[Railroader]

[I'll admit right away, I found this on digg.com]

This site mp3 or not - Don't you hear it? has two mp3 files for people to listen to and then pick the higher bit rate encoded files. The difference was very clear to me.

The digger makes this claim:

Test yourself! On my B&W 685 speakers I could only hear a little difference between the two clips, never knew that 128 kbps mp3's where so good!

I find that comment sad. Even using my cheap iPod earbuds connected to my MacBook that the two files are clearly different. The person must be nearly deaf! The 128 file sounds awful.

[Oisín]

The difference was audible, but I wouldn’t say the 128 file sounded awful. If I hadn’t had them set up right next to each other, if it had just come on as a regular song playing, I would never have noticed that the quality was sub-par.

And my hearing is perfectly fine.

[Mastrap]

The vast majority of people can't tell the difference between 128 and higher encoding. Indeed, at a recent test, the file encoded at 128 frequently won out over the higher quality files.

[Doofy]

Have to admit that on a set of Soundsticks I can't hear the difference. Since I can normally tell the difference immediately, this points to a number of explanations:

1) The original recording quality was so dire that it didn't contain any of the high frequency pointers ("air") that my ears usually use to measure such differences.

2) That I really, really, really hate opera.

3) That I was drunk when I did it.

4) That I'm getting old and should perhaps consider a career change from "music biz" to "idle rich".

[Eug]

Originally Posted by Mastrap 

The vast majority of people can't tell the difference between 128 and higher encoding. Indeed, at a recent test, the file encoded at 128 frequently won out over the higher quality files.

Using what?

Most studies I've seen have shown very consistently demonstratable differences with 128 and say 192, at least with certain material.

Indeed, after my own tests, with my own music, I decided to do all my iTunes MP3 encodes at 256. I couldn't tell the difference between 224 and CD on the stuff I tested, but went with 256 just in case. I dropped the rate to 192 for AAC though, because 192 AAC seemed basically equivalent, and I wanted to save a bit of space.

[osiris]

On my Xeon's crappy builtin speaker I heard the difference, though anything from 192+ sounds decent to me.

[- - e r i k - -]

I have Grado cans and I couldn't tell the difference. I think Doofy is on to something with the example used being not very representative for the typical quality losses one would expect at 128Kbs encoded MP3s though.

And yes, at 26 my hearing is still technically impeccable. Ie. I can still clearly hear the "stealth" mosquito ringtone sound that keeps getting bandied about as "adult proof".

[Mithras]

I reloaded the page a bunch of times, and the higher-quality clip always stayed in the same position. That's pretty lousy experimental design!

[Eug]

When I was doing my testing, I found it a lot harder to tell the difference on certain material than others.

One example is the utility of Joint Stereo. Joint Stereo is fine if there isn't that much use of stereo separation with significant stereo effects. I'd guess that describes about 98% of Top 40. However with clips which do make use of significant stereo separation, esp. if if there is movement of the audio from one side to the other, Joint Stereo with 128 Kbps MP3 sounds like you're playing it under water.

[Railroader]

Originally Posted by Eug 

Most studies I've seen have shown very consistently demonstratable differences with 128 and say 192, at least with certain material.

Indeed, after my own tests, with my own music, I decided to do all my iTunes MP3 encodes at 256. I couldn't tell the difference between 224 and CD on the stuff I tested, but went with 256 just in case. I dropped the rate to 192 for AAC though, because 192 AAC seemed basically equivalent, and I wanted to save a bit of space.

Seriously, this is EXACTLY my experience and what I do. 192 AAc or 256 MP3. Anything less is nearly unlistenable anywhere but in the car while driving.

[Railroader]

Originally Posted by - - e r i k - - 

I have Grado cans and I couldn't tell the difference. I think Doofy is on to something with the example used being not very representative for the typical quality losses one would expect at 128Kbs encoded MP3s though.

And yes, at 26 my hearing is still technically impeccable. Ie. I can still clearly hear the "stealth" mosquito ringtone sound that keeps getting bandied about as "adult proof".

I am 36 and have been to a few rock concerts a year. More than half of those in the first ten rows. And I can clearly hear the difference in the first few seconds even in the primarily vocal portions of the music. On Apple iPod Ear Buds no less.

I would NOT call your hearing "technically impeccable".

Some music obviously is wasted at higher bit rates. And I am not speaking about my music tastes, but music that is recorded with intentional distortion or recorded from a bad source. Like the Dead Milkmen. I like the Dead Milkmen's music, but they weren't exactly trying to get perfect recordings when they made their music.

[Railroader]

Originally Posted by Mithras 

I reloaded the page a bunch of times, and the higher-quality clip always stayed in the same position. That's pretty lousy experimental design!

My thoughts too. the site would be much better with rotating music clips and switching the locations around. Perhaps adding 160kbs bit rate sample as well.

[Weyland-Yutani]

I've never done this kind of test before, nor am I that particular about my music. Most of my mp3s are encoded in 128kbps and they sound fine to me.

What was surprising to me is that I heard the difference between the two sound clips - I only listened twice to each because there is nothing more godawful than *this* kind of music performed by windbags such as that - but I immediately picked up on it.

The 320kbps sample sounded better to me. Slightly clearer, but enough for me to pick up on it easily.

[Eug]

Originally Posted by Weyland-Yutani 

I've never done this kind of test before, nor am I that particular about my music. Most of my mp3s are encoded in 128kbps and they sound fine to me.

What was surprising to me is that I heard the difference between the two sound clips - I only listened twice to each because there is nothing more godawful than *this* kind of music performed by windbags such as that - but I immediately picked up on it.

The 320kbps sample sounded better to me. Slightly clearer, but enough for me to pick up on it easily.

In truth, 128 Mbps (AAC or MP3) usually sounds fine. Not necessarily exactly the same as the original, but good enough for general iPod usage and background music, especially if you don't own the original CD to compare it to in the first place. That's one of the reasons iTunes can be successful.

To put it another way, if HD DVD and Blu-ray releases are considered top notch hi-def, then 160 Mbps MP3 should be considered "hi-def" as well, because in reality it's actually pretty reasonable. It just isn't perfect (just as most HD DVD and Blu-ray releases aren't perfect either).

In my case, I chose 256 Mbps MP3 and 192 Mbps AAC cuz I was listening to them on my stereo, and was ripping everything from my original CDs and figured I may as do it (relatively) right the first time, with the caveat that I don't have unlimited space. (If I had unlimited space I would have just used WAV files.)

But for the average person who isn't necessarily as anal, 128 Mbps is fine. "Fine" just doesn't mean "transparent to the source", and I think most people, if told what to listen for, CAN hear the differences.

[Doofy]

OK, just tried again. Beyer DT150s into an MPB this time. Sober.

Nope, can't tell the difference.

So, tried some old Motley Crue. Day and night.
I conclude that my ear just isn't tuned to this kind of music.

[Jawbone54]

Definitely heard the difference when listening through my Sony in-ear headphones. Through my MacBook Pro speakers, the difference was almost negligible (but who uses those for music anyways?)

I've been encoding most of my stuff in 256 kbps AAC. It does the job, but it's not some huge issue for me. If I want quality out of my music, I listen to the CD (my favorite stuff is bought on CD; the dispensable stuff is bought online).

[Oisín]

I would NOT call your hearing "technically impeccable".

Is there actually such a thing as ‘technically impeccable’ hearing? Hearing is made up of so many different things that it seems an irrelevant goal to set.

Example: I have absolutely no trouble hearing very high sounds. When we did those tests in physics class in school and high school where an, erm, machine-thingy emitted a constant sound that increased by 500 Hz every five seconds and you took your hand down when you could no longer hear it, I was always the last person left. I can’t remember exactly, but I think I usually lost it somewhere between 20,000 and 21,000 Hz. So in that area, I have above-average hearing.

However, I’m very bad at separating various loud sources of sound. I can never hear what people are saying to be at parties, even though they’re basically screaming in my ear—all I hear is the music. And if I’m trying to talk to someone on a mobile phone, even regular traffic noise will usually almost completely drown out what the person on the other end is saying. Same goes for the iPod—to drown out traffic so that I can hear the music properly, I always need to have it near maximum volume.

There are many other factors that need to be taken into consideration for someone’s hearing to be anything that might be called ‘technically impeccable’, and I don’t think anyone’s hearing scores perfectly in all categories.

[- - e r i k - -]

I'm strongly suspecting an "emperor's new clothes" type experiment here.

[Railroader]

Originally Posted by Doofy 

OK, just tried again. Beyer DT150s into an MPB this time. Sober.

Nope, can't tell the difference.

Not even the percussion parts?

[Railroader]

Originally Posted by Jawbone54 

Definitely heard the difference when listening through my Sony in-ear headphones. Through my MacBook Pro speakers, the difference was almost negligible (but who uses those for music anyways?)

I've been encoding most of my stuff in 256 kbps AAC. It does the job, but it's not some huge issue for me. If I want quality out of my music, I listen to the CD (my favorite stuff is bought on CD; the dispensable stuff is bought online).

When I commit to a listening session, I pull out the CD as well. Mostly because I like the liner notes/lyrics in front of me.

[Eug]

For the record, I didn't actually do this test. I figure my own tests are good enough. Furthermore, I don't know what the original sounds like, as I don't know who they are (and I definitely don't own the CD).

And where I am, I only have a crapola set of speakers (built into the monitor), so it would all sound like crap anyway.

[Doofy]

Originally Posted by Oisín 

However, I’m very bad at separating various loud sources of sound. I can never hear what people are saying to be at parties, even though they’re basically screaming in my ear—all I hear is the music. And if I’m trying to talk to someone on a mobile phone, even regular traffic noise will usually almost completely drown out what the person on the other end is saying. Same goes for the iPod—to drown out traffic so that I can hear the music properly, I always need to have it near maximum volume.

Some reading for you:
Sensory Integration Dysfunction - Wikipedia, the free encyclopedia

Not saying you're dysfunctional - everyone has this "dysfunction" to a certain extent. Just thought you might be interested in it.

[Railroader]

Originally Posted by Oisín 

Is there actually such a thing as ‘technically impeccable’ hearing? Hearing is made up of so many different things that it seems an irrelevant goal to set.

Example: I have absolutely no trouble hearing very high sounds. When we did those tests in physics class in school and high school where an, erm, machine-thingy emitted a constant sound that increased by 500 Hz every five seconds and you took your hand down when you could no longer hear it, I was always the last person left. I can’t remember exactly, but I think I usually lost it somewhere between 20,000 and 21,000 Hz. So in that area, I have above-average hearing.

That's why I used the quotation marks. I agree with you that "technically impeccable" is a misnomer. A lot of animals have more acute hearing than humans, but I would call their hearing "technically impeccable". Hearing is either very good or not very good (or deaf).

Originally Posted by Oisín 

However, I’m very bad at separating various loud sources of sound. I can never hear what people are saying to be at parties, even though they’re basically screaming in my ear—all I hear is the music. And if I’m trying to talk to someone on a mobile phone, even regular traffic noise will usually almost completely drown out what the person on the other end is saying. Same goes for the iPod—to drown out traffic so that I can hear the music properly, I always need to have it near maximum volume.

There are many other factors that need to be taken into consideration for someone’s hearing to be anything that might be called ‘technically impeccable’, and I don’t think anyone’s hearing scores perfectly in all categories.

Same here (hear?). Doofy gave me the technical term for this, but alas I forgot it. Doof? [EDIT: Thanks Doofy!]

[Railroader]

Originally Posted by - - e r i k - - 

I'm strongly suspecting an "emperor's new clothes" type experiment here.

Certainly not.

[Doofy]

Originally Posted by Railroader 

Not even the percussion parts?

Nope. I keep picking the lower rate one as "best".

Maybe my subconscious interprets it as "best" because there's less Pavarotti in it.

[Oisín]

Originally Posted by Doofy 

Some reading for you:
Sensory Integration Dysfunction - Wikipedia, the free encyclopedia

Not saying you're dysfunctional - everyone has this "dysfunction" to a certain extent. Just thought you might be interested in it.

Thanks.

That article seems to imply, though, that SID usually affects several aspects of perception and has an effect on behaviour, which doesn’t seem to be the case with me. I have no trouble detecting distance or direction of sounds, for instance, I only have trouble separating loud sounds that intermingle with each other.

I don’t know, might be a mild symptom of SID, or it might not. Doesn’t matter much in my everyday life.

[Eug]

Originally Posted by Doofy 

Nope. I keep picking the lower rate one as "best".

Maybe my subconscious interprets it as "best" because there's less Pavarotti in it.

Well, I say you can't really know which is "best" unless you've heard the original. IMO, in some situations it's easier to hear what is "different" than what is "best.

[Doofy]

Originally Posted by Oisín 

That article seems to imply, though, that SID usually affects several aspects of perception and has an effect on behaviour, which doesn’t seem to be the case with me. I have no trouble detecting distance or direction of sounds, for instance, I only have trouble separating loud sounds that intermingle with each other.

Yeah, it's a crap article. Ploughing through the plethora of info out there suggests that it could be one symptom or many, affecting one or more senses.

Originally Posted by Oisín 

I don’t know, might be a mild symptom of SID, or it might not. Doesn’t matter much in my everyday life.

I've got a form of it which does affect everyday life. For example, a pub with loud music and wood floors is a hostile environment to me - what with all the reflections coming off the floor and everything hitting me at the same time I can't hear a word anyone says to me. And I have to live at least a hundred metres away from other households.

[Oisín]

In a pub, I’m usually okay, since the music isn’t so loud. In a club, though, conversation is nearly impossible. But since I set foot in a club perhaps once a year at the most, that’s not an enormous loss. The bigger problem for me is the mobile phone in traffic scenario, which is annoying. I suspect it also has something to do with my phone being old and bad at filtering out sounds, though.

And I have to live at least a hundred metres away from other households.

Yes, but this was about hearing, not your primadonna caprices.

[analogika]

Originally Posted by Doofy 

Nope. I keep picking the lower rate one as "best".

Maybe my subconscious interprets it as "best" because there's less Pavarotti in it.

This came up in that other iTunes flame-fest with Tetenal and voodoo, and while your explanation is, of course, facetious, the point it illustrates is valid:

If 320 kbps is a noticeably more accurate reproduction of the original material than 128 kbps, then it's quite possible to consistently pick the lower bitrate as "better" - or "less unpleasant", as it were.

My point in that thread was that pure A/B tests are completely worthless without a full-resolution lossless control sample - otherwise, you're just asking people a) whether they can hear a difference, and b) which they find more pleasant (not truer to the source).


I find that it's usually pretty easy to tell 128 kbps over longer passages - usually, it's hi-hats and cymbals that are completely ****ed up and artefacty - and that I consistently get an earache after about ten minutes of 128 kbps noise (though I'm sure the encoding engine - WMP, LAME, iTunes, etc. - makes a difference).

As for this particular test - the recording is terrible. Annoying, shrill, and muddy. Still managed to tell the difference correctly, though.

[Shaddim]

On my laptop speakers it wasn't obvious, so I ran the test through a cheap set of Sennheisers and it was noticeable.

[voodoo]

Originally Posted by Doofy 

Maybe my subconscious interprets it as "best" because there's less Pavarotti in it.

How true

[scaught]

I blame watching bands playing in basements with marshall halfstacks for all those years. I think my ears have been ringing to some extent or another for 15 years.

[sek929]

On a shitty Aiwa stereo connected to my computer I was able to tell when the percussion kicked in.

Pretty easy to tell, I hate the swishy washed-out sound of the highs in 128Kbps MP3s. All of my songs are 192 and higher.

[subego]

Originally Posted by - - e r i k - - 

And yes, at 26 my hearing is still technically impeccable. Ie. I can still clearly hear the "stealth" mosquito ringtone sound that keeps getting bandied about as "adult proof".

If we're talking about the same ringtone, not only can I hear it, I find it excruciatingly painful.

This is relevant because my hearing is technically for ****.