Theoretical Sample per seconds does not match
Theoretical Sample per seconds does not match
Theoretical Sample per seconds does not match error is quite clear, but I could not find anywhere as how to debug this. I am using TCP connection so I guess this could be a problem. Under the hood what does happen ?
Re: Theoretical Sample per seconds does not match
Dear Nicolas,
we recently got a few issues on several drivers that did not send the number of samples they promised to. We moved the monitoring of the sample count to the acquisition server (was only on Nexus32 driver previously) and it seems that other drivers are affected. What driver / device are you using ?
There are several reasons to make this message appear (apart from a bug in the implementation of the OpenViBE driver itself) :
- you receive an abnormal sample count at start time due e.g. to an internal buffer being full before start and emptying as soon as some samples are requested
- the device is not sending the samples correctly regarding to the requested sampling rate. It could be a weird implementation on the device side or it could just be that the internal clock of the device is not exactly synced with the computer clock
We will move the acquisition part to a separate thread (independent of GTK) in the hope that it helps.
If it doesn't help enough, then each driver will have to adapt its behavior to its own prorblem...
Before twicking anything, be sure you provide the correct information (sampling rate, buffer sizes etc...) to your device and that you are not missing some samples...
Hope this helps,
Best regards,
Yann
we recently got a few issues on several drivers that did not send the number of samples they promised to. We moved the monitoring of the sample count to the acquisition server (was only on Nexus32 driver previously) and it seems that other drivers are affected. What driver / device are you using ?
There are several reasons to make this message appear (apart from a bug in the implementation of the OpenViBE driver itself) :
- you receive an abnormal sample count at start time due e.g. to an internal buffer being full before start and emptying as soon as some samples are requested
- the device is not sending the samples correctly regarding to the requested sampling rate. It could be a weird implementation on the device side or it could just be that the internal clock of the device is not exactly synced with the computer clock
We will move the acquisition part to a separate thread (independent of GTK) in the hope that it helps.
If it doesn't help enough, then each driver will have to adapt its behavior to its own prorblem...
Before twicking anything, be sure you provide the correct information (sampling rate, buffer sizes etc...) to your device and that you are not missing some samples...
Hope this helps,
Best regards,
Yann