Closed Foucl closed 7 years ago
I presume this is coming from some kind of webcam? I've sen this kind of sampling irregularity before. If memory serves, I think that you don't even need to set can_drop_samples in order to smooth out this irregularity. It should be that the normal time smoothing routines in load_xdf can handle what you describe. I would recommend testing all configurations before coming to a final conclusion. You can edit an XDF post-recording with XZDFBrowser. Also, you can load the XDF with HandleJitterRemoval set to false. This will prevent both jitter removal and, if present, the can_drop_samples routine is triggered by the flag in the XDF from happening.
On 11/11/2016 07:37 AM, Christopher wrote:
I am using LSL to forward samples of a recording device (facial expression recognition software) generates samples at about 31.3 Hz so that usually a new sample is generated every 32 ms. However, exactly every 12th sample is delayed by 16 ms (so the timespan between the 12th and 13th, 24th and 25th etc sample is 48 ms instead). Is it enough to just set can_drop_samples to true or is there a better way to accomodate that in the stream settings?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/sccn/labstreaminglayer/issues/151, or mute the thread https://github.com/notifications/unsubscribe-auth/ADch7ojQNv5bfrtzYW7mBPFQ1Be9GOAHks5q9Iu4gaJpZM4Kv3Dl.
yes, it's a logitech webcam that is connected to a PC running iMotion's Facial Expression Recognition Module. It seems impossible to fix the webcam's framerate (at least on Windows and with the official drivers). thank you for your suggestions - I hadn't thought about testing using the XDFBrowser. If it's alright, I'll close this issue and post the results of my testing as soon as I get to it.
I am using LSL to forward samples of a recording device (facial expression recognition software) generates samples at about 31.3 Hz so that usually a new sample is generated every 32 ms. However, exactly every 12th sample is delayed by 16 ms (so the timespan between the 12th and 13th, 24th and 25th etc sample is 48 ms instead). Is it enough to just set can_drop_samples to true or is there a better way to accomodate that in the stream settings?