Re: LW compiler optimizations
"David McClain (as dbm at refined-audiometrics dot com)" <[email protected]>
| Newsgroups | gmane.lisp.lispworks.general |
|---|---|
| Message-ID | <[email protected]> |
When it comes to perception, you also have to distinguish between instantaneous peak levels, and something akin to RMS levels. Today’s Loudness meters use a frequency weighted RMS measurement that deemphasizes the bass frequencies, and slightly amplify those above 1 kHz. That’s for normal hearing, of course. But it is an RMS measurement, which averages out instantaneous peaks that occur. The sample words, of course, have to deal with instantaneous levels, not RMS levels. So those 24 bits have to accommodate peaks: 24 dB headroom for that. But the RMS loudness won’t be anywhere near to +24 dBVU. > On Jan 14, 2025, at 06:03, David McClain (as dbm at refined-audiometrics dot com) <[email protected]> wrote: > > No, not exactly what I was stating… > > 8 bit samples will have a noise floor at -48 dBFS, assuming the samples peak at 0 dBFS. Now where you place that dynamic range is totally up to you. > > Given some immediate sound level, you probably can’t hear more than -36 dB down from that current level. But it is entirely possible that you do hear the noise floor in those 8-bit samples, so long as their present level is -12 dB down from their absolute peak levels. Your ears have adaptive dynamic range. So the audible range will ride down as the sounds become quieter. > > But you also got it right. Those 8 bit samples are inferior to wider bit-width samples, only in that their noise floor will be higher. There is no ratchety sound coming off 8 bits samples. No staircase effect. They simply have a higher noise floor. > >> On Jan 14, 2025, at 05:49, Yuri Davidovsky (as work at disclosure dot ie) <[email protected]> wrote: >> >> >>> On 14 Jan 2025, at 13:30, David McClain <[email protected]> wrote: >>> >>> While it is supposedly true that you can hear 96 dB of dynamic range, you can’t hear it all at once. At most, human hearing can accommodate about 6 bits of dynamic range at any one time. >> >> Hm, I am not so sure about that. If you listened to the ancient tracker modules with 8-bit samples, you would hear plenty of hiss in those, and 8 bits give you 48dB of dynamic range. >> >> https://www.youtube.com/watch?v=ncEBmfJ0ekQ >> >> Or do you mean that you can only hear the noise when the difference between the noise signal and the actual audible payload signal is < 36dB? >> >> >> _______________________________________________ >> Lisp Hug - the mailing list for LispWorks users >> [email protected] >> http://www.lispworks.com/support/lisp-hug.html > > > _______________________________________________ > Lisp Hug - the mailing list for LispWorks users > [email protected] > http://www.lispworks.com/support/lisp-hug.html _______________________________________________ Lisp Hug - the mailing list for LispWorks users [email protected] http://www.lispworks.com/support/lisp-hug.html