<d1b2>
<xfol_> could someone please clarify for me - can we use ngscopeclient to play back waveform recordings as though they are a live device? I had trouble getting either of my scopes to work reliably (MSO5074, RTB2004) and figured I could at least do my measurements using the scope's waveform record function as an interim solution but when I load a waveform file into the filter graph and "trigger" with the play button I don't see any waveform displays. The
<d1b2>
demo source is all I've been able to get to work so far.
<d1b2>
<azonenberg> There's no need to trigger when doing offline analysis
<d1b2>
<azonenberg> a waveform recording is a single waveform that should show up as soon as you load the file
<d1b2>
<azonenberg> If nothing is showing up you might be not importing it correctly?
<d1b2>
<azonenberg> (better error reporting during file load is something we need to work on)
<d1b2>
<xfol_> Is there an example waveform file and filter graph somewhere I could use to sanity check
<d1b2>
<azonenberg> How are you exporting the waveform and what import filter are you using?
<d1b2>
<azonenberg> in particular for the RTB2004 I don't know what the native waveform format is, we may not yet have a loader for it
<d1b2>
<azonenberg> we have loaders for Rigol, LeCroy, and Tek binary waveforms off the top of my head
<d1b2>
<xfol_> all I could get to work without crashing on import is csv export on my Rigol MSO5074
<d1b2>
<azonenberg> I don't know if R&S uses the same file format or not
<d1b2>
<azonenberg> It definitely should not crash
<d1b2>
<azonenberg> if you have a file that crashes the importer, even if it's malformed / the wrong type, that's a bug we want to know about
<d1b2>
<azonenberg> importers should always either successfully load the file or fail and display nothing, plus logging to the console until we get a better error reporting mechanism
<d1b2>
<azonenberg> crashing is never acceptable
<d1b2>
<xfol_> actually they may not have all crashed, but only thr Rigol CSV produced a filter node that had an output to connect to other nodes.
<d1b2>
<xfol_> Can't check at the moment, i could share some waveform files at work tomorrow with debug info
<d1b2>
<azonenberg> ok yes please do
<d1b2>
<azonenberg> This sounds like either you've come across a new waveform format we don't yet have a loader for, we have bugs in the importer, or both
<d1b2>
<azonenberg> in any case, CSV should be the least-common-denominator format that any scope can work with
<d1b2>
<azonenberg> it's just bulky
<d1b2>
<azonenberg> our CSV importer expects one line per sample, first column is timestamp in seconds, subsequent columns are channel data
<d1b2>
<azonenberg> which is afaik what most scopes put out natively
<d1b2>
<xfol_> Thanks I'll check that and share a bug report if the problem is not on my end
<d1b2>
<azonenberg> if you arent already, run from a console so you can see stdout, and add --debug to the arguments
<d1b2>
<azonenberg> this will dump more status information and perhaps shed light on where it's failing