Netcam Studio Crash

Congrats, it looks very good! At least in my constallation. The Logitech C270 provides audio in uncompressed/linear PCM, 16-bits, 1 channel with a 48 kHz sampling rate.

The latest 1.3.2 have been running for 10h with no problems at al. I have made several stress tests with combinations of motion and audio detection and what else I can do. No .net runtime error. No error in the Event Logs. I added an USB cam and used the audio source from a USB headset which also worked. This is actually a very interesting way for adding audio.

So to summarize I think this last version of NCS 1.3.2c is very stable and have solved the problem with audio. I am running this on win 10, .net framework 4.6.1 and in service mode.

The D-link 942L use AAC for audio and that do not work properly, but for the moment that is a minor problem.
-Henrik

Ok, good and strange at the same time.

While investigating your issue I have fixed a problem with USB audio handling, it was working when recording / streaming over web but not working when listening in Netcam Studio X (straight without converting to mp3 / unconverting as the other modes do due to the fact it considered the source as 8-bit audio instead of 16-bit audio).

But that’s the only thing and I cannot explain why the previous build failed in the things that were working correctly on my end and that did no change in this compared to the previous 1.3.2 and 1.3.1 so it’s really strange…

I unfortunately have no more AAC camera, I had an Axis M1031W which I use to perform AAC tests but it died a few months ago. I think I still have a M1054 somewhere but couldn’t find it, it’s somewhere in my basement :smile:

I was able to get AAC working at the time by using FFMPEG’s default AAC library in a public FFMPEG build and it worked great, the problem is that this library is GPL licensed which means that it cannot be included in a public build of Netcam Studio.

Netcam Studio includes a custom LGPL build of FFMPEG which I can redistribute with the software without having to open the source code of Netcam Studio and it also includes a very crappy and limited alternative of AAC library with all kind of issues and strange behaviours.

That’s basically why there are still problems with AAC but since all the cameras I knew that are providing AAC (they are pretty rare because the manufacturer also needs to pay a license for this) were offering alternative (G711, G726) then yes I came to the conclusion that it’s not the priority and that users will have to switch manually to G711 or G726 in the camera settings (as long as they are aware of this limitation which I agree, I’m probably the only one to know).

Are you sure that your Dlink camera do not offer such setting ?

Btw, I hate releasing official updates just before going on holidays but I think I will still do it if it also performs well on my test system until tomorrow. It would be great to know if the blob detection problem is also fixed, I have found a few memory management issues related to motion detector so I hope that was the problem but didn’t manage to see the problem by myself neither…

Here’s the changelog of 1.3.2:
Improved memory management
Added some checks in Blob detection to prevent memory issues
Fixed token issue may allow access to undesired cameras
Fixed Onvif detection when using multiple cameras with the same hardware identifier
Fixed audio playback in Netcam Studio X with local / usb audio

“so it’s really strange…” - thats why we are in this business :slight_smile: .
I checked new cams and as you say AAC is very rare. The 942L is only AAC. I can change a lot on the video, but not audio. For about 4 years ago they made a big deal about now supporting AAC. I found a new firmware, but no. This is not my main camera anymore so not a big deal.

I have had some crashes with blob detection before, but I discovered in general that when changing some parameter it is important to have an idle camera, i.e. motion detection should be off!

1.3.2 will be a release with very important improvements, but it is up to you. Hopefully, people keep older version in case of …
Happy holidays!
-Henrik

I haven’t been testing as intensively blob detection as the default frame difference mode so I will now knowing that there were issues. The frame difference mode I had it running 24/7 on an outdoor cam for about 45 days on my test system (it’s ± the longest I have achieved without having to reboot for windows update).

Apparently it was a handle / memory leak in blob detection and this one seems fixed now. I hope it will also fix the reported issues. I will send you PM with links to previous versions if some users have issue with 1.3.2 while i’m not here.