a question about mp3 bitrate.....
- Thread starter mito
- Start date
Bouncing Zabaglione Bros.
Legend
Why? All it's saying is that there's more improvement to be had on low quality files than on high quality files. If you read the rest of that paragraph, they say it's not necessarily better, just different.
I got it... I understood differently. That a source encoded at 128kbps could sound better than the same source encoded at a higher bitrate.
Bouncing Zabaglione Bros.
Legend
It's not all bad. Even wav files are technically discrete samples, so are missing some of the info in true analogue recordings. Given their greater size, WAVs can need more memory and being that much larger need more processing. I've seen movies that stutter like mad due to the extra overhead of a wav audio track, compared to the compressed audio version.
I much prefer ogg vorbis over MP3 because the former was designed to hit quality targets, whereas the latter was made to hit bitrate targets.
Bouncing Zabaglione Bros.
Legend
That's not what they're saying. They are particularly referring to the effects of Crystalizer on audio tracks.
There's lossless audio compression formats in case you want to praise at something for your existance, such as APE(Monkey's Audio) or FLAC. The last i'v been using on my personal encodings for over a year already, and it's also open source based.
FLAC
APE (Monkey's Audio)
There's also the older Shorten project apparently read more at the wiki
Last edited by a moderator:
Are they saying that a source poorly encoded may be worsened by Creative's Crystalizer?
They might, but I'll say it can in specific cases. To me when I had a X-Fi it was largely not better but different. I think it would have pleased people with lacking equipment when it comes to headphones/speakers but those who have decent equipment it made it a bit worse as to me it brought out the encoding artifacts a bit more. It basically tries to artificially increase dynamic range, which works decently well for lower end equipment, not so much for my Sennheiser HD-580 headphones though...
I also think the wording of your practical snippet in the OP meant more that lower bit rate encoded tracks noticed more of a improvement, which would be true. That does not mean however that a lower encoded with crystallizer is better than a higher bit rate track, not even close IMO.
Personal opinion
The increase in storage is rendering audio compression less important. Sales of music online is giving compression another lease of life because bandwidth is at a premium, but that too is becoming less important.
You can store around 1400 CDs on one of the new 1TB harddisks available.
The 44KHz 16 bit format of CDs isn't perfect, but it's pretty darn good.
If the precision was increased to 24bits you'd have 144dB SNR for the entire frequency range the human ear can detect. That's from the smallest whisper the ear can detect all the way up to lethal amounts of sound pressure, all at the same time. And you'd still be able to cram >1000 hours of music onto one of the new TB harddisks.
Why compress at all ?
Cheers
The increase in storage is rendering audio compression less important. Sales of music online is giving compression another lease of life because bandwidth is at a premium, but that too is becoming less important.
You can store around 1400 CDs on one of the new 1TB harddisks available.
The 44KHz 16 bit format of CDs isn't perfect, but it's pretty darn good.
If the precision was increased to 24bits you'd have 144dB SNR for the entire frequency range the human ear can detect. That's from the smallest whisper the ear can detect all the way up to lethal amounts of sound pressure, all at the same time. And you'd still be able to cram >1000 hours of music onto one of the new TB harddisks.
Why compress at all ?
Cheers
Bouncing Zabaglione Bros.
Legend
Creative claims that Crystalizer improves the audio quality of a file. The article is saying that if there is any improvement (and they are not really sure there is) that it is the low bitrate files that show the most improvement, and the high bitrate ones don't improve as much. Obviously they are saying that as there is more to "fix" on a low bitrate file, Crystalizer can actually have more of an effect.
They are not saying that Crystalizer working hard on a low bitrate file will make it sound better that Crystaliser not doing much work on a high bitrate file. In the end, bitrate and equipment makes the most difference, and there's not much that Crystalizer can do about that.
Compressed always beats uncompressed in quality per bits. As such, I consider mp3 a higher quality format than wav, because if you saved your source audio into wav file of the same size as the mp3 its quality would be awful.
Also, the bottleneck is usually I/O. Reading a compressed file from the disk and decompressing is often faster than loading an uncompressed file.
Even with today's relatively very fast SATA II disk systems? Not to mention RAID.
For instance, a WAV file that is 50Mb compared to its 3Mb encoded MP3 counterpart file. For today's system, loading a 50mb is very fast.
Quote:
Generally speaking, tracks encoded at low bitrate seem more likely to sound better than those with higher bit rates.
that should really be :
Generally speaking, tracks encoded at low bitrate seem more likely to sound better than those with higher bit rates when using crystalizer
Generally speaking, tracks encoded at low bitrate seem more likely to sound better than those with higher bit rates.
that should really be :
Generally speaking, tracks encoded at low bitrate seem more likely to sound better than those with higher bit rates when using crystalizer
No, it should be:
Generally speaking tracks encoded at low bit rates notice a larger improvement over those at higher bit rates when using the crystalizer, however lower bit rate encodings are still below the quality of higher encodings.
It's still faster to read a 3MB file and decode it in memory than loading a 50MB file.
You're right of course. A psycho acoustic codec would have better fidelity than straight PCM given a 1411Kb/s bitrate.
I just think that we're at a point storage-wise and computationally where hard compression of audio (128Kb/s) makes no sense at all. All IMO.
Cheers
Absolutely. Say your wav file is 50MB. The mp3 might be say 5MB. Reading in 50MB takes slightly less than a second on a fast disk assuming it's continuous on the disk. Decoding the mp3 to audio should take less than that. My best benchmark of that is to run winamp for a while and see how much CPU time is accumulates over a few songs. After it started it said 2 seconds CPU time in the task manager. I'm now in the 7th song since I started it and it's still 2 seconds CPU time total. About 25 minutes of mp3 audio apparently took less than a second to decode. Not sure how much less though, I'm still waiting for it to toggle to 3 seconds.
Edit: It took around 35 minutes. So you could essentially decode a full CD of music (70 minutes) in two seconds. Reading in the same in wav format would take 10-15 seconds from a fast drive.
I would almost see it the other way around. I mean, I could see why they went with uncompressed samples for CDs back in the 80s. Nobody wanted to build in an advanced decoder in every CD player produced. That would have raised the cost massively back in the days given where technology were when CDs were invented. Today it would probably have been better to take studio audio at 24bit 96Khz and encode it into the same space with some form of compression. That would have given us better quality and the additional hardware cost for CD players would have been small given that even cellphones can play mp3s with no problems these days and you can buy cheap players that hardly cost more than the flash memory inside it.
Of course, we're at a point where we don't "have to" compress audio, but it doesn't mean uncompressed is "better".
Last edited by a moderator:
Similar threads
