manual section 2.5.1 - there's clipping on macOS

[porres]

someone made this remark at a video, see 4:23 https://www.youtube.com/watch?v=G3CIJ3l5Fd8

Apparently, macOS doesn't clip the output and we should note it.

[umlaeute]

i haven't looked at the youtube video, but the macOS audio backend is known to

• use floating-point values natively (so you can easily send values to the [dac~] that exceed the -1..+1 range)
• have a built-in limiter (so when sending out-of-range values, they are gracefully rescaled and sound OK)

other audio backends might use floating-point values (like JACK; where you can pass out-of-range values between applications), but I don't know of any other backend that has an automatic limiter before the actual DAC (so you do get clipping effects).

in any case this is highly backend specific and I wouldn't get into details too much (as the backend behaviour is outside Pd's control).

[ben-wes]

wouldn't get into details too much

i agree. but i guess that these words might be misleading:

Inputs all appear between the values of -1 and 1;
and output values will be clipped to that range.

i checked again here on macos. it sounds to me like it clips output beyond -100..100, which means that even with the system volume set to 0.01, i can still produce full volume output from Pd (or any other app using the same backend that doesn't clip), which can get painful on headphones. this never caused any trouble for me - but i guess, we might just change this line anyway? it also doesn't seem correct for inputs. i can easily produce [adc~] inputs with blackhole beyond the -1..1 range.

how about something like this?:

Regular input appears between the values of -1 and 1;
and output values are expected in the same range.
Values may be clipped beyond this range depending on OS
and audio backend.

[ben-wes]

just adding one more little detail for the record after testing some more: the range of Pd's output gets clipped to -1..1 for my bluetooth headphones (even with macos' volume set very low for these). so what i observed above is mainly true for output through the macbook speakers and headphone jack. i'd still keep the adapted text as is though - or should we also mention drivers and hardware?