SBIG ST 2000 XM with Banding

Discussion in 'Legacy Models - Community Support' started by astroedo, Nov 4, 2018.

  1. astroedo

    astroedo Standard User

    Joined:
    Nov 4, 2018
    Messages:
    2
    Hi all,
    I have a SBIG 2000 XM Camera and I use normally on my old PC (with 32 bit windows XP).

    Now I've Bought a used mini PC for astronomy.
    It is a Intel Atom 1,46 GHz with Windows 10 64 bit with 8 GB Ram controlled via WiFi witth a remote desktop session.

    The PC is quite "slow", but it works fairly so, I expected no problems with the camera.
    With great surprise I've found that all my frames suffers from strong, variable, horizontal banding.

    In the attached image you can see 3 examples of BIAS frames, got in the same session, stretched with the same Histogram transformation.

    I'v tried different settings for the SBIG Driver (the latest version) without success.
    The only thing that I've noted is that, during the download of the frame, the processor usage goes up to 100% and the download time is variable: when the time is longer, the banding seems stronger.

    Any suggestions?
    Thank you

    If you want to see the original raw BIAS you can download them here
    https://www.dropbox.com/s/4hgbw010ar8h9eo/Target 1_0sec_1x1__bias_frame2.fit?dl=0

    https://www.dropbox.com/s/j97hcwae19am03v/Target 1_0sec_1x1__bias_frame1-1.fit?dl=0

    https://www.dropbox.com/s/ldhf4m6tzeqs24f/Target 1_0sec_1x1__bias_frame1.fit?dl=0
     

    Attached Files:

  2. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    7,961
    Location:
    Earth
    I'm just another customer.
    There are a few possibilities. Could be a number of things - like a hardware/firmware issue, bus timing issue, or driver issue.

    What kind of mini PC is it?
    Things to check:
    - Do you have the latest BIOS installed from the manufacturer web site?
    - Do you have the latest drivers for all peripherals and chipset? These may be on the manufacturer web site.
    - Go get the Intel Driver Support Assistant and have it check that you have all the latest Intel drivers.
    https://www.intel.com/content/www/us/en/support/intel-driver-support-assistant.html
    - Make sure you have all the Windows 10 updates.

    Are you using a USB hub? If so, try connecting the camera directly to the PC, on the workbench (kitchen table).
    Try a powered USB hub (one that has it's own power supply eg that plugs into the wall).
    Try replacing the USB cable with a shorter one.

    What is running on the mini PC?
    Shut down all unnecassary applications. Run task manager to see what is eating resources.
    If you are running planetarium software like TheSkyX, it can be resource hungry, and so you need to turn off Video Hardware Acceleration and slow down the frame rate to 1 in the Advanced Prefs.
     
  3. Colin Haig

    Colin Haig Staff Member

    Joined:
    Oct 27, 2014
    Messages:
    7,961
    Location:
    Earth
    I forgot to mention that since the camera doesn't have a frame buffer, if the PC is too busy/slow to get the data out fast enough, later rows will sit longer before getting clocked out.
     
    astroedo likes this.
  4. Doug

    Doug Staff Member

    Joined:
    Sep 25, 2014
    Messages:
    10,313
    If the computer cannot keep up with the camera's download, then you will get artifacts. The dark current accumulates more as the pixels are sitting in the sensor - and especially those sitting in the horizontal readout register.

    It's weird to think that a modern computer, even a fairly low-powered one like that, would not be able to keep up with a 486-era camera... but it looks like that is what is happening. If you are running via WiFi it's possible that the TCP-IP stack is consuming a lot of computer resources, and that's slowing down the readout. Really this vintage of camera needs a thread to itself, because it lacks a frame store. If you're running via Remote Desktop or similar that might be gobbling up so much CPU that it's causing trouble.

    The old Windows 98 style workaround was to increase the thread priority of the application, but that can't be done on modern computers. Multiprocessor / multithreaded CPUs may well do better.
     
    astroedo and Bill like this.
  5. astroedo

    astroedo Standard User

    Joined:
    Nov 4, 2018
    Messages:
    2
    Thank for the replies: this confirm my suspect an explains the observation that the banding is much more visible on BIAS frame than in darks and much more visible in narrow band imaging than in broad band ones.

    Clear skies

    Edoardo
     

Share This Page