[bigshot]

  no matter how many times I and others already wrote that we never thought mp3 to be as good as lossless, his answers are always telling that we do

 
For the purposes of listening to music on a home stereo (even an excellent one) with human ears (even excellent ones) MP3 320 LAME is just as good as lossless. Why are we unnecessarily equivocating just to make people who refuse to do listening tests happy? This is the truth. To human ears, lossy can be just as good as lossless for the purposes of listening to music.
 
I've been trying for over a year to find someone who can tell the difference between high bitrate lossy and redbook. If someone thinks they can hear a difference, I'd be happy for them to step up to the plate and prove it.

 

[headwhacker]

Why not put a banner or a message a the top of SS forum saying that "DBT Required" or at least a link to a study/test. Just like what the cables forum has. 

 
Heck, anyone who brings up DBT anywhere outside the SS forums gets killed right away or issued a warning from the mods/members. I remember there was a proposal to that SS should be moderated separately from other forums. This will keep threads from spiralling out of control here as it does outside the SS forum.

 

[Ruben123]

I think ss needs an own mod.

 

[castleofargh]

 

  no matter how many times I and others already wrote that we never thought mp3 to be as good as lossless, his answers are always telling that we do

 
For the purposes of listening to music on a home stereo (even an excellent one) with human ears (even excellent ones) MP3 320 LAME is just as good as lossless. Why are we unnecessarily equivocating just to make people who refuse to do listening tests happy? This is the truth. To human ears, lossy can be just as good as lossless for the purposes of listening to music.
 
I've been trying for over a year to find someone who can tell the difference between high bitrate lossy and redbook. If someone thinks they can hear a difference, I'd be happy for them to step up to the plate and prove it.

arf, now you prove that I'm wrong to say "we" by saying it's as good ^_^.
ok so "I" never said that mp3 is as good as lossless will be my official statement then

.
 
I fail badly my abx even with vbr256, and that's what I use in my DAPs. but some passages of some tracks with a headphone that actually goes passed 10khz without massive roll off(I would have to spend some real time looking for one second inside several songs most likely) do let me tell something is different. I still wouldn't bet on which one is the mp3 though. and I must be clear I might find out a difference only by using foobar's abx tool and repeating some short passages again and again. nothing like what Bookmark claims to be obvious and bad.
but even so if I wanted to prove I can do it, all I would have to do is just listen to that half a second of some weird track with a very special sound and do 20sets of the same exact part. that would be it, I would pass an abx almost without cheating. obviously it wouldn't be significant of anything, but it does stop me from claiming total transparency.
and because I can't really seem to be able to use the same trick on 16/44 vs 24/96 at normal listening levels, I do claim transparency for me with 16/44. 
 
that's the difference. I can't blame bookmark for abusing terms and facts all day long, and then myself just dismiss some ever so small ones just to win my argument.

 

[castleofargh]

  Why not put a banner or a message a the top of SS forum saying that "DBT Required" or at least a link to a study/test. Just like what the cables forum has. 

 
Heck, anyone who brings up DBT anywhere outside the SS forums gets killed right away or issued a warning from the mods/members. I remember there was a proposal to that SS should be moderated separately from other forums. This will keep threads from spiralling out of control here as it does outside the SS forum.

that wouldn't help IMO. we can fake results of almost any test if we put our mind to it. blind testing should serve to help people open their own eyes on how much our brain and other senses impact audio. not to prove stuff to others.
now making claims without any mean to back it up, that's a little different, and people should know better than to turn gut feelings into claims.

 

[KT66]

Dbt talk should be banned, like other sites do for good reason, every test is flawed in some way.

If I hear a difference in some kit or files, I hear it. I certainly don't need to prove anything to anyone else.

 

[RRod]

Dbt talk should be banned, like other sites do for good reason, every test is flawed in some way.

If I hear a difference in some kit or files, I hear it. I certainly don't need to prove anything to anyone else.

 
So every test is flawed, except the test of non-blind listening...

 

[Ruben123]

I think you're missing some sarcasm

 

[RRod]

I think you're missing some sarcasm

 
One can never tell in this crazy place ^_^

 

[krismusic]

One can never tell in this crazy place ^_^

I make you right on that! Place is full of rabbit holes.

 

[cjl]

 
arf, now you prove that I'm wrong to say "we" by saying it's as good ^_^.
ok so "I" never said that mp3 is as good as lossless will be my official statement then

.
 
I fail badly my abx even with vbr256, and that's what I use in my DAPs. but some passages of some tracks with a headphone that actually goes passed 10khz without massive roll off...

I wouldn't count on that to be the giveaway. High frequency rolloff is a good way to detect low bitrate MP3, but by the time you get up to -v0 or 320CBR, the high frequency extension of MP3 is all the way up to 19 or 20kHz, which is not audibly different from lossless. Instead, you'd have to look for specific artifacts in samples designed to be especially difficult for the particular perceptual algorithms used. It's an incredibly subtle difference, and it's only audible in a select few samples when compared side by side while knowing exactly what to look for.

 

[limpidglitch]

  If you're starting a test with the idea that it 'shouldn't be audible' the test is already flawed. I see two kinds of biases going on here . . .

 
Quite the opposite
If you do a test and assume a particular outcome to be probable when it in fact is not, you will introduce bias.

 

[bigshot]

  but even so if I wanted to prove I can do it, all I would have to do is just listen to that half a second of some weird track with a very special sound and do 20sets of the same exact part. that would be it, I would pass an abx almost without cheating.

 
What does "almost without cheating" mean?
 
I don't think there's any music track that would allow you to spot LAME or AAC 320 in an ABX. At that bitrate, they don't roll off the high end to any audible degree. I'd even like to find a piece of music that can be spotted in 256. I've found a track that clearly artifacts at 192, but I've never found anything that causes trouble at higher bitrates.

 

[castleofargh]

 

 
arf, now you prove that I'm wrong to say "we" by saying it's as good ^_^.
ok so "I" never said that mp3 is as good as lossless will be my official statement then

.
 
I fail badly my abx even with vbr256, and that's what I use in my DAPs. but some passages of some tracks with a headphone that actually goes passed 10khz without massive roll off...

I wouldn't count on that to be the giveaway. High frequency rolloff is a good way to detect low bitrate MP3, but by the time you get up to -v0 or 320CBR, the high frequency extension of MP3 is all the way up to 19 or 20kHz, which is not audibly different from lossless. Instead, you'd have to look for specific artifacts in samples designed to be especially difficult for the particular perceptual algorithms used. It's an incredibly subtle difference, and it's only audible in a select few samples when compared side by side while knowing exactly what to look for.

oh sure I was mentioning the frequencies only because most of the IEMs I use are super rolled of so that I can listen to the worst crap and still enjoy it. so for any kind of test, making sure I go over 10khz is like my routine to avoid missing too much stuff. 
I didn't even think about the cut of low rate mp3 in the high freq when I wrote my post TBH. I myself come closer and closer to not hearing above 16khz anyway.
and yeah it's clearly not often and not obvious when something does sound different, I don't know the reason, maybe something to do with some of those intersample clippings we were talking about the other day? maybe a little hard time for the psycho acoustic algo thinking something will be masked but isn't entirely? maybe the choice of the encoder itself(but I've been on lame for years now)?  I really don't know enough about mp3 and how it works to say.
 
now I feel like I'm trying to say that mp3 sucks when I noticed those stuff maybe 3 times in my life, and only when doing some tests, never actually just by listening to music.
I'll try to find an example if I can, but right now I use mp3 only on my DAPs and they all have some crossfeed+eq+replaygain+ sometimes a little de-esser, so I would need to encode some clean mp3s and abx them just to try and find a little passage. while I'm not against doing it to save the planet, it's not super high on my "how to waste my time?" wishlist ^_^.
if somebody remembers of such a little something on a song, plz save me a great deal of time and let us know.

else I'll sacrifice myself as the legitimate price for opening my mouth.

 

[Terja]

 

  If you're starting a test with the idea that it 'shouldn't be audible' the test is already flawed. I see two kinds of biases going on here . . .

 
Quite the opposite
If you do a test and assume a particular outcome to be probable when it in fact is not, you will introduce bias.

 
What I'm saying is: don't assume anything ...