Encoders: Nero 6 Digital AAC and OggdropXpd (libogg 1.1)

After some encoding tests with pop/rock and classical music, I've come to these conclusions:

Ogg vorbis offers in my opinion a better compressibility ratio for general music at around 128 kbps (q4 vs internet high), and does a good job preserving high frequencies without much coarseness with pop/rock music while AAC seems to smooth out a bit.

With classical music and midrage resolution at higher bitrates (q5 and transmision high at around 170 kbps), ACC sounds more natural to me with violins and tonal instruments while Ogg vorbis tend to sound less defined and higher pitched.

Concealing sound quality with optimal bitrate usage, I think ogg vorbis at around q4 (128 to 140 kbps) would be a good choice for pop/rock music while Nero AAC would fit better for classical music or transparency at around 170 kbps in transmission high preset.

It is noteworthy to say that, at similar bitrates Nero ACC makes larger filesizes than Ogg, what results in higher bitrate encoding in ogg to match the actual filesize, which is important in order to compare the two formats, instead of only bitrate.
That's why I prefer ogg for optimal filesize compression, because it manages to sound well compared to ACC at the same filesize. On the other hand for faithful resolution Nero AAC seems to reach transparency first at higher bitrates with the same filesize than Ogg, mostly due to resolution at midrange. It wouldn't even get better by lowpass filtering, for instance at 16 KHz and by increasing bitrate: the violins still seemed less defined than AAC. Anyway, the two codecs sound similar to me here, and OggdropXpd provides more flexibility and fine tuning than the Nero AAC plug-in encoder.

Related samples and comments: http://www.hydrogenaudio.org/forums/index....ndpost&p=295678

Note: The degree of confidence of my asessment is based on conclusions after repeated listening of several samples. May not strictly conform to a real blind test, but is noticeable for me anyway..

ABX test? TOS #8 violation? Yadda yadda yadda...................

wyup, please read terms Of Service # 8.


If not, people won't take notice about your post.

my thought exactly.

where do I get more details on ABX test, would like to better understand...thx...

http://www.hydrogenaudio.org/forums/index....showtopic=16295

dankschön Sebastian

Hello,

Are your tests based on these two samples you've uploaded, or are they based on other samples? I asked this, because I'm a bit annoyed by the tone of your conclusions (very general conclusions for -maybe- one classical sample tested).

I did a similar test (with ABX confirmations though) some months ago, and it was based on 12 different classical samples. It's probably still not enough to make strong conclusions, but if you look on separate results, you will notice that vorbis was better on four samples, but lower quality (for my taste) on average.

Therefore, if you want to have a good idea about general performance of any encoder with any musical genre, don't limit your test to one or two sample: it's clearly not enough.

Anyway, thanks for sharing impression (despite the lack of ABX test - I hope to see in the next future).

Hello,

I've done more tests with other samples, only that for this post I just included two, summing up my impressions for each bitrate. I didn't intend to do a full abx test at first, because I didn't know about TOS #8, but after other's replies I added the samples and think it is still valuable. In my opinion there's a light coarseness with ogg vorbis in tonal instruments compared to AAC, but it manages to keep high frequencies at low bitrates, which makes it sound bright with general music.