Faster iTunes 8.1 with new import and Genius features due soon - Page 2

Originally Posted by quinney

Exactly right in both cases. For the highs, even though the harmonic frequency, by itself,
may not be audible, its presence in addition to the fundamental and lower harmonics can
add "coloration" to the sound. I only brought this up, because it seems to contradict your
assertion that people cannot distinguish between compressed and non-compressed
versions of the same music. You even say that you can tell the difference yourself, if
your system is capable of reproducing lows.

Not trying to be confrontational. I enjoyed talking to you.

Originally Posted by FJRabon

eh, there's a lot of debate about how harmonics that the ear can't pick up effect frequencies you can hear. I tend to come down on the "they don't do anything" side. However, its controversial enough that it causes engineers to not mess with them (well, that and the fact that its a royal PITA to mess with harmonics). As to the lower range stuff, Id certainly not take them off if I was listening through speakers with a LOT of power. However, nothing you're going to listen to in a residential area is going to produce enough power for you to really feel 10 htz waves.

Originally Posted by FJRabon

you generally don't cut 10-20 htz, because you can "feel" those waves, and it makes things sound a touch fuller if you are listening through a very large sound system (think 15 inch subwoofers powered by at least 10k watts). Also, you get harmonics outside of 20k, because engineers don't like messing with harmonics, and they have no real reason to anyway.

Originally Posted by quinney

I think it has to do with additive properties of waves and subtle phase cancellation effects.
Thank goodness my neighbors don't have that system. What they do have is almost a
military weapon as it is.

Originally Posted by ragamffn

Pure BS.

Originally Posted by FJRabon

I am guessing 256kbps will just be the defualt now. Most people don't know that you can import at pretty much whatever bitrate you want and a whole slew of options.

I wonder why they don't use VBR as the standard though. I guess people feel more assured if they see a single number next to all their files instead of a varying number.

Originally Posted by FJRabon

One last thought, then I'm done. If you're listening through stock apple earbuds, you just plain aren't going to hear the difference between 128 and 256. Those things won't re-produce the differences.

Originally Posted by FJRabon

I get the feeling that many people tend to think that AAC and mp3 encoders just randomly throw away digital audio information. They tend to think that 256kbps is "twice as good" as 128 kbps, etc

It doesn't really work this way. AAC (and mp3) uses a psycho acoustic model when it converts. 70% of what it ditches is nothing at all. And by nothing, I mean nothing. .wav puts down the same amount of information for pure black silence as it does for a raging guitar solo. mp3 and AAC don't work this way (well, they do if you don't use VBR, but that just means that they throw away less info when less info is required to encode the main audible parts) AAC amd current mp3 encoders throw out a bunch of stuff that's well outside the human audible range. The first thing they do is throw out black silence. There is a surprisingly large amount of pure non-info encoded in .wav. They then dispense with the higher frequencies that are inaudible and then the lower inaudible range (since sometimes the lower inaudible frequencies can be felt, they're cut second). Then it starts cutting out extremely light noises within nano-seconds of very loud noises, since thousands of studies have shown that humans can't hear very quick and large changes in dynamic level (ie it takes some time for the ear to "rebound" after a loud noise). Just these changes alone will often times get you into the 256 kbps range. And if you burned this to .wav and then re-ripped, you'd lose no additional audible information, since it would realize that "hey, I can encode this exact file with just 256 kbps by throwing away all this info that is either black silence or looks exactly like an oddly .wav'd up 256 kbps file."

Originally Posted by teckstud

Look up any test comparison- there are other factors involved. The playback of a CD simply sounds better- sampling rates, D/A converters, etc.

Originally Posted by FJRabon

I'm not saying its necessarily right, just saying that engineers don't cut those frequencies for the most part. Almost every CD you can find has a lot of stuff going on above and below 20-20000 htz.

And you can definitely feel a 15 htz wave with enough power.

Originally Posted by kim kap sol

There's actually a great deal more thrown away than what you're claiming in your first two paragraphs. The *lossless* formats usually have zero problem compressing the silence and that's how *they* achieve the bulk of their compression. The lossy formats, as you mention in that last paragraph and beyond, throw away frequency ranges that would usually not be heard by most people.

That said, 256 kbps would usually retain most of the important information. But it's still throwing away data. Silence is data. 0 isn't null. Every self-respecting music enthusiast would rip its music in a lossless format to retain *ALL* the information.

Of course, people that just don't care, just don't care, so I don't care what they do.

Originally Posted by bloodstains

Thanks for the heads up. I'll be sure to adjust my life to fit your needs.

Originally Posted by kim kap sol

That said, 256 kbps would usually retain most of the important information. But it's still throwing away data. Silence is data. 0 isn't null.

Every self-respecting music enthusiast would rip its music in a lossless format to retain *ALL* the information.

Originally Posted by foljs

Audio enthusiast != music enthusiast.

Originally Posted by DaveDaveDave

"Coke! No, Pepsi NO! Coke!"

IMHO, there's a simple solution to this whole debate.... test YOURSELVES!

Rip one of the most scintillating pieces of music you have with different codecs and different bit rates. Do a blind audio test or better yet, a double blind test. If you don't have someone to tell help you with testing, you can "blindfold" yourself by playing the pieces on shuffle, and take notes. After you're done, flip back through what you've played and compare to your notes. Voila, you've found your very own personal bit-rate threshhold.

I actually do this every now and then with test tones, hoping that one day I'll actually hear 17.5k again.

Cheers,
Dave

Originally Posted by bslaght

how about importing songs....will that be sped up...???

Originally Posted by teckstud

Is your name Carol Anne and aren't you a character from Poltergeist?

Originally Posted by Virgil-TB2

This has to be a mistake in the article in that I have been doing it at 256 kbs for ages.

256 would actually be "better than (Lossless)" however in that your ears cannot tell the difference and you would save oodles of space. Unless you are purposely creating a digital archive of your CD collection as a backup, Lossless makes little sense as a format.

Originally Posted by FJRabon

I am guessing 256kbps will just be the defualt now. Most people don't know that you can import at pretty much whatever bitrate you want and a whole slew of options.

I wonder why they don't use VBR as the standard though. I guess people feel more assured if they see a single number next to all their files instead of a varying number.

Originally Posted by Silverflyer

For you, maybe.

For me, not very likely, see, I use my Mac Pro as a Media server and have every single CD I own, and some vinyl too, ripped to it as a lossless AIFF file so I can play it back through an obscenely expensive DAC, and then off to the rest of my obscenely expensive audio equipment.

I would bet 10,000 US Dollars you would be able to hear the difference between an AIFF file and the same song at 256kb within 30 seconds.

Originally Posted by FJRabon

I'd take that bet. Easily. Unless you're 14 years old and have never been to a rock concert. I know people who have had years of training in knowing EXACTLY what to listen for when it comes to compression artifacts (a couple of them helped encode LAME and OGG which are probably the two standards for lossy encoding) and at 256kbps, they don't get it right more than 65% of the time.

EditL and they're listening through Sennheiser Orpheus's, which would blow away pretty much anything out there with regards to the amount of detail you can year.