On Nov 05 11:32:49, hans at stare.cz wrote:> On Nov 05 11:05:34, hans at stare.cz wrote: > > > Did you also try to listen at the beginning, shortly before the real tone appears in the audible spectrum? While significantly larger, Opus had produced significant ghost noise (much less than Vorbis did)...I experience the "same" low level noise even in a wav file, even on frequencies a bit higher, when playing on shitty enough speakers. I don't think it is a relict of the compression (it cannot be, here). Every speaker (or headphones) is designed for a certain frequency range. Outside of that range, all bets are off. I wouldn't bother with what you "hear" outside of the audible spectrum. My current bet is that the very low frequencies are correctly represented in the Opus file, but your speaker/headphones are not capable to play them properly (which is no wonder). Jan> > Yes. But given that these are not audible sounds, > > the various codecs might use various strategies of throwing them away. > > I have no solid knowledge of what e.g. Ogg vs Opus does, but I would > > just throw away any frequency below, say, 15 Hz. Maybe they do/don't, > > and this might be the result. > > Have you also tried creating a swepp only up to those frequencies? > Say, 1 to 20 Hz? Is the low-freq noise still there? > > Jan >
On Nov 05 18:00:06, hans at stare.cz wrote:> On Nov 05 11:32:49, hans at stare.cz wrote: > > On Nov 05 11:05:34, hans at stare.cz wrote: > > > > Did you also try to listen at the beginning, shortly before the real tone appears in the audible spectrum? While significantly larger, Opus had produced significant ghost noise (much less than Vorbis did)... > > I experience the "same" low level noise even in a wav file, > even on frequencies a bit higher, when playing on shitty enough speakers. > I don't think it is a relict of the compression (it cannot be, here). > > Every speaker (or headphones) is designed for a certain frequency range. > Outside of that range, all bets are off. I wouldn't bother with what you > "hear" outside of the audible spectrum. > > My current bet is that the very low frequencies are correctly > represented in the Opus file, but your speaker/headphones > are not capable to play them properly (which is no wonder).Attached I send the spectrogram (vic SoX) of the first 20 seconds for the wav file and the opus file. Indeed, there is extra noise for the low frequencies, but somewhere around -100 dB. Jan -------------- next part -------------- A non-text attachment was scrubbed... Name: spectrogram-wav.png Type: image/png Size: 34311 bytes Desc: not available URL: <http://lists.xiph.org/pipermail/opus/attachments/20181105/ce02fd90/attachment-0002.png> -------------- next part -------------- A non-text attachment was scrubbed... Name: spectrogram-opus.png Type: image/png Size: 167785 bytes Desc: not available URL: <http://lists.xiph.org/pipermail/opus/attachments/20181105/ce02fd90/attachment-0003.png>
On Mon, Nov 5, 2018 at 11:01 AM Jan Stary <hans at stare.cz> wrote: Attached I send the spectrogram (vic SoX) of the first 20 seconds> for the wav file and the opus file. Indeed, there is extra noise > for the low frequencies, but somewhere around -100 dB. > > Jan >That might be entirely due to SoX treating it as a 16-bit file, which it is not; -100dB is almost exactly the limitation of 16-bit. All Opus files are infinite-precision, and they'll encode the input at whatever precision is fed to them, but they do have a silence-detection mechanism which defaults to 16-bit in opusenc. SoX is either reading that value and erroneously assuming that's that internal precision, or just falling back on a 16-bit default because it has no information, but it would be more proper to _always_ decode to 24-bit to eliminate that broadband noise low bit depths create (except in the case of hardware limitations). As you've found, Opus is always 48kHz, never more, never less. Its resampler is very accurate, and should never introduce noise. My speakers and headphones definitely have issues with the sweep, so it's hard to isolate any differences. (I do love it though, it really gets my floor shaking; I'll keep it around for testing purposes.) The only remaining issue is the way that Opus doesn't even come close to respecting the requested bitrate for this sample. For instance, encoding it at 130 gives me a file of 210kbps. Over a wide corpus of music, I've noticed libopus almost invariably overshoots the bitrate, rather than averaging out close to it; only speech is consistently at or below it. Em -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.xiph.org/pipermail/opus/attachments/20181105/8cd1cdd6/attachment.html>
Ulrich Windl
2018-Nov-06 06:48 UTC
[opus] Antw: Re: Antw: Re: Antw: Re: Possible bug in Opus 1.3
>>> Jan Stary <hans at stare.cz> schrieb am 05.11.2018 um 18:00 in Nachricht<20181105170002.GA72174 at www.stare.cz>:> On Nov 05 11:32:49, hans at stare.cz wrote: >> On Nov 05 11:05:34, hans at stare.cz wrote: >> > > Did you also try to listen at the beginning, shortly before the real tone > appears in the audible spectrum? While significantly larger, Opus had > produced significant ghost noise (much less than Vorbis did)... > > I experience the "same" low level noise even in a wav file, > even on frequencies a bit higher, when playing on shitty enough speakers. > I don't think it is a relict of the compression (it cannot be, here). > > Every speaker (or headphones) is designed for a certain frequency range. > Outside of that range, all bets are off. I wouldn't bother with what you > "hear" outside of the audible spectrum. > > My current bet is that the very low frequencies are correctly > represented in the Opus file, but your speaker/headphones > are not capable to play them properly (which is no wonder).Hi! I actually doubt it: I have a sound system with a sub-woover, and I can see the subwoover move before I can hear anything; looks kind of funny... Similar for my headphones: It's not the cheap dirt they sell you everywhere (but also it is not high-end). Still, I'm trying to investigate. Regards, Ulrich> > Jan > > >> > Yes. But given that these are not audible sounds, >> > the various codecs might use various strategies of throwing them away. >> > I have no solid knowledge of what e.g. Ogg vs Opus does, but I would >> > just throw away any frequency below, say, 15 Hz. Maybe they do/don't, >> > and this might be the result. >> >> Have you also tried creating a swepp only up to those frequencies? >> Say, 1 to 20 Hz? Is the low-freq noise still there? >> >> Jan >> > _______________________________________________ > opus mailing list > opus at xiph.org > http://lists.xiph.org/mailman/listinfo/opus
Ulrich Windl
2018-Nov-06 06:49 UTC
[opus] Antw: Re: Antw: Re: Antw: Re: Possible bug in Opus 1.3
>>> Jan Stary <hans at stare.cz> schrieb am 05.11.2018 um 20:01 in Nachricht<20181105190138.GA78488 at www.stare.cz>:> On Nov 05 18:00:06, hans at stare.cz wrote: >> On Nov 05 11:32:49, hans at stare.cz wrote: >> > On Nov 05 11:05:34, hans at stare.cz wrote: >> > > > Did you also try to listen at the beginning, shortly before the realtone> appears in the audible spectrum? While significantly larger, Opus had > produced significant ghost noise (much less than Vorbis did)... >> >> I experience the "same" low level noise even in a wav file, >> even on frequencies a bit higher, when playing on shitty enough speakers. >> I don't think it is a relict of the compression (it cannot be, here). >> >> Every speaker (or headphones) is designed for a certain frequency range. >> Outside of that range, all bets are off. I wouldn't bother with what you >> "hear" outside of the audible spectrum. >> >> My current bet is that the very low frequencies are correctly >> represented in the Opus file, but your speaker/headphones >> are not capable to play them properly (which is no wonder). > > Attached I send the spectrogram (vic SoX) of the first 20 seconds > for the wav file and the opus file. Indeed, there is extra noise > for the low frequencies, but somewhere around ‑100 dB.OK, that looks very much convincing; I'll investigate...> > Jan
> > My current bet is that the very low frequencies are correctly > > represented in the Opus file, but your speaker/headphones > > are not capable to play them properly (which is no wonder). > > I actually doubt it: I have a sound system with a sub-woover, and I can see the subwoover move before I can hear anything; looks kind of funny...Yes, I can see my bass speaker moving too. That does not mean it is representing the low frequencies correctly. I am not convinced there is a bug in the opus codec that would mis-represent the very low frequecies. At any rate (ha, pun!), the point is moot: I don't care how well a codec compresses a 10 Hz sound, because I will never hear it anyway. Jan