My PARAM knob goes from 0 to 16340 (and not 16383 as it should).
The one hesitation/issue I have with calibration is that it can be a really hard thing to do the UI for. It usually requires 2 values to be specified, and it’s not always that easy to communicate where those values come from.
Also, will it be computationally cheap enough to use the calibration data? My gut says yes, as OPs are never run that often (computationally speaking).
what to load on startup, blank scene, last saved, or a specific scene
Maybe… but having a configuration option just avoid a single key press at power on might set a bad precedent for how many things end up in the menu. You might struggle to complain about other similarly trivial items if this ends up in there.
I expect than @tehn will veto any number of things, so I’ll just aggregate everything in the top post for now.
There is a trade-off with the config menu and setting basic user expectations. The menu permits flexibility when a design decision would otherwise be rigid. I see things being bumped into the config menu as catering to the community, but there is a lot to be said in favour of just making good decisions where they need to be rigid.
I see your point - in fact that was my first thought about Startup Metro on/off. But then I could not think of another example that would not be as specific but still obvious. Until I engaged in a discussion about if the variable monitor should be on or off at startup. Why not saving a single key press on every startup as long as it aims to some very defined features.
Also the variable monitor has no OP and can not be toggled from an initial I script.
Yeah, and I think it makes sense. It’s a hardware interface feature as opposed to a teletype core feature. I think @tehn has a good vision for where to draw the boundary, and was right to cut those aspects of the feature out.
So… thinking about calibration a bit more. Assuming all the calculations are done at runtime and no pre-calculation is done (e.g. pre-calculating N values). Maybe it would be better for CV out calibration to be an OP… in an ideal world calibration is done to some universally agreed upon standard (i.e. 1V/oct), but in practise it’s to particular set of VCOs (or VCFs).
Assuming you’ve got the values worked out in advance and kept somewhere safe, you could then add a bunch of CV.CAL calls to your I script setting up each of your oscillators.
Just a thought.
I’m a bit in 2 minds about which is better. I know in reality I just won’t be bothered to make a list of values for each of my oscillators (and the values would be different for each CV output used). But I might be bothered to do it once for each of the outputs to get them as near to 1V/oct as I can be bothered.
(Also I would really like my PARAM knob to hit the full range…)