Testing audiophile claims and myths

[bigshot]

a case of chronic “Smartest Man In The Room” syndrome when discussing audio

Not that uncommon around here!

His concept of audible is numbers he circles in yellow highlighter taken out of context from a book.

I think it might be wise to take this to PM. It's already spawned another thread, and if we don't watch out, we'll summon the demon and he will be all over us again. For me, the matter is settled. I don't need any more of that kind of stuff.

 

[Blackwoof]

--

 

[magicscreen]

Take the .wav files, create the mp3 files from them yourself, and to be absolutely sure that the .wav and .mp3 are not treated differently somehow somewhere in the playback system: convert the mp3 back to the same wav format as the original wav, level matched, and then compare the 2 wav versions in a blind abx test...

Thank you!
But I do not need this test. I can hear the differences between mp3 and flac files.
Wav files sound better than 16bit/44kHz flac files.
But the best are the 24 bit/192kHz files.
It is unfortunate there is no higher resolution files than that. Maybe in the future if we had enough storage space.

 

[sander99]

Thank you!
But I do not need this test. I can hear the differences between mp3 and flac files.
Wav files sound better than 16bit/44kHz flac files.
But the best are the 24 bit/192kHz files.
It is unfortunate there is no higher resolution files than that. Maybe in the future if we had enough storage space.

Then you definitely need to do the test. Do you know and understand that you can hear differences also when there is no audible difference? Your brain does most of the hearing, not the ears.
By the way: wav files and flac files do not sound at all, but they can be used to make a DAC produce an output signal. If the flac is decompressed you get the exact same bits as are in the wav. If the same bits are sent to the DAC (with the same timing) there can be no difference. Imagine the following playback system: first decompress the flac to the wav (that obviously is identical to the original wav, assuming no mistakes were made and neglecting meta data that could be different), then play the wav. How could this sound different from playing the original wav? So what I am saying: if (big if) there was audible difference between playing a wav or a flac then something is done wrong.

If you are able to hear the differences that you claim based on the audio alone, without knowing what you are hearing, ergo in a well controlled double blind ABX test, then you would probably be the first human being in history. Then you should go to a university to have yourself properly tested. You would become famous and probably there are rewards for you to be claimed.

 

[hakunamakaka]

mp3 quality can vary a lot. I could hear a difference between good 320kbs and the bad one. I could even link you with the track where spotify 320kbs LP rip sounds worse than youtubes 128kbs ripped and uploaded by user

 

[sander99]

I meant at least 320kbs good quality encoded. And of course the ridiculous claims about wav versus flac, and "even higher resolution" versus 192/24 versus 44.1/16.

 

[bigshot]

But I do not need this test. I can hear the differences between mp3 and flac files.

Sound Science is the one forum in Head-Fi where we get to say, "Do a controlled test or we can say you're wrong." And guess what? If you say you don't need to do the test, I get to say you don't know what you're talking about. You are a walking, talking poster child for expectation bias and placebo effect.

I could hear a difference between good 320kbs and the bad one. I could even link you with the track where spotify 320kbs LP rip sounds worse than youtubes 128kbs ripped and uploaded by user

Have you got a controlled test to back that one up?

 

[Bozzunter]

Since this thread has partially degraded to things like recognizing mp3 vs wav, I was wondering if it’s possible to discuss headphones.

I mean, I’m not sure it’s possible to use a strictly scientific approach as with other topics, but I’d love to hear some opinions here. Namely, I purchased more and more expensive headphones (from Sennheiser 650 to 800s, Focal Elegia, Audeze LCD i4) and I would love to read about the subject outside the dedicated forums which reach a level of auditive hallucinations to be even funny. Well, not really.

 

[castleofargh]

Thank you!
But I do not need this test. I can hear the differences between mp3 and flac files.
Wav files sound better than 16bit/44kHz flac files.
But the best are the 24 bit/192kHz files.
It is unfortunate there is no higher resolution files than that. Maybe in the future if we had enough storage space.

Can't tell if you're trolling us or not...


People with a strong belief are always confident. But some will be right and some will be wrong about their beliefs, so why would we accept your self confidence as only measure of truth? Doesn't seem like a very good idea.
You're obviously free to do(not do) what you want, and believe whatever you wish to believe. But don't expect others to trust your blanket statements under those circumstances. What is freely asserted is freely dismissed.

 

[bigshot]

I was wondering if it’s possible to discuss headphones. I mean, I’m not sure it’s possible to use a strictly scientific approach as with other topics, but I’d love to hear some opinions here.

Transducers are always the wild card because they are mechanical, and are subject to the limitations of the physical world; and because they are greatly affected by external factors- the shape of your ear canals, your room, etc. There is no one-size-fits-all solution to transducers. All you can say for sure is, "It depends." Since solid state electronics and digital audio are so precise, it's safe to use them as a baseline and adjust to suit your particular transducers and taste using EQ. As long as a set of cans can reproduce a full spectrum of sound loud without distortion, you should be able to massage the EQ to sound like anything you'd probably want them to sound like. That means that a decent mid range set of cans with EQ can sound as good as expensive ones. Start by approximating the Harman curve (see the other thread for info on that) and then adjust to personal taste. There is no absolute hard and fast calibration in home audio, like there is in pro audio. Consistency isn't an issue, pleasing yourself is. Feel free to add salt and pepper to taste. There's a point where science ends and personal preference begins. But you can approach determining your preferences scientifically, by experimenting, analyzing results and keeping track of what works and what doesn't.

 

[PhonoPhi]

...What is freely asserted is freely dismissed.

Is it also applicable to people who claim that 128-160 mp3 are indistinguishable from CD-quality flacs?

 

[bigshot]

Is it also applicable to people who claim that 128-160 mp3 are indistinguishable from CD-quality flacs?

Sure! The way to do it is to point to published studies or do a controlled test yourself. The way to counter a statement that isn't backed up is to answer it with a statement that is.

 

[Blackwoof]

Is it also applicable to people who claim that 128-160 mp3 are indistinguishable from CD-quality flacs?

With lame, MP3 can be tranparent for most music and people. V4 Lame = 128k AAC/Vorbis.

Even hydrogenaudio has defended that claim.

That means that a decent mid range set of cans with EQ can sound as good as expensive ones

Yep i don't bother with headphones anymore since i got the ER4SR/ER3SE. If i want a bassier sound for EDM/Metal/etc i just do a +5 db 125Hz and few fixes on 2 dips/peaks I'm all set.

 

[Bozzunter]

But you can approach determining your preferences scientifically, by experimenting, analyzing results and keeping track of what works and what doesn't.

Thanks for the answer, much appreciated. I still have doubts that a mid level headphone can reproduce the level of detail of a pair of HD 800s, as well as the soundstage. But I read so much BS in every headphone’s forum that I feel intoxicated and I’m not sure of anything anymore.

By any means, I don’t think my 1300€ spent for the HD800s (open) and 700€ for the Stellia (closed) were too much and I tend to agree that, based in my taste, anything else is completely pointless for me.

 

[bigshot]

With lame, MP3 can be tranparent for most music and people. V4 Lame = 128k AAC/Vorbis. Even hydrogenaudio has defended that claim.

For the average casual listener, that is certainly true. But the devil is in the details... and the words "most music and people".I spent a couple of weeks testing Fraunhofer, LAME and AAC. I found that what you say is generally true, but it wasn't uncommon for some tracks to artifact at that level. I found one particular recording with a string tone that was very difficult to encode. It required 320 LAME and 256 AAC to safely encode. (Frau was *almost* perfect at 320.) Since I didn't want to have to check each and every file I encoded, I chose 256 AAC and added VBR to it, just to be safe. Encoded like that, I think it's safe to say "all music and people".