nelgau has quit [Remote host closed the connection]
nelgau has joined #scopehal
nelgau has quit [Ping timeout: 268 seconds]
kbeckmann has joined #scopehal
mxshift has quit [Ping timeout: 245 seconds]
mxshift has joined #scopehal
nelgau has joined #scopehal
nelgau has quit [Ping timeout: 256 seconds]
<_whitenotifier-8>
[scopehal] azonenberg pushed 1 commit to master [+0/-0/±5] https://git.io/JDr5T
<_whitenotifier-8>
[scopehal] azonenberg 503efe0 - TouchstoneParser: Added support for loading files with arbitrary port counts. Fixes #496. Fixes #497.
<_whitenotifier-8>
[scopehal] azonenberg closed issue #496: TouchstoneParser: support .s1p files so we can do TDR emulation with single-port measurements - https://git.io/JZch9
<_whitenotifier-8>
[scopehal] azonenberg closed issue #497: TouchstoneParser: support arbitrarily many ports - https://git.io/JZz8f
<azonenberg>
Ok so, i'm about to start a more broad refactoring of how i handle s-parameters in libscopehal and glscopeclietn
<azonenberg>
rather than mag and angle being separate channels, a single vector response will be considered a channel
<azonenberg>
with sub-streams for mag and angle
<azonenberg>
This will require a more fundamental refactoring of how we handle units and scaling because that will no longer be the guaranteed same for all streams in a channel
<azonenberg>
So probably almost every filter and scope driver is going to require some changes lol
<azonenberg>
I guess i know what i'm doing this weekend but it's for the best
nelgau has joined #scopehal
nelgau has quit [Ping timeout: 256 seconds]
nelgau has joined #scopehal
nelgau has quit [Ping timeout: 240 seconds]
nelgau has joined #scopehal
nelgau has quit [Ping timeout: 240 seconds]
<azonenberg>
ok i got about halfway through the most recent refactoring
<azonenberg>
which is a prerequisite to a lot of additional fun coming down the road
<azonenberg>
hoping to finish tomorrow
* GyrosGeier
wonders when people will abuse it for SDR
<GyrosGeier>
(and how to bring that about)
bvernoux has joined #scopehal
nelgau has joined #scopehal
nelgau has quit [Ping timeout: 240 seconds]
nelgau has joined #scopehal
<Degi>
Is there a reason for using mag and angle instead of I and Q?
nelgau has quit [Ping timeout: 240 seconds]
<GyrosGeier>
my expectation would be "no one can look at an I/Q graph and tell what is wrong"
<Degi>
Hm, but for a display render we could still convert it, but for filters etc. isn't I/Q more useful?
<GyrosGeier>
probably the best thing would be to convert depending on what the next block needs
someone-else has joined #scopehal
nelgau has joined #scopehal
nelgau has quit [Ping timeout: 268 seconds]
nelgau has joined #scopehal
nelgau has quit [Ping timeout: 256 seconds]
nelgau has joined #scopehal
river has quit [Ping timeout: 240 seconds]
bvernoux has quit [Quit: Leaving]
<azonenberg>
GyrosGeier: Yeah. Mag/angle is the most display friendly representation
<azonenberg>
Then i convert to real/imag if needed for specific applications
<azonenberg>
GyrosGeier: i mean i've already got some SDR features already, like digital downconversion and waterfall plots and FIR filters
<azonenberg>
and my scope-as-VNA project essentially uses the scope as a direct sampling receiver and digitally downconverts
nelgau has quit [Remote host closed the connection]
nelgau has joined #scopehal
nelgau has quit [Ping timeout: 268 seconds]
nelgau has joined #scopehal
nelgau has quit [Ping timeout: 240 seconds]
nelgau has joined #scopehal
nelgau_ has joined #scopehal
nelgau has quit [Read error: Connection reset by peer]
nelgau_ has quit [Client Quit]
nelgau has joined #scopehal
florolf has quit [Remote host closed the connection]
bvernoux has joined #scopehal
<_whitenotifier-8>
[scopehal-apps] robot-army opened issue #394: Segfaults on empty golden channel for Clock Jitter (TIE) - https://git.io/JD6on
<Degi>
Hmm, is mag/angle stored as 32 bit floats or so?
someone-else has quit [Quit: Connection closed]
florolf has joined #scopehal
bvernoux has quit [Quit: Leaving]
<azonenberg>
Degi: yeah
<azonenberg>
well
<azonenberg>
two separate fp32 channels
<azonenberg>
so you can plot them individually
<azonenberg>
the refactoring i'm doing allows that to be improved a bit so they're separate streams of one channel
<azonenberg>
right now, sub-streams don't support different y axis units
<Degi>
Nice
<Degi>
I guess as long as its at least fp32 it shouldn't matter whether its amplitude / phase or I / Q
<azonenberg>
Yeah. For some filters I do convert to real/imag internally
<azonenberg>
but the native format optimized for display is mag/angle, and the angles are in degrees
<Degi>
(Though with like 8 bit values it would be a problem since both would cover different spaces but with fp32 the points should be close enough to not matter)
<azonenberg>
When i'm doing de-embedding or channel emulation i have to resample the s-parameters anyway from the frequencies in the touchstone file to those of the FFT
<azonenberg>
so i can convert to a more computation friendly format then
<azonenberg>
that's only done once when changing the s-params in question or when the fft length changes