Re: vorbis quality - quality scale vs bitrate
Gunter Königsmann <[email protected]> Mon, 27 Nov 2017 07:27:54 +0100
| Newsgroups | gmane.comp.multimedia.ogg.vorbis.devel |
|---|---|
| Message-ID | <[email protected]> |
--===============4841377562634769367== Content-Type: multipart/alternative; boundary="----W1J5LAK9KEVXDB4M3PFC62XB44U6VJ" Content-Transfer-Encoding: 7bit ------W1J5LAK9KEVXDB4M3PFC62XB44U6VJ Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Additionally: do you really need exactly this amounts of bits in every seco= nd? Or is the file length more important? In that case you can specify an a= verage bitrate and the codec might save bits when there is silence and no m= ovement for times there is much complexity that requires many bits to expla= in=2E Am 27=2E November 2017 01:55:25 MEZ schrieb Tuomo Latto <djv@iki=2Efi>: >On 27=2E11=2E2017 02:00, YIRAN LI wrote: >> I'm using libvorbis in my program and need to encode to target >bitrate=2E I >> know libvorbis prefer to use quality scale but I can't use it=2E >>=20 >> I've found something at faq http://vorbis=2Ecom/faq/#quality >>=20 >> *For now, quality 0 is roughly equivalent to 64kbps average, 5 is >roughly >> 160kbps, and 10 gives about 400kbps=2E Most people seeking >> very-near-CD-quality audio encode at a quality of 5 or, for lossless >stereo >> coupling, 6=2E The default setting is quality 3, which at approximately >> 110kbps gives a smaller filesize and significantly better fidelity >than >> =2Emp3 compression at 128kbps=2E* >>=20 >> so 0 =3D=3D 64kbps >> 3 =3D=3D 110kbps >> 5 =3D=3D 160kbps >> 10 =3D=3D 400kbps >>=20 >> Could anyone give me a complete mapping from quality to bitrate =EF=BC= =9F > >It doesn't really work that way=2E Those are about right nominally >(and I know 6 =3D=3D 192kbps), but in practice they tend to produce >slightly larger bitrates=2E The bitrates vary quite a bit, since >quality implies VBR or ABR=2E For example, with q 6 you might >get an average bitrate anywhere from 180 to 220 kbps with >significantly more variation within the encoded file=2E >(And, of course, quality is not limited to integer values=2E) > >It might be a good idea to start by looking at oggenc command-line >arguments to get an idea of what you need to for CBR encoding=2E >You at least need to use bitrate management and set the minimum >and maximum bitrates, I think=2E AFAIK oggenc is essentially just >a fairly straightforward mapping from command-line to libvorbis, >so its source is probably a good place to see how it's done=2E > > >--=20 >Tuomo > >--- >This email has been checked for viruses by Avast antivirus software=2E >https://www=2Eavast=2Ecom/antivirus > >_______________________________________________ >Vorbis-dev mailing list >Vorbis-dev@xiph=2Eorg >http://lists=2Exiph=2Eorg/mailman/listinfo/vorbis-dev --=20 Diese Nachricht wurde von meinem Android-Ger=C3=A4t mit K-9 Mail gesendet= =2E ------W1J5LAK9KEVXDB4M3PFC62XB44U6VJ Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: quoted-printable <html><head></head><body>Additionally: do you really need exactly this amou= nts of bits in every second? Or is the file length more important? In that = case you can specify an average bitrate and the codec might save bits when = there is silence and no movement for times there is much complexity that re= quires many bits to explain=2E<br><br><div class=3D"gmail_quote">Am 27=2E N= ovember 2017 01:55:25 MEZ schrieb Tuomo Latto <djv@iki=2Efi>:<blockqu= ote class=3D"gmail_quote" style=3D"margin: 0pt 0pt 0pt 0=2E8ex; border-left= : 1px solid rgb(204, 204, 204); padding-left: 1ex;"> <pre class=3D"k9mail">On 27=2E11=2E2017 02:00, YIRAN LI wrote:<br><blockqu= ote class=3D"gmail_quote" style=3D"margin: 0pt 0pt 1ex 0=2E8ex; border-left= : 1px solid #729fcf; padding-left: 1ex;"> I'm using libvorbis in my program= and need to encode to target bitrate=2E I<br> know libvorbis prefer to use= quality scale but I can't use it=2E<br> <br> I've found something at faq = <a href=3D"http://vorbis=2Ecom/faq/#quality">http://vorbis=2Ecom/faq/#quali= ty</a><br> <br> *For now, quality 0 is roughly equivalent to 64kbps average= , 5 is roughly<br> 160kbps, and 10 gives about 400kbps=2E Most people seeki= ng<br> very-near-CD-quality audio encode at a quality of 5 or, for lossless= stereo<br> coupling, 6=2E The default setting is quality 3, which at appro= ximately<br> 110kbps gives a smaller filesize and significantly better fide= lity than<br> =2Emp3 compression at 128kbps=2E*<br> <br> so 0 =3D=3D 64kbps= <br> 3 =3D=3D 110kbps<br> 5 =3D=3D 160kbps<br> 10 =3D=3D 400kbp= s<br> <br> Could anyone give me a complete mapping from quality to bitrate = =EF=BC=9F<br></blockquote><br>It doesn't really work that way=2E Those are = about right nominally<br>(and I know 6 =3D=3D 192kbps), but in practice the= y tend to produce<br>slightly larger bitrates=2E The bitrates vary quite a = bit, since<br>quality implies VBR or ABR=2E For example, with q 6 you might= <br>get an average bitrate anywhere from 180 to 220 kbps with<br>significan= tly more variation within the encoded file=2E<br>(And, of course, quality i= s not limited to integer values=2E)<br><br>It might be a good idea to start= by looking at oggenc command-line<br>arguments to get an idea of what you = need to for CBR encoding=2E<br>You at least need to use bitrate management = and set the minimum<br>and maximum bitrates, I think=2E AFAIK oggenc is ess= entially just<br>a fairly straightforward mapping from command-line to libv= orbis,<br>so its source is probably a good place to see how it's done=2E<br= ><br></pre></blockquote></div><br> -- <br> Diese Nachricht wurde von meinem Android-Ger=C3=A4t mit K-9 Mail gesendet= =2E</body></html> ------W1J5LAK9KEVXDB4M3PFC62XB44U6VJ-- --===============4841377562634769367== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: base64 Content-Disposition: inline X19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX19fX18KVm9yYmlzLWRl diBtYWlsaW5nIGxpc3QKVm9yYmlzLWRldkB4aXBoLm9yZwpodHRwOi8vbGlzdHMueGlwaC5vcmcv bWFpbG1hbi9saXN0aW5mby92b3JiaXMtZGV2Cg== --===============4841377562634769367==--