So I downloaded the vorbis_tools_1.1.1_aoTuVb4.51 for Mac OS X from rarewares.org, but wonder if the OggEnc v1.0.2 is still the one I should use?


I'm changing a minor project from MP3 to OGG due to distribution royalty issues. Since the purpose is to code an OGG/CD-ROM now, the filesize/quality ratio would be most important in here. I don't mind if the encoder is slow, just that the quality would be transparent with the smallest possible file size.

So the question is, should I actually encode the files on another platform? Though not convenient, I might be able to pull it through if that ment better encoding result.

I'm pretty sure OggEnc 1.0.2 is the most recent version. Using a different version of OggEnc (2.83, for example) won't make any significant difference in the resulting files, unless you use a different version of Vorbis. The most important thing is that it uses aoTuV b4.51. I doubt you'll be able to get much better results with an encoder on a different platform.

Somehow the version number gave the impression of being obsolete. Ok then, I'll start experimenting with the sound files (lectures from old tapes) and looking for the suitable -q settings.

I compiled these binaries using these Vorbis-tools, libvorbis, etc. versions.
* Vorbis-tools 1.1.1
* aoTuV b4.51
* libogg 1.1.2
* libflac 1.1.2

http://www.hydrogenaudio.org/forums/index....showtopic=38920

libogg have been updated to 1.1.3, but other than that it should be up to date.
aoTuV have released pre-beta5, but the sources are not available...

Wow, Ogg Vorbis sounds clear! It has the same freshness as the MP2 had, and frankly what I was missing in the MP3 mostly all the time.

Well at 192k none of them will sound stale.

I mentioned that in the wiki, when I was helping to assemble the page.



You should just use Speex.

I tried that, but it seems that when the recording isn't in studio quality, Speex sounds occasionally very weird. Rustlings, bangs, coughs and even speaking aside the microfone makes the encoded result more eccentric than hoped for.

When recording live lectures, it's a bit different from studio. Speex seems to do wonderful job when there isn't any static.

plosives? I wonder if that's what the echo canceller is being designed for? or is designed for I should say. Jean-Marc just added it no less then a month ago.

No, not the stop consonants, but other interference coming out of the surrounding environment. The tapes aren't professionally recorded, even though they are worth making it an Ogg/CD-ROM.

I tried encoding the Ogg Speex with Max 0.6.1 which uses the speex-1.1.12 as the encoder. There is the Complexity slider, from 0 to 10, though. Should that make things better if one uses both the Quality and Complexity as a value of 10?

Yes, I am familiar with Max. I would use between 2 and 4 on the complexity scale as recommended. You could try higher settings at the cost of increased CPU usage.

I tried to encode a real studio recorded commercial audiobook for a change to see what the Ogg Speex format is capable of. The result sounds occasionally metallic with the complexity 10 setting as well. It sounds just like if the bitrate wouldn't be enough for the information to be encoded, even with its highest settings.