No worries, any of the new styles or functions have the template info at the top of the .h file. ColorSelect<> is primarily for Edit Mode but it takes functions so it can be used like Brian did with WavNum<>. Although some functions probably wouldn’t work so well with it if they update too often.
Feel free to test whatever you want, I beat up all of the styles and functions pretty heavily and most are already built into the Edit Mode styles but more eyes never hurt. I don’t think they’re holding up Alpha as they’ve already been pretty well tested out.
Some will only really be used in Edit Mode as they were specifically added to support the arguments so the best method to test is Edit Mode.
Very cool! This definitely shows that the audio filter is working pretty much as intended. However, I think we might need a wider sweep, because it seems like we’re already starting in the down-slope. I suspect that if we started at 1000Hz or maybe even 4000Hz, we’d see that the response at 40Hz is much weaker than what it appears in this graph. If I’m reading the graph right, the sensitivity is falling about 15dB per octave even at 300Hz.
I’m not sure what you mean by this. The idea with the filters is to reduce the amount of electricity that the board has to output, while changing the output sound as little as possible.
PS: Screenshots and videos are perfect for graphs like this. It’s not screenshots that I have a problem with really. It’s screenshots of text.
I guess I meant it has me rethinking whether filtering out “unnessesary” freqs helps, or at least where the line should be drawn, as apparently there’s a good amount of bass possible?
I was just amazed that freqs that low were even reproduced. I guess I stand corrected preaching how these little 28’s are pointless under ~100.
I’ll do some larger sweeps in a bit.
Unfortunately, My GitHub-fu is weak and splitting up seems abundantly more difficult than closing it and instead just editing small bits of related parts as individual PRs.
If you’re wanting to see accurately how well the filter is working, you need direct audio recording. Earballing it with a speaker should be done after to A/B using the raw output as a baseline.
We could also just add some code that saves the output to the SD card and analyze the data from the SD card.
However, I think the fact that it sounds right coupled with the fact that the spectrogram clearly shows a dip in the right place is pretty strong evidence for it working reasonably well.
hey can anyone point me in the right direction of the descriptions for each of the defines? there’s quite a few in this OS that I find myself forgetting what some of them are meant to do. as an example, I was trying to find the descriptions to FETT263_HOLD_BUTTON_OFF versus FETT263_HOLD_BUTTON_LOCKUP.
Eventually that’s the plan, after everything is built but there’s already a ton of documentation available which is why I built that Countdown website to put everything together so you just need to read through what is there.
Yes, during Alpha especially you need to pull down the latest OS anytime updates are merged. There’s still more coming, to help with Alpha you always need the current version from github, in some cases it’s being updated several times a day.