[Ferrariman]

Because by looking at a grid I can easily tell it sounds better. With my eyes.

I can't tell the difference between 320k and FLAC most of the time, and I can easily enjoy ~175 kps.

It doesn't really matter.

 

[xnor]

Quote:

Originally Posted by leeperry /img/forum/go_quote.gif it's pointless to go for the best headphones/source

true

 

[AtomikPi]

Quote:

Originally Posted by leeperry /img/forum/go_quote.gif well, look at the THD/THD+N rates, it's pointless to go for the best headphones/source when you settle down for distorted audio files. ibuds FTW

Sure, why don't you listen to lossy files on both a pair of iBuds and an HD800 and tell me the iBuds are just as good. MP3 is specifically made with psychoacoustic considerations so that the absence of data and increases in distortion are inaudible 90%+ of the time at 320. Even when they are audible, they're usually not that large and don't detract much from the music (at high bitrates). Measurements don't matter if they don't correspond to actual differences so you should listen for yourself in an ABX and then come back with a log and tell us the differences are blatantly obvious. For me, ABX proved to me that it is extremely difficult to tell 320 from FLAC with not very many exceptions, and that's the case with nearly all of the population, even among audiophiles with good equipment who have tried for themselves.

I'm not saying I would archive or rip in mp3, but there's nothing wrong with it for listening, even with high quality equipment.

 

[leeperry]

Quote:

Originally Posted by AtomikPi /img/forum/go_quote.gif MP3 is specifically made with psychoacoustic considerations so that the absence of data and increases in distortion are inaudible 90%+ of the time at 320. Even when they are audible, they're usually not that large and don't detract much from the music

well I got pretty low end equipment, Asus STX>DT770/600Ω...and MP3 kills the fun! I'd rather not listen to any music than to wooshy wooshy mp3 artifacts..I don't like that underwater feeling

listening to mp3's on a HD800 is like giving jam to a pig...as they say where I come from

of course mp3's will sound better on a HD800 than on ibuds, what is that supposed to prove? mp3 distorts your audio, if that's fine w/ you...great! more power to you

 

[Mediahound]

Quote:

Originally Posted by leeperry /img/forum/go_quote.gif of course mp3's will sound better on a HD800 than on ibuds

Not necessarily. The HD800s can reveal more flaws than cheap earbuds/headphones so the HD800s could actually sound worse since they are more faithful no matter the source material..

 

[revolink24]

They just wouldn't sound worse. There are so many other considerations besides pure reproduction, ie bass impact, detail, clarity, punch, rolloff, soundstage, headstage, dynamic range, speed, etc. There is no way with any music an ibud would sound better than an HD800.

 

[Ferrariman]

Quote:

Originally Posted by leeperry /img/forum/go_quote.gif well I got pretty low end equipment, Asus STX>DT770/600Ω...and MP3 kills the fun! I'd rather not listen to any music than to wooshy wooshy mp3 artifacts..I don't like that underwater feeling

I beg that you try an ABX test and say this with a straight face.

 

[AtomikPi]

Quote:

Originally Posted by leeperry /img/forum/go_quote.gif well I got pretty low end equipment, Asus STX>DT770/600Ω...and MP3 kills the fun! I'd rather not listen to any music than to wooshy wooshy mp3 artifacts..I don't like that underwater feeling listening to mp3's on a HD800 is like giving jam to a pig...as they say where I come from of course mp3's will sound better on a HD800 than on ibuds, what is that supposed to prove? mp3 distorts your audio, if that's fine w/ you...great! more power to you

I do understand what you're saying, and I certainly rip to FLAC, but I don't bother replacing old high bitrate mp3's in my library and I don't mind using v0 on my mp3 player.

 

[Vel]

Quote:

Originally Posted by xnor /img/forum/go_quote.gif @Vel: As we can see you were able to differentiate 2 out of these 7 test samples. Some of the test files aren't even what I'd call music (heh, just my opinion), especially the "WelcomeToDrexciya 2 second in easy" which as the name suggest was easy to test (think you've mentioned that). 2/7, thats about 29%. I really wonder how low this number would sink if you'd redo the test with some random files of your music collection. (I'm not saying there's something wrong with this test.) Of course there's a difference between the original and mp3 file because you're trading file size for quality. And that difference can be heard in some cases. Thus I'd never archive music with mp3, or any other lossy format. Oh, did I mention that iTunes mp3 enc. is inferior to lame?

Thanks for the comments

Haha, I agree about some of the samples not really being music. Human_Disease in particular made me cringe and it was quite hard forcing myself to listen to it and try to pick out details. How can anyone like sound like that (I won't even call it music)??

Who knows how well I would do if it was real music. I now know what to focus on, short sharp sounds like snare drums, so I figure I wouldn't be too much worse provided the music had sounds like that. Perhaps now I will gain even more of an appreciation for lossless music due to focussing on things I learned.

For me, as long as I can tell the difference at 320 (2/7 isn't too bad), it's enough of a reason to have a lossless audio collection. As I said in the blog, it feels better to listen to lossless music and now I have some justification in my preference without just relying on feeling (which I hate doing). Lots of my music is in various bit rates so I will of course be upgrading the lower bit rates first and not worry about the higher quality mp3s till much later.

 

[MrOutside]

sigh.

 

[Ferrariman]

By the way, you have a classy blog.

 

[Vel]

Cheers, and it's ridiculously easy to make with iWeb!

<3 Apple

 

[badmonkey]

Quote:

Originally Posted by Vel /img/forum/go_quote.gif So anyway, this may be a blatant advertisement, but you can find all the details of my experiment at my blog (links below). My first post was an introduction to lossless vs lossy in which I went into great detail with how lossless audio is encoded (which may interest some people), then I did a materials and methods post (how I would do it and what would I would use) and finally a results and discussion post. I have included all the raw output from Foobar for those that are interested along with graphs of my results. Introduction Materials and Methods Results and Discussion

I think you should use LAME to generate the MP3s, not iTunes which is held to be crappy

, and make sure you're using a VBR mode to be fair. Unless your test is for some reason focusing on CBR, in which case it is not applicable to most real world scenarios.

 

[Vel]

Quote:

Originally Posted by badmonkey /img/forum/go_quote.gif I think you should use LAME to generate the MP3s, not iTunes which is held to be crappy , and make sure you're using a VBR mode to be fair. Unless your test is for some reason focusing on CBR, in which case it is not applicable to most real world scenarios.

Maybe I should have used LAME, I read it's better at higher bit rates and worse at lower so it's give and take regardless. (Ref: Wikipedia - MP3) Don't even know what I'd use to encode into LAME on Mac OS

The point was to see at which point I stop telling the difference between lossless and mp3, if I used VBR it would get really confusing because there is no constant rate. My results would then turn into a yes/no rather than yes here, no there, etc. Regardless of that, 320 kbps is the maximum for MP3 anyway, so if I can tell at 320 kbps 2/7 of the time, then VBR wouldn't have changed that as it can only go up to 320 kbps itself.

The majority of my MP3's are not in VBR so personally it is a "real world scenario". I don't actually know how VBR works anyway so can't comment on it too much. Perhaps it can be a future blog post or something

 

[Mediahound]

Quote:

Originally Posted by badmonkey /img/forum/go_quote.gif I think you should use LAME to generate the MP3s, not iTunes which is held to be crappy , and make sure you're using a VBR mode to be fair. Unless your test is for some reason focusing on CBR, in which case it is not applicable to most real world scenarios.

iTunes on a Mac is great. On Windows, I believe it's a bit different so the quality is not quite the same.