[trains are bad]

Quote:

Originally Posted by Patrick82 192 kbps mp3 is good enough, anything higher is insignificant

I can actually abx 192 kbps LAME and .FLAC. Barely. With select recordings. Based on that, I would say that anything higher than that is indeed transparent, and 192 kbps LAME is practically transparent.

Since I don't do portable audio, I only listen to .flac at the moment. If I had a DAP, I would certainly fill it with high quality lossy compression, just for capacity and battery life reasons.

 

[Edwood]

Quote:

256kb mp3 indistiguishable from lossless?

Yes, that is absolutely true.












[size=xx-small]When you are listening to iBuds.[/size]

-Ed

 

[russdog]

Quote:

Originally Posted by trains are bad I can actually abx 192 kbps LAME and .FLAC. Barely. With select recordings. Based on that, I would say that anything higher than that is indeed transparent, and 192 kbps LAME is practically transparent.

That was exactly my experience. So I created mp3's at "alt-present-Extreme", which seems to average 224 kbps. I thought about using the highest level of presets, "alt-preset-Insane", but decided that doing that would be, well, insane. I love its name, though

 

[DarkJC]

I use lossless formats not only to make sure there's no possible quality loss, but also to keep an archive of my albums so I'm future-proofed for [insert the next great lossy codec here]

For me it's not about being able to tell the difference. If I can avoid reripping my entire collection to a new superior format later on (and instead use a simple batch process to reconvert all the FLAC files to that format), while benefiting from possible quality improvements in the process, so be it.

 

[Skylab]

Edwood and DarkJC, great posts (Edwood I was worried about yours until I read the fine print

)

 

[ijan]

In my short life, i wasnt able to distinguish a 128kbps mp3 and a 192..after sum time, 128 sounds crappy, then 192..then 256..but sum 192 sounds great (cant AB to lossy..yet). but sum of my lossless is really bad compared even to the 192k and sum 256k is 99% tat of a lossless in my ability of hearing properly. My POV, a properly ripped record is a major factor and the compression when lossy is used. BTW, i rarely am able distinguish a perfectly ripped 256k mp3 compared to the original, but below that bitrate, i can tell.

 

[pedxing]

For me, 192 kpbs is the breaking point where most music sounds transparent in respect to the original CD. Most pop recordings for the past three decades have been so bad, encoding them at 192 kbps doesn't degrade the quality in any noticeable way. Just to save space, I even use like LAME setting -q3 or -q4 for such recordings. I can hear a difference, but the slight savings in drive space is a nice trade off.

Anything below 192 kbps, the differences becomes gradually noticeable. I cannot tolerate most mp3s encoded at 128kbps. I prefer to use ogg at that point.

 

[Laptopia]

Quote:

Originally Posted by russdog Because I can't tell the difference between CD's and alt-preset-Extreme, that's why. I have done ABX tests. Why do you assume people who prefer this somehow don't understand? If I'm happy with it, why do you care? Why do you want to imply that I'm somehow making a mistake? I am not claiming that nobody can tell a difference. I am claiming that I can't, and since it's my money, that's what I care about. Size does matter when it come to portable devices. And, no, I don't want to maintain separate databases of lossy and lossless. What about this is hard to understand? I have a golden-eared friend who has more than $30K in his stereo. He was adamant that he could tell. We did a test. He couldn't. He was depressed for days. He still doesn't want to talk about it. If you can tell a difference, that's fine, I'm happy for you. But please don't tell me I'm wrong.

You are cruel indeed! But you know what? I would bet that a number of oenophiles (wine connoisseurs), if presented with a blind test between their beloved $60 bottles and the middling $10 bottles, would suffer the same result. My experience is that when anything becomes that exacting, there is some puffery and pseudoscience present. As Paul Fussell wrote, when you see the word, "Gourmet!", someone is about to get fleeced . . .

 

[K2Grey]

Quote:

Originally Posted by Bob ♫ You are cruel indeed! But you know what? I would bet that a number of oenophiles (wine connoisseurs), if presented with a blind test between their beloved $60 bottles and the middling $10 bottles, would suffer the same result. My experience is that when anything becomes that exacting, there is some puffery and pseudoscience present. As Paul Fussell wrote, when you see the word, "Gourmet!", someone is about to get fleeced . . .

I'm not sure about that.

Lossy codecs employ psychoacoustics in an effort to mimick lossless as well as possible. Much effort has gone into it and many tests are done in order to gauge how successful the effort has been so far and problem samples are found and fought with. Also all lossless codecs are supposed to sound exactly alike and all lossy codecs seek to sound exactly like lossless codecs (unless you have hearing problems or something like that and the psychoacoustic tricks don't work as well).

On the other hand $10 wine makers do not spend years attempting to imitate a specific brand of $60 wine, and $60 wine makers don't attempt to all imitate each other as closely as possible. So there will be variation between the tastes of different wine, simply because they don't try to taste exactly like each other.

 

[Svperstar]

Oh yeah another reason I went with .APE is that APE with high compression encodes much, much faster than LAME -alt preset extreme

 

[pheonix991]

I can tell a diff from 320kbs mp3 to FLAC. So, Yeah, you can take that for what it is worth... Either I have sensitive ears, or, something.

 

[Svperstar]

Quote:

Originally Posted by pheonix991 I can tell a diff from 320kbs mp3 to FLAC. So, Yeah, you can take that for what it is worth... Either I have sensitive ears, or, something.

100% of the time? You have ABX tested this? How many trials? How many different tracks?

 

[trains are bad]

Quote:

Originally Posted by Svperstar 100% of the time? You have ABX tested this? How many trials? How many different tracks?

I don't think it matters. If he can tell a difference at all, with any recording, that counts as nontransparent.

 

[wanderman]

Quote:

Originally Posted by trains are bad I don't think it matters. If he can tell a difference at all, with any recording, that counts as nontransparent.

or it could be the fact that he is guessing. The number of trials performed compared to the number correct can show if he is guessing or just good at hearing the difference.

 

[russdog]

Quote:

Originally Posted by pheonix991 I can tell a diff from 320kbs mp3 to FLAC. So, Yeah, you can take that for what it is worth... Either I have sensitive ears, or, something.

If you like FLAC, great.

If you want to know what's true, then you need to determine whether you can tell a difference via appropriate methods. The appropriate method requires that neither you nor the person assisting you in the comparison test knows which one is which. It also requires an appropriate number of trials. If you can tell a difference without that level of control in the procedure, then the results are highly suspect. But, like I said, if you are well satisfied using FLAC, that's fine. I don't think you "should" reconsider that, I think you should do whatever you like.

I do think it's important to distinguish between the personal decisions we each make vs. statements about "what's true". Around here, these 2 things get confused all the time, which in turn causes even more confusion for people who are trying to learn.