[vorbis-dev] Delphi Header by Matthijs Laan vs RC3 = bitrates not working!

[John Zitterkopf]

I'm confused.

I have just about complete in adding support for Vorbis to my application
CDMaster32. It uses Matthijs Laan's Delphi header library which is available
at:
http://delphi-jedi.org

Apparently the headers / code is built upon the RC2 implementation of
Vorbis.

With Matthijis' compiled libraries, I've been getting divide by zero errors
in one of the DLLs (don't remember which one). But when it worked... it
worked as expected... in that when I set a bitrate of say 160kbps; it would
encode at near that bitrate (ie 177 - 150).

Hoping to fix the / by zero problems; I downloaded the RC3 SDK from
www.vorbis.com , backed up the original DLLs, and replaced them with the
one's in the RC3 SDK.

I re-ran CDMaster32... and encoded the same files which where giving me /
zero errors. The encoded with no issues. <cool>

Then I looked at the files and they were ~10MBs... which is quite a bit
bigger than I expected.

Upon checking the files with Winamp; all of them were encoded at over
300kbps. The Winamp plugin reports that the nominal bitrate is 160kps... but
the average is over 300kbps.

So... Would the API change discussed below be affecting me?
I had assumed that the API was nearly finished and that backwards
compatiblity wouldn't be broken.
Did I assume incorrectly?

Anyone have any ideas?

John

----- Original Message -----
From: "Magnus Holmgren" <lear at algonet.se>
To: <vorbis-dev at xiph.org>
Sent: Wednesday, December 19, 2001 11:10 AM
Subject: Re: [vorbis-dev] thoughts about MP3 -> OGG convertor

<p>> Jack Moffitt wrote:
> >
> > > Small note: Did a quick check of the oggenc man file, and it seems to
> > > not match the source. Man page says quality ranges from 0 to 10, code
> > > uses 0.0 to 1.0. Though using a 0.0 to 1.0 is more flexible, it can
also
> > > confuse the user initially, if he happens to use the wrong decimal
> > > separator (as I did... :).
> >
> > Why does everyone assume the manpage is wrong :)
>
> Well, *I* didn't assume anything (though the last sentence could be
> taken as if I did, I guess). >;-)
>
> Regarding bitrate estimates: the nominal bitrate is (currently) filled
> in as follows:
>
>  0  - 64
>  1  - 80
>  2  - 96
>  3  - 112
>  4  - 128
>  5  - 160
>  6  - 192
>  7  - 224
>  8  - 256
>  9  - 320
>  10 - 498
>
> Good enough for me.
>
> Btw, did a quick test using applause.wav (available at the Lame site, I
> think; "standard" test sample, AFAIK). At quality 4, there is an easily
> heard "swashing" (a bit like maracas) sound that shouldn't be there. At
> q 5, I didn't hear anything (didn't ABX, but at least there wasn't any
> obvious fault).
>
> --
> Magnus Holmgren
>
>
> --- >8 ----
> List archives:  http://www.xiph.org/archives/
> Ogg project homepage: http://www.xiph.org/ogg/
> To unsubscribe from this list, send a message to
'vorbis-dev-request at xiph.org'
> containing only the word 'unsubscribe' in the body.  No subject is needed.
> Unsubscribe messages sent to the list will be ignored/filtered.
>
>

<p>--- >8 ----
List archives:  http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-dev-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.