I am trying to understand how the flats I have been using for the past months were not accepted last night in a EAA session by Sharpcap 4.09538 Pro. Message was that resolution did not match. Flats 2072 x 1410, camera resolution was 2072 x 1411 . I was last night, and have for a long time been successfully using the asi294mc at Bin 2 and these same flats. But not last night. I have not changed the flats. Can anyone provide insight into what happened?
Thanks for your help.
Randy
zwo asi294mc flats are one pixel less than lights - resolution mismatch
Re: zwo asi294mc flats are one pixel less than lights - resolution mismatch
Hi Randy,
I am finding something likely related, not EAA, standard processing. Old calibration frames are not working with new captures. File sizes are slightly different and it looks like resolution is off by one pixel. I suspect the new version, 4.0.9538 vs .9521. Did you change versions?
It is also possible that a new zwo driver is at fault. I vaguely remember loading a new driver recently. Lousy weather, haven't done much imaging.
It is snowing, but if I am not feeling too wimpy I will head out to the observatory and test this. I think I have both versions still on my laptop.
Steve
I am finding something likely related, not EAA, standard processing. Old calibration frames are not working with new captures. File sizes are slightly different and it looks like resolution is off by one pixel. I suspect the new version, 4.0.9538 vs .9521. Did you change versions?
It is also possible that a new zwo driver is at fault. I vaguely remember loading a new driver recently. Lousy weather, haven't done much imaging.
It is snowing, but if I am not feeling too wimpy I will head out to the observatory and test this. I think I have both versions still on my laptop.
Steve
Re: zwo asi294mc flats are one pixel less than lights - resolution mismatch
I didn't have the older version, 4.0.9521 but I have plenty of captures from that version.
Captures from the old version always had the "hide small pixel mode" checkbox in file->settings checked.
All the old captures have resolution listed in the FITS metadata as 2072x1410.
Using the new version,4.0.9538, whether or not the small pixel mode checkbox is checked, resolution is listed as 2072 x 1411 in the metadata. Which actually seems more correct than the old resolution, which loses a pixel row. But unfortunately, makes the processing fail as incompatible if using calibration files from the old version. Requires new calibration frames. Not a big deal.
So this could be looked at as a bug fix, I suppose.
I was more concerned with another bug in the metadata, which in the "hide small pixel mode" yields incorrect pixel size. Robin knows about that.
It is possible that a new ASI driver is at fault but I don't see the asi driver version listed anywhere. Good chance it is SharpCap. Sorry Robin. This camera must be a PIA for you!
Steve
Captures from the old version always had the "hide small pixel mode" checkbox in file->settings checked.
All the old captures have resolution listed in the FITS metadata as 2072x1410.
Using the new version,4.0.9538, whether or not the small pixel mode checkbox is checked, resolution is listed as 2072 x 1411 in the metadata. Which actually seems more correct than the old resolution, which loses a pixel row. But unfortunately, makes the processing fail as incompatible if using calibration files from the old version. Requires new calibration frames. Not a big deal.
So this could be looked at as a bug fix, I suppose.
I was more concerned with another bug in the metadata, which in the "hide small pixel mode" yields incorrect pixel size. Robin knows about that.
It is possible that a new ASI driver is at fault but I don't see the asi driver version listed anywhere. Good chance it is SharpCap. Sorry Robin. This camera must be a PIA for you!
Steve
Re: zwo asi294mc flats are one pixel less than lights - resolution mismatch
Steve - thanks for your insight. I see that I am running Zwo driver 3.18. On Zwo driver page I see new driver listed as 3.21. I guess I should try that.
I loaded in an old version 4.09511 of Sharpcap and the same issues exists. It did not help.
I can retake flats but I really like to understand things like this. It cost me a night of EAA in a dark sky site trip.
Maybe Robin has a suggestion.
Thanks
Randy
I loaded in an old version 4.09511 of Sharpcap and the same issues exists. It did not help.
I can retake flats but I really like to understand things like this. It cost me a night of EAA in a dark sky site trip.
Maybe Robin has a suggestion.
Thanks
Randy
Re: zwo asi294mc flats are one pixel less than lights - resolution mismatch
I just tried new Zwo driver 3.21. It did not help any. Same issue.
Randy
Randy
Re: zwo asi294mc flats are one pixel less than lights - resolution mismatch
Looks like a SharpCap issue then. Too bad for Robin!
BTW another buglet: When the "ignore small pixel" is not checked (so I get correct binned pixel size in the FITS metadata) the binning is not preserved when SharpCap is exited and restarted. The remainder of the format entries are preserved.
Lastly, even when "unchecked" and hence getting correct binned pixel dimensions, ASTAP never seems to be invoked. I don't see the ASTAP progress window or a green bar reporting anything about plate solving. But I have not tested that extensively yet, only used 15 sec exposures. Which did produce 20 or so stars visible stretched.
Steve
BTW another buglet: When the "ignore small pixel" is not checked (so I get correct binned pixel size in the FITS metadata) the binning is not preserved when SharpCap is exited and restarted. The remainder of the format entries are preserved.
Lastly, even when "unchecked" and hence getting correct binned pixel dimensions, ASTAP never seems to be invoked. I don't see the ASTAP progress window or a green bar reporting anything about plate solving. But I have not tested that extensively yet, only used 15 sec exposures. Which did produce 20 or so stars visible stretched.
Steve
Re: zwo asi294mc flats are one pixel less than lights - resolution mismatch
Randy,
Sorry, I misread your last post. If the issue is with the ZWO driver, you would need an old version, maybe really old. I dont know how old, not something I change frequently. Unfortunately I think I got the new one about the same time I got the new SharpCap version. So these are confounded. Don't know what ZWO version I had been using but it could have been quite old.
I don't see the ZWO driver version listed either in the FITS metadata or the SharpCap image data file. Maybe I'm just missing it. If that version could be associated with a capture file we could untangle this without loading an old driver and an old SharpCap.
Steve
Sorry, I misread your last post. If the issue is with the ZWO driver, you would need an old version, maybe really old. I dont know how old, not something I change frequently. Unfortunately I think I got the new one about the same time I got the new SharpCap version. So these are confounded. Don't know what ZWO version I had been using but it could have been quite old.
I don't see the ZWO driver version listed either in the FITS metadata or the SharpCap image data file. Maybe I'm just missing it. If that version could be associated with a capture file we could untangle this without loading an old driver and an old SharpCap.
Steve
Re: zwo asi294mc flats are one pixel less than lights - resolution mismatch
Randy,
I checked a different laptop that I used for imaging up until a few months ago. All my calibration subs used it. It has ZWO driver 3.17, and SharpCap 4.0.9478. I can check that to see what resolution it gets for the 294. Then upgrade either the driver or SC.
Steve
I checked a different laptop that I used for imaging up until a few months ago. All my calibration subs used it. It has ZWO driver 3.17, and SharpCap 4.0.9478. I can check that to see what resolution it gets for the 294. Then upgrade either the driver or SC.
Steve
-
admin
- Site Admin
- Posts: 13344
- Joined: Sat Feb 11, 2017 3:52 pm
- Location: Vale of the White Horse, UK
- Contact:
Re: zwo asi294mc flats are one pixel less than lights - resolution mismatch
Hi folks,
this was indeed a bug fix in SharpCap for cases where the rules on what resolutions are available to ZWO cameras were wrongly being applied by older versions of SharpCap (particularly in binned modes). This made SharpCap often have smaller resolutions by a few pixels than those used by other applications. As far as I am aware, SharpCap 4.0 now applies the rules correctly for normal (software) binning, but is still slightly wrong for ZWO hardware binning mode. SharpCap 4.1 fixes hardware binning, and since few people use it, I am not intending to put that change into 4.0.
Unfortunately the change has led to an *increase* in pixel size in some cases, which leads to the flat/dark mismatch. If it was a small reduction in size then it would have been seamless since SharpCap can deal with an ROI capture when using a full resolution flat/dark
I am reluctant to put things back the way they were, since that would simply perpetuate the bug of slightly wrong resolutions in SharpCap. You could go back to SharpCap 4.0.9529 as an interim measure as that version was the last before the fix for ZWO resolution calculations. I don't think you will loose anything significant for ZWO camera use, since the later versions only contain some small bug fixes and SDK updates for other brands.
cheers,
Robin
this was indeed a bug fix in SharpCap for cases where the rules on what resolutions are available to ZWO cameras were wrongly being applied by older versions of SharpCap (particularly in binned modes). This made SharpCap often have smaller resolutions by a few pixels than those used by other applications. As far as I am aware, SharpCap 4.0 now applies the rules correctly for normal (software) binning, but is still slightly wrong for ZWO hardware binning mode. SharpCap 4.1 fixes hardware binning, and since few people use it, I am not intending to put that change into 4.0.
Unfortunately the change has led to an *increase* in pixel size in some cases, which leads to the flat/dark mismatch. If it was a small reduction in size then it would have been seamless since SharpCap can deal with an ROI capture when using a full resolution flat/dark
I am reluctant to put things back the way they were, since that would simply perpetuate the bug of slightly wrong resolutions in SharpCap. You could go back to SharpCap 4.0.9529 as an interim measure as that version was the last before the fix for ZWO resolution calculations. I don't think you will loose anything significant for ZWO camera use, since the later versions only contain some small bug fixes and SDK updates for other brands.
cheers,
Robin
Re: zwo asi294mc flats are one pixel less than lights - resolution mismatch
Thanks Robin. I had just verified this...should have checked the post first!
Not a big deal, more correct is more better. Just need to update calibration data going forward.
It would be nice if the binning was preserved over shutdown in the "unlocked" mode as well as the "ignore small pixel" mode since the former has the more correct binned pixel size in the FITS header so ASTAP can work reliably.
I'm still not seeing SharpCap actually calling ASTAP and doing a resync and will look at that when there is some clear sky
Steve
Not a big deal, more correct is more better. Just need to update calibration data going forward.
It would be nice if the binning was preserved over shutdown in the "unlocked" mode as well as the "ignore small pixel" mode since the former has the more correct binned pixel size in the FITS header so ASTAP can work reliably.
I'm still not seeing SharpCap actually calling ASTAP and doing a resync and will look at that when there is some clear sky
Steve