I have a source audio WAV/PCM (705kbps) with a sample rate of 22050hz that I wish to convert to MP3 with Lame3.98 encoder. Since MP3 doesnt support that sample rate I have will choose to re-encode with 44100hz.
Now, my question is, does it make sense to encode at 320kbps (vs 160kbps for example) or is that just a waste of space?
My thinking is, since mp3 is not a lossless container ANY bitrate increase would result in less quality loss. Point of debate would be whether or not that reduced loss is worth the space but more importantly; is my reasoning correct or not?
Darn good question. Can we assume a mono show and the show opens up in Audacity? You have a restricted fidelity show to begin, so some of the conventional rules don’t work right. Yes, MP3 is a destructive processor and produces permanent damage. The trick is to make the damage invisible. Also, all compressors and rules start out with the assumption that you have a technically perfect show. Compressing a compressed show rapidly produces garbage.
32 is considered the lowest you can go for a mono MP3 show before everybody can tell you did something funny to the quality. 64 for a stereo show. Audacity defaults to 128 and you can select much higher rates which will produce a nearly undamaged production. Listen to the show on your high quality sound systems or very good headphones and I bet you can’t tell the difference between the original show and 160 or higher. Further, I bet you can’t tell the difference between the original restricted quality show and much lower rates like down to the default 128 or lower.
If you step outside the standard rules, sometimes ears are the way to go.
The source file is a 16-bit stereo and has undergone several decompressions (maybe even conversions) to even come to its current PCM state. However, the ouput sounds wonderful/flawless so the damage -if any- is extremely limited. Audacity can handle it without a hitch.
I guess I’ll play it by ear, as you suggested. But because I’m a nerd ill probably end up going for max bitrate anyway since I’m already biased to believing (and probably hearing!) that’ it’s better.
Not true. MP3 does support 22050 Hz sample rate. There is no advantage in resampling to a higher sample rate - just export as MP3 and select the bit rate in the normal manner.
Yes, I tested it again and you’re right! Its’s just that certain sample rates are incompatible with certain bitrates, it seems. Since 22050-320kbps didnt work I assumed it wouldnt work for all bitrates. Thanks for pointing that out.
