Do you make music transcription? Some audio processing plugins manipulate the phase of stereo signals to emphasize or minimize some instruments panned on a specific position to make transcription easier, the most common one is the so-called "vocal removal" technique which assumes vocal is mono and centered on the stereo field.https://youtu.be/WvqcrAWflJkWith mp3 or some other lossy codecs such techniques could yield poorer performance.
Do you make music transcription?
(To my ears, LAME V0 does sound identical to the original WAVs, but I haven't done enough testing to put any confidence in my own conclusions.)
Update:Sorry guys, I really should have done this before posting the thread, but…I have just done some blind testing (for the first time in many years) and I am horrified to discover that my ears can actually hear the difference between LAME V0 and WAV! To me, the mp3s sound like they're missing high frequencies, and thus they sound slightly duller (and narrower). In contrast, the WAV sounds less like a recording and more like "real life".
Did you ABX a lot of files within two hours and suddenly noticed the nontransparency?
BTW which version of Lame are you using, what are your exact encoding parameters, and what is your age?
Do you mind sharing a 30sec. snippet of that track with us? An ABX protocol is also welcome.
You must understand that to most people here it sounds strange when someone claims that he can ABX a track @V0 which is not known in advance to be difficult to encode. Especially as you aren't talking about artefacts but about missing highs (which is the usual reasoning of people judging from spectograms). ABXing successfully an ad hoc track can happen, but more evidence is extremely welcome.
For an alternative to mp3 try opus --bitrate 192 (or more) and/or lossyWAV+FLAC at a ~400 kbps (avg.) setting.
I am more interested to hear some of your transcriptions. It is not totally off-topic. By hearing your transcriptions (not rearrangement or improvisation) and compare with the original piece, it is possible to evaluate your ability to hear individual components of a mix.
As for ABX protocols, I don't know what that is so you'll have to explain to me how I can do a proper test? The method I used wasn't particularly scientific:What I did was I opened both the original WAV and the mp3 in Audacity, renamed the two tracks so that the origin of each track was hidden, jumbled up the two tracks a lot (not always knowing if I had dragged it properly or not). Then I simply played excerpts on loop whilst toggling the "solo" button to switch between the two versions. Certain excerpts sounded the same, but in most cases one would sound clearer, and the one that sounded clearer was always the same track, which turned out to be the original WAV. I did this with a few tracks and the same happened in every case.
Thanks for your interest but I don't tend to share my cover versions, I only make them for my own amusement. I sometimes think about sharing them on YouTube, but I'm in no rush to. And they are not supposed to be 100% perfect copies of the original, usually I am seeking to retain the original music but with subtle improvements and embellishments, especially to the production.
In the meantime, may I ask: are there people who are known to be able to genuinely hear the difference? Or would I be the first - therefore it's much more likely I have tested wrong?
Because the probability that you could get the same result or better by flipping a coin is below 1 in 20 (your score, for lack of a better term, was 3.8%) this is considered a passed test.
I would like to know whether or not you can hear a difference also with my lame3995o variant using -V0 as well as -Q0.5 (same average bitrate). Can you please try it?
Do you mind sharing the critical spot with us?
The track I tested was Shola Ama - Much LoveOutput:DS : Primary Sound Driver
Unfortunatly this makes things more complicated for the OP.I guess also with the straightforward procedure he will have a hard time finding something he can ABX.In case he should really find someting we can have a look at the provided sample and, if the track peak value should give rise to the suspect that the overall playback chain will change volume, we can ask him to manually edit track replaygain values of the original (converted to FLAC) and the mp3 file and ask him to redo the ABXing using replaygain. We can help him to any amount with that in case it should be necessary.
the WAV sounds slightly more "wholesome", "natural" and "old fashioned". It seems "fuller" and "richer" and has slightly more of a "crackle" to it [snip] the WAV can even sound a little louder at times (I have tested the files in question and there is absolutely no difference in gain.)