Starlight Xpress SX-694

Peter, I’ll check my darks tonight. I have 20 minutes at -20C, but not 30 minutes. (Though I really need to shoot new ones. These are from last winter.) I thought this was just a QSI issue, so it’s interesting to see it in Joel’s camera as well. I’ll post my dark master. Should it be bias calibrated or not?

Kevin

Hi Kevin,

Stack of 20 minutes darks is fine. I would also appreciate another stack of 20 minutes dark that’s calibrated with bias. In other words, I would like to see stack of uncalibrated and calibrated with bias darks.

Thank you very much.

Peter

Peter,

I have an Atik 460EX and I don’t see any glow in the 30m master dark. If you are interested please contact me offline and I will share it with you so you can inspect it and draw your own conclusions.

I ordered a QSI 6120 but I canceled it after reading the CN post. I tried to contact QSI support but they didn’t reply at all.

CS,

Jose

A year ago I had an SX H18 and the ASCOM driver had issues, I contacted Bret but he didn’t have an H18 to reproduce the issue so basically he was in the dark on top of that he code/debug the driver as a hobby since he doesn’t work for SX and the issue never got solved. The camera worked fine with native dirvers (maximdl) tho.

I sold the camera and got an sbig STF8300, it works perfectly with SGP.

CS,

Jose

Hi Jose,

Thanks for the offer. I also used to have Atik 460EX and I didn’t see any glows either. I saved all darks from Atik 460EX camera and checked that all darks look fine. Also, several people I know of so far have not reported any glows from this camera as well.

I am not worried about it. The glows are not amp glows and have very little ADU or electrons values and just barely above the rest of the image without glows. Dark subtraction has worked very well.

Peter

Hi Peter,

Uncalibrated -20C darks (only 10 x 20 mins. It looks like I have yet to use this particular set. I didn’t have much cold weather last winter, and all my shots have been at -10C or 0C for quite a while.):
https://dl.dropboxusercontent.com/u/58468743/Firmware%20G%20-20C%20with%20delay_1200sec_1x1__frame10.zip

Master bias:
https://dl.dropboxusercontent.com/u/58468743/MasterOffset_ISO0.tif

I’d be interested to hear what you find.

Kevin

Hi Kevin,

My preliminary observation is both your bias and darks have pretty high ADUs (about 3000) and Joel’s master dark is about 800 ADU. Why the big difference from same cameras?

I am not seeing any glows from your darks but it could be high ADU values masking out the glows.

From Pavel’s blog, the average gain from your camera is about 0.3 e/ADU so that’s about 900 electrons from your dark and 240 electrons from Joel’s dark. Calibrating with your bias resulted about 25 ADU or 7.5 electrons.

My 30 minutes dark from QSI 660wsg is about 540 ADU and gain of 0.15 e/ADU which comes to 81 electrons. It shows about same amount of glows as Joel’s master dark.

My 30 minutes dark from my former Atik 460 EX is about 330 ADU, gain of about 0.25 which translates to 83 electrons. It does not show any glows.

I am not sure what all this means? I think we should start collecting darks from other people’s SX-694 and draw conclusion.

Peter

Joel and I probably have different camera firmware. You may remember Terry sent me several sets of firmware to try to solve some issues with patterning. The version I now have probably just has a higher pedestal than Joel’s version. I don’t know why, but maybe that helped solve the issue, because as long as I use darks I don’t see any problems now. The extra pedestal gets subtracted out by the bias, so it all comes out in the end.

I’m glad you didn’t see any glows. I’m surprised Joel’s Trius shows this and my H694 doesn’t. Yet another mystery of CCDs.

Yes I remember Terry’s customized firmware. Seems a lot of pedestal added. I thought pattern noise are subtracted out by bias. Did you also open the back of the camera to adjust the gain?

Peter

I have contacted Terry at SX to see if there is an updated firmware. I’m not at all worried about this because the glow gets calibrated out, but I thought it couldn’t hurt to ask. It would worry me a bit if the background ADU was 3000 vs. 800. I’m curious to hear from Kevin too about the gain setting.

Just for future reference Brett McKee (SX ascom author) told me that SX cameras have no avenue for reporting the percentage of TEC power being used, so the ascom driver cannot report it in SGP or any other control software.

I have not adjusted the gain at all.

Edit: I should note, though, that I really don’t have any deep understanding of the technical issues with CCD calibration. I would be delighted if you discovered some obvious error on my part, and then tell me the magic button to press to get much better images. :smile:

Hi Kevin,

I created a new thread at CN requesting others to upload their darks from their Sony ICX-694/814 based cameras to investigate further. I am hoping enough people will provide their darks from all camera brands.

You might want to ask Terry Platt exactly what kind of firmware customization he did.

Peter

Guys,
I’ve been in contact with Terry Plat and he has provided me with a solution that upon initial tests has completely eliminated the glows with my SX-694. I’d like to do some more tests though. I’d like to get Terry’s permission before I comment further, but the solution is not firmware related but hardware. After I take several darks I’ll post a download link so you can take a look for yourself.

Cool, thanks!!!

Peter

Here’s a single 20min dark, -25C:

And here’s a link to a stack of 3-20min darks:

As far as I can tell the glows have been completely eliminated.

HI Joel,

Your single dark sub looks great but something went wrong with your stack of three subs. The single sub showed median of slightly over 800 ADU but the stack showed over 16,000 ADU and much higher standard deviation.

I am curious about what was done to remove the glows. I hope whatever you did will work on other cameras including QSI.

Peter

Still learning PI…I think I applied a histogram transformation?
Where do you find the adu value an other info in PI?

Try this stack of 5 20min darks:

Hi Joel,

Your new Master Dark looks much better.

To look at ADU values with PI, use “Statistics” tool. For some reason it does not default to include Standard Deviation but you can add it. Click on Statistics Option button (Wrench shape) at upper right, it will display another window and you can add more tools including Standard Deviation.

Peter

I thought it was statistics but didn’t see the right values. Thanks for that.

As soon as I hear back from Terry Plat I’ll comment on the procedure. It involves opening up the camera and I don’t know if he wants that info out there.

Sounds like the fix is specific to SX cameras and many not work for QSI or other cameras.

Peter