13.33 grid

[blablack]

Hi,

If I have the time-scale configuration set to 1s, the time-scale info shows the screen width to be 13.33s which makes me believe the grid is 13.33 cells wide.

If I display a LFO with a frequency of 1 Hz, the graph keeps steping back at each run (I don't know if I'm clear, what I mean is that the graphs doesn't superpose one run from another).

It would be great if the number of cells in the grid is an integer number so that given a frequency of a LFO, I can set the time-scale configuration any way I want to get the graphs to superpose at each run.

I don't know if I'm clear of confusing :)
Would you need more information, please let me know.

Aurélien

[x42]

The data is [supposed to be] continuous. There's no gap right-edge -> left edge. I'll verify this again.

The number of grid-cells depends on the width of the window and sample-rate. It is only limited by the fact there needs to be an integer number of samples per pixel.

[x42]

Works just fine here (after resizing the window). 48KSPS, 720px wide effective display

[x42]

PS. there's a define WITH_TIME_ADJ in gui/sisco.c:29 which, if enabled, adds a dial to to variably adjust and fine-tune the time-scale.

It's still limited to granularity of 'integer number of samples per pixel' but could get you want you want. It's not yet enabled by default because I don't like the plain dial without annotation and it looks odd in the layout..

Anyway, if you want a standing wave, just use a trigger..

[blablack]

Should I be able to resize the GUI?
In Jalv or Ingen, maximising the window look like the attached screenshot below.
The screenshot is done with Ingen and the ams-lv2 LFO set at a tempo of 30 pulsations/minutes...

[x42]

Looks like you've got an older version.
Resizing was added just after the v0.5 release and v0.6 is not yet tagged.

[blablack]

Hi,

Confirmed, I tried with the latest git of sisco.lv2.

Regarding the WITH_TIME_ADJ dial, would it be possible to type in the time-scale in the box direclty?

[x42]

Regarding numeric entry of the time-scale: Maybe.

I'm still pondering how to best facilitate variable time-scale in general.

As documented under (1) and (2) at http://x42.github.io/sisco.lv2/ the time-scale configuration is just a hint. What matters is the value displayed bottom left - not the value that one selects or enters.

Currently the actual display-grid-spacing is the closest approximation of the selected time-scale which yields an integer number of samples per pixel (for the given sample-rate and display width). An irregular number of samples-per-pixel would skew the display as well as trigger position. The same holds true if cairo sub-pixels were used instead of an integer grid. The only good solution [that I'm currently aware of] would be re-sampling; and that still imposes a limit to integer ratios. Accurate arbitrary time-scale digital scopes are far from trivial.

Anyway, I think a better solution is to add support for 'long term' measurements: overlaying waveforms from multiple cycles in combination with triggering. Similar to