azonenberg changed the topic of #scopehal to: libscopehal, libscopeprotocols, and glscopeclient development and testing | https://github.com/glscopeclient/scopehal-apps | Logs: https://libera.irclog.whitequark.org/scopehal
Degi_ has joined #scopehal
Degi has quit [Ping timeout: 246 seconds]
Degi_ is now known as Degi
<_whitenotifier-e> [scopehal] azonenberg pushed 1 commit to master [+0/-0/±1] https://github.com/glscopeclient/scopehal/compare/2cfb77d42d43...ed9b3ac12cda
<_whitenotifier-e> [scopehal] azonenberg ed9b3ac - EmphasisRemovalFilter: fixed crash if data rate is zero
bvernoux has quit [Ping timeout: 260 seconds]
bvernoux has joined #scopehal
<_whitenotifier-e> [scopehal] azonenberg opened issue #603: Weird numerical issue in emphasis removal filter - https://github.com/glscopeclient/scopehal/issues/603
<_whitenotifier-e> [scopehal] azonenberg labeled issue #603: Weird numerical issue in emphasis removal filter - https://github.com/glscopeclient/scopehal/issues/603
<_whitenotifier-e> [scopehal] azonenberg labeled issue #603: Weird numerical issue in emphasis removal filter - https://github.com/glscopeclient/scopehal/issues/603
<_whitenotifier-e> [scopehal] azonenberg pushed 1 commit to master [+0/-0/±1] https://github.com/glscopeclient/scopehal/compare/ed9b3ac12cda...33e09a7e587a
<_whitenotifier-e> [scopehal] azonenberg 33e09a7 - DeEmbedFilter: fixed 90 degree rotation in legacy/portable (no AVX2) filter implementation
<_whitenotifier-e> [scopehal] azonenberg commented on issue #603: Weird numerical issue in emphasis removal filter - https://github.com/glscopeclient/scopehal/issues/603#issuecomment-1125747080
<azonenberg> wellll
<azonenberg> i'm beginning to think this is not my code's fault
<azonenberg> i just spent the last hour completely rewriting the emphasis code and got identical results with the new implementation
<azonenberg> and i just tried using the LeCroy EyeDr software package on my SDA 816
<azonenberg> ... and those little squares are still there
<azonenberg> So it's either really in the signal, or it's an artifact of the scope capture
<azonenberg> or me and lecroy made the same mistake processing the waveform
<azonenberg> either way i think i'm done for the night
<_whitenotifier-e> [scopehal] azonenberg commented on issue #603: Weird numerical issue in emphasis removal filter - https://github.com/glscopeclient/scopehal/issues/603#issuecomment-1125819678
<_whitenotifier-e> [scopehal] azonenberg closed issue #603: Weird numerical issue in emphasis removal filter - https://github.com/glscopeclient/scopehal/issues/603
<azonenberg> yeah it's definitely ISI
Stephie has quit [Ping timeout: 260 seconds]
Stephie has joined #scopehal
<Degi> Little squares? The jaggy looking appearance in the first pic?
bvernoux has quit [Quit: Leaving]
<azonenberg> the little + shaped mark of 4 squares
<azonenberg> i thought it was a rendering/processing artifcat
<azonenberg> upon closer inspection it's ISI
<Degi> Ah, between two eyes in the middle
<_whitenotifier-e> [scopehal] 602p opened pull request #604: Allow looser One-Wire timings, marking them as errors but still decoding - https://github.com/glscopeclient/scopehal/pull/604
<d1b2> <louis> RE: above PR: I'm open to a different way of doing this (configurable sloppiness parameter for decoder?) but this was quick + dirty and matched the RESET behaviour
<azonenberg> @louis: in general
<azonenberg> i want to make our decoders understand malformed protocol data to the extent possible
<azonenberg> as long as it's clearly marked as invalid
<azonenberg> this will greatly aid in debugging
<d1b2> <louis> OK. Maybe we want to add specific COLOR_DATA_WARN and COLOR_CONTROL_WARN for this (and for invalidly-timed RESET pulses)? Or perhaps an orthogonal flag for "something is wrong" (could do away with CHECKSUM_xxx then and just have CHECKSUM that may be bad)?
<d1b2> <louis> (PR as-is works fine and fixed my "in anger" use-case)