Anybody Out There Who Hears a Difference Between 320 kbps MP3 and Red Book CD? What Differences Do You Hear?

[fpitas]

Yep.

@fpitas what to you think? Does that warrant a whac a mole?

Let's give him a few more posts to see if he needs that treatment

 

[HarmonicTHD]

Let's give him a few more posts to see if he needs that treatment

You are cruel. Playing with the prey

 

[fpitas]

You are cruel. Playing with the prey

*evil cackle, rubs hands together in anticipation*

 

[ads_cft222]

If that is the ethos , and you were running a business it would be bankrupt

 

[antcollinet]

If that is the ethos , and you were running a business it would be bankrupt

Your post seems to lack context: you might want to quote who you are repling to. Hit the reply button rather then simply adding a new post.

 

[paudio]

You should also do the tests in my signature.. Remove the doubt.

I tried 5 trial 320 lame test and got....

Hrm

You probably can't hear the difference between the lossy and lossless samples (p >= 0.10)

You got 56% correct

There is a 69% likelihood of getting this or a more extreme score by chance

Track / Correct / p-value
The Killers 40% (p >= 0.020)
James Blake 80% (p >= 0.020)
Daft Punk 80% (p >= 0.020)
The Eagles 40% (p >= 0.020)
Dixie Chicks 40% (p >= 0.020)

Edit: Thought there might be some small differences on James Blake track with some decay on the percussion but I still got it wrong once apparently. Maybe I will try again on my studio monitors or closed back planars another day. It's about what I expected though considering past tests.

 

[antcollinet]

I tried 5 trial 320 lame test and got....


Edit: Thought there might be some small differences on James Blake track with some decay on the percussion but I still got it wrong once apparently. Maybe I will try again on my studio monitors or closed back planars another day. It's about what I expected though considering past tests.

To get a significant result, you need at least 10 tests with 9/10 correct. Even then, there is a 5% chance of getting that by guesswork.

 

[Blumlein 88]

1st off MP3 was never meant to be fully transparent. In fact during testing it was not. The testing was to manage small files with the least degradation. I think if you want to find out if you hear a difference start with 64 kbps MP3. Then go to 128 kbps. So on and so forth. See using Foobar where you no longer hear a difference. 320 kbps is pretty hard to hear, and may not be heard with most music. And don't forget the original MP3 algorithm has improved from the initial version.

And really, 10 is not enough trials. I wouldn't put much stock in anything less than 20, and really 30 trials. You don't have to do them all in one session. Of course more than 30 or two sessions of 15 gets very tedious. If it takes more than that to show a difference it was a damned small difference anyway.

 

[paudio]

To get a significant result, you need at least 10 tests with 9/10 correct. Even then, there is a 5% chance of getting that by guesswork.

1st off MP3 was never meant to be fully transparent. In fact during testing it was not. The testing was to manage small files with the least degradation. I think if you want to find out if you hear a difference start with 64 kbps MP3. Then go to 128 kbps. So on and so forth. See using Foobar where you no longer hear a difference. 320 kbps is pretty hard to hear, and may not be heard with most music. And don't forget the original MP3 algorithm has improved from the initial version.

And really, 10 is not enough trials. I wouldn't put much stock in anything less than 20, and really 30 trials. You don't have to do them all in one session. Of course more than 30 or two sessions of 15 gets very tedious. If it takes more than that to show a difference it was a damned small difference anyway.

If I can't tell clearly enough in 5 there is no need for more imho. At this point both versions were enjoyable even if I thought I heard some sort of difference. Tests were done on my Surface Pro 8 3.5mm out to a pair of Sennheiser PX100-II. I was falling asleep trying to pick them out.

 

[Blumlein 88]

If I can't tell clearly enough in 5 there is no need for more imho. At this point both versions were enjoyable even if I thought I heard some sort of difference. Tests were done on my Surface Pro 8 3.5mm out to a pair of Sennheiser PX100-II. I was falling asleep trying to pick them out.

The need for more is to make it statistically meaningful to convince others. And to show you aren't fooling yourself or just getting lucky.

 

[ahofer]

If that is the ethos , and you were running a business it would be bankrupt

Don't know what you are replying to, but most of us when we are engaged here are aiming to be skeptical and educated consumers, not running a business. Those who are running a business are obliged to wear the 'vendor' tag so that we can weigh their remarks accordingly.

In my business I have to politely engage some clients' fantasies and then work towards achieving a solution in their interests. I have no interest in engaging fantasies (other than my own, I suppose) in my spare time.

 

[ads_cft222]

Don't know what you are replying to, but most of us when we are engaged here are aiming to be skeptical and educated consumers, not running a business. Those who are running a business are obliged to wear the 'vendor' tag so that we can weigh their remarks accordingly.

In my business I have to politely engage some clients' fantasies and then work towards achieving a solution in their interests. I have no interest in engaging fantasies (other than my own, I suppose) in my spare time.

I am not referring to you, just some members where I specifically called out for and think they will get the medal of honour if they disrespect in herd like behavior which is totally incompatible with the attitude of many knowledgeable members here .

I am sceptical also and I acknowledge the limitations of hearing perception. I am not against the arguments I am against the totally unnecessary parroting and scolding without discrimination in the name of a misplaced perception of ASR gatekeeping.

If you want to enlighten and educate you do so with arguments and just that . Eventually reality wins at least for those open to it . The other behaviors achieve the entirely opposite result . As If you had a business you would lose all your customers if you would be disrespectful to them…

 

[Julf]

If you want to enlighten and educate you do so with arguments and just that . Eventually reality wins at least for those open to it . The other behaviors achieve the entirely opposite result . As If you had a business you would lose all your customers if you would be disrespectful to them…

Unfortunately a lot of people find pointing out facts disrespectful.

 

[Zensō]

Unfortunately a lot of people find pointing out facts disrespectful.

Yup. And others have no interest in examining their own preconceived notions.

 

[ahofer]

I suppose another bit of context is that many of us are refugees from sites like Audiogon (in my case) where bringing up controlled testing results in a massive pile-on and magical thinking is applauded. Rightly or wrongly, we tend to view this as one of a few places, or the only place, where unscientific audio woo is kept at bay by the majority of users. We like it that way.

So we aren’t always trying to convince people so much as keep the forum the way we prefer it. Slightly different objective. Convincing audiophiles that they are ignoring science easily accessible to them, at least in the moment, is almost impossible anyway. (See the Shaw quote in my signature)

 

[Ken Tajalli]

MP3 is donkeys years old. Why would anyone still use it.
But if you are asking about audible differences between a knowingly inferior codec and the redbook, then depending on material, it could be hard to tell.
On complex orchestral pieces that were well recorded, then perhaps one could tell, not easily.
Days of MP3, AAC and the like are over. rest in peace.

 

[Zensō]

MP3 is donkeys years old. Why would anyone still use it.
But if you are asking about audible differences between a knowingly inferior codec and the redbook, then depending on material, it could be hard to tell.
On complex orchestral pieces that were well recorded, then perhaps one could tell, not easily.
Days of MP3, AAC and the like are over. rest in peace.

A very large majority of people are still listening to either AAC or Ogg Vorbis through the big streaming services.

 

[sq225917]

Just try the clapping one, find a track, encode it at various rates with different encoders, burn it to disc with random mix of clips and see how you do.

I've tried it many times, level matched, blind, with some else controlling track choice.

It's easy, the mp3 versions sound like you're underwater, totally amorphous, sounds more like noise than distinct multiple claps.

It's pretty much a pathological test for most encoders.

 

[Ken Tajalli]

A very large majority of people are still listening to either AAC or Ogg Vorbis through the big streaming services.

I hadn't considered lowfi streaming services.
My bad.
But why? I suppose, when you are listening on your phone's speaker, or wireless buds, you can't be choosy.

 

[danadam]

It's easy, the mp3 versions sound like you're underwater, totally amorphous, sounds more like noise than distinct multiple claps.

Here's 10s of applause compressed with different lame presets: https://mega.nz/folder/N1ATCA7b#7-LD4wjieN121UvrTZfDvg

I'm not sure I'd ABX that at 192 kbps and above.

So am I deaf or do I have a poor gear or is the clapping I chose not good enough to show the differences?