Lossy audio questions

[Omnigamer]

Hello,

I'm currently starting to look into the world of high-fidelity audio, but want to make sure that some of my bases are covered. I realize that most audio equipment is used to play directly from CDs since the waveform is perfectly preserved on them. However, the vast majority of my current music collection is in the form of lossy files (mp3s etc). I could gradually rebuild a good portion of it with FLACs and other lossless formats, but it would take a significant investment in time and searching since my musical tastes are far from mainstream. Which brings me to a couple questions:

1. My current concern is investigating new/better speakers. For the sake of argument, let's say I invest in a pair of NHT's Classic Three (connected to receiver fed with digital optical). The flaws in the lossy audio will certainly come to the surface over a set of typical "computer speakers", but to what extent? Is there still some benefit to having better overall speakers if the audio source isn't perfect?

2. For higher quality lossy audio (320 kb/s mp3 etc), will I see any significant benefits from resampling to, say, 96kHz or 192 kHz? How about with lossless files or audio discs?

3. Is there any form of lossy audio that I would still be able to get the most out of my speakers? Ex, are 320 kb/s mp3s of appreciable quality that most of the flaws will go unnoticed? How about other formats, such as ogg vorbis?

Just trying to clear some of this stuff up before I jump into a big purchase that I might not be able to get the best out of. Thanks for any help.

Edit:
I guess what it all comes down to is this:

Would I still be able to have an improved listening experience with a new audio setup even with using lossy audio sources?

 

[phide]

For the sake of argument, let's say I invest in a pair of NHT's Classic Three (connected to receiver fed with digital optical). The flaws in the lossy audio will certainly come to the surface over a set of typical "computer speakers", but to what extent?

Impossible to determine. Your ability to hear artifacts is dependent primarily upon how well you've trained yourself to spot these artifacts and secondarily on how good your playback equipment is. You can easily determine your artifact perception threshold by doing some simple ABX tests, which you can do while you do an in-home speaker audition.

Is there still some benefit to having better overall speakers if the audio source isn't perfect?

Of course, given there is no such thing as a 'perfect' source. We can only do the best we can.

For higher quality lossy audio (320 kb/s mp3 etc), will I see any significant benefits from resampling to, say, 96kHz or 192 kHz? How about with lossless files or audio discs?

No to both questions, unless there's a specific clocking-related need for upsampling (won't be unless you're using antiquated converters). Dithering may be helpful, but only in specific circumstances.

Is there any form of lossy audio that I would still be able to get the most out of my speakers? Ex, are 320 kb/s mp3s of appreciable quality that most of the flaws will go unnoticed? How about other formats, such as ogg vorbis?

Sure. There's lossyWAV, Vorbis, AAC and various others. At the same bit rates, AAC and Vorbis typically fare better than LAME MP3, and both support higher bit rates. lossyWAV is a bit different -- it doesn't use a perceptual model, and it's something you use in combination with FLAC or TAK to reduce the total bit rate (as lossyFLAC or lossyTAK).

 

[obobski]

just to add to phide's advice, WMA is another good codec to consider if you use Windows (especially if you download the Pro Encoder from MS (its free))


as far as "using evil lossy audio sources all the time", no your ears will not melt off, no you will not die, no your favorite 80's pop singer will not die, no god will not come down and smite thee, and so on, as long as the encoding isn't garbage (this varies per track, 192kbps can, and does, sound fine on a number of albums, 320kbps is usually excellent for most things)

basically, you'll notice FAR more by changing your speakers than you will by bothering with encoding, unless your current stuff is ripped at like 32kbps, you'll probably be perfectly content with 192 or higher for the vast majority of media

 

[Sycraft]

1. My current concern is investigating new/better speakers. For the sake of argument, let's say I invest in a pair of NHT's Classic Three (connected to receiver fed with digital optical). The flaws in the lossy audio will certainly come to the surface over a set of typical "computer speakers", but to what extent? Is there still some benefit to having better overall speakers if the audio source isn't perfect?

Depends on what rate it is compressed at how much you notice. At 256k+ you find that compressed music is essentially transparent even on the highest end setups in almost all cases.

However even with lower rates, good speakers can be nice. While you'll notice the flaws more, I'd still say 128k MP3 sounds better on $1500 home theatre speakers than on $50 computer speakers. I listen to lower bitrate MP3s at times and it doesn't drive me crazy or anything.

2. For higher quality lossy audio (320 kb/s mp3 etc), will I see any significant benefits from resampling to, say, 96kHz or 192 kHz? How about with lossless files or audio discs?

In my opinion, no, never. I have never been able to hear the difference between higher Hz recordings and not. I am also not aware of any scientific evidence that humans can hear anything up there. Now do note, people disagree with me, however in my judgment, which is the only one I have to use, there is no benefit, at all, to higher sampling rates.

3. Is there any form of lossy audio that I would still be able to get the most out of my speakers? Ex, are 320 kb/s mp3s of appreciable quality that most of the flaws will go unnoticed? How about other formats, such as ogg vorbis?

Yes, as I mentioned 256k or above for any format (MP3, AAC, OGG, WMA) seems to be totally transparent for 2 channel sources meaning you can't tell the difference from the original in blind tests. There have been real, empirical studies done on this using reference grade gear that confirm it. That isn't to say there will NEVER be a problem, but more or less you aren't going to have an audible difference, even on great gear.

Now that said, I've started doing all lossless these days since HD space is cheap.

Would I still be able to have an improved listening experience with a new audio setup even with using lossy audio sources?

With out a doubt. Like I said, I listen to compressed music and I love my high end setup ($1500 tower speakers). I like it better with higher quality music, but it is fine with lower quality stuff. I play games on it, which always seem to use pretty heavily compressed music, listen to lower bit rate MP3s and so on. I for sure notice a step up over the cheap stuff I have at work.

With better source material you will get better sound for sure, but you still get more satisfying sound out of lower end sources.

 

[pwnrhasta]

For me, I have dozens of albums in FLAC format, and compared to 320kbps MP3 format, I could hardly tell the difference.

So i switched to MP3 format.

Also anything above 192kbps is fine with me as well.

Anything below, you will notice a decrease in sound quality.

 

[Fryguy8]

Hard drives are really cheap, so might as well just encode lossless. If you need a lossy version for a player that only plays X format, just make a transcode on the fly. Doing an album from flac->mp3 takes <1 minute on anything resembling modern cpu hardware.

Even if it is only for peace of mind.

Vorbis comments (the comment type that flac files use), tend to be easier to work with than id3 tags as well