While I'm certain I'm not the first person to discover this, I stumbled upon an easy way to demonstrate the audibility of MP3 compression artifacts.
My living room system is a monophonic hi-fi console built out of an old cabinet and some salvaged Magnavox drivers. The speakers are wired in parallel with a single capacitor protecting the tweeter. The whole mess is driven with a cheap JVC stereo amplifier and the on board EQ provides the necessary mid-cut. I am running the amp in a "stereo" bridged configuration by inverting one channel and allow the L+R mix to happen in the speakers. I also happen to enjoy listening to my music in Out of Phase Stereo (OOPS), which is easily accommodated by this setup by simply bypassing the phase inverter. My digital source is a PS3, and while I usually listen to CDs, I do have some MP3s stored on the hard drive.
At any rate, when moving things around my living room this weekend, I was puzzled by some odd behavior exhibited by the console after reassembling everything. The music was present, but it sounded like it was overlaid with all sorts of digital chirps and squeaks. I realized that I was in OOPS mode, but that alone shouldn't have accounted for the bizarre sounds I was hearing. I finally realized that the difference was that, quite out of laziness, I had been using the MP3 tracks as test material and that the strange digital noises I was hearing were compression artifacts usually masked by listening in normal stereo mode. I verified this by comparing the CD tracks to the MP3 versions.
So, while lossy audio compression certainly has its place, it's indisputably audible under some (admittedly unusual) circumstances.
My living room system is a monophonic hi-fi console built out of an old cabinet and some salvaged Magnavox drivers. The speakers are wired in parallel with a single capacitor protecting the tweeter. The whole mess is driven with a cheap JVC stereo amplifier and the on board EQ provides the necessary mid-cut. I am running the amp in a "stereo" bridged configuration by inverting one channel and allow the L+R mix to happen in the speakers. I also happen to enjoy listening to my music in Out of Phase Stereo (OOPS), which is easily accommodated by this setup by simply bypassing the phase inverter. My digital source is a PS3, and while I usually listen to CDs, I do have some MP3s stored on the hard drive.
At any rate, when moving things around my living room this weekend, I was puzzled by some odd behavior exhibited by the console after reassembling everything. The music was present, but it sounded like it was overlaid with all sorts of digital chirps and squeaks. I realized that I was in OOPS mode, but that alone shouldn't have accounted for the bizarre sounds I was hearing. I finally realized that the difference was that, quite out of laziness, I had been using the MP3 tracks as test material and that the strange digital noises I was hearing were compression artifacts usually masked by listening in normal stereo mode. I verified this by comparing the CD tracks to the MP3 versions.
So, while lossy audio compression certainly has its place, it's indisputably audible under some (admittedly unusual) circumstances.
Try your test again.
First set the Polarity/Phase switch correctly.
Adjust the volume to an appropriate level.
Then reverse the Polarity/Phase switch, but don't touch the volume control and listen again.
Please report back on your findings.
First set the Polarity/Phase switch correctly.
Adjust the volume to an appropriate level.
Then reverse the Polarity/Phase switch, but don't touch the volume control and listen again.
Please report back on your findings.
I listen to 320kbit MP3 in the car only. It drives me nuts everywhere else. So can I hear the loss in fidelity?
Oh yes.
But as stated in the right environment where there are a great many other problems. It is convenient.
Oh yes.
But as stated in the right environment where there are a great many other problems. It is convenient.
The only thing I think that test could show is that there is a difference between channels. And as speedy mentions, it could be due to level only.
Does the same track in loseless do the same thing?
Does the same track in loseless do the same thing?
It is proably because the stereo channels are "joined" on frame to frame basis during encoding and he is listening mono?
MP3 - Wikipedia, the free encyclopedia
MP3 - Wikipedia, the free encyclopedia
My point was probably obscured by the needlessly verbose description of my system, but what it comes down to is that the fidelity of digitally compressed files (at least those encoded at standard bit rates) breaks down when listened to in this unconventional manner.
When neither of the two stereo inputs inverted, my setup essentially becomes a big differential amplifier. This feature is intentional, since the Out of Phase Stereo effect cancels any content mixed down the middle and emphasizes anything mixed farther to the left or right. It is just a fun way of listening to familiar recordings in a different way (note to Hendrix fans: the acapella version of Foxy Lady is hilariously lascivious). When one of the inputs is inverted, it becomes a normal summing amplifier and everything plays normally, just in mono (digitally compressed files sound perfectly acceptable this way).
To address one of the responses, no changes were made to anything but the source material (digitally compressed file vs. same song direct from CD). I'm assuming the standard PS3 encoding method is 128 kbps MP3, but I could be mistaken.
At any rate, no matter how you accomplish the OOPS effect (at least within the analog domain) I would expect the result to be the reproducible and not at all subtle. I realize this doesn't in any way impact normal listeners, but I find it interesting that such a simple test can expose the ugly underpinnings of the digital compression process.
When neither of the two stereo inputs inverted, my setup essentially becomes a big differential amplifier. This feature is intentional, since the Out of Phase Stereo effect cancels any content mixed down the middle and emphasizes anything mixed farther to the left or right. It is just a fun way of listening to familiar recordings in a different way (note to Hendrix fans: the acapella version of Foxy Lady is hilariously lascivious). When one of the inputs is inverted, it becomes a normal summing amplifier and everything plays normally, just in mono (digitally compressed files sound perfectly acceptable this way).
To address one of the responses, no changes were made to anything but the source material (digitally compressed file vs. same song direct from CD). I'm assuming the standard PS3 encoding method is 128 kbps MP3, but I could be mistaken.
At any rate, no matter how you accomplish the OOPS effect (at least within the analog domain) I would expect the result to be the reproducible and not at all subtle. I realize this doesn't in any way impact normal listeners, but I find it interesting that such a simple test can expose the ugly underpinnings of the digital compression process.
No, but inverted mono isn't too pretty.
As a test I took a random stereo file and inverted the left channel, then mixed it to mono. Not a lot there, but you can hear the resulting difference - it's basically what a Dolby decoder would pull out for the rears.
I took the same file, saved it to MP3 @ 160Kbs. Opened it, inverted the left channel and mixed to mono. About the same resulting level, and no weird squeaks or pops or anything. However, the sloppy, slushy, phasey MP3 artifacts are easy to hear. Easier than in the non-altered MP3 version.
So it may depend on the file, the encoder and who knows what else. I can hear the artifacts in the inverted MP3, but no chirps or squeaks.
EDIT: Also tried it at 128Kb joint stereo. The results are noticeably worse.
well of course if you cancel the music that is masking the artifacts then they are audible - the point of lossy preceptual coding is that you can rely on masking
- Status
- Not open for further replies.