While this may be true for a few popular encoders which have screwed up Joint Stereo implementation (FHg, Xing, etc), this is not true for LAME. This was asked before many times, mostly by people who have been 'brainwashed' by some audio ripping community that joint stereo is a bad thing and simple stereo is always better. Vinyl mastering often employs the same technique, just not for the same reasons. If the M and S channels get encoded separately, high-passing the S channel (essentially removing out-of-phase information at the low frequencies) is not going to be noticed perceptually. There is no standard for how to actually implement joint stereo, so developers have a lot of leeway with this. But it is true that the joint stereo process CAN be lossless in and of itself, but there is no requirement to do so. No MP3 encoding is lossless, and those links do not state that it is. It's also explained at various places you can find via here. This is explained at the link lagerfeldt cited (I was planning to cite the same link). My understanding is that joint stereo is lossless. Adding to that, the various methods used for determining low end don't work particularly well … the joint stereo setting actually "forces" a shuffling of some low end to a mono signal. The downside is that the separation of multi-channels is performed on the mid- to high frequencies, leaving the low end essentially monaural. Or use Apple Lossless, which is much smaller (roughly half) than AIF, but with zero loss in quality. I realize the test is a little unfair because most musical material isn't panned hard that way, but still.Ĭonclusion: let Logic produce the AIF, and then let Max (or some other quality encoder, which means not iTunes) convert it to mp3 for you. Joint Stereo is widely used and is generally a good idea. WTF? I suppose this might have something to do with Joint Stereo. If I do the encoding with iTunes (at the same settings I used in Logic), there is some noise, but much less than the noise produced when the encoding is done by Logic. If I bounce to AIF instead of mp3, and then encode to mp3 with my favorite encoder ( Max), using the same settings I used in Logic, the silent channel is dead silent. Drag the file into Logic, and you can use the Sample Editor to see the noise spikes. Now audition your bounce using any method you like (Quick Look, Quicktime Player, etc). Also, realtime vs offline made no difference.) And doubling the bit rate to the maximum will reduce the problem, but the problem is still there. (Feel free to turn Normalize off, but it won't matter. Check MP3, and use all the default settings. (I set the project end marker to 3 to speed up the next step.) Select Bounce (⌘B). Pick the first item that appears ("12-8 Jazzy Drumset 01") and drag it into Arrange (I said Yes to "Import Tempo Information," but I doubt that it matters). Open the Apple Loops browser ("O"), and press the "Kits" button. Logic's mp3 encoder made it look like this:Īs you would expect from the photo, the noise is quite audible. I tried re-installing the MP3 component in the DAE folder, and this did not help.Ĭan anyone help me out here? Has anyone else had the same problem? I'm on a Mac G4 400 MHz, using ProTools LE 5.I was messing around with panning and accidentally ran into a surprise. I checked, and earlier MP3s I created do not have this problem. My sister's (a PC-based player) also comes up with the wrong time, but when it gets to the end of the wrong time, it still plays what is left. On RioPort, it gets to the end of what it thinks is the end of the song, and skips the rest and goes on to the next song. Not sure what software my sister uses, but she notices the same thing. This is true for both mine (RioPort SoundJam) and my sister's. When I play them using Quicktime player they are fine, but when I use an MP3 player, the time that the software reads as being the length of the song does not match the length of the actual song. I've been bouncing to MP3, and I've noticed something odd. Did a search for this problem and didn't find anything, so I'm creating a new post.
0 Comments
Leave a Reply. |