Slashdot Mirror
Audiophiles Test MP3, EPAC and MWMA
An anonymous reader wrote in to tell us that "Sound&Vision has tested three different "codecs" and compared the sound quality to a normal CD. The three are MP3 system, Lucent's EPAC, and Microsoft's Windows Media Audio V2. None could give full cd quality but MP3 was the over all winner."
LPs do sound better than CDs if the quality of the hardware is good. An audiophile mag did a test a couple of years ago to find out if it was true. They used high end equipment for both systems and used new LPs and new CDs. IIRC, the LPs had a richer, fuller sound. The reason for this is that CDs, being binary, can't reproduce as well all of the gradients in a musical tone. It's the old Digital vs Analog thing again. The same thing can be experienced when dealing with tube based amps vs solid state amps. I've actually heard this one firsthand. The tube amps have a better sound. Unfortunately, that quality comes at a price. To realize the superiority of analog systems in music, you have to pay much more for new equipment. Plus, most people either can't tell a difference (I can only in the right circumstances) or it's not worth it to them. CDs are still the best system for 99% of us.
Chris
What about VQF? I've heard good things about VQF but it just never caught on. Anyone know anything about it?
...
Bitchslapped? Give Rob a bitchslap from bitchslapped.com.
Comment removed based on user account deletion
... your stereo components and your so called 'audiophile' magic ears ?
No, No, No!!!
2N samples will reproduce a frequency N perfectly given decent playback hardware. The DAC simply outputs a level and holds it until the next one comes in. This would produce the sort of output you're talking about. The trick is that a low pass filter is put after the DAC and it will let through all frequencys below and including N and not much else. Any distortion in the original was simply added frequencies. The low pass filter removes these and creates a perfect sine wave!
Think of reproducing a square wave (which both CDSs and LPs are horrible at doing). A square wave contains LOTS of high frequencies. It has to in order to have the sharp rise and falls. THe low pass filter takes out these and creates simply sine waves.
But not all distortion sounds "bad" - that's why guitarists use valve amps, and some audio "purists" swear by them. While the measured distortion figures of expensive valve amps are higher than all but the cheapest of solid-state amps, the characteristics of that distortion are "softer", and much more acceptable to the human ear.
Umm, none of them. They all suffer at low bitrates, and even 128k is a low bitrate. Personally, I feel that those people who claim to be able to hear no difference between a 128k MP3 and the original source have cloth ears... Listen to the sibilants, guys! But try it yourself. Some testing I did a while ago suggested to me that VQF was best on music at bitrates below 128k (subjectively, the VQF 96k was roughly equivalent to MP3 128k, although the distortion effects are different).
But it's your soundcard, your speakers, and your ears - try testing with them!
This is a good point - I have obtained much better results from ripping with my plextor SCSI than with any IDE drive out there (and for you windows folks, Plextor provides a nice ripping tool with the drive)... As for soundcards - many times you have more noise intrduced by other components on the motherboard or in the case than by the soundcard itself - hiss and hum in your audio out can sometimes be fixed by relocating harddrives, changing the slot for your sound card, or finding a motherboard with better power regulation (not to mention the power supply in your case!)
"It's tough to be bilingual when you get hit in the head."
I beg to disagree.. Take a song written in scream tracker.. The entire song is, say, 800k. Now produce a WAV output from scream tracker.. the size of the WAV is BIG (lets just say annoying).. now mp3 compress it.. you loose quality and you get a file that is by no means 800k (may it's 4meg).. My point is that you can track just about anything. I have the sheet music for a number of songs.. they take about 3, maybe 4 pages to represent the entire song.. if mp3 is the best that we can get then I think we're missing something. I don't think codecs are the entire field of audio compression.
How we know is more important than what we know.
Of course, oversampling and filtering will smooth the signal back to its original state, within the signal to noise ratio offered by your bitrate (16 bit=96db, 20 bit=120db, 24 bit=144db) and the dynamic range provided by your sampling rate (44.1kHz sampling=->22.05kHz, 48kHz sampling=->24kHz bandwidth--of course, if you use an analog filter, you lose some bandwidth due to the rolloff, but oversampling should take care of this problem). However, within those constraints, the signal will be recreated faithfully.
The human ear is commonly accepted to have a signal-to-noise ratio of about 120db. I'm guessing that there are some with more sensitive hearing. Furthermore, humans are commonly accepted to have a hearing range of 20Hz-20kHz, excepting any hearing loss. Again, some can hear beyond this range. And filtering, which is done in analog on all but the most expensive of cd players, still incurs some audible problems with phase and intensity with existing cd players.
However, there are several standards proposed for audio DVD. The differences between them center primarily about compression, most standards will send stereo as a 96kHz or 192kHz stream, sampled at 20 or 24 bits. Will these produce ideal sound? Probably not. But they will sound noticably better than existing CDs. And, I feel they will sound good enough to supplant analog as the highest quality recording medium. Most studio masters are already recording using these faster streams and higher bit rates. Furthermore, by reducing difficulties encountered during the filtering process, even el cheapo $200 dvd audio players will sound excellent. Assuming, of course, that the analog pre-amplification stage is not done poorly (not a safe bet, but at least the problems are minimized).
--Be human.
this is true, a good DAC should take care of jitter, but I'd guess that anybody who is using an SBLive! for serious music doesn't have an Apogee PSX-100/DA-2000/etc... lying around :)
You've managed to forget one of the 'other' things about the MP3 format. Not only does it nuke extraneous information from the actual audio stream, but it ALSO compresses the resulting data which is, by the very nature of audio, extremely compressible. (Have you ever tried gzipping a WAV file? It's amazing. Now try the same with an MP3. No banana. Therefore, your statement of losing 9/10ths of the original audio stream is entirely incorrect.
Only on slashdot can a posting be rated "Score -1, Insightful".
NO! NO! NO! NO!
You can reproduce a bandwidth limited signal exactly through 2x the highest frequency with perfect sampling.
So what is the catch? "Perfect Sampling". Basically to 'sample' as it applies to the theory you discuss, you need to filter the input at 1/2 your sampling frequency (a perfect filter, that is, which is unattainable in real-time). Then you multiply the input by a train of impulses at the sampling frequency. An impulse is described as a signal with infinate magnitude, and no width, but has an area of exacly 1.
To get the signal out, you simply multiply the pulse string by a low-pass filter, and ta-da, instant, perfect output (the LPF is one of the previously mentioned non-realtime jobs, that are practically worthless)
You posted that you would get out a triangle wave with only the 3 samples taken. This is incorrect. A triangle wave is made up of frequency components in the fundamental, 2nd, 3rd, 4th...and on with exponentially decreasing values. The perfect filter would filter out all the 2nd, and higher harmonics, leaving a sin wave.
OK...so that's theorhetical....what's reality? You take you signal, high pass filter it (1st order) to remove DC. Maybe lowpass filter it with a 2nd order filter. Take and Analog sample (sample and hold circuit), and then convert to digital. What you end up with closely approximates the signal being multiplied by a string of PI functions (basically a string of rectangles who's area in this case is 1, just like the impulse). To play back, you send the sampled input through a filter that acts like a LPF. The problem is that the frequency components of the resulting signal are not the same as what came out. When you run the math (and please forgive me for not giving the proof), you need to boost the middle in order to get back what you started with.
The point I'm making is that sampling is a bit of an art. The real trick to getting what you want out is the input/output conditioning. I could have a 32-bit ADC, and DAC, and get ass-like performance if I start using bad op-amps, crappy caps, and 1st order filters. If the filtering IS bad, then ya, you WOULD get a triangle wave out.
OK...I think I've gone on enough.
- Dan
How does "MP3" have a higher sampling rate than CD? As for analog sources, radio?, except of course at the radio station there is a cd player or a dat. And there already are plenty of "better than cd" mastering options. Like 24bit/192khz, or Sonys SDDS or just plain old analouge. Believe me, using a data-reduced standard for studio mastering is a very very bad idea.
Bladeenc will go all the way up to 320 kbps, IIRC. However, most folks (that I know of) don't go about 256 kbps. And personally, I use 192 (although I'd NEVER go anywhere below at least 160... at 128, bladeenc has noticeable high-frequency distortion... at least to my ears :) But really, it's a matter of personal preference... encode a couple tracks at 128, then 160, 192, etc, and with different encoders to see which you prefer... it takes a bit of time, but if you intend to do a lot of encoding, it's worthwhile. I know. After encoding about 8 of my CD's, I realized that my 128 kbps MP3's sounded like crap and I had to re-encode. :)
So the test would be meaningful wouldn't suck.
Yerp.. and then you can have a GPL'd SDMI..
How we know is more important than what we know.
I don't think that MP3 is going to be (or is right now) very popular with people who are very concerned with audio quality. 128 doesn't even cut it for me, and I'm not very picky (as in, my whole stereo system probably costs about $250, excluding my computer). People with two, three, or four thousand dollar stereo systems just don't want anything "compressed" at all.
The weak link is my system sounds like it's the sound card (I have "gold" SB 64) from what you say, but I don't really notice it and I don't really care. I use MP3 (mostly at 192), because it sounds all right and it's very convenient to me.
If you go beyond simple A/B testing, i think a lot more people migth have heard the differences. When my father had various $2k+ CD players on loan from a couple of local shops, the most revealing thing was often going back to our old CD player after a couple of weeks. And this was on a $10k system. Five minutes does not, alas, tell the whole story.
Portable mp3 makes possible a world that I would love to live in. Imagine
.. They just didn't seem to be able to grasp the concept. I'm not advocating a pirate radio station.. there are legitimate mp3's.. LOTS of them.. and this post is about distributing them. Funny, misc.int-property didn't seem to take it too well either........
you have your portable mp3 player that costs less than bread and has a
wireless receiver system capable of 11Mbs.. now add to that tommorrow's
compression standards that allow you to transmit a complete song in about
1/2 a second. You have a number of base stations that are constantly
transmitting songs, in the order of 180 songs a minute and there's
multiple stations in any one area. You turn on your portable mp3 like
device and pick up a song, listen to it for as little or as long as you
like and go grab another one.. maybe you really like this song, so you
send it to a friend and he sends it to a friend.. songs get transmitted
around the world in a few hours and we have a hit.. a hit caused by people
actually liking the music.. not by record companies buying up CD's to get
on the charts to get more air time.. When air time is unlimited and more
than any one person can ever listen to, music will be free. (that's free
like freedom folks.. but who doesn't like free beer?)
This caused a bit of a rucass on rec.music.makers.piano
How we know is more important than what we know.
Sure, the best LPs sound better than redbook CDs, but that does not mean that "digital sound" per se is somehow lacking. Crank up the data rate, and it will sound better. Indeed, this is what the dvd-audio standard is all about. However, i do remember reading in an audiophile mag some years back that the sound quality of LPs equal about a 400khz sampling rate in detail (not in frequenzy respons though). This means not even 196/24 will be enough. Maybe SACD will.
Are you sure? I use grip and it not only adds ID3 tags, it has CDDB support to automatically get the tags. In fact it's a pretty nice CD player even without the MP3 capabilities :) Daniel
As was noted, sound card manufacturers want to make money more than music, so things like the SBLive don't exactly follow the specs. The SBLive says it has a S/PDIF over coax output, but its output is not properly terminated internally, and leaves open the possibility of a ground loop, also it outputs at too high of a voltage, according to S/PDIF specs. Also, I would not describe the Live! as a low jitter device, and large amounts of jitter do become audible on beter output systems.
No, you need exactly 2N samples/sec, and any more are superfluous. This is what makes Nyquist's Theorem so useful.
What you are doing here is (incorrectly) interpolating values between samples. If you don't do that, and assume that it is a sine wave, you don't have this problem, and the signal is exactly reproduced with no errors. Re-read Fourier :-)
Only if you are interested in inaudible frequencies. I'm all for overkill in this regard, but even 50-60kHz is overkill. 96kHz was probably chosen because it is an even multiple of 48kHz (the DAT sampling frequency).
Bear in mind that a very high end ADC might give you a 120dB range, which is almost 20 bits. Also, simply "chopping" the bits isn't a very good idea either. Apogee's web site have some quite good info on this; check it out for more information.
Well so far through all my musical experiementation I have discovered that anything of 20kbps or below really sucks for anything other than previews. However, there is a nice zone between 24 and 64 kbps for 22khz mono/stereo, especially with the cheap-assed equipment all around my home. Y'see, I was brought up mainly on tapes, and vinyl via a naff turntable. CDs were a latecoming technology (hell, you should see the computer i'm writing this on... :). Even a good quality audio tape peters out around 12khz frequencies, so when I first heard MP3s I wasn't so much struck by the quality drop from CD as the lack of noise compared to tapes! All of my MP3 collection is made up of 22khz files, mainly 56 or 64 kbps stereo, a couple 48 stereo and a few older 24 & 32 mono files (2mb for 10 minutes of OK sound!) back from when my hard disk was laughably tiny. I would suggest to many people to try 56k, especially with a good encoder such as L3Enc from frauenhofer (it is hella slow though, and to get the shareware to work you need your system clock to be in 1996), and move up to 64 if the quality needs a boost, or down to 48 if you can't afford the space. Instead of 10 or 11 hours on a CD, why not 20? Why not a whole 24?! :D Try it, you may like it. Especially if you have cloth ears like me. (Though that said, mine appear to be like fine satin compared to the sack-canvas of my family and friends... I can pick up much more distortions than they can). 64k seems to be an almost bomb-proof codec, in the past when I think I've run up against artefacts in the encoding a check of the original CD shows that they were always there! Only problem is, my only 64k-capable encoder (not L3enc) keeps crashing all the damn while. So, has anyone else gone down this route yet? How was it for you? Be interested to know if I'm the only soul on the net trying it... works well coming out of my sb16 clone into a $80 hi-fi all-in-one with it's funky 10-watt speakers (a step up from an older 5W system).
MP3 got first? Big surprise there.
However, I thought they'd be a bit closer to
CD quality..
Oh, shit, a preview button. Why the hell didn't I notice that before? DOH! (Slaps head). Apologies everybody. Is that a new feature or have I been stupid for a lot longer than I think? Hironymous Howard
I wonder how the music industry is going to dance around this one?
If I rip a CD to WAV, encode it into MP3, and then compare the sound of the WAV vs the MP3, I can hear the difference. It's often subtle, but it's noticeable (my soundcard is a "consumer class" Turtle Beach Montego, which while much better than a lot of consumer cards is still in a "consumer" range of price, and was included in some Dell machines as "premium sound"). Using computer speakers, I doubt one could tell the difference, but even into my Aiwa mini-system or good headphones, there is a difference.
Distinguishing between different codecs can be a little more difficult but not impossible.
People talk about mp3 being so low of a quality, but I think that the quality isnt too bad... and I have my computer hooked up to my stereo which is fairly high quality. What I use to make my mp3's the best quality is audio catalyst with the bitrate set to variable at the highest quality and use stereo, none of the joint stereo crap. It comes out pretty good with almost no difference from the cd itself.
Too bad they only tested at 128 kbps, I encode at 256.
Although the artical states that the test CD's were set out to testers who used their own equipment to perform the tests, it doesn't state what sort of equipment these testers had. This could make a large difference as the more expensive the components of this set-up, particularly the CD transport and amp the better they will be at reproducing the sound. This will make any differences and distortions more noticable. Fair enough, but most people do not have super-expensive equipment (
I use a much cheaper stereo than sugested above (£400), and am in the process of encoding most of my CD's to MP3, which provides a much quicker and easier way to browse the tracks, as well as amazing random and programable play functions. These advantages are one of the main reasons that people will move over to MP3. On my system there is a noticable difference, but only just. And off my PC you get 4 speaker enhanced stereo :).
What's a good free (as in no money) Linux mp3 maker? bladeenc as I recall only converts wav files to mp3, whatabout ripping CDs?
You have a valid point when comparing the analogue outputs, but external sound modules such as Yamaha's MU-10 get around the noise problem by externalizing the card. Admittedly, the price is higher than a consumer card, but not by much.
The SoundBlaster Live has a surprisingly good analogue output, even in the value edition. If you rip your tracks from CD directly you will get more than acceptable results from even a Value. For the audiophiles, the SBLive (full) has an SPDIF digital output, which is a fairly standard interface on high end amps/powered speakers. Third party manufacturers (such as Hoontech) also make optical digital output daughterboards for the Value for exceptionally low prices. Coupled with the Live's internal 6-point sample interpolation, this gives a consumer level card professional quality output for an amazingly low price.
The only drawback is that the sample rate of the digital output is fixed at 48kHz, which is not a standard rate (CD's play at 44.1kHz) - you'll need to make sure your equipment can handle this rate before splashing out. Hoontech also manufacture an affordable digital amp which can handle this, and I'm sure if you looked hard you could find plenty of others.
One reason I can tolerate Windows for the
timebeing is because BeOS doesn't yet have
support for isynchronous USB, which means they
can't support USB digital speakers.
For those who haven't used these yet, they're
great! You don't even need a sound card, the
audio goes straight from your computer to the
speakers, no analog translations at all.
But be prepared to need a bigger hard-drive.
You'll want to encode your CD's no lower than
192kbps, or you can really hear the problems
in the encoding. I have found certain CD's need
to be encoding at the max bit rate in order to
get rid of the high-pitched 'whispy' noise.
But you'll be blown away by the sound quality
with USB speakers... I don't even mind that mine
have "Microsoft" printed on them anymore...
-WW
--
Why are there so many Unix-using Star Trek fans?
When was the last time Picard said, "Computer, bring
It'll never happen. The concept of SDMI cannot work without either infringing upon:
(a) the right to privacy of the listener
or
(b) the freedom to write open source software
or
(c) the freedom to use _any_ digital audio equipment of any kind
Here's why:
Either (a) the player will need to 'report' back to a central licensing authority (over the Internet or phone lines), or, if that proves unworkable, (b) the file format and players will be proprietary secrets so that a programmer can't just write a program to remove the copy-protection information, or, finally, (c) the record companies, realizing that (a) and (b) are insufficient, will lobby the government to make _all_ computer digital audio programs illegal unless they incorporate SDMI.
Don't laugh and think it won't happen-- it already has in the case of hardware (BY LAW, you can't sell a DAT recorder to the public that doesn't incorporate copy protection). Don't think they won't try to get the government to make non-SDMI programs illegal as well.
So, no thanks, SDMI is a broken concept that can work only at the expense of my freedom. I think I'll stick to buying CDs at the store-- the sound quality is better anyway.
I've encoded at 128, 192 and 256; first, to hear any difference, you must have a good (shielded?) soundcard hooked up to a full stereo system with full DSP (digital surround) capability, or a really nice soundcard (like SB Live) hooked up to an amp with surround speakers. In tonally rich music, say, Bach on church organ, you can totally hear the difference up to 256; for pingy dance music, however, even 112 can suffice.
Just my experience.
> IMHO: A good move for the industry is to move to MP3 for the recording and mastering. Any ideas? A
> multi-track MP3 player?
Sorry, this is a horrible suggestion. The idea of studio music production today is repeated mixing/layering, which requires that the same signal be played and re-recorded. Doing this with MP3 would require re-encoding the same signal many times, which would result in a considerable loss of quality.
Also, since MP3 involves converting back and forth from the frequency domain to the time domain, the temporal accuracy of MP3 is rather poor. What this means is that if you tried to make a multi-track recording of a performance, you'd probably find out that the beats don't line up between the tracks. (non-compressed audio can synchronize to a sample, which is either 1/44100 or 1/48000 of a second)
Finally, for today's computers, the processing required to do MP3 encoding is so great that even encoding _one_ track in real time is impossible on all but the fastest machines. This would drive up the cost of multi-track recorders greatly. On the other hand, even the old 680x0 Macs were able to handle many tracks of non-compressed audio easily.
Yes, MP3 can be very good. The problem is that, due to the fact it is patented, the availability of good encoders is poor. (you have to shell out a lot of money to get a good encoder; the free/cheap encoders are all crippled)
Most people are using shitty encoders to make MP3s today, and the quality shows.
There's a difference between "better sound" and "faithful reproduction". The test for a storage medium should be whether or not a trained listener can distinguish between the signal going into it, and the signal coming out of it.
Now, a test for LP/CD is to:
1. Play an LP on your player, digitize its output signal, and burn a CD of it.
2. Connect the LP player and the CD player to an AB switchbox going into your amp.
3. Conduct a double-blind test to see whether differences are detectable.
4. Repeat steps 1-3, but starting with a CD and mastering an LP from the player's output.
A test like this will help to show whether the CD actually loses information that was present in the LP, or whether it's just that the characteristic distortion of an LP sounds "good" to your ears. You could do a similar "live" test with a back-to-back ADC and DAC at various resolutions and sampling rates, to see exactly where the quantization errors became audible. Somebody's probably done this already, but I don't have any links.
Aureal's reference quadzilla has it, so do some others. Zero noise.
:)
Anonymous Coward, get it?
Anonymous Coward, get it?
Not bad spelling, bad typing
...their testing methodolgy was very good. (double blind, random sample etc) Did you even go and read the entire article? :) I did :P
:)
Anonymous Coward, get it?
Anonymous Coward, get it?
Not bad spelling, bad typing
Dunno what 'other' conector you could be referring to.
;)
:)
I recently aquired a reference aureal quadzilla though, which has spdif on toslink. Now if oly I had a stereo that had this
Anonymous Coward, get it?
Anonymous Coward, get it?
Not bad spelling, bad typing
It's pretending that it's better than it really is - so teh default bitrates are half those of mp3... So everyopne's using the default setup and finding it's not as good as Mp3...
For best sound quality use FhG at 160Kb/s+ (at bitrates less then 160K, it uses joint stereo at a 16khz low pass filter). If you can't afford (or dont want to support) FhG, use LAME (a recent beta). Lame is almost as good as the Fhg encoder, and does a bit better at higher bitrates sometimes. Never use Bladeenc or ANY of the other unfixed ISO based encoders at ANY bitrate. The ISO encoder has a number of BAD bugs (check LAMEs changelog) with will degrade sound quality even at 320Kb/s! Lame has these bugs fixed, and you can tell it not to use Jstereo and 21SFB clipping (it defaults to them upto 160Kb/s) which is what the bladeenc homepage complains about. I'd suggest you dont use VBR with ANY encoder right now, because none of them do a very good job of it. The latest lame beta stuff is alot better then the other stuff out there but needs source tweeking to work corectly. Xing's (audioactive in windows) VBR makes the same mistakes that the old lame does (overrelience on the psyco model for bitrate selection). Btw- Lame (and FhG) at 224Kb/s come up in listening tests as the same as the orignal CDs. The ISO encoders don't even meet that at 320Kb/s
Both Xing's and Lame's current VBR sucks. They both cause audiable distortion in some cases because the psyco model thinks it can go lower then it can. If you really want good VBR: Get the latest lame beta (3.25) and change #define RHXX to #define RH in loop.c, then change -X3 to have a max_noise of -1db (also in loop.c) and then encode with: lame -b128 -v -V0 -X3 -ms -k infile outfile This should produce ok results. But it would be better if you just waited for a few weeks for better VBR stuff. Also, keep in mind that encoding w/ VBR REALLY slows the encoding process. It's much simpler to always encode at 224 or 256.
Mose windows users are using the FhG encoder, they pirate it. :) Unfortunatly, 128 w/ MP3 is not enough. I don't know what crack these people were taking, but professional listening tests only show results like this at 160Kb/s. FhG performs a 16Khz lowpass when running at 128Kb/s.. If their listeners couldn't hear that then they all had BAD audio systems or broken ears. Also, current versions of lame are as good (sometimes better) then the FhG encoder at high bitrates (160+) and almost as good at lower bitrates (96-128), but are noticably worse below that (not that bad though, still alot better then blade)
VQF didn't catch on for two main reasons: Encoder is SLOW (slower then AAC), and the software (even player) is patented and propritary. There are additional more subtle problems with VQF. They spec it at too low a bitrate: It sounds A LOT better then MP3 at 96Kb/s, but this is because it's quant noise isn't as noticeable. It actually carrys slightly less phase and spectral information then mp3 at the same bitrate, and a lot less at 96K VQ vs 128K mp3, but the way is screws up the audio sounds better then MP3's nasty artifacts. So VQ really screws up sound stage and feeling. If they made a high quality 128K or 160K mode for it, it would be vastly superior to mp3, except for speed.
Did you read the article? Most people, in *scientifically controlled* tests couldn't really tell the difference.
When I first hooked my computer to my new stereo, I could really tell the difference between Mp3 and CD's. A song on CD just sounded better then a song in mp3, (even if they were different songs). When I decided to test a song that I had in both MP3 and on CD, I found that the audio coming out of my CD-ROM drive was about %20 louder. After adjusting the volume to be about as close to the same as I could, I could barely tell the difference
In the article, the people they had reviewing the codecs were professionals, people who worked for speaker makers, etc. (the three best listeners). They said that you could tell the difference with careful, multiple listening. In the 'general' listening sample (all audiophiles) some people even rated the compressed audio *above* the CDs (meaning that they couldn't really tell the difference).
Let me guess, your listening to the MP3's on your crapy PC speakers, and the and the CD's on your nice CD player...
Even if your running them both of your computer, you've probably got your CD-out louder the Wave Out. (even though the bars on my volume control at the same place, the sound was defiantly louder).
to get a really accurate sample, burn some MP3s to CD, and then compare, so you're using the same audio system.
"Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
ReadThe ReflectionEngine, a cyberpunk style n
why would you need an amp with 'soround sound' when both MP3s and CDs are only sterio encoded?? The raw audio comming out of the soundcard shouldn't need any more tweaking.
acutaly I have a tuner that can do that, and pure 'sterio' mode sounds better then 'simulated suround' It sounds like it's echoing in a large room or somthing, witch is exactly what it's supposed to sound like. (it sound's 'cooler' but not 'better')
"Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
ReadThe ReflectionEngine, a cyberpunk style n
> or whether it's just that the characteristic distortion of an LP sounds "good" to your ears.
You hit it right there. Digital distortion sounds like sh*t. All sorts of unpleasant sounding ringing (quantization errors) and other undesireable artifacts. Where as when an analog source distorts it don't sound as grating and sometimes sounds good.
Hardware is definitely more of a factor than MP3 for me. I ripped a couple of CDs (using Grip, CD Paranoia, and lame) and heard faint static in the background. I increased the bitrate, but it didn't help. Finally, instead of producing MP3s I ripped a track to a wav file and played that. Sure enough, the wav file which should have been a perfect digital copy of the track on the CD had faint static and clicks in the background.
This is on an Ensoniq AudioPCI 1371 that I bought specifically because several people on Slashdot had recommended it in a previous MP3 discussion.
The biggest problem for MP3s or any other compressed audio format is the lack of good hardware and the lack of reliable recommendations. If a sound card specifically recommended by Slashdot MP3 fanatics as a very low noise card can't even play uncompressed audio acceptably, how is the average consumer, who buys a computer without even knowing what kind of sound card it has, going to be able to get decent quality compressed audio playback?
The Fraunhofer codec is generally regared to be the best at encoding at 128KBps.
The claims of the audio bigots all fall apart when confronted with the scientific method of the double blind test. I would bet that the article claiming that LPs sound better than CDs was not a double blind, statistically valid test. I recall at least one audio-bigot magazine claiming that because they could not prove their audio-bigot claims in double blind testing, then it was obvious that double blind testing was invalid. CD player and LP player quality varies greatly. CD and LP recording quality varies greatly. Some LPs sound better than some CDs on some players to some people. However, all LPs sound better to the audio-bigot that paid $3000 for their LP audio equipment, because if it didn't, they would be proven to be fools for paying $3000. QED. The truth is that CDs have less noise, wow, flutter, etc. LPs may have a higher transient response, but you probably can't hear frequencies that high. To me, the positive qualities of the CD win by a long shot. Anyone want to buy some old LPs?
If you're concerned about audio quality and you're encoding your files with anything other than FhG or Radium, you're just wasting your CPU time. It was entirely appropriate for them to use FhG for this test.
What if it's a fact. Perhaps the encoding and compression of some music actually makes it "better" (remember that better in this case is defined as "less annoying for the majority of the population".
This reminds me of a similar article I read (probably about 12 years or so ago) in which a magazine ran double-blind test to determine the quality of T-120 vido tapes; IIRC the material was recorded on various manufacturer's tapes, and each recording was done three times (one for each speed, SP, LP, EP.)
The interesting thing (and the only reason I remember the article at all) was that the "regular joes" viewing the tapes frequently rated the EP recordings as giving the highest quality picture (EP always gives the lowest-quality picture.)
I think it's kind of like wine-tasting; if you get Joe Blow off the street to do a "blindfolded taste-test" with $100/bottle wines vs $10/bottle ones, you probably shouldn't place too much emphasis on the results if s/he picks the $10 bottle.
ATRAC encoding (used on MiniDiscs) sounds a helluva lot better than any other lossy audio compression I've heard, I wonder why they didn't review that?
In fact, on my system at home (Cyrus Amp Pair, Apogee DA-1000E, obscenely thick cabling, and home-assembled ear-tuned speakers), I find it hard to discern between the MiniDisc and the original! (Ok, this is definitely flamebait in the audiophile crowd but I can probably get away with it on /.)
Perhaps with one of those soundcards that feeds, say, a mini-jack to RCA connector, yes. But not with my Awe64 Gold. I just used double male RCA jacks to cross connect this with my stereo, and the sound is very, very clear. No distortion at all. Even the GUS MAX I have (when using the line out) is fairly decent, if not ultra-awsome. :-)
--
Internet Explorer (n): Another bug -- that is, a feature that can't be turned off -- in Windows.
Since MP3 is the method I have choosen for my newly founded music/distrobution label, I am happy that it is the highest quality.
But the real question is, is the sound actually better (ie, closer to the source), or just more appealing to the ear?
I often hear people claim that analog media outperform digital for reasons like "a binary signal cannot possibly reproduce all the gradients in a musical tone", usually these are people that have not encountered Fourier Waveform Analysis and Nyquist's theorem (which states that if you want to exactly reproduce a signal of bandwidth H, you only need 2H samples per second).
Let's face it - digital mastering levels (24-bit, 96kHz) give a theretical S/N ratio of 144dB (using Shannon's equation), and faithful reproduction of sound up until 48kHz. You are telling me your ears are more accurate than that? Wait, let me put that in perspective - noise from a Harrier Jet engine at 1 metre is roughly 140dBW/m^2, and a silent room is usually around 20-30dbW/m^2. (For those curious, the figures for CDs are 96.3dB and 22kHz).
The phenomenon you describe is due to the fact that when tubes distort, the sound is nicer than that from a bipolar amplifier. This is because the distortion harmonics are greater on even harmonics rather than odd harmonics. For various psychoacoustic reasons, that sounds "better". MOSFET based amplifiers also have even harmonics when they distort, but are more difficult to get as linear as a tube. But they do make a top class bass amplifier.
For a soft sound, I like to mount my CD player on sorbethane. For a sharper sound, I use metallic spikes. Mounting it on a Rimu table I found gave a solid sound. My favourite is folded hundred dollar notes under each foot, which gives a very rich sound. And don't forget to circle the edges of the CD with a green pen to dull the internal reflections from the laser!
You can hook up your outputs to your soundcard
and use something like wavrecord (I use it
with its X GUI wrapper xltwavplay) to record
the tracks into a WAV file, and encode that.
You're going to have to go through an analog to
digital stage no matter what, and presumably
that will be with your soundcard.
In theory, theory and practice are the same; in practice they're different. (Yogi Berra & A. Einstein)
What if it's a fact. Perhaps the encoding and compression of some music actually makes it "better" (remember that better in this case is defined as "less annoying for the majority of the population". Possible, but this test doesn't allow you to say that. Notice that none of the error bars clear the 0 line in the positive direction, while there are clearly those that do so in the negative direction. Yes, these are 95% confidence intervals, but keep in mind that n=16 after a selection process that threw out about the same amount of other listeners. So you can say that some codecs actually enhance audio, but this test doesn't back you up. (It doesn't convincingly refute you either, though. :)
Just plug the Record player into the back of the computer (mic port, line in) and record a .wav file. Make sure you use 44.1khz, but Records arn't sterio so you can you can record in mono (the mp3s will take up 1/2 the space).
check the wav file to make sure it sound nice first though
if you want a really good burn, you might try a CD-recorder, so that nothing hits the computer untill it's already digital. since FM interferance inside a computer can be pretty bad
"Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
ReadThe ReflectionEngine, a cyberpunk style n
The noise from the rest of the computer is pretty bad. On my box I can hear the HD and even video memory out of the speakers if I turn the volume up enough.
I think the best solution would be to use a S/PDIF(sp?) for digital output to an amp, does anyone know how to do this well? my stereo has fiber optic inputs, are there any sound cards that have these?
"Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
ReadThe ReflectionEngine, a cyberpunk style n
Not all bootlegs are low quality. Some bands (perhaps the Dead) allow fans to plug straight into the mixing console at live events. An experienced audio engineer with a DAT and a good equalizer could make a top quality disc from a live event.
--
Gosh, I'm going to get nostalgic here. I started
.5KHz.
reading usenet back in the mid-80s, and I remember
a sure way to start a long flame war in the audio
newsgroups was to say "digital's better than
analog" (or vice versa) or "solid state's better
than vacuum tubes" (or vice versa). The way I see
it, the only real way to test is to have LIVE
performers in a room to be listened to, then
listen to recordings made by different equipment
and or methods and see which sounds most like the
LIVE recording.
While I'm not an advocate of analog in particular
I always found one flaw in the sampling rates
proposed on the basis that humans can only hear
up to 20K cps, and that is that presumably a human
can hear beat frequencies generated by sounds above 20K Hz. I.e. a 22KHz and 22.5KHz sound
would produce an audible beat frequency of
In theory, theory and practice are the same; in practice they're different. (Yogi Berra & A. Einstein)
I'll stick with my $600 stereo setup, with a 100watt 5.1 Dolby digital receiver, thank you :). Sure, the sound card puts in some line noise but it's not *that* bad. Your USB speakers are probably just better speakers.
Another problem is with USB speakers is that you can't do really high quality mixing for sounds, so if your listening to sound from a game that uses multiple sounds mixed together, they will get muddled by CPU mixing (sound cards have special hardware to mix sounds) if your just listening to MP3s it should be OK though.
btw, you ears *can* hear digital waves, IE 1bit samples. this is the way a lot of old PC demos used to output sound through the PC speaker. (you mentioned this in another post)
"Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
ReadThe ReflectionEngine, a cyberpunk style n
Beware, I had to replace an SB64 with the ES1371 because of the same static noise you mention. The card was crap. The new SB64 with ES1370 chipset I got is perfect: no static at all in the same PC. I think the static was generated by the 3D feature of the ES1371.
As I understand it, the war between vacuum tube and solid-state amplification has nothing (well, very often nothing) to do with the merits of analog vs. digital sound. Solid-state amplification doesn't mean digital sampling, processing, and conversion is taking place, only that tubes aren't used to drive the output. Both signal paths are analog; some people just love the smooth, warm tone from tube amplification and the small artifacts it embeds in the signal.
Of course, these days, it's possible to buy an amplification system that digitizes the input signal (at some point) for equalization or other signal modification, and then converts the signal back to analog for output. Many guitarists argue that digital effects suck the tone out of your music. I don't necessarily agree; bad effects suck the tone out of your music.
--
There are over 65,000 sound levels available on a CD (16 bit), they can produce any wave that a record player can (and probably more). Records are not only analog, there physical, and the degrade everytime you play them. Also, do they have sterio capability? (and if so, how does it work).
The only reason that anolog sounds better then digital is people have convinced themselves of it. I'd be willing to bet that if it was a 'double blind' test the results would be the same, if not better for the CD
"Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
ReadThe ReflectionEngine, a cyberpunk style n
Bzzt. No cookie. This is a common misconception - people can't understand why anyone would encode at 192kbps until they realize that a CD is 1411kbps - 44.1khz x 2 stereo channels x 16 bits per channel = ~1411kbps. So, encoding at 128kbps means losing over 90% of your audio data. Now you see why every bit counts, and why 192kbps is so much better than 128kbps.
I think there is a world market for maybe five personal web logs.
>Also, do they have sterio capability? (and if so, >how does it work).
The second time I read this question here, unbeliavable.. there were quadrophonic, true 4-channel LPs already in the 70s, and OF COURSE the 33 rpm LP format does stereo.
>There are over 65,000 sound levels available on a >CD (16 bit), they can produce any wave that a >record player can (and probably more)
Pardon?? Analog, by definition, means an infinite number of values/states. Cutting that down to a limited number can never produce "more".
>The only reason that anolog sounds better then >digital is people have convinced themselves of >it.
Haha, first this is simply nonsense, second it comes from someone who doesn't even know the obvious fact that LPs are stereo.
Anyway.. in the second half of this decade, some very careful investigation has been done to find proof for what people were saying about the differences between LP and the CD.
It turned out that the engineers in the 80s were far too much concerned with frequency response, and too less concerned with impulse response. CDs have a brilliant freq response, but impulses get seriously "weakened" mainly because of the abprupt filter cutoff between 20 KHz (max. recorded frequency) and 22 KHz (1/2 sampling rate, more than this must not be fed to the D/A converter or it will distort badly).
The effect can be heard particularly well with percussions and acoustic guitars when the strings are plucked. Impulses get also "flattened" by speakers very badly, but solutions for this have a t least partially been found (transmissionline principle and others).
When it was noticed that impulse reproduction might be more important than an accurate frequency response at any cost, D/A converters with a Spline algorithm were developed that trade off the latter for the first. The quality enhancement is supposed to be fantastic, however the last time I checked, the cheapest Sony CD player with a Spline converter retailed for at least $700.
By the way, whats still "analog" about CD is the timing.. only pretty high-priced players have a small built-in RAM to read the data to and then play it from there. Most players just rely on the drive motor itself for timing; the out-of-sync, now invalid data that pops up occasionally then shows up as distortion in the output signal.
So the CD format is theoretically pretty good, the problems stem from the A/D and D/A conversion processes. Most CD productions from the 80s were of poor quality and even until a few years ago, bands that were not blessed with big bucks from a major label frequently complained about bad CD mastering.
Reciting the suggestive "digital==better" crap from the ads is just naive.
nj@realworld.de
You assume that the "quality" is LINEAR with the
compression rate.
That's not true.
Even the Fraunhofer Gesellschaft says that
beyond 128 kbit/s there's no much difference anymore.
Frank
Because when you encode you can transfer files to other people on the internet so that they can burn Audio CDs without having access to the original CD.
--------
42% of statistics are made up on the spot.
1. picked 3 different tracks from different CDs
2. ripped them to hard drive
3. encoded them to 128 kbps and 256 kbps (that makes 9 tracks total)
4. decoded the MP3s to wav
5. burned the tracks at 2x to reduce jitter
6. listened
Hardware/software is important in this sort of thing, so here's the list:
CD-R drive: Plextor 4/12 (great digital audio extraction)
DAE software: Exact Audio Copy v0.85
CD-R media: Mitsui silver
Encoding software: BladeEnc
Decoding software: Winamp 2.5C (not optimal, but had it handy)
Burning software: CDRWIN 3.7E
My stereo:
JoLida JD302B integrated tube amplifier with Svetlana Mullard copy EL34 tubes
Marantz CD48 used as transport
MSB Technology Link DAC, connected with a Canare digital cable, Audioquest Jade interconnects
Triangle Zephyr MkII loudspeakers
Harmonic Technology Melody cables (8'), single-wired
The tracks:
1, 2, and 3: Johnny Frigo with Bucky & John Pizzarelli - "Stompin' and the Savoy" - Live from Studio A - Chesky Records
4, 5, and 6: Widespread Panic - "Chilly Water" - 8/8/99 (audience taping, source: Schoeps M222 > MK-4Vs > Lunatec V2 > HHB DAT > Zefiro ZA2 > CD)
7, 8, and 9: Widespread Panic - "Party at Your Mama's House" (aka "That Thang") - Til The Medicine Takes - Capricorn Records
The Results:
First, my disclaimer. This test is far from scientific, and I don't claim to be an audiophile. My room acoustics leave much to be desired. Plus, this test is not blind, although I'm not going to lie to you...
First I listened to the CD track, then the 128 kbps, then the 256 kbps. Of course, the CD track was used for the reference. Listening to the CD track and then the 128 kbps track was bordering on pain! The difference was huge. I won't attempt to use any of the audiophile jargon, I'll just leave it at that.
I expected the difference between 128 kbps and 256 kbps to be fairly small, but I was a little surprised. I had to listen carefully to tell the difference between 256 kbps and CD audio.
Conclusion: for those who don't listen to music seriously and/or who listen on mediocre systems, MP3-sourced discs should fit the bill if you're looking for convenience/cost effectiveness.
But did I really need to go to all this trouble to convince myself of something I already knew?
-Drew Boyles-
dboyles@resnet.gatech.edu
-- "Complacency is a far more dangerous attitude than outrage." -Naomi Littlebear
"The fact that some average scores came out positive, implying that the codec version was consistently less "annoying" than the original, is probably the result of the averaging procedure"
What if it's a fact. Perhaps the encoding and compression of some music actually makes it "better" (remember that better in this case is defined as "less annoying for the majority of the population".
This is not completly unbelivable since if a codec removes "noise" and other stuff from a track.
So, the next time some annoying audiofreak "vinyl-is-better-than-cd-is-better-than-mp3" speaks you can say I always encode my music with MP3, it's simply more enjoyable that way...
128kbps is bad, I agree. However, there isn't too much reason to encode at 256kbps. A CD itself is only 172kbps - so an MP3 at 160 or 192 should be fine. I don't think you really add anything between 192 and 256.
People seem to forget that the only way to distinguish the codecs is with very good equipment. Consumer class soundcards certainly don't fall into that category. So as long as you don't burn the mp3's (or whatever) to CD and listen with a Good player it doesn't really matter.
CD-Rate is 44.1 kHz x 2 x 16 Bits = 1411.2 kbps.
There are no consumer sound cards on the market that even come close to the output quality of a halfway decent component CD player. One of the main reasons for this is tons of electromagnetic noise inside the case, but also just because sound card manufacturers like to make money. I can pretty much guarantee if you've got a halfway decent stereo hooked up to your computer the weak link is the sound card.
That said, the best way to try to improve sound quality is make sure you've got one of the better cards on the market. Some good tests can be found here.
Also, for some Linux specific issues, the Audio-Quality-HOWTO is a good source.
with your own studio, and band :)
"Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
ReadThe ReflectionEngine, a cyberpunk style n
any standard lossless data compression scheme should be able to compress audio to 60%, or even more zip, gzip, rar, whatever
"Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
ReadThe ReflectionEngine, a cyberpunk style n
eom
I'm a little confused, why encode you mp3's at all if your just going to be burning tracks to a CD? just copy tracks strait from one disk to the other (unless you're burning mp3 CDs)
"Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
ReadThe ReflectionEngine, a cyberpunk style n
I agree mp3s are good for burning onto cds, but I wish Sony or Sharp,et al. would make a MiniDisc "drive" that would appear as a normal fat filesystem on your ide chain but would take mp3s or wavs and encode them into atrac(sony's md compression) I can't see carrying a discman around with me and right now copying mp3s onto minidisc is inconvinient. Imagine another 74min of music for $2, compared to what $50 for 32mb flash for the rio. Sony would kill the mp3 players if they bundled this "drive" with a portable MD player for ~$250
Shine on, you crazy diamond.
Xing is by far the fastest software encoder out there... and the current 1.5 version doesn't cut out >16 kHz frequencies unless you tell it to. I generally use -V 100, giving me a VBR file that generally averages about 192 kbps.
LAME isn't as fast, but it works quite well (and it can use pipes for input and output).
--
Oh, no! You have walked into the slavering fangs of a lurking grue!
In a test this in the swedish magazine Mikrodatorn this spring (sorry, no link AFAIK), a whole batch of cards where tested (even including a pricey Turtle-Beach card), and all cards scored 50-60 dB, except for the live!, that scored somewhere above 70 dB.
A s/n ratio of X dB can be explained like this: The strongest sound that can be produced is X dB loader than the unwanted noise. So if we set the noise to 0 (remember logarthims, this doesn't mean no noise...), the loudest sound we can play is X dB. If my memory serves me correctly (not always the case, please feel free to correct me on this one) 50 dB is something like a quiet street, 70 dB more like a load conversation.
A signal to noise ratio of 50 dB means that the noise will be noticable even when listening at rather low volumes, and is simply unaceptable. 70 dB is a pretty decent score, but not a good one by far.
This means that people are probably mostly judging the quality of their soundcards, NOT of the codec. Personally, I bought a live! value, and a Hoontech (possible misspelling!!!) daugterboard, used the Hoontech's optical out, and plugged it in to my MiniDisc player, which has a pretty good D/A converter. This is not the simplest/most elegent solution, but it works...
The reason why this should sound better is that there aren't as many EM-fields as inside of a computer case, an EM-fields are a NIGHTMARE when dealing with D/A convertion.
If anyone has more test data of s/n ratio of modern soundcards, please post them, I know that the article I am referring to is unavailable to most, not to metion rather old. More data on the subject would be nice.
Try out fish, the friendly interactive shell.
The mp3 format is hardly the limiting factor of my music enjoyment. Consider:
1) I only have the cheap harmon/kardon speakers that came with this machine.
2) I Never just listen to music. Music just plays in the background of the rest of my life.
3) I don't care enough about music to invest tons of money. I don't know what sound card I have, only that it came with this machine and it works under linux. When I buy a cd, it's usually from some used cd store here in town.
Music is just pleasent noise to listen to as you do something else.
I haven't seen any discussion of LOW bitrate encoding. From what I can tell, all 3 codecs do a fairly good job at 128kbps. Good. Whoopidy do. I could have told you that. It's nice to know that MP3 is a little bit better, but it was just a little bit. All three will produce output from my computer's speakers that sounds identical to the CD if encoded at 128. So what?
I'm interested in how low the different codecs can go and still keep that characteristic. Which codec can encode at the lowest bitrate and still sound like a CD on my computer's speakers? And which codec sounds the best when streamed over a 56k or 28.8k modem? This is probably more important for the overall computer industry. Audiophiles will probably almost always keep the audio on true CD's, and may now occasionally be willing to do some encoding with 192kbps MP3's for greater convenience. But the rest of us want to know which codec to pick when we want to check out a song over the internet and don't have T3's going into our home computers. At 32kbps (56k modem speeds), who sounds best? And which codecs can take a hit from a late/lost packet (graceful degradation)?
The test was valuable. 128k is great for my personal collection. But most of the Internet music industry is running at a much lower bandwidth. Which codec is right for them?
Time flies like an arrow. Fruit flies like a banana.
You're as bad as the fellow saying 'LPs aren't as good, do they even have stereo capability?'
The fact is, LPs can substantially outperform CDs in certain ways and decidedly not in other ways. In order to so blatantly outperform CDs, you have to completely overhaul your system- in particular, the playback system has to do something to get a handle on the low frequency inadequacy of 99% of turntables, and it has to have high frequency extension to waaaaay over 20K because among the additive distortion in that region is important information. Finally, you have to play a suitable record- it is very easy to find records that push the extreme high end, but much more difficult to find ones that attempt to present low end accurately, and half the time you're better off looking for the minimalist recording techniques of the 50s and 60s.
Does this go some way toward explaining why you don't know what you're talking about? It's very unlikely that you have ever experienced an analog playback system worth listening to. Have you listened to openreel tape playback at 15 or 30 ips, or is your experience with tape likewise limited to cassettes?
ObCompression: I can get better results out of mp3 than I've heard from any other codec including the Quicktime Qdesign codec. That's not to say I always _have_... I need to make some more equipment to do this... but IMHO as a hardcore highend system designer audiogeek mp3 is as good as anything. To maximize its audio quality, feed it an enhanced analog signal that precompensates for the known losses in the process: give it the analog over heavy cables with good equipment, you want to be giving it the hottest transients possible and not softening and blurring them. Doing this means the transients can be glossed over by the compresssion losses without coloring the rest of the sound- using shoddy cables for your analog paths is a really _horrible_ idea especially if you mean to record only 128k: as compression quality goes down, you have to feed the software a _purer_ signal to try and compensate for it. Finally, in order to deal with the known losses of codec compression, you need to give the analog source a minor amount of _audio_ compression because details like faint high frequency sounds are going to be lost in the codec. Ideally you want to be giving it multiband compression. Don't expect comparable results from digitally effecting CD audio- the point of this is to bring more of the original source into the 16bit 44K range of digital audio, and then to leave it alone in purist fashion. I'm not talking consumer level here, or CD ripping (just use the bits from the CD), I'm talking digital audio mastering especially for producing mp3s.
Hopefully some other people who are not too easily pleased will also take to mp3 mastering as a serious artform comparable to the old vinyl mastering. Going 'it's already perfect' is NOT enough.
Another weak link in MP3 creation is the drive used for ripping from the CD. As a general rule, SCSI CD-ROM and CD-R drives are better for CD ripping than IDE drives. My SAF CD-R 4012 (SCSI, xcdroast reports it as a TEAC mechanism) and SB32 PNP do quite well. I have the sound card as far away from the power supply as possible, and get very little noise except at extreme, blow-your-eardums-to-bits volume levels. At listenable volume, I don't notice any noise, even with high-quality headphones.
--
Oh, no! You have walked into the slavering fangs of a lurking grue!
Makes a huge difference to me. If you're like me, you're constantly convincing others as to why you need a computer with XXX MB of RAM, or a car that does 0-60 in under 5 seconds, etc. It usually centers around their belief that something inferior is "good enough." Nothing wrong with that, life's about compromises, and budgeting your needs and wants with respect to reality.
A little while ago some guy posted in another audio related article that there was some free (gpl?) audio compression program that could compress about 60%, and on decompression reproduced the original song bit for bit (if anyone knows where this is available, could you point me to it?).
Sounds like that guy is me. The compression is Shorten format (shn), and it's how I prefer to trade my CDs. A full show download (3 discs) can run up to 1 GB. Not a problem on my T3, but modem users need not apply. It's much easier/cheaper than trading traditionally, through the mail. I have the files needed on my FTP, let me know if you need them.
The odd thing about it, though, was he said that it was the only compression scheme they used to trade Grateful Dead bootlegs.
Yep, MP3s are shunned. Some folks have MP3-sourced stuff on their lists, and I don't trade with them. The serious traders won't touch them.
A bootleg itself will be low quality (I would assume), so that would be the limiting factor for compressed sound quality, not the compression itself...
In a way, that's true. However, all of these shows were taped legally (meaning the taper doesn't have to stealth, and can place mics optimally), or recorded via a soundboard patch. I have a Grateful Dead '77 show (5/21/77) that absolutely smokes, and the quality is excellent. I've got a '71 show that was taped with mics, and the sound is still great - gives more of a "you are there" sound (more crowd noise).
BTW, they're technically not bootlegs, since they're legal (not sold for profit).
-Drew Boyles-
dboyles@resnet.gatech.edu
-- "Complacency is a far more dangerous attitude than outrage." -Naomi Littlebear
And in this sample group, they only found half who could statistically tell the difference between the original CD and compressed data.
This conclusion cannot be drawn from the original quote. You cannot tell how many of the unusable responses were due to those who couldn't follow instructions, and those who couldn't tell the difference.
You may be right about "golden ears" being very rare, and MP3 being indistinguishable from CD. Then again, you may be wrong. This particular quote neither supports nor invalidates your argument.
Um, no.
If you look deep down at an actual CD player playing a CD, and compared it to the real stream, you'd find numerous errors. Most players don't use the EDC (error-detection code) at the end of each audio block, instead relying on the filters beyond the DAC and oversampling to take care of the bit-errors.
Generally, modern players read a block multiple times to accomplish a form of "error correction." There are two modes in a CD-ROM drive of DAE, though, "raw" and "cooked." If your ripper uses RAW mode (generally, faster), it reads the bit stream and makes no attempt at correction (hence, 'raw'). If your ripper uses cooked mode (which it should, for *good* ripping), the CD-ROM will try to use the limited error-correction data and multiple reads to get a good copy of the bits.
(sidenote: DAE is much harder on a CD-ROM drive than reading a data block because in a data block, the block number gets encoded into both the block buffer and the control buffer, whereas in an audio stream, the block number only goes to the control buffer - the drive can only look at the DAE and make a "best guess" at the block number. That's where the 'jitter correction' etc come into play)
Well, tubes tend to have two things going for them:
...), but hopefully, those higher order terms (these tend to be harmonics, etc) are very very small compared to the linear term - for a very small range of Vin. Once Vin exceeds certain limits, the straight-line approximation doesn't hold, and you get your lovely "overdriven" sound. Now, tubes tend to have the straightest transfer curves within the input range, while MOSFETs have a similar curve, it's not as straight as we would like.
1) More linear
2) Easier to handle higher powers.
1) IIRC, the on a linearity scale, from least to most goes: BJT transistor, MOSFET, Tubes. Linearity is important in amplification, since you want to have a transfer function as close as possible to H = n, where n is a constant [such that Vout = nVin]. Most amplifiers don't have such a linear relationship (higher order terms, Vout = nVin + n1*Vin^2 + n2*Vin^3 +
2) Tubes are much easier to work with at higher powers - they have lesser cooling requirements (generally, tubes can reach 250+C safely (surface), while transistors generally max out around 125C. Also, high power transistors are generally *very* expensive.
This is from an EE point of view, I've not heard a tube amp yet so I can't comment on the "warmness" of the sound (hmm. Pink noise?)
I've had folks check out some of my QDesign2 stuff and literally fall over in shock...especially when they see the file sizes.
I know its very un Politically Correct to say this around here, but I personally think the QDesign portables (yep, they are on the way) will be the ones to make a difference on the consumer market.
With the same amout of FLashRAM as your run of the mill mp3 player, you get 3x the music and better quality.
The (sigh) down side of this is that they intend to 'play ball' with the ever-lovin' RIAA...
-K
One day, you'll learn to watch what you post...
The music is only as good as your ears...
I personally like mp3 though. Isn't mp3 open standard and the other two proprietary?
Even worse than OS bigots. Remember this is the crowd that has convinced themselves that LP's actually sound better than CD's. They couldn't stand the fact that when CD's hit the scene for a few hundred bucks you could put their $$$K systems to shame.
I find it strange they didn't test the QDesign music codec. It's available in Quicktime (no Linux, I know, but the same can be said for Microsoft's format) and it produces *VERY* high quality files at very low bitrates. Just listen to the sound of the last posted Starwars trailer.
Obviously ripping (assuming a good ripper eg. cdParanoia :) should be perfect. You should end up with the same bits on your HD as on the CD. Of course the encoder has something to do with this too (since mp3 is inherently lossy).
You've got it right when you say the real problem is the sound card -- even an SBLive! gold is pretty noisy in analog mode, and the inside of your computer might as well be a freaking radio station. But with the Gold (and I believe certain versions of Diamond's MonsterSound, tho I'm not sure) you can output in digital, which in theory could be hooked up directly to your stereo.
At least one of the digital out's on the SBLive! Gold is a proprietary (i believe, does anyone know the name of the format/standard it uses?) plug that works with the FPS2000 digital speaker set. It's good sound for a computer but not exactly audiophile stuff. I'm not sure if there are any standard (eg. Dolby digital? SPDIF? you tell me...) outs on the SBLive! Gold.
There are however definitely Dolby digital outs on many DVD players, and you can hook those up in digital to a good component system.
All these comparements of "MP3" against other audio compression technologies are rather meaningless.
There is simple no "MP3" at all. MP3 is using a psychoacustic model for data reduction, and this model is not specified in the MP3 patents and therefore there are different models out there with varying results. I know of at least 6 models at this time:
- DIST10 The acustic model used by the ISO reference source. Said to be rather bad.
- BLADEENC Is basically the DIST10 model, but with few improvements and fixes.
- FRAUNHOFER Used by Producer, l3enc etc. Said to be one of the best.
- GPSYCHO GPL-model used by LAME. Apperently also quite good quality.
- XING/OLD The old Xing Encoder used this. Cuts the frequencies at 16 kHz. Increndibly fast compared to others, but bad quality.
- XING/NEW Apparently the new Xing Encoder (at least the linux version) use a new model, as there is a new switch for changing between cut at 16 kHz and not cut. To my tests the quality is ok.
So you see, testing just one MP3 encoder is not meaningfull. All these encoders have different qualities, different speeds. Some encoders have better sound at 128 kbps than other at 160 kbps or more. Use a bad encoder, and the result will be bad. Use a good encoder, and the difference to a CD will be heard only by trained people (these people who helped developing the psychoacustic models).
Additional every psychoacustic model will not match on all people. The human ear is just too complicated and different for a catch-all model. So even different persons may rate the encoders different in quality.
If i may offer a advise for MP3-Encoding: Use the new Xing-Encoder for Linux or LAME. Make use of variable Bitrate-Encoding. Fixed Bitrate-Encoding is bad, as the bitrate will always be to low at some very special pieces of the audio and very often just to high. Variable bitrate encoding tries to use the Bitrate just needed. I've made very good experiences using VBR and got smaller files which sounds better.
And I quote:
"All of the codecs were set to produce an encoded data rate of 128 kilobits per second (kbps), the most common data rate for MP3 files and greater than a 10:1 reduction in bit rate compared to a CD. Different data rates likely would have produced drastically different test results, especially at rates below 128 kbps. To mimic a record-company-sanctioned download, which would presumably be encoded at the highest quality available at a given bit rate, I ran all three codecs in their top sound-quality mode. "
Answer this one for me, though I think I am correct in assuming. Isn't the major drawback to MP3 the fact that most encoded from an inferior source? Most of the traded MP3s are recorded from CDs, so of course they cannot outperform the CD in quality. Most are recorded at a bit rate that will optimize downloads. Even the CDs that we buy are mastered at 16 or 20 bit. If you had an analog source directly to MP3 it should sound better than the same source to CD given that you use a higher sampling rate.
IMHO: A good move for the industry is to move to MP3 for the recording and mastering. Any ideas? A multi-track MP3 player?
pronoblem
Unless the compression is seriously changing the sound output, aren't there just too many factors to make any type of conclusion? There's the matter of codecs, soundcards, and speakers in the first place. Combinations of these are going to give you different sounds.
But besides that, some people really can't tell the difference between true cd quality, and mp3 quality. For those who can, though, what difference does it make? For example, how exactly do you know what the music is supposed to sound like? You're taking the cd as an example, but who's to say that's exactly what the artist had in mind... and most likely that's not how it would sound if the artist played it live.
A little while ago some guy posted in another audio related article that there was some free (gpl?) audio compression program that could compress about 60%, and on decompression reproduced the original song bit for bit (if anyone knows where this is available, could you point me to it?). The odd thing about it, though, was he said that it was the only compression scheme they used to trade Grateful Dead bootlegs. Does that make any sense? A bootleg itself will be low quality (I would assume), so that would be the limiting factor for compressed sound quality, not the compression itself... I get the feeling that there are many misconceptions about compressed digital audio.
He said, "You'll be able to tell your grandchildren that you helped assemble the first NT supercomputer," and I cringed.
I have been using BladeEnc because it is the only free encoder I could find...
--
An esoteric scratched itch:
Homeworld Map Maker Tool
The second time I read this question here, unbeliavable.. there were quadrophonic, true 4-channel LPs already in the 70s, and OF COURSE the 33 rpm LP format does stereo.
Well, I didn't know, that's why I asked. Also *how* does it work? are there multiple needles? some sort of Dolby-prologic type thing (I wouldn't think it would be posible in the 70's/80s). If it's multiple heads, how do you get the right head on the right grove?. I really do want to know how this works.
>There are over 65,000 sound levels available on a >CD (16 bit), they can produce any wave that a >record player can (and probably more)
Pardon?? Analog, by definition, means an infinite number of values/states. Cutting that down to a limited number can never produce "more".
Is that why a tape always sounds better then a CD? Analog sources will always have some degidation. The more degridation, the less of a 'eqivalent' samplerate/sample size they would have. As the needle moves up and down it will exsist in an infinet number of positions, however there is probably some limit as to what the human ear can tell the difrence between. I don't know if that number is more or less than 65,000.
>The only reason that anolog sounds better then >digital is people have convinced themselves of >it.
Haha, first this is simply nonsense, second it comes from someone who doesn't even know the obvious fact that LPs are stereo.
You admit later in your post that early CDs were masterd poorly. This might have caused some problems with the orgional tests in the 80s
"Subtle mind control? Why do all these HTML buttons say 'Submit' ?"
ReadThe ReflectionEngine, a cyberpunk style n
hmmmm..
there is a point to note about how you mention of "Tomorrow's Compression Standards"..
Compression of a certain piece of data is strictly limited by the entrophy of the data.. That's the actual "amount" of information conveyed by the data stream. Without loosing information you CAN NOT compress further than the limit imposed by the entrophy. All (losless) compression tehcniques are bound on this limit without any exemption.
I note this because I want to point to the fact that "One can not expect a leap in the performance of compression technology".. What can be done, has already been done. Even if there may be slight improvements indeed, they will be really small . Never ever shall we see performance increases such as 2fold or 10fold..
Unless of course, you choose to sacrifice some of the information in the data..which is the case for MP3.
MP3 or similar audio compresion techniques are called "LOSSY COMPRESSION". What they "loose" in order to save bits is the information that human ear does not differentiante. By compressing at 128kb you already hit the limit of differentiability by the human ear.
So, introducing the phrase "Tomorrow's Compression Standards", and leading to dreams of 2 fold or 10 fold increases in compression performance from there is a big mistake. The only thing that has room for dreaming is possibly the increase in available bandwidth. That's a good thing, and surely it will be there.
HSY
huseyin_serkan@hotmail.com
The explaination you were looking for...
Lets face it, tapes always sounded bad, but they were easier to use and more portable than LPs. MP3s are more portable than CDs, so.........
...your (fill-in-annoying-person-in-your-life {for me, it's "brother"}) can't scratch your MP3's like CD's by leaving them lying around your computer desk...
The magazine made a major effort to find serious audiophiles: people at speaker companies, members of "user groups," even employees at their own publication.
And in this sample group, they only found half who could statistically tell the difference between the original CD and compressed data.
Even when listening carefully, in back-to-back tests, to snippets designed to push the encoders to their limits.
I don't doubt that there's a tiny fraction of the population, well-trained, who really does get occasionally annoyed by MP3 or other encoding. But "golden ears" are very rare even among those who like to pretend they're experts. What this article really says is that, in real-world usage, encoded audio is effectively indistinguishable from the original.
Jamie McCarthy
Jamie McCarthy
jamie.mccarthy.vg
Is there a limit to how many kbps bladeenc can encode? And at what point does it not make enough of a difference?
Why was the above post moderated down?
If you're a big audiophile, you're not going to be using MP3 anyway. MP3 is for a whole different group of pepole.
The issue with MP3 is its portability. The idea is that I can encode 10 or 11 of my CDs, burn it on a CD-R, and have a nice wide selection of music I can play on my desktop PC at work. Rather than haul around (and possibly scratch) a whole stack of CDs, I just have to carry around one CD-R that if I scratch up, I just have to burn a new copy from the files on my hard drive. I don't care about quality - heck, I listen to MP3s on a Sun with 8-bit audio. But it's great to be able to stick in one CD, fire up xaudio, and have 10 or 11 CDs to pick from.
Or, MP3 is nice because of convenience -- I can get my music quickly. Suppose at home, and I'm sunk deep into coding and don't want to be distracted. If I want music, I'd have to go to my CD collection, pull out a CD, walk over to the CD changer on my stereo, wait for it to turn on and spin up, and then play the CD. OR, I can just click over to the directory of MP3s on my PC, queue up a whole bunch, and have great music without even getting up from my chair.
Sure, MP3 is nice for downloading too, but unless you have a fast connection, who really wants to sit around and download a whole bunch of 4-meg files?