As stated in this thread, we're going to have another MP3 encoder blind test.
This is how it works. I am hosting 5 WAV files, each a little over 1MB. Each of them is named after a color for anonymity. Download each one and rank them in order of preference. One of them is the original WAV (a segment of "Sun" by Live's The Distance to Here) while the other 4 were encoded to a lossy format and then decoded back to WAV. To make things more "scientific", I won't tell you what formats I used...I may have used MP3 for all 4 or I could have thrown in some WMA or I could have used all Ogg Vorbis...basically, you don't know.
For the poll above, select which of the 5 is the original WAV. Then reply to the thread, listing your preference, adding comments if desired.
e.g.
Orange
Brown
White
Purple
Grey
(Yes, I used different names above.)
blue.wav - MusePack 0.90s standard (170kbps VBR MPC)
red.wav - LAME 3.91 -b 128 -q2 (128kbps CBR MP3)
yellow.wav - LAME 3.91 --alt-preset insane (320kbps CBR MP3)
green.wav - LAME 3.91 --alt-preset standard (265kbps VBR MP3)
black.wav - Original WAV
I will reveal the true identities in a few days, so don't rush.
---
The correct answer is black.
Those who picked yellow were very close since 320kbps MP3 is almost always transparent.
Those who picked red need some listening lessons because that file sounds the worst by far (I picked it out 16/16 on ABX). You can hear the swishy artifacts throughout the clip.
I'm surprised most people did not like blue, which uses my new preferred encoder, MusePack (MPC/MP+). I think it sounds remarkably transparent given the bitrate and it has the advantage of being a time-domain encoder, not a frequency-domain encoder like LAME MP3. This means when you decode the MPC back to WAV, the filesize is the same. Notice how blue and black (the original) both are 1135KB while the other three are 1139KB? MP3 adds frames to the beginning of each file and this throws off the track-by-track timing if you reassemble them into an album, for instance. MPC is the best when it comes to preserving transients and other time-sensitive acoustical events and I suppose my ears prefer the artifact masking of MPC to MP3. MusePack has been dubbed "the audiophile format".
Anyway, I did not really expect anybody to accurately pick which one of the 5 files was the original and I personally would have picked "unable to determine". Unless you have special hearing, it is very difficult to hear any differences between the files, except for red or perhaps green.
This is how it works. I am hosting 5 WAV files, each a little over 1MB. Each of them is named after a color for anonymity. Download each one and rank them in order of preference. One of them is the original WAV (a segment of "Sun" by Live's The Distance to Here) while the other 4 were encoded to a lossy format and then decoded back to WAV. To make things more "scientific", I won't tell you what formats I used...I may have used MP3 for all 4 or I could have thrown in some WMA or I could have used all Ogg Vorbis...basically, you don't know.
For the poll above, select which of the 5 is the original WAV. Then reply to the thread, listing your preference, adding comments if desired.
e.g.
Orange
Brown
White
Purple
Grey
(Yes, I used different names above.)
blue.wav - MusePack 0.90s standard (170kbps VBR MPC)
red.wav - LAME 3.91 -b 128 -q2 (128kbps CBR MP3)
yellow.wav - LAME 3.91 --alt-preset insane (320kbps CBR MP3)
green.wav - LAME 3.91 --alt-preset standard (265kbps VBR MP3)
black.wav - Original WAV
I will reveal the true identities in a few days, so don't rush.
---
The correct answer is black.
Those who picked yellow were very close since 320kbps MP3 is almost always transparent.
Those who picked red need some listening lessons because that file sounds the worst by far (I picked it out 16/16 on ABX). You can hear the swishy artifacts throughout the clip.
I'm surprised most people did not like blue, which uses my new preferred encoder, MusePack (MPC/MP+). I think it sounds remarkably transparent given the bitrate and it has the advantage of being a time-domain encoder, not a frequency-domain encoder like LAME MP3. This means when you decode the MPC back to WAV, the filesize is the same. Notice how blue and black (the original) both are 1135KB while the other three are 1139KB? MP3 adds frames to the beginning of each file and this throws off the track-by-track timing if you reassemble them into an album, for instance. MPC is the best when it comes to preserving transients and other time-sensitive acoustical events and I suppose my ears prefer the artifact masking of MPC to MP3. MusePack has been dubbed "the audiophile format".
Anyway, I did not really expect anybody to accurately pick which one of the 5 files was the original and I personally would have picked "unable to determine". Unless you have special hearing, it is very difficult to hear any differences between the files, except for red or perhaps green.
I did think that yellow.wav sounded a little better to me. Other than that one I couldnt tell much difference between any of them really.
Facebook
Twitter