Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Psytel AACEnc 1.5 (Read 3232 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Psytel AACEnc 1.5

Psytel AACEnc 1.5 (stable) has been released.
Improvements over v1.2:

- Much improved CBR mode (128 Kbits/s or less!)
- New ATH formula (much better sensitivity)
- Improved VBR modes, bugfixes on some critical items
- Significantly better 128 Kbps mode
- Implemented window grouping
- Almost all coding tools are checked and improved.
- Noiseless coding module (lossless step) is also improved

http://www.inf.ufpr.br/~rja00/
Juha Laaksonheimo

Psytel AACEnc 1.5

Reply #2 –
Ivan had some additional comments he emailed to me:

Quote
In general, biggest improvement is elimination of "dropouts" or "holes", like one in bassdrum.wav - also, weird effects like that on start of "peaceful.wav" are also completely eliminated.

New ATH is also MUCH better than in any older version!

-normal and -extreme VBR modes also work very well, in some cases with smaller bitrates than V1.2 but with higher quality.



I did some quick tests on a few test samples using the 128 kbit/s setting he recommends (-br 128 -qual 9 -production).  This encodes very slowly, by the way.

castanets:  a little bit of a flange on the initial guitar, and the slow castanets sound like they are a bit more reverberant than the original (ABX 15/16).  Liquid Audio streaming 128 is still the one to beat on this sample.  Psytel 1.5 is definitely not bad -- at 128, it has less smear than Dibrom's rev 8a --alt-preset standard, for example.  Ogg pre-RC3 at 128 (10/26/01 compile) sounds kind of noisy on the initial guitar strum and not clear (double hits?) on the slow castanets, so yours is very competitive.  On castanets, Psytel definitely sounds better than at least two of the commercial aac encoder samples you sent me.

41_30sec:  just barely noticeable smearing on the intitial fast cymbal hits (ABX 15/16).  Otherwise, very good overall, at least to my ears.

rach_original:  can't tell the difference

wayitis:  glitch at very beginning.  After trimming the first 35 milliseconds at the beginning, I can't tell the difference from original.  Clearly better than Liquid Audio on this sample, which surprisingly makes some of the keyboard notes sound thick.

ff123