oaCapture 0.5.0 release

[JamesF]

I have no Chameleon camera to test with I'm afraid, so I don't know how much I can help.  I'll see what I can find out about it.  It should not be necessary to install libdc1394 unless I have compiled it incorrectly.  That should be statically linked into the executable.  I'll check for the 0.6.0 release.

I was hoping to get the next release out before I went on holiday, but work took too much of my time before we left and I didn't get all the testing done that I wanted.

James

[Arroway]

Hi James,

 Spanish user here. I've been struggling with my QHY5-II camera for days with firecapture and QHY drivers. Horizontal lines everywhere when raising gain to 100. Totally useless.

 Until I met your software. I installed it on an Ubuntu 14.04 yesterday and it simply works. No noise at all and thats FANTASTIC. Here is my feedback and questions :

- Gain threshold is from 1 to 47 (or 48, can't remember now). Is 47 the equivalent to 100 at firecapture?. Why that particular value?.

- Changing the filter to, ie. green, totally changes the image. It shows almost burn-out, like if gain was set to a much much high value than the actual one. Is that the intended way to be?. I ended capturing with RGB without touching filter settings so I was able to keep gain in control.

- How is it possible that you managed to make that camera work better than their own drivers?.

- It would be nice to add gain and exposure to the filename.

- I had to start oacapture with sudo to make it work properly. No problem with that anyway.

Thanks for your work!

Francisco

[JamesF]

Hi James,

 Spanish user here. I've been struggling with my QHY5-II camera for days with firecapture and QHY drivers. Horizontal lines everywhere when raising gain to 100. Totally useless.

 Until I met your software. I installed it on an Ubuntu 14.04 yesterday and it simply works. No noise at all and thats FANTASTIC. Here is my feedback and questions :

I'm pleased to hear that (in the main) it seems to be working well for you.

- Gain threshold is from 1 to 47 (or 48, can't remember now). Is 47 the equivalent to 100 at firecapture?. Why that particular value?.

The way gain is set in the camera is quite odd.  It may be that I've misinterpreted how that works, but reading the spec sheet for the sensor persuaded me that the QHY implementation didn't look right at all.  The gain setting is a look-up table of 48 values, so each position on the gain control corresponds to one of those.  Presumably that means the highest gain level corresponds to 100 in the Windows driver.

I'll make a note to have another look at this though.

- Changing the filter to, ie. green, totally changes the image. It shows almost burn-out, like if gain was set to a much much high value than the actual one. Is that the intended way to be?. I ended capturing with RGB without touching filter settings so I was able to keep gain in control.

Gain can be set independently for each filter and should be remembered each time you change filters. I'd guess the default for green is too high. If you change it, then switch back to the red filter the gain should change to whatever it was last time you used the red filter. Then changing to green again should restore the previous green gain value.

At least, that's the way it is supposed to work. I've not had very much feedback on the filter management side of things, so there may well be bugs here.

- How is it possible that you managed to make that camera work better than their own drivers?.

I really have no idea   I went through their code in great detail because I couldn't get it to work properly and because their code is pretty much the only source of documentation for the current QHY5 range. There were places where the code just made no sense to me at all.  Some of their QHY5-II driver code appears to have been ripped from the QHY5L-II driver even where it really doesn't look like it should have been.  I did the best I could to create a driver that worked based on reading QHY's code and the sensor spec sheet.

- It would be nice to add gain and exposure to the filename.

That shouldn't be hard. I have added it to my list of features to implement.

- I had to start oacapture with sudo to make it work properly. No problem with that anyway.

That shouldn't happen. As the firmware has loaded correctly the udev rules are clearly working. It may be that you need to add yourself to the "users" group and log in again. That's not as neat as I'd like it to be, but the alternative is to make the camera device globally writable, which I don't like either. I'll sort this out more neatly once I can get everything packaged up properly as a .deb file.

Thanks for your work!

You're very welcome.

James

[Arroway]

Hi James,

 Thanks for taking time to answer here. Just to add some additional info about the filters; lets say I have set the maximum gain value (48) to my QHYII. When I select green filter afterwards, the gain increases!. I mean, even when the gain setting keeps showing 48,  the gain is boosted up and the image is much much brighter without touching any other configuration. It is like if I magically was able to raise the seting from 48 to 78 for example.

Is it like if you add more gain? Or the actual gain is really increased?.

Thanks!

Francisco

[JamesF]

I'll try to have a look at this over the next couple of days to see if I can find out what's up.  It sounds like some sort of bug to me.

James

[jmccormac001]

Hi James,

Sorry if you get this twice, you are the only person I can find online who is trying to track down the history of the 199e:1801 and 1802 cameras. 

I recently bought an all sky camera from Alcor in France and I want to use it under linux. The manufacturer said they had "a linux SDK" which is why we bought it from them. It turns out to be a bit more trick than a simple SDK. 

I've interogated the camera and found the following:

ops@ubuntu:~/data/skycam/IMG$ lsusbBus 001 Device 004: ID 199e:8102 The Imaging Source Europe GmbH 

I then found a device info script from the unicap author and find the following properties of the camera:

ops@ubuntu:~/alcor/skycam$ ./device_summaryUnicap device list:Device: DMK 51AU02.AS (/dev/video0)   Properties[5]:      Gamma      Gain      Exposure, Auto      shutter   Formats[1]:      Greyscale 8-bit (Y800) ( GREY )

Unicap lists 5 properties but only displays 4. Looking at the python wrapper that Alcor suggest I can see I am missing the frame rate property. I was hoping to use this python wrapper around unicap that they call their SDK, but it doesn't work. Have you been able to run an e:8102 camera under linux? 

Any advice would be great, thanks in advance. 

Cheers

James

[JamesF]

TIS do indeed have a Linux SDK, but I agree it's not quite what one might expect.

Sadly I think unicap is fairly moribund these days.  I started tinkering with it when I was first looking at capture for TIS cameras on Linux and cast it aside fairly rapidly because it didn't appear to support things I needed and some of the documentation looked badly outdated.

Getting the colour versions of the DxK51 working on Linux requires a little hoop-jumping because although it looks like a UVC device that would be supported by V4L2, but I believe your mono camera should "just work", in that the kernel should recognise it as a UVC camera and set up a V4L2 device for it.  That may provide you with a way forward.  If you plug the camera in and run

$ v4l2-ctl --list-devices

I'd hope to see it show up there.

If that doesn't work with unicap then you may still be able to use some of the gstreamer stuff.  Or you could of course use the oacapture camera interface libraries should you wish.  One of my plans has always been to write some sort of all-sky version of the capture application and I'd be more than happy to pull that into the sources if someone else were to write it

Anyhow, see where you get with recognising the camera as a V4L2 device and we can take things from there.  I'll drop you an email so you can reply directly.  In fact you could try controlling it with oacapture as a test.  That may not be a solution you want to use in the long term, but it might well prove that the camera can be made to work.

James