Ivan released version 1.61ß of his AAC Encoder.

You can grab it at my page, as usual...

Regards;

Roberto.

There is an ongoing development and tuning in 1.61 and it is not considered to be finished or stable... this version is uploaded to the temp. directory for the beta testers and core experiments and therefore it will be changed a lot in days to come...

I'll keep the version at my page as up-to-date as possible, if Ivan doesn't mind...

Regards;

Roberto.

No, of course I don't mind, but the internal beta builds might not give best possible quality! For those interested in the "internal" work, here are some news about what is going on currently:


I have uploaded latest AACEnc 1.61 (dated 12/25/01 23:51) - The purpose
of the next experiment is to verify if the following claims are true or
false:

1. Variable TMN/NMT is better solution for CBR 128 and 96 modes
2. Peak Filter chaos measure gives better results than standard euclidic
distance chaos measure

For this experiment, following switches in AACEnc are used:

-vtmn (enable variable TMN/NMT)
-cht <type> Chaos measure type, 0: euclidic distance, 1: peak filter

So, the following options are to be tested:

1. aacenc -br 128 (96!) -if input.wav -cht 0 [euclidic chaos measure,
fixed TMN/NMT]
2. aacenc -br 128 (96!) -if input.wav -cht 0 -vtmn [euclidic chaos measure
but with variable TMN/NMT]
3. aacenc -br 128 (96!) -if input.wav -cht 1 [peak filter chaos measure,
fixed TMN/NMT]
4. aacenc -br 128 (96!) -if input.wav -cht 1 -vtmn [peak filter chaos
measure but with variable TMN/NMT]