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: Vorbis and dual POKEY chiptunes (Read 4829 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Vorbis and dual POKEY chiptunes

Introduced in a separate thread, the following samples illustrate poor performance with Vorbis in stereo Atari POKEY chiptunes. The example piece is Grzegorz Kwiatek's "Alive" in SAP format. Please use ASAP if you wish to hear the source:

[attachment=3544:attachment]
In WavPack format is a five second crop from the song's introduction:

[attachment=3538:attachment]
Each of the lossy tracks are encoded from the above WavPack file and labeled with their quality; the "kbps" numbers indicate the average bitrate of the entire encoded song at two minutes twenty-nine seconds. Encodes were performed in foobar2000 v0.9.4.3 with WavPack 4.41, aoTuV Beta 5, and LAME 3.97 final, the latter two from RareWares. All tracks have ReplayGain applied.

Distortion type is most audible, naturally, with lower bitrates. I recommend first comparing the WavPack file with aoTuV q0.0. One will quickly notice the scratchy distortion.

[attachment=3539:attachment]
With the scratchy sound in mind, now listen to q4.0. Distortion is subdued but still easily identifiable. At this level I can still ABX roughly at the rate it takes me to move the cursor between buttons.

[attachment=3540:attachment]
Comparison starts to become difficult at q5.0 but not so much that I cannot reliably differentiate.

[attachment=3541:attachment]
I would deem 5.0 as minimum acceptable level for Vorbis with this style of music. For comparison, LAME -V6 produces an encoding indistinguishable from source to me.

[attachment=3542:attachment]
The Vorbis sample becomes indistinguishable to me somewhere between 5.0 and 5.9, but I suspect 6.0 is the true general transparency point.