Question about FLAC encoder behaviour

--===============1474496223206438015==
Content-Type: multipart/alternative;boundary="----=_20240201224850_72784"

------=_20240201224850_72784
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable

=0A=0A=0AHello all! =0A=0A=0AI have a question about the FLAC's reference i=
mplementation encoder.=0ASo I have the following use cases:=0AI have tested=
 both flac 1.3.2 and flac 1.4.3 on reencoding existing .flac=0Afiles with t=
he -f option. There are cases in which I try to reencode .flac=0Afiles whic=
h are compressed by other FLAC implementations (ffmpeg's one for=0Aexample)=
 which seems to deliver better compression on some material than=0Athe refe=
rence implementation.=0A=0A=0AIn these cases, flac 1.3.2 gives an error for=
 failed compression ("ratio=0Amust be =E2=89=A4 1.000" message) and writes =
bigger file than the original one,=0Awhile flac 1.4.3 recompresses without =
giving the error , otputting ratio=0Abigger than 1.000.=0A=0A=0AMy question=
 is whether these behaviours are normal and is there a risk for=0Acorruptin=
g the audio material because of some error? Because it seems to me=0Athat r=
egardless giving an error or compression ratio change or not both=0Aflac 1.=
3.2 and 1.4.3 are actually finishing the compression?=0A=0A=0AI've tested t=
he output files with -t and none of them had problems so=0Afar.=0A=0A=0ATha=
nk you in advance for your time! Greetings! :)=0A=0A=0AHave a nice day! :)=
=0A
------=_20240201224850_72784
Content-Type: text/html; charset="utf-8"
Content-Transfer-Encoding: quoted-printable

<p valid=3D"true" style=3D"line-height: 16px; font-size: 13px; font-family:=
 Tahoma, sans-serif; margin: 16px 0px;">Hello all!&nbsp; <br><br>I have a q=
uestion about the FLAC's reference implementation encoder.<br>So I have the=
 following use cases:<br>I have tested both flac 1.3.2 and flac 1.4.3 on re=
encoding existing .flac files with the -f option. There are cases in which =
I try to reencode .flac files which are compressed by other FLAC implementa=
tions (ffmpeg's one for example) which seems to deliver better compression =
on some material than the reference implementation.<br><br>In these cases, =
flac 1.3.2 gives an error for failed compression ("ratio must be =E2=89=
=A4 1.000" message) and writes bigger file than the original one, while fla=
c 1.4.3 recompresses without giving the error , otputting ratio bigger than=
 1.000.<br><br>My question is whether these behaviours are normal and is th=
ere a risk for corrupting the audio material because of some error? Because=
 it seems to me that regardless giving an error or compression ratio change=
 or not both flac 1.3.2 and 1.4.3 are actually finishing the compression?<b=
r><br>I've tested the output files with -t and none of them had problems so=
 far.<br><br>Thank you in advance for your time! Greetings! :)<br></p><p va=
lid=3D"true" style=3D"line-height: 16px; font-size: 13px; font-family: Taho=
ma,sans-serif; margin: 16px 0px;"><span style=3D"font-family: Tahoma,Nimbus=
 Sans L,sans-serif; font-weight: bold; color: rgb(255, 0, 0);" valid=3D"tru=
e">Have a nice day! :)</span></p><br>
------=_20240201224850_72784--



--===============1474496223206438015==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
flac-dev mailing list
[email protected]
http://lists.xiph.org/mailman/listinfo/flac-dev

--===============1474496223206438015==--