Maybe this is true. I thought B and Y were the lossless ones, was wrong in the case of Y, but thought differences between X and Y easier to spot than those between A and B. What made me think Y was lossless, mainly was better clarity at crescendos - and yes, I listened through the motherboard's sound system and a pair of cheap headphones
It has happened before!!
A couple of years back, there was a similar test on a Greek forum: A single music file was recorded in many different bitrates and people were called to spot the .wav. Guess what, most preferred a 256Kbs MP3! (or was it a 320Kbs?)
I did not pay much attention then, but now maybe its worth the fuss to engage my good sytem for some serious listening...
It's certainly made me think. The good point is that, although I was happy with the choice it has confirmed that listening to VBR Mp3s at work and on the go is a sensible choice. But it's also made me realise that I have to spend a bit more time investigating something I noticed recently. My work playback system, using etymotic er4p with comply tips is pretty ideal for my needs (good music, block out the shouty man who sets next to me), and having eq'd to be close to the Harman curve it relaxing and pleasing to listen to for hours on end.
However, a couple of recent releases on the excellent Dutton label got me worried as that label does superlative recordings but these sounded distant and dull. Confused I changed the wax protectors, but wasn't that. It was only when I ran the foobar DR plugin that I got a pointer. The DR on the CDs I have was 17 and 19 respectively, which is a lot higher than any of the other music I have tested. This made me realise I was just listening at too low a level. Cranking up a few notches and things were a lot better, but still not as expected. I have not tried these CDs on speakers yet to compare but the differences between speakers and headphones in terms of perception is greater that I had expected.
I now wish I had paid more attention to the articles the late John Crabbe wrote on this subject. At the time I was young and stupid and wrote him off as old guard and not relevant. hey ho.
Picked up a cold over the weekend so critical listening is pointless, but when it clears will listen some more.
However, a couple of recent releases on the excellent Dutton label got me worried as that label does superlative recordings but these sounded distant and dull. Confused I changed the wax protectors, but wasn't that. It was only when I ran the foobar DR plugin that I got a pointer. The DR on the CDs I have was 17 and 19 respectively, which is a lot higher than any of the other music I have tested. This made me realise I was just listening at too low a level. Cranking up a few notches and things were a lot better, but still not as expected. I have not tried these CDs on speakers yet to compare but the differences between speakers and headphones in terms of perception is greater that I had expected.
I now wish I had paid more attention to the articles the late John Crabbe wrote on this subject. At the time I was young and stupid and wrote him off as old guard and not relevant. hey ho.
Picked up a cold over the weekend so critical listening is pointless, but when it clears will listen some more.
I was listening with headphones.
Needing to vote for one only, I was thinking that probably 3 of the files were compressed.
I had the impression that each of the recordings had a certain amount of digital hash. File A seemed very bad, but B had enough hash so that I thought it was more compressed than file Y, although better than A.
Maybe a little 256 mp3 compression can reduce the hash from a wave file to give it a more comfortable sound.
I listened again after I knew the results, and I did find more high frequency material in X which I interpreted as digital hash, but I could not convince myself the X sounded better than Y.
I am looking forward to more comparisons with 16 bit vs 24 bit. Have not had much chance to do this yet.
Thanks, I enjoyed this little test. Really got me thinking critically.
In the UK now there is so much classical music on Radio 3 website at 320 mp3 that sounds better than the B file in this test. I am thoroughly enjoying this.
The Radio 3 sound is frequently better than many of my CD's,
I rate 192 mp3 as intolerable, 320 mp3 as tolerable. But ultimately my musical enjoyment, it is the recording and the quality of the artists that matter the most.
Needing to vote for one only, I was thinking that probably 3 of the files were compressed.
I had the impression that each of the recordings had a certain amount of digital hash. File A seemed very bad, but B had enough hash so that I thought it was more compressed than file Y, although better than A.
Maybe a little 256 mp3 compression can reduce the hash from a wave file to give it a more comfortable sound.
I listened again after I knew the results, and I did find more high frequency material in X which I interpreted as digital hash, but I could not convince myself the X sounded better than Y.
I am looking forward to more comparisons with 16 bit vs 24 bit. Have not had much chance to do this yet.
Thanks, I enjoyed this little test. Really got me thinking critically.
In the UK now there is so much classical music on Radio 3 website at 320 mp3 that sounds better than the B file in this test. I am thoroughly enjoying this.
The Radio 3 sound is frequently better than many of my CD's,
I rate 192 mp3 as intolerable, 320 mp3 as tolerable. But ultimately my musical enjoyment, it is the recording and the quality of the artists that matter the most.
IME people are prejudiced against MP3 because the low bitrates DO sound bad. Listening to low birate jazz streaming can be awful, with all the phase noise on the splashy cymbals. A lot of cymbals in jazz. But higher bitrates really aren't bad. To my ears, they just tend to erase ambiance details, making the recording more forward or dry. A lot of people like that.
A few yeas ago I was at a DIY meet in Canada. We were standing around listening to some Open Baffle speakers. All were praising the sound, until it was mentioned the tracks were MP3. That brought great guffaws, grumbles and "I hate MP3" comments. Everyone but me and the guy with the MP3 player wondered off. "That's funny" he said, "they loved it until they knew it was MP3".
A few yeas ago I was at a DIY meet in Canada. We were standing around listening to some Open Baffle speakers. All were praising the sound, until it was mentioned the tracks were MP3. That brought great guffaws, grumbles and "I hate MP3" comments. Everyone but me and the guy with the MP3 player wondered off. "That's funny" he said, "they loved it until they knew it was MP3".
Lol, I can well believe the comments Pano. Ignorance is bliss... everyone is happy until they know.
What I am trying to understand is what we have been comparing and how the different equipment (CD player/computer/dac) are processing the files.
It seems to me that we have been evaluating audacity's ability to convert formats and not the playback of the different formats.
It seems to me that we have been evaluating audacity's ability to convert formats and not the playback of the different formats.
why do you think that? The test was to see if people could tell mp3 from wav and if 192kbit/s was significantly worse than 256. It seems most (just not me at work) could tell a difference, just the wrong way round. I for one find this interesting, esp based on experience. I need to do some more open minded tests now.
audacity is likely to use a std codec so it is comparing mp3 with wav, not audacity with wav.
audacity is likely to use a std codec so it is comparing mp3 with wav, not audacity with wav.
Non of the files are mp3. They are wav or wav converted from mp3. Is there no difference in the way a playback system handles the different file formats?
File B.wav is 13667 kB at 1411 kbps and after I converted it to mp3 it became 1861 kB at 192 kbps.
File B.wav is 13667 kB at 1411 kbps and after I converted it to mp3 it became 1861 kB at 192 kbps.
Last edited:
Just had to post this "The Ghost in the MP3": Making Art With MP3's Detritus | AudioStream art made from thrown away bits. daft, but fun.
Hi Pano.
I have been with a DIY group comparing two different systems.
The group decided that one of them was better than the other. I disagreed. I thought the less favoured one sounded more like the original live performance.
I was told that comparing a system against a live performance was too harsh.
It appears that the ultimate criteria of the goodness of a system is other systems.
I have been with a DIY group comparing two different systems.
The group decided that one of them was better than the other. I disagreed. I thought the less favoured one sounded more like the original live performance.
I was told that comparing a system against a live performance was too harsh.
It appears that the ultimate criteria of the goodness of a system is other systems.
The one I have is from the original CD that shot Suzanne Vega to stardom back in the 80's. Haven't played it for many years tbh but it did come to mind as a test track.
Lol, no 🙂
I'll be digging the CD out next to refresh my memory of it. It used to be one of my favourites, nowadays its more the likes of Beethoven and Brendel.
I'll be digging the CD out next to refresh my memory of it. It used to be one of my favourites, nowadays its more the likes of Beethoven and Brendel.
I didn't vote and I didn't compare the second two. My hearing is pretty jacked up. I've been to some loud concerts in the 70's, My own bands in the 80's and I have constant allergies. My topend stops at 13.5khz. Yet I could hear the soundstaging collapse when going from A to B.Firstly, thanks to all who listened and voted. The poll shows a good mix so perhaps differences weren't so obvious after all, or ,they were and you voted thinking your favourite has to be the WAV. That's something only you know 🙂
So then...
A = 44.1kHz MP3 at 192 kbs
B = 44.1kHz WAV Lossless
X = 44.1kHz WAV Lossless
Y = 44.1kHz MP3 at 256 kbs
So X and Y had the distinction of using what we might consider a half respectable bit rate.
A = 44.1kHz MP3 at 192 kbs
B = 44.1kHz WAV Lossless
My overriding impression is just how good these compressed files can be (honestly).
Tubelabs comments earlier were I think what make these tests useful... not knowing what you are listening too.
The 'percussive triangle' is exactly where the limitations of compression should show... and it looks like you have heard something along those lines. X should have the higher overall hf energy content according to the way compression seems to work, and from what you guys have shown with FFT analysis.
So a bit of swings and roundabouts going with these. The direct rip (X) seems to stretch the system more and I'm inferring that Y allowed you to listen at higher level without things getting subjectively uncomfortable.
A vs B you would think would have more glaring compromises, perhaps the musical content has a lot to do with how we perceive the overall result.
Very very interesting. Thanks again to all that listened.
http://www.diyaudio.com/forums/loun...ch-preamplifier-part-ii-6154.html#post4221703
Thanks for giving them a listen. Identifiable differences yes, but the vote (admittedly a small sample) seemed to show a preference for the compressed.
(I'll leave the files up until the end of this month in case anyone else is curious)
(I'll leave the files up until the end of this month in case anyone else is curious)
Why? on another thread was suggested no evidence anyone can hear 24 bit vs 16 bit? Or are you suggesting we try and test that?
I'm surprised that nobody has brought up the 'nice to listen to' vs 'accurate reproduction' debate in response to this.
Clearly, we aren't comparing re-production systems, so accuracy of reproduction isn't compared in that way, but we are seeing that people prefer a little less dynamics and treble... could be why those $XXXX audiophile devices that turn out to be tiny inductors before the speaker were so popular ....
Clearly, we aren't comparing re-production systems, so accuracy of reproduction isn't compared in that way, but we are seeing that people prefer a little less dynamics and treble... could be why those $XXXX audiophile devices that turn out to be tiny inductors before the speaker were so popular ....
- Status
- Not open for further replies.