The radmon.org radbox-20

Ok, this could be an issue with RadLog. I have some similar dips in my data, take a look at my shared data. This is probably due to Windows crunching due to something, I will investigate when I finally get back onto v0.42. The fix may well be simple, like putting the audio sampling thread on a slightly higher priority.

Just gotta get this damn server onto a VM on the big machine...

My development machine is Vista x64 and low on memory so it should suffer from this problem. The damn thing crunches and swaps all the time. The last few sampling periods in shared data don't show any problem. But on 8/04 16:06 there is a significant decrease in count detection. This could be a problem we share or windows updates doing its thing.

I wonder if you could find a system performance logger and match up any drops in counts with heavy memory load?

I'm pretty sure I can detect when the audio recording buffer stalls so could put in some logic to halt counting until the audio system is stable. What is your machine spec/OS ?

Win 8 is definitely slimmer and faster than its predecessors, so should be better than 7 on your machine. I don't know what progress has been made with audio in windows since Vista, where they move audio into userspace and all the extra bloat caused a load of audio stuttering problems with weaker CPUs. Surely it must be improved?

(Gotta get my RPi setup for the radbox! I am going to try moving the RadLog development machine onto a VM on the big server so the PC, which also runs RadLog for my detector, can be retired to uC development only and powered off. I have an audio interface for it, but would really like to try one of those audio-serial USB-TTL converters to see if it will work without the add on audio interface.)