Absolute best sound format?

[---X---]

What is the absolute highest quality sound format? I don't care how much space it takes up, I want quality :)

[john240sx]

320kbps for MP3 and AC3 for video. i don't have any music in other formats so it might not be as good as i think.

[---X---]

WAV is better than 320 kbps MP3. I was just wondering whether or not there's something better than that.

[Penguin4x4]

The best sound is acheived through LP's, as they are uncompressed, whereas CD's are compressed to 320Kbps, The best way to check is to see the amount per side you can achieve on an LP, and translate that onto a CDs size. If there are 4 songs a side, then 4 songs a CD-R with highest Kbps and KHz encoder size available. A sorta uncompressed digital disc, the best of both worlds.

[john240sx]

the bit rate on a music cd is well over 1000kbps, mp3's highest compression is 320kbps.

EDIT: i just checked a cd, it's 1411kbps

[Illah]

I don't believe that whole LP's are best hype. For turntablism and mixing, yeah, but as far as sound quality, no. They work like this, a groove is dug out of a slab of vinyl representing the sound. Think like this, you know how a speaker moves in and out real fast to make sound? The printing needle of a record does the same thing, and that's the groove in the vinyl. As the playing needle rides those grooves the slight variations are amplified into a speaker to make them audible. That's not necesarily uncompressed, it's nineteenth century. Sure it's analog, so purists are like, "It's more organic." I don't know about that.

Enough ranting. While doing two semesters of sound work with a sound engineer who's worked with Janet Jackson among others, here's what he uses. AIFF. It's basically a Mac version of WAV files. As long as you have a raw format like WAV or AIFF at 16 bit 48K sample rate that's about as good as digital sound gets (there's a new 24bit format availiable, but it's essentially unused). Any more would be unnecessary. Once you get over 41K samples it's pretty much crystal clear, 48K is kinda just making sure.

--Illah

--Illah

[Penguin4x4]

CDs are 1411Kbps? :eek: My bad. @Illah, what I meant was it is more of a raw sound than CD format, as it is not compressed whatsoever. I have no idea of wtf 16 bit 48K sample rate means. I only know Kbps means a certain number of kilobits per second per track.

[---X---]

I sorta meant something on a hard drive, so CDs/LPs wouldnt matter ;). Im not using a Mac anytime soon, so I guess its 48k WAV for me :D

[MospeadasDark]

Wavs are the best because they're raw data. You can also do lossless compressed wavs that are the same size as "high quality" MP3.

[Penguin4x4]

24 bit, 96KHz is also doable, provided you use a TerraTec DMX 6Fire 24/96 or DMX 6Fire LT ...........www.terratec.net........just a lil heads up, :)

[---X---]

I knew about that a while ago, thats goin in my new rig :)

[Richard]

The "best" is lossless encoding.

I have an 80 GB drive dedicated to music and a good portion of it is encoded using flac.

http://flac.sourceforge.net/
I encourage you to give it a try.

http://www.monkeysaudio.com/
Another popular format.

For what its worth. I use lame for mp3 encoding.
http://www.mp3dev.org/mp3/
http://www.r3mix.net/ is an excellent website. I advise that you read through that page pretty thoroughly if you're interested in encoding mp3s.

[flaming gerbil]

ok here we go. as far as digital audio goes .wav and .aiff are the best quality because they are uncompressed. all mp3's are compressed meaning some data is omitted. some compression schemes are better than others, but none contain as much data as a .wav or .aiff file. there is no such thing as lossless compression, compression by definition is omitting data and therefore data is "lost". now a lot of people can't hear the difference between decent file compression and standard .wav files but some of us can. i can. but then i work with audio for a living and listen acutely all the time. 16 bit 44.1 khz is the standard red book cd specification. 16 bit means the quantization and 44.1 indicates the sampling rate. this means when the analog audio was converted to digital, sixteen voltage measurements (approximations)were taken 44,100 times per second. the bit depth is exponential, so 24 bits has 256x the resolution as 16 bit. and 96khz means roughtly twice the resolution as 44.1. (doubling the number of samples per second) in the real world though differences between 24/96 and 16/44.1 are subtle. it sounds much better to me though, and probably will be the audio spec once dvds replace cds. (cd's don't hold enough data) going to 24 bit increases your file size exponetially and doubling the sample rate doubles your file size. it's expensive to store all that stuff on hard drives. i've have probably 50-60 hard drives sitting in my closet right now. kind of a simple explanation but i hope it helps:)

[---X---]

Storage is no problem for me *evil grin* hehehehehe:D

But seriously, thanks everyone sound is something i used to not know but now I do :)

[Avatar28]

Originally posted by flaming gerbil
ok here we go. as far as digital audio goes .wav and .aiff are the best quality because they are uncompressed. all mp3's are compressed meaning some data is omitted. some compression schemes are better than others, but none contain as much data as a .wav or .aiff file. there is no such thing as lossless compression, compression by definition is omitting data and therefore data is "lost". now a lot of people can't hear the difference between decent file compression and standard .wav files but some of us can. i can. but then i work with audio for a living and listen acutely all the time. 16 bit 44.1 khz is the standard red book cd specification. 16 bit means the quantization and 44.1 indicates the sampling rate. this means when the analog audio was converted to digital, sixteen voltage measurements (approximations)were taken 44,100 times per second. the bit depth is exponential, so 24 bits has 256x the resolution as 16 bit. and 96khz means roughtly twice the resolution as 44.1. (doubling the number of samples per second) in the real world though differences between 24/96 and 16/44.1 are subtle. it sounds much better to me though, and probably will be the audio spec once dvds replace cds. (cd's don't hold enough data) going to 24 bit increases your file size exponetially and doubling the sample rate doubles your file size. it's expensive to store all that stuff on hard drives. i've have probably 50-60 hard drives sitting in my closet right now. kind of a simple explanation but i hope it helps:)

Actually, that's not true. You CAN compress without losing data. Your modem does it all the time if you're on dialup. Ditto if you use winzip, winrar, etc. Also, there IS a type of audio compression that DOES allow for lossless compression at roughly 2:1. it's ulaw compression I believe, but it's been a little while since I read up on it. I suppose I could look for references if someone really wants to know. But that compression will let you compress audio and when you uncompress it, you get back every single bit that went into it.

[flaming gerbil]

that's different than audio compression. you can't replace the bits of let us say a guitar solo unless you have those original bits on your end of the download. i would like to see a link, cause something amazing might be going on i don't know about.

[Maxvla]

best i've heard is AC3... much superior to mp3 or wav files. but good luck finding songs in ac3 format. they aren't plentiful.

[---X---]

Originally posted by Maxvla
best i've heard is AC3... much superior to mp3 or wav files. but good luck finding songs in ac3 format. they aren't plentiful.

What about encoders?

[MospeadasDark]

Originally posted by flaming gerbil
that's different than audio compression. you can't replace the bits of let us say a guitar solo unless you have those original bits on your end of the download. i would like to see a link, cause something amazing might be going on i don't know about.

Two of the more popular ones.

http://www.monkeysaudio.com/

http://www.wavpack.com/

[flaming gerbil]

oh geez, i just realized i screwed up streaming formats with data compression. sorry dudes.:beer:

[Richard]

EDIT: I've been suggesting lossless as the "best", but I highly recommend you investigate the lame mp3 encoder. lame --alt-preset extreme renders CD quality. Don't just take my word for it. Test it yourself.

To give you idea on filesizes...

The track in question is, "World in my Eyes, by Depeche Mode's Violator Album."

CD rip: 45 MB
FLAC encoding: 28 MB
MP3 "insane" 320 Kb/s : 10.5 MB
MP3 "extreme" VBR : 8.2 MB

All of these sound identical. In fact, there have been tests with $15,000 equipment in which a number of audiophiles have not been able to discern a difference between 256 Kb/s and up mp3s.

The best recommendation is to be judicious in your encoding. Not every piece needs lossless encoding. For those that do (or your favorite CDs) go for it. For the rest. You have some good alternatives.

http://www.3dsoundsurge.com/ has a good article on creating better mp3s. I suggest you take a look around.

--

As was mentioned previously by myself and MospeadasDark lossless encoders are what you're after.

You'll be hard pressed to do better than lossless encoding. (If you find better please share it.)

I can't tell the difference between the original CD and the flac encodings. This is with rather expensive audio equipment as well.

Something I think bears mentioning: Ensure that your CD/DVD drive is up to snuff. If the original rip isn't perfect, even the best encoder will be useless. For windows I recommend EAC. Exact Audio Copy.

http://www.exactaudiocopy.de/

If you're in linux. A program named Grip will be more than adequate.

http://nostatic.org/grip/

Happy listening.

Originally posted by ---X---


What about encoders?

[JokerHCG]

AAC is good and WMA is promising at low bitrates but my vote goes to midi(although its not a very good comparison) :D . Just as long as your midi playback is excellent ;)

[FunkDaMonkMan]

---X--- what are you going to do with this really high quality sound you store?

[CrystalMethod]

It's better than changing CD's everytime you feel like listening to something. I have all my CD's ripped to MP3 and are stored on my file server. If I want to play anything, I just stream it over my network to either my first or second machine.

[Penguin4x4]

Originally posted by flaming gerbil
16 bit 44.1 khz is the standard red book cd specification. 16 bit means the quantization and 44.1 indicates the sampling rate. this means when the analog audio was converted to digital, sixteen voltage measurements (approximations)were taken 44,100 times per second.

So you take the quantization x sampling rate, i.e. 24/96= 24 x 96 = 2,304Kbps, to get the data rate, right?

[Avatar28]

I was unable to find the exact link where I read about it earlier, but if you go to Google and do a search for lossless audio compression, you'll find a LOT of hits. There are apparently a number of programs out there that will do lossless audio compression.

[Penguin4x4]

Which means 24/96 encoded .wav's would average 70MBytes.

[Nismo Nate]

Great Thread Guys!!!!!!

I feel smarter now that I've read it, and these are great links....

On a side note, anyone tried Music Match Jukebox? I used it.. Encoded some CD's at 160kbs, that was the highest it went. But to me it was a resource hogger, I won't be using it again.

IMHO..... I smell Sticky......:beer:

[Penguin4x4]

@Avatar28 - Would this be the link? (http://www.firstpr.com.au/audiocomp/lossless/) WavZip seems to be the "losslessest", no?:beer:

[MospeadasDark]

Originally posted by Nismo Nate
Great Thread Guys!!!!!!

I feel smarter now that I've read it, and these are great links....

On a side note, anyone tried Music Match Jukebox? I used it.. Encoded some CD's at 160kbs, that was the highest it went. But to me it was a resource hogger, I won't be using it again.

IMHO..... I smell Sticky......:beer:

Ugh! The only thing MMJ is good for is playing music. Don't rip with it. It keeps your albums, artists, et-c together, unlike winamp. I like using WMP7.1 for it's SRS in older songs to bring it out.

Also for some reason winamp doesn't like playing all my chinese and korean songs. The "shuffle" mode is broken and will always play certain songs over and over. Sure, it's small..but if you have 256MB/512MB of ram what is 13MB to that?. WMP works, and works well. If I have ram, I will use it.

[Nismo Nate]

Yeah, I didn't like it at all, the only reason i used it, is that since my new sys build, for some reason my audio grabber dosen't work with my machine, i think it might be XP compatability, i just havent' goten around to looking for an update.

I happened to have MMJ on a Max PC cd so i tested it out

I hated that it always wanted to be conected to the internet , that sucks when you are on dial-up.

I'm gonna look at all these new links here i find a new one

IS there any "better" format for certain gere's of music?

For example i listen to Rock and Trance primarily...

[Richard]

As was mention, Music Match Jukebox is trash for ripping and encoding. I'm not a fan of all-in-one utilities. They tend to be Jacks of all trades - Masters of None.

---

I've done some experimenting with the Ogg Vorbis encoder and I must say I rather like it.

It renders very high quality encodings. Plus, its an open format. It seems to have already gained a large acceptance.

If you're into HQ audio compression. It's worth a look.
http://www.vorbis.com/

Very easy to encode extremely nice sounding audio. Read through the FAQs and you'll get an idea about lossy vs lossless encoders as well.
--

Onto music players... The vast majority of players are not very good. Winamp's output can be improved easily. Give the MAD plugin a listen.
http://www.mars.org/home/rob/proj/mpeg/mad-plugin/
--

When I use windows I use a different program for MP3 listening. Known as coolplayer. It has been around for a long time, but has undergone some recent improvements. If you give it a try make sure you go into options and change the directsound output to coolwavemapper.

http://coolplayer.sourceforge.net/
http://sourceforge.net/projects/coolplayer/

---

[Avatar28]

Originally posted by Penguin4x4
@Avatar28 - Would this be the link? (http://www.firstpr.com.au/audiocomp/lossless/) WavZip seems to be the "losslessest", no?:beer:

No, that's not it. It was an article talking about different compression methods. I'm not certain, I'm thinking it was either Huffman encoding or u-law (technicall the greek mu).

[Penguin4x4]

mu-Law (http://www.hr/josip/DSP/FAQ/27.html)

[flaming gerbil]

sorry guys i was reading my original post and see part of it may could be possibly read incorrectly. i stated a 16 bit system takes 16 voltage measurements which is incorrect. i takes a 16 bit word length voltage measurement which is exponential. for example: 1 bit system would take two voltage measurements, ei.e. 1 and 0. a two bit system would take four 00,01,10 and 11. 3 bit is eight: 111, 110, 101, 100, 011, 010, 001, 000 and so on. so 16-bit (standard cd specification) would take 65,536 voltage measurements. 24-bit is 16,777,216. so you can see the difference in sound quality between 16 and 24 bit. so 16/44.1 (again stand cd quality) is 65,536 voltage measures ments taken 44,100 times per second (the sampling rate).

standard 16/44.1 files take 5 megs per minute per track. the songs you listen to off a cd are in stereo, i.e. a left channel and a right channel. so this is two tracks making it 10 megs per minute. so a three minute song off a cd as a standard .wav file would be 30 megs (at 16/44.1). increasing the sampling rate going from 44.1 khz to 96khz doubles the storage space. increases bit-depth increases storage space exponentially.

[Richard]

Thoughts about encoding quality...

The general consensus seems to be that 128 Kb/s .mp3 is rather poor. (In relation to CD quality) I tend to agree with that. Granted, if you do your own ripping and encoding you can obtain rather good results with 128 Kb/s. However, it will still sound lacking in comparison to CDs.

160 Kb/s is my cut-off for encoding. If there's a piece of music that I don't intend to listen to a lot; I'll encode it at 160.

192 Kb/s renders near CD quality. Some encodes will be nearer in quality to the original depending on the source. It is pretty much universally accepted that 192 Kb/s is the best tradeoff between quality and size.

Higher than 192 Kb/s is for all intents and purposes CD quality. An .ogg or .mp3 at 256 Kb/s or better will be nearly indistinguishable from the original.

The most demanding music tends to be symphonic music. The frequency range tends to go lower and higher than any other genre. If you're big into classical guitar, jazz, orchestral recordings, etc. Go for broke. Spare no size expense. The decay of a note on a piano or the crash of a cymbal will sound significantly worse on anything lower than 192 Kb/s. Consider .ogg quality level 9 or 10 or .mpg lame extreme or insane.

This brings me back to lossless encoding. Lossless should be considered a convenient method of archiving your music. If you have a rare CD that you would like to preserve for all time. It is a great way of storing it for future listening since the original bits are all neatly compressed. If you have an extra 100 GB hard drive laying around. Why not preserve your CD collection? :) The added advantage is you can seal those CDs and keep them pristine for when 1 TB hard drives are all the rage. :D

[Richard]

Originally posted by flaming gerbil

standard 16/44.1 files take 5 megs per minute per track. the songs you listen to off a cd are in stereo, i.e. a left channel and a right channel. so this is two tracks making it 10 megs per minute. so a three minute song off a cd as a standard .wav file would be 30 megs (at 16/44.1). increasing the sampling rate going from 44.1 khz to 96khz doubles the storage space. increases bit-depth increases storage space exponentially.

Indeed it does. Which is another nice thing about good lossless compression. As it stands with 24/96 audio or 24/192, it wouldn't take long at all to fill up a hard drive. Sure, with DVD writers coming down in price all the time, and even better mediums on the horizon storage space is becoming less of an issue. For the time being though, compression still has its place.

Very soon I'll be adding an SACD player to the mix, and picking up either a LynxTwo or DAL Deluxe to do some listening tests on really high end audio. I've listened to some SACDs in the past, and it really does sound worth the added investment if you're big into audio.

[Penguin4x4]

So a 24/44.1 would be 256x the size of a 16/44.1, correct?

[flaming gerbil]

yep that is why i have around 60 hard drives in my closet. storage for current projects and some past projects. to give you an idea of what us audio engineers need in terms of storage (at 24/96) take the example above how how much megs per minute per track. the current project i'm working on at my home studio is 42 tracks. the last one i did was only 9 though. of course in reality the difference between let's say 24 bits and 16 bits is subtle. but i definitely hear a difference. even between 44.1 and 96. some studios are doing 128 and even 192. the quality of the digital to analog (d/a) and anolog to digital (a/d) also plays a significant role in the sound quality. 16 bits over really good converters sounds much better than 24 bits over mediocre converters. at my workplace (i.e. the "real" studio) we still mostly track to analog (reel to reel) which to my ears sounds better. yep richard is right about the classical thing. i almost always go digital with classical and extremely good converters are needed with the highest bit depth and resolution you can get. those last piano notes and reverb dying out can be lost without really good resolution and converters.

[flaming gerbil]

penguin you seems pretty interested in this file size thing so here you go. i took an mix done at 16/44.1, which ultimately at some point everything has to be mixed down to to get it onto a cd. it is a 3 miniute and 52 second song with a file size of 40.05 megs. i then resampled at and converted it to 24/96. it's file size increased to 737. 4 megs. now imagine a 40 track mix. hard drives fill up very fast.:( this is the conundrum with dvd audio, we don't know if consumers would want 24/96 or surround sound. it's got to be one or another, not enough storage space for both. you either have to have a standard "cd" done at 24/96 or six channels of sourround at 16/44.1 i'd be interested in what people think, since i know people involved in making these decisions. me as an engineer i'd prefer 24/96 but most of us think consumers would prefer surround sound. personally i don't like mixing surround.

[MospeadasDark]

Originally posted by flaming gerbil
penguin you seems pretty interested in this file size thing so here you go. i took an mix done at 16/44.1, which ultimately at some point everything has to be mixed down to to get it onto a cd. it is a 3 miniute and 52 second song with a file size of 40.05 megs. i then resampled at and converted it to 24/96.

Here's where I will recommend an upsample to 24/88.2 rather than 96. 44.1 goes into 96 an odd number. You'll get some weird multiple and your songs will sound a bit weird. Noticable esp if you play classical.

My wavs at 24/88.2 is about 184MB for a 4 minute song.

[flaming gerbil]

agree. i did it on a copy just to demonstrate the files size. when i go digital i always use 24/88.2 to keep a nice even resample.

[flaming gerbil]

well i shouldn't say i always go 88.2 because i do 96 a lot. 88.2 because i can do it at work. my home studios converters will only sample at 44.1, 48 and 96.

[Penguin4x4]

Wait, wouldn't a 30MB file @16/44.1 be 16.7GB @24/96?

[Emericana]

jeesh lol my 192 birate mp3's i rip sound great lol i can hardly tell a difference between that and 128 birates and i cant tell the diff between that and a cd

[MospeadasDark]

Originally posted by Emericana
jeesh lol my 192 birate mp3's i rip sound great lol i can hardly tell a difference between that and 128 birates and i cant tell the diff between that and a cd

You sir, are in need of new speakers. Most likely a better soundcard as well.

[Check-X]

Originally posted by Illah
I don't believe that whole LP's are best hype. For turntablism and mixing, yeah, but as far as sound quality, no. They work like this, a groove is dug out of a slab of vinyl representing the sound. Think like this, you know how a speaker moves in and out real fast to make sound? The printing needle of a record does the same thing, and that's the groove in the vinyl. As the playing needle rides those grooves the slight variations are amplified into a speaker to make them audible. That's not necesarily uncompressed, it's nineteenth century. Sure it's analog, so purists are like, "It's more organic." I don't know about that.

Enough ranting. While doing two semesters of sound work with a sound engineer who's worked with Janet Jackson among others, here's what he uses. AIFF. It's basically a Mac version of WAV files. As long as you have a raw format like WAV or AIFF at 16 bit 48K sample rate that's about as good as digital sound gets (there's a new 24bit format availiable, but it's essentially unused). Any more would be unnecessary. Once you get over 41K samples it's pretty much crystal clear, 48K is kinda just making sure.

--Illah

--Illah

a brand new CLEAN LP on a GOOD clean record player with a hi-fi needle will produce far superior sound than any CD system will. The only reason records sound bad is because down the road they collect dust, get small niches in them and distort the original recorded sound. if Analog sound wasn't better.. why are mobo's now comming with vacuum tubes on them?

[Emericana]

sb audigy w/ klipsh 4.1 cant get much better than that ;)

[parkan]

You cannot imagine how lucky you are to only work with sound. Working with uncompressed video AND high bitrate sound makes you wish you had hard drives as far as the eye can see... 160GB is barely enough for a simple low-botrate project scratch space.

[Avatar28]

No, I don't think it is an exponential increase either. More of a linear increase.

Example, if you make two recordings of the same song, one at 8-bit and one at 16-bit, the 16-bit song is twice the size of the 8 all other things being equal. 24-bit would be 1.5 times the size of 16-bit.

8 bits x 44,100 samples x 2 channels = 705,600 bits/sec
16 bits x 44,100 samples x 2 channels = 1,411,200 bits/sec
24 bits x 44,100 samples x 2 channels = 2,116,800 bits/sec

Edit:
Doubling the sampling rate will also doble in size again. So a 24/88.2 file would be 4,233,600 bits/sec. 8 bits to the byte will give you 529,200 bytes/sec or 31,752,000 bytes per minute (compared to 10,584,000 bytes/min for CD quality audio). That said, I'm rather skeptical that


Oh, and I also checked and I believe that it WAS Huffman encoding that I was thinking of earlier.

[Nismo Nate]

What programs do you guys use to burn your Audio CD's?

My buddy and me jsut went to fry's and he got a sony burner 40x but it came with B's recorder gold, and it is having problems, i tried Nero and it didn't reconize the drive, anyone have some sugestions?

[Avatar28]

Yeah, try downloading a newer version of Nero. The programs basically have a list of CD burners within the program. When a new drive comes out, it's not in the list and won't be recognized. Go get a newer version of Nero and it may fix that problem. Or use the burning software that comes with the drive.

[Nismo Nate]

Originally posted by Avatar28
Yeah, try downloading a newer version of Nero. The programs basically have a list of CD burners within the program. When a new drive comes out, it's not in the list and won't be recognized. Go get a newer version of Nero and it may fix that problem. Or use the burning software that comes with the drive.

Thanx, but I persuaded him to just take it back and get a lite-on from newegg, like mine ... no problems ever....:cool:

[Penguin4x4]

After crunching numbers and rereading this thread, I'd say use 18/48 .WAV's. They seem to be the best of both worlds, space effeceint, uncompressed, and digitally enhanced(like the picture quality difference between DVD's and VHS's)sound. Am I right?

[Azzkiller]

Originally posted by Check-X


a brand new CLEAN LP on a GOOD clean record player with a hi-fi needle will produce far superior sound than any CD system will. The only reason records sound bad is because down the road they collect dust, get small niches in them and distort the original recorded sound. if Analog sound wasn't better.. why are mobo's now comming with vacuum tubes on them?

I disagree with your statement. The dynamic range on LP's simply is not as great as a compact disc. I have a turntable and listen to tracks on the record and then on a cd. Each has its advantages and disadvantages. Records have that cool retro feel to them, and to me, really that is their only advantage over cd. Cds on the other hand have a higher dynamic range, and greater frequency range. Now technically, I suppose records could have greater in both, but I have never found that to be true.

The reason Vacuum tubes are used on a motherboard is for gimmick purposes only. If you ask me, transistors are better than tubes. They are more efficient and introduce less distortion to the original signal. Now that is not to say that you cannot have a good tube design, but why, other than for the fun factor? I mean you could even have a tube design that outperforms a comparable transistor based design. But has anyone noticed how few amplifiers there are out there that are vacuum tube based and have more than 100w of output? They really stink in high output situations due to their inefficiency.

Oh, and a transistor based amplifier is still analog. The only amplification designs that could be considered digital (but its all a technicality if you ask me) are class D, T and their variants.

[Azzkiller]

Originally posted by flaming gerbil
penguin you seems pretty interested in this file size thing so here you go. i took an mix done at 16/44.1, which ultimately at some point everything has to be mixed down to to get it onto a cd. it is a 3 miniute and 52 second song with a file size of 40.05 megs. i then resampled at and converted it to 24/96. it's file size increased to 737. 4 megs. now imagine a 40 track mix. hard drives fill up very fast.:( this is the conundrum with dvd audio, we don't know if consumers would want 24/96 or surround sound. it's got to be one or another, not enough storage space for both. you either have to have a standard "cd" done at 24/96 or six channels of sourround at 16/44.1 i'd be interested in what people think, since i know people involved in making these decisions. me as an engineer i'd prefer 24/96 but most of us think consumers would prefer surround sound. personally i don't like mixing surround.

Gerbil:

I think I may be a bit confused. Are you saying that you have a sound file that was originally encoded at 44.1khz/16bit and you upsampled it to 24/96, or that your original file source was of higher resolution? What I am trying to understand as to why upsampling anything would do anything. First, wouldnt this be akin to trying to make an mp3 into a redbook cd, the redbook cd could only be as good as the mp3 file, no better. Second, how can any of you hear past 20khz anyways for 96khz audio to matter? Third, how do you know your loudspeakers and amplifiers have a frequency response to 48khz?

I can understand why the new sound technologies could use higher sampling rates of 24bits, but increasing the bandwidth seems to be rather pointless, perhaps you can enlighten me?

[Richard]

Alot of the arguing about which sounds better, is purely subjective. And as such, will never arrive at a conclusive answer.

I will agree with you though. Given equal quality components. CDs will yield the most accurate sound.

I still maintain that speakers are THE most important and decisive factor in determining music quality. You could have a $20,000 set of components, but if the speakers are lacking, you might as well have a cassette.

Touching on the tube vs transistor amplification. I like a good tube amp. It's a very sweet sound. You rarely see a cheaply made tube amp. Of course, that is probably large in part of the reason why they tend to sound so good. That's not to say that tubes are more true. Tubes, introduce a level of distortion. (Pleasant sounding distortion maybe, but distortion nonetheless.)

Transistor based amps are a dime a dozen, and many of the popular brands are very low quality in build. A disturbing trend, because 20-30 years ago, transistor based amps were fantastic in quality. Yeah, they were big and bulky by todays standards, but I have some components that are 25 years old that still sound wonderful. They truly don't make things like they used to. Now most things are cheap plastic, with weak power supplies. Oh, but they look cool. <sigh>

[Lochreas]

Originally posted by funkdamonkman
---X--- what are you going to do with this really high quality sound you store?
Yeah, I was wondering this also. Are you going to be recording music or something? I plan on upgrading my system soon for recording my original music and some of my band's stuff, so I'm also going to need the best quality. I haven't did much research on it yet because in about a year I will be going to the Recording Workshop in Ohio, but it will be done on my pc. Are those TerraTec cards good for digital recording on pcs?

[Azzkiller]

Originally posted by Richard
Alot of the arguing about which sounds better, is purely subjective. And as such, will never arrive at a conclusive answer.

I will agree with you though. Given equal quality components. CDs will yield the most accurate sound.

I still maintain that speakers are THE most important and decisive factor in determining music quality. You could have a $20,000 set of components, but if the speakers are lacking, you might as well have a cassette.

Touching on the tube vs transistor amplification. I like a good tube amp. It's a very sweet sound. You rarely see a cheaply made tube amp. Of course, that is probably large in part of the reason why they tend to sound so good. That's not to say that tubes are more true. Tubes, introduce a level of distortion. (Pleasant sounding distortion maybe, but distortion nonetheless.)

Transistor based amps are a dime a dozen, and many of the popular brands are very low quality in build. A disturbing trend, because 20-30 years ago, transistor based amps were fantastic in quality. Yeah, they were big and bulky by todays standards, but I have some components that are 25 years old that still sound wonderful. They truly don't make things like they used to. Now most things are cheap plastic, with weak power supplies. Oh, but they look cool. <sigh>

Absolutly, I believe speakers play the most important role. I am glad you see the limitations of tube amps in relation to fet based ones. A lot of people dont realize that there are tradeoffs in every amplifier, and you have to choose between each carefully.

If you ask me though, I would say that even those cheap things with a weak power supply sound the same as the high quality amplifier at low power, it is only when we start getting to the limits of the linear range of the amplifier when they truly start sounding different.

P.S. What do you think of the expensive audio cables that cost hundreds of dollars? ;)

[james.miller]

kindof off subject.anyway...

has anybody mentioned mp3pro? ive got to versions of a song:

1) encoded in mp3 - 192k
2)mp3PRO - 64k
they are very nearly,and i meny 'VERY' nearly identical.
encoded in mp3pro it was just over 2mb. incredible

[Penguin4x4]

MP3PRO is the MPEG-4 of music compression. You have to buy software, as the technology is patented.

[james.miller]

to encode it yes i know. but playing it doenst cost a penny

[james.miller]

ooops i was wrong:
http://www.thomson-multimedia.com/gb/00/mp3.htm
you can olny choose 64k. but just listen. its bloody good for 64k

[Mpegger]

Oh gawd... dont get me started on everything wrong with Mp3"pro":rolleyes:

[james.miller]

like what?

[Mpegger]

Well the #1 problem which is caused mainly by the manner that it encodes audio is the gawd awful audio it produces. And if you can't tell the difference, then either; (a) You need better equipment; or (b) Better hope you still have whatever current hearing you still have left in a couple of more decades.:D

But seriously, the manner in which it encodes audio, there is just no way it can compare to an mp3 file, much less the original CD. The codec was aimed mainly at the limited memory portable mp3 players so people can fit more music onto a 64mb player. If you play the mp3"pro" with a mp3 player that doesn't support the "pro", what you hear is a low bitrate, 22kHz file. Guess what.... thats whats been encoded. The highs that you hear are basically faked by the decoder. The encoder just adds in some bits to each frame that tells the decoder what bands to extend into the upper frequency range and how far. And since most of those portable players come with crappy earbuds, one would "swear it sounds just like a cd!"

I've tested it out myself. I encoded a rip from one of my CDs with mp3"pro" and LAME. I used the highest bitrate that the demo Thompson encoder would allow, 64kbps. I then used Lame in CBR mode @ 128kbps joint stereo with a Q value of 0 and a lowpass of 19.5kHz. The mp3"pro" couldn't even compare.

The highs that the "pro" file produced weren't even present on the original wav file! I dont know how to explain it, but it basically sounded like the player was just making up the high frequency range as it went along. In one particular area of the track, it actually sounded different everytime it was played. Not to mention all the warble and artifacts present throughout the whole song.

I tried using Lame @ 96kbps and STILL, the "pro" file just couldn't cut it. And at 64kbps, the Lame file still edged out the "pro" file (although not by much).

Ever since that little test run, I wouldn't touch that file format with someone elses computer, much less any of mine.

A curse on mp3"pro"!:mad:

[james.miller]

Originally posted by Mpegger
(a) You need better equipment; or (b) Better hope you still have whatever current hearing you still have left in a couple of more decades.:D
thats a little rude for my liking. oh, by the way, i have a marantz 6.1 surround system with a mission speaker setup and a marants monobloc powering a lighting audio 12inch subwoofer. think i need better equipment? WTF ARE YOU ON?


"bloody good for 64k" is what i said

[Azzkiller]

Originally posted by james.miller

thats a little rude for my liking. oh, by the way, i have a marantz 6.1 surround system with a mission speaker setup and a marants monobloc powering a lighting audio 12inch subwoofer. think i need better equipment? WTF ARE YOU ON?


"bloody good for 64k" is what i said

Yes, but you also said that 64K mp3pro was very identical to 192K mp3. Perhaps you use a terrible mp3 encoder? That is the only difference that I can think of. I personally have not used mp3pro, but their is only so much a compression algorithm can do. After reading what Mpegger said, I find it odd that there was not a substantial difference. Another issue could be that your ears cannot hear above 10khz or so. That would explain why you could not hear the extra high frequency garbage that the mp3pro decoder adds.

Look, I am not trying to "be a big meanie" I just sometimes read something, and I not understand how someone else came to the same conclusion. For instance, when you said your lightning audio was in a box tuned to 18hz, in such a small enclosure with such a large surface area on the vent, I cannot logically see how that would work. In my years of speakerbuilding, having such a long vent would lead to excess noise and pipe resonances. I would know too, I designed a 5cuft enclosure tuned to 20hz for a 12" driver. I used 2 4" vents. They were extremely long, and I still got noise, even though I had plenty of vent surface area. I had to tear the bottom off the enclosure and redo the entire box.

[james.miller]

the enclosure is 3.5 cubic feet. 38inch wide, around about 15 depp and about 13 high WTF is that conciidered small?

and yes there is noise - at 120herts - my box is electronically limited (by an active cross-over) to 60 herts. YOU DONT HEAR THE PORT NOISE

was that 24 inch or 2 4inch? either way that is not 'extreamly long'. any you of all people should know that you get more noise when increasing the port length

and another thing - i use franhoufer or however its spelt
the dimentions of my subwoffer are not in question - its off topic so DROP IT

[flaming gerbil]

orginally posted by azzkiller

I think I may be a bit confused. Are you saying that you have a sound file that was originally encoded at 44.1khz/16bit and you upsampled it to 24/96, or that your original file source was of higher resolution? What I am trying to understand as to why upsampling anything would do anything. First, wouldnt this be akin to trying to make an mp3 into a redbook cd, the redbook cd could only be as good as the mp3 file, no better. Second, how can any of you hear past 20khz anyways for 96khz audio to matter? Third, how do you know your loudspeakers and amplifiers have a frequency response to 48khz?

I can understand why the new sound technologies could use higher sampling rates of 24bits, but increasing the bandwidth seems to be rather pointless, perhaps you can enlighten me?


wow this thread is stil around! you ask some very good questions. you are correct with your first question. the upsample will not improve the sound quality because it was orginally recorded at 16/44.1. you cannot regain lost information by upsampling. i did this merely to demonstrate the file size differences between 16/44.1 and 24/96. in your second question you are confusing sampling rate with frequency response. sampling rate measured in khz is how many samples taken per second, frequency response indicates what frequencies the ear hears. so A440 is a concert A pitch with a frequency of 440 hertz. a sampling rate of 440 hz would be 440 samples per second. make sense? using the nyquist theorem, you have to sample at twice the frequency you are wishing to reproduce. so if you want to reproduce a 20khz signal you have to sample at 40khz. that's why cd's are sampled at 44.1 to leave a little head room. it is commonly agreed the range of human hearing is 20hz to 20khz. so a 96khz sampling rate would reproduce frequencies up to 48khz. this is a good point you make and is being hotly debated in the audio feild right now. some engineers think these higher sampling rates are not needed, others sample at 128 khz and even 196! recent studies have shown (not conclusively) they we may not outright hear signals beyond 20KHz but may "sense" them. we can tell the audio is different but don't know why. this is also being hotly debated but is gaining more credibility. also george martin (the famous beatles engineer) has been clinically tested hearing signals up to 39 khz. giving you a needed sampling rate of 78khz. so some of us can hear beyond this accepted range. that is why a lot of us sample higher, just to be safe. myself and others can also hear the difference between 44.1 and 96. sounds smoother to me. more analog in character. i have a subwoofer and 3 speaker monitoring system with built in amplifiers for the woofer, midrange, and tweeter speakers. gives me a frequency response from 15hz to 22 khz. i have measured this with pink noise and software based analyzers. highend mastering equipment can go beyond this range.

[flaming gerbil]

orginally posted by azzkiller

I think I may be a bit confused. Are you saying that you have a sound file that was originally encoded at 44.1khz/16bit and you upsampled it to 24/96, or that your original file source was of higher resolution? What I am trying to understand as to why upsampling anything would do anything. First, wouldnt this be akin to trying to make an mp3 into a redbook cd, the redbook cd could only be as good as the mp3 file, no better. Second, how can any of you hear past 20khz anyways for 96khz audio to matter? Third, how do you know your loudspeakers and amplifiers have a frequency response to 48khz?

I can understand why the new sound technologies could use higher sampling rates of 24bits, but increasing the bandwidth seems to be rather pointless, perhaps you can enlighten me?


wow this thread is stil around! you ask some very good questions. you are correct with your first question. the upsample will not improve the sound quality because it was orginally recorded at 16/44.1. you cannot regain lost information by upsampling. i did this merely to demonstrate the file size differences between 16/44.1 and 24/96. in your second question you are confusing sampling rate with frequency response. sampling rate is how many samples taken per second, frequency response indicates what frequencies the ear hears. so A440 is a concert A pitch with a frequency of 440 hertz. a sampling rate of 440 hz would be 440 samples per second. make sense? using the nyquist theorem, you have to sample at twice the frequency you are wishing to reproduce. so if you want to reproduce a 20khz signal you have to sample at 40khz. that's why cd's are sampled at 44.1 to leave a little head room. it is commonly agreed the range of human hearing is 20hz to 20khz. so a 96khz sampling rate would reproduce frequencies up to 48khz. this is a good point you make and is being hotly debated in the audio feild right now. some engineers think these higher sampling rates are not needed, others sample at 128 khz and even 196! recent studies have shown (not conclusively) that we may not outright hear signals beyond 20KHz but may "sense" them. we can tell the audio is different but don't know why. this is also being hotly debated but is gaining more credibility. also george martin (the famous beatles engineer) has been clinically tested hearing signals up to 39 khz. giving you a needed sampling rate of 78khz. so some of us can hear beyond this accepted range. that is why a lot of us sample higher, just to be safe. myself and others can also hear the difference between 44.1 and 96. sounds smoother to me. more analog in character. perhaps this is why some engineers prefer analog reel to reel machines. a well maintained good machine can have up to a 140khz frequency response. this would be 280 khz sampling rate for a digital machine to reproduce. no digital does this yet. i have a subwoofer and 3 speaker monitoring system with built in amplifiers for the woofer, midrange, and tweeter speakers. gives me a frequency response from 15hz to 22 khz. i have measured this with pink noise and software based analyzers. highend mastering equipment can go beyond this range.

[Penguin4x4]

And DVD's can reproduce up to 80KHz if I'm not mistaken

[flaming gerbil]

well remember it is just digital information so a dvd can reproduce what you want it to depending on how much information ("memory") you want to use. you can fill a dvd up with 16/44.1 if you want. we can do a cd's woth of music at 24/96. how we are going to dvd hasn't really been decided yet. we can either do 6 channels of 16/44.1 giving you "cd" quality surround sound, or do an "albums" worth of material at 24/96 with vid clips and short interviews, etc. we are not shure yet if consumers want better quality or surround. looks like we may be heading towards 24/96 which is what most engineers want. surround is pretty cool, but i personally really dislike mixing it. i like surround on classical music though.

[Penguin4x4]

Here it is, Sony Home Theater Speaker Set
The five voice-matched satellites are solidly built and, each with dual 2" drivers and a 1" tweeter, offer a frequency response of up to 50k Hz — perfect for handling extended-bandwidth multichannel audio, and delivering a seamless surround sound experience.

BS?

[james.miller]

the extended responce 'seems' to be better at producing a more convincing sound field (better, more presise highs, better seperation, blah blah blah), from what i have heard about it. i dont know though, im not convinced untill i hear it myself

[o770]

ive not followed this thread for awhile now so i hope this is not much off the topic.

how much of quality difference do you ppl notice between the music on a original CD and the same music that was encoded to .mp3 and then decoded and burnt to another CD? considering the encoding quality is optimal and so is the mp3 bitrate.
thanks.

[james.miller]

the is quite a noticable lack in quality - sounds almost harsh compared to the originals