I’m impressed with MP3’s compression rate, but there has got to be something simple I’m missing in the math I’m doing, and I just can’t see it:
Say I have a recording that is 76 min., 34 sec. long. Thats (76*60)+34=4594 seconds.
So, converted to MP3 at a bit rate of 128 would be 128*4594=588032 bits? I’m assuming “bit rate” means bits per second, and I’m suspicious that that’s where I’m getting mixed up.
A good rule of thumb, as demonstrated by your corrected calculations, is that a 128 kbps bitrate will use up 1 Mb per minute. That’s not an exact figure, but it’s close enough for a good approximation without crunching the numbers.
So, 64 kbps will take up about 0.5 Mb per minute, 256 kbps will take up about 2 Mb per minute, etc.
I disagree that 128kbps is anything like approaching CD quality.
I know that the compression algorithm makes a differance to the sound quality, however, playing a CD and the MP3 of the same material is revealing, you’d need 240kpbs for the differance to be small enough to consider it CD quality, and to be honest, 320kbps is really when MP3 and CD quality are similar enough to make no differance.
If you are using some little portable kit, that can play both CD and MP3 then you may find that 128k is similar on such a poor device, but put that CD in a decent machine and the differance shows.
I sometimes wonder how many younger folk haven’t experienced decent music through a decent CD player, they have no idea what they are missing.
Thought of another way, 128Kbps is just about 10x smaller than its original, uncompressed audio. Generally speaking, the factor increases as the bitrate decreases, and vice-versa. For example:
And so on. It’s a general estimate but a lot easier than working out the exact math involved. These are based on the usual assortment of bitrates that most encoders make available – though you are not specifically limited to one of these factors. VBR (variable bitrate) encoding complicates things, too.
Yeah, I was wondering about that too. How can they justify saying that? I thought CDs, unlike MP3’s, were lossless. Does “CD quality” even mean anything specific?
I’m probably going in over my head given the ignorance I displayed in the OP, but I think terms like “lossless” and “lossy” only apply to compression algorithms. CD Audio files aren’t compressed per se, but their sample rate of 44.1 KHz can vary by bit depth and is far from perfect.
FWIW, I would only ever assume that “CD quality” meant an overall bit rate approximating that of 44.1 KHz at 16 bits that is (IIRC) standard for Audio CDs. I will leave the math on that to someone who’s arithmetic is working today.
This is correct. And it’s quite good for reproducing audio signals. At 16-bits (two bytes per sample) that gives you a range of 65,536 possible values to represent one sample (1/44,100th of a second) which allows for a very descriptive audio signal. There is no compression – it’s raw pulse code modulation.
It isn’t absolutely perfect, but it’s more than adequate. Most studio mastering these days is done at 96KHz in 24-bit, which means three bytes per sample, or 16.7 million possible values to represent 1/96,000th of a second of audio. While it’s certainly possible to go 32-bit or bump the sample rate, I don’t think there would be any need to go above current standards – unless you’re writing symphonies that only dogs can appreciate the subtle nuances of.
Compression does not necessarily mean lossy audio, however. The Free Lossless Audio Codec (FLAC) format manages to compress audio data to roughly 2/3ds to half of its original size but does not remove any information from the audio stream. It’s kind of like the ZIP compression method, but tailored for audio streams. It still isn’t ideal because the files are still quite large – about 5-7 megabytes per minute of audio – but real audiophiles with space to spare on their MP3 player can appreciate the lack of artifacts typically present with lossy compression schemes at lower bit rates.
128k is only considered “near-CD quality” and it is capable of coming fairly close, provided the audio source isn’t tremendously rangy or loud. I prefer 192k at the very least when it comes to MP3. It seems to hit the sweet spot between quality and size.
Don’t ask me about the math, but I’ve recorded at bit rates much higher than 44.1 KHz at 32 bits, and there is a BIG difference. For the most part, though, it doesn’t matter unless you’re a serious audiophile.
To respond to casdave, in my opinion and experience, the quality of the CD player doesn’t matter as much as the quality of the speakers/headphones and the sound card (if you’re using a computer). Either or both of these will compensate for a lower bit rate. In fact, I once connected a portable CD player to a set of decent speakers and had a good experience. I’ve also had the unfortunate experience of listening to a professional-quality CD player through bad speakers; to me, it was like nails on a chalkboard. Ditto for iPods; the ones that come standard suck, but I’ve used my studio headphones and the sound quality is magnificent. I’ve also bought a set of relatively not-bad earbuds for daily use, and that improved things considerably. YMMV and so forth.
Yeah, i really like FLAC, and with the ever-decreasing price of hard drive space, it’s actually becoming feasible to put a FLAC-encoded collection on your home computer. With 500 and 750 gig hard drives going for less than $150 or so now, space on your computer is cheaper than ever.
Of course, there’s still the mp3 player itself to consider, and encoding in FLAC means far fewer songs on your iPod. But for home systems, i think FLAC is quite a good solution now.