Can you really tell the difference in sound quality?

12/31/2005 10:53:05 PM

in the car, no. on a very good home theater setup, yes.


I use 192 for all of mine.

12/31/2005 11:06:54 PM

192-centered VBR

1/1/2006 5:08:05 AM

on an ipod with headphones, could you tell the difference between 128 and 192?

i am getting one soon, and so i am starting to rip my cds.

should i do 128 or 192? any audible static with 128?

1/1/2006 6:38:13 AM

i notice distortion with 128 compared to 192, even on cheap earbuds.

1/1/2006 9:54:08 AM

0EP: do 192.

especially on the music you like, you'll hear the crackling on the bass hits.

1/1/2006 10:40:57 AM

192 is the way to go. 320 uses too much memory and you get no appreciable audio difference. 128 is the good middle of the road, but I would go no lower.

1/1/2006 10:47:55 AM

192, mp4 -- if you are going to be using an ipod

1/1/2006 1:23:29 PM

VBR is the way to go

1/1/2006 2:35:09 PM

^
^^
^^^
^^^^

thanks

1/1/2006 3:17:39 PM

yep 192 is good enough for most people, 320 is over kill.

1/1/2006 4:35:15 PM

I use the LAME MP3 codec's "extreme" preset, which is about 200-240kbps

1/1/2006 5:50:59 PM

^^ yep.

1/1/2006 6:00:13 PM

32 above what you need VS way more than you need

thats what the thread title SHOULD be

1/1/2006 7:16:08 PM

maybe if you have shitty headphones, read use the apple headphones, but if you nicer head phones or a nicer stereo, I wouldn't do less than 224kbps. I typically do vbr with an average bitrate of 224 or 256

1/2/2006 2:07:33 AM

it's very unlikely that people can truly distinguish between 192 and CD quality.

Quote :

"Judging from our tests, we think most people will find music compressed at higher bit rates indistinguishable from the original versions. Music fans will probably find even moderately compressed songs acceptable, especially if the compression means they can fit more tracks onto a portable player. However, this compromise becomes less tolerable when it involves compressing the music to the lowest bitrate."

For general use, 128 kbps is fine (or 192 kbps if you have really good ears).

1/2/2006 12:14:58 PM

Most people who claim that they can tell the difference between 192 and anything higher is full of shit and is just trying to pretend that they are some kind of audiophile.

1/2/2006 12:31:56 PM

listen to anything with 18kHz+ frequencies and you'll immediately hear the difference between 192 and cd on good speakers/headphones. Is disk space really such a premium that you can't afford an extra meg or 2 per song?

1/2/2006 1:19:32 PM

when MP3 was first developed... yes.

1/2/2006 1:47:48 PM

like smoothcrim said, it is quite easy to notice the difference between 192 and higher bitrates when you're listening to good music.

1/2/2006 1:50:41 PM

i call bullshit

1/2/2006 3:10:16 PM

192 was chosen as a standard for a reason - stick with it

1/2/2006 3:17:36 PM

Anything but 320k is noticeably inferior.

1/2/2006 5:30:55 PM

I used to listen to 192 kbps MP3s. All my new music is 320kbps. There is a MUCH MUCH notable difference between the quality.

That said, you should check out the aacPlus codec.

1/2/2006 6:00:18 PM

if you want the highest quality and lossless, check out FLAC

1/2/2006 6:58:27 PM

Is FLAC really better than APE?

1/2/2006 6:59:49 PM

i'm not familiar with APE, but I know that FLAC is one of many different lossless formats. Microsoft and Apple both have a lossless formats, they all compress the files down to nearly the same size anyhow.

I dont think you can really argue that one lossless format is better quality than another, only the filesize and compatibility matter really.

1/2/2006 10:18:46 PM

what's the deal with lossless formats?

there is no loss of data but file size is reduced (to what?)? how?

Quote :

"Anything but 320k is noticeably inferior."

yup cd quality is inferior to 320k!

1/2/2006 10:21:46 PM

listen to anything with 18kHz+ frequencies and you'll immediately want to kill yourself hear the difference between 192 and cd on good speakers/headphones.

For an mp3 player, you'll probably be happy with WMA VBR 85-145Kbps or 50-95 Kbps. For a top dollar stereo system use VBR 135-215 Kbps or WMA lossless.

[Edited on January 2, 2006 at 11:01 PM. Reason : ?]

[Edited on January 2, 2006 at 11:02 PM. Reason : ?]

1/2/2006 10:58:05 PM

when i rip, i go 320, full stereo, crc's...when i re-encode for the mp3 player in my car, i go 128-center vbr for maximum sound quality/size

that being said, i've been checking out re-encoding all my car tunes to 64kbps wma...half the file size and i challenge y'all to listen to a 128kbps mp3 and a 64kbps wma...i was only BARELY able to notice a difference, and that was with full earphones and with a song that has parts that were acapella voice and full symphony pieces

just something to consider

1/2/2006 11:00:17 PM

i usually rip stuff to 256, seems like a good middle ground to me.

1/3/2006 12:31:42 AM

Why would you want to kill yourself upon hearing 18khz?

1/3/2006 1:26:31 AM

http://en.wikipedia.org/wiki/MP3

Yeah yeah, wikipedia isn't the greatest of official sources, but...

Quote :

"Because MP3 is a lossy format, it is able to provide a number of different options for its "bit rate"—that is, the number of bits of encoded data that are used to represent each second of audio. Typically rates chosen are between 128 and 256 kilobit per second. By contrast, uncompressed audio as stored on a compact disc has a bit rate of about 1400 kbit/s."

Quote :

"Good encoders produce acceptable quality at 128 to 160 Kibit/s and near-transparency at 160 to 192 kbit/s, while low quality encoders may never reach transparency, not even at 320 kbit/s. It is therefore misleading to speak of 128 kbit/s or 192 kbit/s quality, except in the context of a particular encoder or of the best available encoders. A 128 kbit/s MP3 produced by a good encoder might sound better than a 192 kbit/s MP3 file produced by a bad encoder."

1/3/2006 5:22:54 AM

Quote :

"Most people who claim that they can tell the difference between 192 and anything higher is full of shit and is just trying to pretend that they are some kind of audiophile. "

not true. i think most studio works are fine at 192, but for live music recordings, you can tell a noticable difference between 192 and 320.


that being said, i use 320 for all my music. it may be overkill for some things, but i have more then enough space, so 320 is fine.

1/3/2006 8:01:33 AM

Yeah i was mainly referring to studio tracks. I don't really listen to jam bands and the like, so I hadn't thought of that scenario.

1/3/2006 9:12:28 AM

True or false:

Quality of 128 AAC is equal to (or maybe even greater than) the quality of 160 or 192 mp3.

If so, why don't a lot of people use AAC as opposed to mp3 (context: mp3 players).

Are there any drawbacks to AAC?

Advantages to mp3?

I am ripping my CDs to put on my iPod, and I had decided to do 192 mp3, but now I am wondering if I should do 128 AAC to save space, if it will give the same quality anyway.

1/21/2006 5:14:06 PM

AAC is a modular codec and is infact better than MP3 at equal bitrates.

1/21/2006 5:21:01 PM

^^ do 192kps AAC VBR

1/21/2006 5:24:49 PM

Quote :

"listen to anything with 18kHz+ frequencies and you'll immediately hear the difference between 192 and cd on good speakers/headphones."

i'll let me dog know

i mean jesus christ, who wants to listen to 18kHz+

1/21/2006 5:44:23 PM

why would you not encode everything 320 VBR? it's not like it takes up too much space with storage these days.

1/22/2006 12:11:01 PM

on some unencoded wav songs i can certainly tell the difference.

1/22/2006 1:24:23 PM

RoidRaginTKE said:

Quote :

"i'll let me dog know i mean jesus christ, who wants to listen to 18kHz+"

You have a poor understanding of both how hearing works and what sampling frequency means. Let me school you.

The human ear is sensitive to frequencies ranging from about 20 Hz to 20,000 Hz, where a Hertz is taken as one change in air pressure per second. So many of us actually are sensitive to frequencies above and beyond 18 kHz.

However, we're not talking about the actual frequency of sound here. We're talking about the rate at which a sound is digitally sampled. A CD-quality digital audio signal is sampled at 44.1 kHz, meaning that the source (often analog) is sampled 44,100 times per second. That's not an arbitrary number, either—the Nyquist–Shannon theorem states that a faithful digital sample of an analog source must be sampled at twice the highest frequency of the source. Since our upper hearing range is about 20 kHz, nearly perfect digital sampling requires a sampling rate of 40 kHz or more to capture all perceptible sound.

You should stop reading here if that answers your question and completely satisfies your curiosity.

Of course, not every source is going to be pushing us to the limit. The human voice produces a maximum frequency of around 3,000 Hz when speaking, which is why your telephone only samples and reproduces at 8,000 Hz. That lower sampling rate means there's less information being sampled, less information to convert to a digital signal, and less information taking up bandwidth.

"But FroshKiller," you may ask, "what does any of this have to do with bitrates of 192 kbps and 320 kbps?" Well, I'm getting to that.

Bitrate is the product of the sampling frequency and amount of information contained in each sample. Going back to our CD audio example, you've got 16 bits of information being sampled 44,100 times per second. Just to throw you a curveball, two-channel stereo sound doubles that number. So you've got 16 bits * 44,100 s^-1 * 2 = 1,411,200 bits per second, or 1,411.2 kilobits per second (kbps). It takes 1,411.2 kilobits—a little over 172 kilobytes—to record a signal second of CD-quality digital stereo sound.

"So how does an MP3 record a high-fidelity digital signal with such a paltry bitrate?" you ask. "How do they cram all that graham?" Well, MP3 compression works on the principle of perceptual encoding. That means (in very basic terms) that if the encoder determines that the human ear can't distinguish a change in sound, there's no sense in recording that change. Can you tell the difference between this shade of black and this one? Then why not treat them both as the same color? Similar concept.

A good encoder goes through your source and determines the maximum frequency contained therein. Since encoders begin with a digitally sampled source, often from the CDs you fags rip to give away to your scummy friends, they're looking through 44,100 samples per second to determine how many times the amplitude of the sound actually changed. Then they trim the fat. If you could reproduce a nearly identical sound in half the sampling frequency, why not do it?

It'd be useful to work with another example at this point. Let's say you're ripping some godawful new song by the Mars Volta that's simply a 3-minute, 16-second stereo recording of a tuning fork that vibrating at concert A (440 Hz). You're starting with a 3:16 two-channel source sampled 16 bits at a time at 44.1 kHz. Your bitrate for CD-quality stereo sound, as you'll recall, is 1,411.2 kpbs. Multiply the bitrate times the length of the track to get the total size of the track: 196 s * 1,411.2 kb/s = 276,595.2 kilobits, about 33.8 megabytes.

How do we go from a 33.8 MB sound with a bitrate of 1,411.2 kbps to an MP3 a tenth of that size with a bitrate of 192 kbps? This is where the MP3 magic comes into play! Remember, MP3s operate on the principle of perceptual encoding: Sounds you can't hear are data that can be discarded. Since our faggy Mars Volta track is just the same 440 Hz tone, we could actually just sample twice that according to Nyquist-Shannon and get optimal results. So our optimum sampling rate for this project is 880 Hz, less than 2% of the actual 44.1 kHz sampling rate dedicated to capturing the sound!

Let's look at what this does for our bitrate: 16 bits * 880 s^-1 * 2 = 28,160 bits/s, or 27.5 kbps! That gives us a total file size of under 0.0034 MB, an impressive improvement over our 33.8 MB start! And how did we get here? We just cut out the overhead of the unnecessarily large sampling frequency!

But it's not over yet. Remember, we were sampling two channels. A good encoder will check both channels to see how much information between them is redundant. If the channels are identical, that instantly cuts our bitrate in half—we go from 27.5 kbps to 13.75 kbps, reducing our total file size to under 0.0017 MB!

Where does this leave us? Well, we could faithfully reproduce the experience of listening to that song on the CD by streaming a paltry 0.0017 MB of data over a 13.75 kbps connection rather than running over 33 MB at a much more demanding 1,411.2 kbps. Of course, MP3 compression is a lot more complicated than just analyzing frequencies and comparing channels, but those two activities are the basis of all audio compression.

"But Frosh," you ask, feeling more comfortable and familiar in our brief acquaintance, "what about that data I can't hear? Doesn't that play a part in the experience of the sound?" Of course it does! Just because we can't HEAR frequencies lower than 20 Hz doesn't mean we can't FEEL those changes in air pressure. That's the whole point of bass, after all: feeling the thump! But MP3 compression privileges perceptible sound over true high fidelity, which is why MP3 audio usually has extremely shitty bass response. The higher the bitrate you encode at, though, the more signal information is preserved. A 192 kbps MP3 will always capture more frequency extremes (and thus more of the sound and feel) than a 320 kbps MP3. The trade-offs: file size and bitrate demands for playback.

So there you have it, a crash course in understanding the basics of digital audio. I apologize for how disorganized this post is, but such is the nature of Sunday mornings!

[Edited on January 22, 2006 at 2:31 PM. Reason : ///]

1/22/2006 2:29:10 PM

maybe i should point out that wasnt really a serious post

way to write a novel dork

1/22/2006 2:31:24 PM

There, there, now. There is no shame in admitting ignorance.

1/22/2006 2:32:53 PM

haha what a nerd

1/22/2006 2:34:18 PM

roidraige you just got pwnt by the best

how does it feel

1/22/2006 2:59:21 PM

there's only one format for audiophiles, lossless

http://www.etrade.org

http://www.sugarmegs.net
http://www.sugarmegs.org

http://www.bittorrent.net

1/22/2006 3:29:02 PM

frequency =/= sampling frequency
dorks

1/22/2006 4:54:44 PM

I've personally noticed a large difference between 192 and 320 when it comes to classical (mainly all I listen to)

192 tends to cut out subtle parts of the audio that are crystal clear if you use 256 or higher

it's also not a terrible amount more diskspace

it's not like a hard drive and/or CD-Rs and DVD-Rs are really expensive anyway

1/24/2006 7:49:02 AM

1/24/2006 12:18:04 PM