageous to use the QHY10 progressive scan sensor.

Re: QHY10 strange behavior
« Reply #12 on: January 20, 2020, 09:19:46 pm »
Hello
I've made a table based on the image you sent, which can more intuitively see the changes caused by the increase in exposure time. Here I use the average ADU of the image. Increased, ADU has improved significantly. When you shoot Light, because the shortest exposure time of QHY10 is close to saturation, it will be difficult to judge the change at this time. If you can shoot in the evening, I think it can achieve the effect of QHY8L in Light.
As for shooting a flat field of 20,000-30000, I think it is a good way to reduce the ambient light.

Re: QHY10 strange behavior
« Reply #13 on: January 21, 2020, 12:11:23 pm »
Hi,

I did look at the ADU characteristics in ImageJ prior to sending you over all the images. I do agree that changes are visible. I'm not understanding why exactly the 10 sensor is so different than the 8L sensor, in terms of how it captures light. They are both from the same Sony family, so I would expect them to behave similarly. I wonder how a 12 would compare?

In order to confirm what you state below, I will try the same experiment in the next day or two. I'll either use dawn or dusk twilight. Once the sky in my area clears up, I will also try to capture a star field (or bright DSO) with both cameras using the same imaging plan in order to compare subs.

Al


Hello
I've made a table based on the image you sent, which can more intuitively see the changes caused by the increase in exposure time. Here I use the average ADU of the image. Increased, ADU has improved significantly. When you shoot Light, because the shortest exposure time of QHY10 is close to saturation, it will be difficult to judge the change at this time. If you can shoot in the evening, I think it can achieve the effect of QHY8L in Light.
As for shooting a flat field of 20,000-30000, I think it is a good way to reduce the ambient light.

Re: QHY10 strange behavior
« Reply #14 on: February 02, 2020, 08:59:33 pm »
Have you tested it? How's the effect。
Although QHY8L and QHY10 use the same Sony chip, they will have different sensitivities due to different readout methods. QHY12 and QHY10 readout methods perform similarly.