timestamps relative to onset of recording

[Luke Strgar]

Hi!

I understand when recording continuous data a timestamps.npy file is generated effectively indexing samples relative to the onset of acquisition. Is anyone aware if it is possible to determine the precise onset of acquisition or the precise onset on recording as a unix timestamp, for example?

I have many continuous recordings where acquisition was occurring for an extended period of time before recording, so the first sample in the timestamps array is >> 0. 

Thanks,

Luke

[Josh Siegle]

Hi Luke,

The settings.xml files saved at the start of each recording contain a “date” entry that provides the start time to within 1 second. Is that accurate enough, or are you looking for something more precise?

Josh

--
You received this message because you are subscribed to the Google Groups "Open Ephys" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-ephys+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-ephys/9aee1b12-6698-425d-aaa5-2a4b2aa1c0c3n%40googlegroups.com.

[Luke Strgar]

Hi Josh. Sorry I should have specified. We need millisecond level resolution. Are you aware of an alternative approach?

Thanks

[Josh Siegle]

Are you trying to recover the millisecond-level start time for existing datasets, or to do it for future experiments? If it’s the former, that might be difficult. If it’s the latter, there are a number of possibilities.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-ephys/1a2e1db7-633d-4920-960b-4b4437fa5375n%40googlegroups.com.

[Luke Strgar]

Unfortunately, the former. To your knowledge, is it possible?

[Josh Siegle]

Are you using Linux? Running “stat settings.xml” will give you the Unix timestamp of when the file was created, with sub-millisecond resolution.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-ephys/425a307f-1a25-48c7-b415-36962b133effn%40googlegroups.com.

[Luke Strgar]

The folks who acquired the data were working on windows 10 apparently and all the data was transferred to external hard drives before being sent to me. If the data ever had sub millisecond resolution it was lost when copying to these drives. 

[Josh Siegle]

Can you describe the problem in a bit more detail? I’m assuming you have some external system that is generating data in parallel to Open Ephys, and the GUI's acquisition was stopped and re-started multiple times while the external system continued to generate data. Are there any shared signals between the two systems, or are they completely independent?

To view this discussion on the web visit https://groups.google.com/d/msgid/open-ephys/655a8e9d-3744-4f1a-8dc5-b57e47437566n%40googlegroups.com.