From: owner-idealcopy-digest@smoe.org (idealcopy-digest) To: idealcopy-digest@smoe.org Subject: idealcopy-digest V4 #69 Reply-To: idealcopy@smoe.org Sender: owner-idealcopy-digest@smoe.org Errors-To: owner-idealcopy-digest@smoe.org Precedence: bulk idealcopy-digest Tuesday, March 6 2001 Volume 04 : Number 069 Today's Subjects: ----------------- RE: [idealcopy] [OFF TOPIC] A Quick Shellac Primer ["giluz" ] [idealcopy] RE: Off-Topic - other lists ["giluz" ] [idealcopy] OT Television! [John Roberts ] Re: [idealcopy] Why are Beef Hearts? [Mark Short ] [idealcopy] [OFF TOPIC] Upcoming digital audio formats [Mark Short ] WG: [idealcopy] Re: Ricks Picks [Woerner Frank ] RE: [idealcopy] Re: Ricks Picks ["giluz" ] [idealcopy] WIRE in The Wire ["ian jackson" ] RE: [idealcopy] [OFF TOPIC] Upcoming digital audio formats ["giluz" ] Re: [idealcopy] Q+A [PaulRabjohn@aol.com] [none] [Chris.Ray@medas.co.uk] Re: [idealcopy] [OFF TOPIC] Upcoming digital audio formats ["Paul Pietrom] Re: [idealcopy] [OFF TOPIC] Upcoming digital audio formats ["Paul Pietrom] Re: [idealcopy] [OFF TOPIC] Upcoming digital audio formats [fernando Subject: RE: [idealcopy] [OFF TOPIC] A Quick Shellac Primer Hi Paul, a few questions about this: > Ah - but Steve would disagree with you. His basic take is that a 16 > bit, 44 kHz CD *cannot* reproduce his music accurately - no matter how > hard you try. Well, how much of all of it is true? I know there are people that can distinguish between digital and analogue sounds, but most of us can't (except for the analogue hiss, which is quite easy to pinpoint). Is it really that significant? What is there in Albini's music which makes it sound bad on digital format? If there is something like that, can you think of a type of music that would sound better on CD than it would sound on analogue? And what about the higher sampling rates - are they better because of their larger frequency range, or because you've got more samples per time unit? Are they noticeably better, or is it the same case where only musicians/engineers, with their extra-sensitive sense of hearing can distinguish between them and lower sampling rate recordings? What does Albini have to say about the future of digital media? Even he has to admit that vinyl presents lots of problems. Can he conceive of any digital format that might sound better than analogue? Can he conceive of any new analogue format which would override vinyl's problems? cheers, giluz ------------------------------ Date: Tue, 6 Mar 2001 02:55:10 -0800 (PST) From: John Roberts Subject: [idealcopy] Re: Another response... Are there any Wire fans here who > dislike Faust,>Beefheart or Shellac?>If so, what was > it about it you didn't like? Can't say that I've heard much Faust. Big Beefheart fan: I've got all the proper albums, some of the outtake comps and a handful of bootlegs. I know some people who saw him on the Doc at the Radar Station tour at Loughborough University and they make me sick with envy. Shellac are alright - I think the 7" singles were better than the albums. Big Black I really liked. I remember hearing Kerosene at a club in Leicester when Atomizer had just been released over here on Blast First: I've never been the same since. I've mentioned this on here before but I saw BBlack when they played on their farewell tour at both Leeds Uni and at the Hammersmith Clarendon. The latter gig was the one where Graham and Bruce appeared for the encore. >If you had to choose one > non-Wire related release to>listen to over & over, > what would it be? At the moment I'd say Vic Godard and Subway Sect's 20 Odd Years compilation. Impossible to get bored of it I would have thought. >What other > music related lists are Ideal Copiers> I'm not. Does the Blue Army Mailing List for Leicester City fans count? John Get email at your own domain with Yahoo! Mail. http://personal.mail.yahoo.com/ ------------------------------ Date: Tue, 6 Mar 2001 13:06:45 +0200 From: "giluz" Subject: [idealcopy] RE: Off-Topic - other lists > When I subscribed to > this list I thought that Wire and King Crimson were two > separate worlds (felt very strange to think of KC when I first heard The > Queen of Ur and the Queen of Um) but right away I noticed one member here > who was also a regular contributor to the KC list, then Charles > subscribed, > and Fripp and KC get mentioned here on a regular basis, especially in this > last week. Does Wire get mentioned on the KC list as well? giluz ------------------------------ Date: Tue, 6 Mar 2001 03:17:21 -0800 (PST) From: John Roberts Subject: [idealcopy] OT Television! I've just realised what the ATP festival is. Its the All Tomorrows Parties thing that Wire played last year. I thought it was a one off. How wrong I was. Am I the only person on this list who wasn't aware that this festival was on? I really ought to get out more often. Anyway the website is up and running still and has the confirmed bill for April 2001. http://www.alltomorrowsparties.co.uk/ At the momemt the bill is Afro Beat | Atmosphere | Autechre | Boards of Canada | Broadcast Calexico | Def Jux Collective | ESG | Lambchop | Mike Ladd | Mr. Lif Neil Hamburger | Prefuse 73 | Rick Rizzo and Tara Key | Sun Ra Television | The Ex | The Sea and Cake | Tortoise | Yo La Tengo It'll cost you tho. It's 100 quid per person but this includes a chalet bed etc. Have a look at the website for further booking details. John - --- Michael Flaherty wrote: > >Date: Wed, 28 Feb 2001 08:34:01 -0600 > >From: "dMc" > >Subject: [idealcopy] OT: cross-polination > > > >i know from past postings there are a few > television fans here: > >from the TV list: > >sounds like a familiar agenda, no? > >"Now there are three shows. The ATP festival in > England April 8, A > >festival in Gijon, Spain on April 12, and at a > festival in Chicago on > >May 10 at the Metro. Hope to see you at these > shows. " > > OK, I can breathe again. Took a minute to recover. > Any ticket info, now > or in the future, please post. > > Michael Flaherty Get email at your own domain with Yahoo! Mail. http://personal.mail.yahoo.com/ ------------------------------ Date: Tue, 06 Mar 2001 11:57:11 +0000 From: Mark Short Subject: Re: [idealcopy] Why are Beef Hearts? Graeme Rowland wrote: > Are there any Wire fans here who dislike Faust, > Beefheart or Shellac? > If so, what was it about it you didn't like? Beefheart - Trout Mask Replica has some good moments, but a lot of it is too loose for my ears. Shiny Beast is probably my fave, and the most disciplined of his albums (at least, of the ones I've heard). Faust - I've only heard the Faust Tapes, and that was a long time ago. I remember it being a rehash of musique concrete ideas. Shellac - Not heard them. > > If you had to choose one non-Wire related release to > listen to over & over, what would it be? Eno - Here Come The Warm Jets. > > What other music related lists are Ideal Copiers > subscribed to? NerveNet - the Eno mailing list. Suprisingly little traffic, and what does turn up is pretty dull - "U2 rock my world" vs. "Bono sucks" ------------------------------ Date: Tue, 06 Mar 2001 12:50:18 +0000 From: Mark Short Subject: [idealcopy] [OFF TOPIC] Upcoming digital audio formats giluz wrote: > > And what about the higher sampling rates - are they better because of their > larger frequency range, or because you've got more samples per time unit? It's not the increased sampling rate in itself that gives better sound, but the fact that it allows the use of less aggressive filters, which have less effect upon sound quality than those which have to be used in today's CD players. > Are they noticeably better, or is it the same case where only > musicians/engineers, with their extra-sensitive sense of hearing can > distinguish between them and lower sampling rate recordings? > Going by the quality of the sound at gigs, most musicians/engineers sense of hearing is impaired rather than extra-sensitive. ------------------------------ Date: Tue, 6 Mar 2001 15:24:57 +0200 From: "giluz" Subject: FW: [idealcopy] [OFF TOPIC] Upcoming digital audio formats - -----Original Message----- From: giluz [mailto:giluz@nettalk.com] Sent: Tuesday, March 06, 2001 3:25 PM To: Mark Short Subject: RE: [idealcopy] [OFF TOPIC] Upcoming digital audio formats > > And what about the higher sampling rates - are they better > because of their > > larger frequency range, or because you've got more samples per > time unit? > > It's not the increased sampling rate in itself that gives better > sound, but the > fact that it allows the use of less aggressive filters, which > have less effect > upon sound quality than those which have to be used in today's CD > players. What do you mean by filtering? Are you talking about the mastering process (i.e. when the master track is transferred to CD format, or whatever other way there is to describe it) or the listening process (i.e. digital-analogue conversion from your CD player to your hi-fi amp)? > > > Are they noticeably better, or is it the same case where only > > musicians/engineers, with their extra-sensitive sense of hearing can > > distinguish between them and lower sampling rate recordings? > > > > Going by the quality of the sound at gigs, most > musicians/engineers sense of > hearing is impaired rather than extra-sensitive. You can't judge anything by the sound quality of gigs. It's too dependant on the local acoustics, PA, no. of people, etc. But there is something in what you say. Engineers definitely have an amazing hearing sense, just because their ears are trained for it. The advantages of them hearing things an ordinary person wouldn't notice can very easily turn into a disadvantage, if they're doing stuff only other engineers can comprehend. I suspect this attitude took a turn for the worse in the 90's, with all the new computerised equipment that allowed you to go into details you couldn't dream of before. Rephrasing my question, then: Most ordinary people, for example, could tell if you played them the same song as an mp3 and later as a CD track that the latter is far superior to the former. Is the fidderence between lower and higher sampling rates that big? cheers, giluz ------------------------------ Date: Tue, 6 Mar 2001 14:42:52 +0100 From: Woerner Frank Subject: AW: [idealcopy] Q+A / about XTC > -----Urspr|ngliche Nachricht----- > Von: Tim Robinson [mailto:timrobinson@cwcom.net] > Gesendet: Dienstag, 6. Mdrz 2001 03:19 > An: idealcopy > Betreff: [idealcopy] Q+A > ... snipped ... > > Well I don't know..... they come between Wire and Yo La Tengo on my CD > racks. Some of Wires more jovial moments go nicely with XTC. Dot Dash, > Outdoor Miner, Map Ref. Like Wire they are also very > English/European, > they have the same kind of loyal fanbase who stick wth them through > various 'challenging' moments and also were also ripped off > like mad by > Blur et al during the Brit-pop wars, and they make some fine tunes and > are still doing their thing outside the mainstream. Their recent stuff > has been a bit...Nick Hornby though....yuck! Oh I really hate it when I don't understand what is written here ... As my mother tongue is German and my capability of speaking/reading English is weak please declare what stuff which is "Nick Hornby" sounds like ... Thanks. Frank from Bavaria ------------------------------ Date: Tue, 6 Mar 2001 14:49:24 +0100 From: Woerner Frank Subject: WG: [idealcopy] Re: Ricks Picks > -----Urspr|ngliche Nachricht----- > Von: Tim Robinson [mailto:timrobinson@cwcom.net] > Gesendet: Dienstag, 6. Mdrz 2001 03:37 > An: idealcopy > Cc: r_j_h@yahoo.com > Betreff: [idealcopy] Re: Ricks Picks > > > Re: Ricks Picks > > Leave Rick alone! Rick is cool! I enjoy Ricks posts because > he isn't a > total music snob unlike some of you. :( > .... > > So....in capital letters *stop dissing Rick* and respect each others > differences! It takes Diff'rent strokes folks! I cannot remember a posting where Rick was "dissed". I myself listen to music from ABBA to Zappa and am very open minded - it should be allowed though to look at Rick's listening list and say ( in a friendly manner ): "Unbelievable !!" Frank from Bavaria ------------------------------ Date: Tue, 6 Mar 2001 16:12:38 +0200 From: "giluz" Subject: RE: [idealcopy] Re: Ricks Picks > > Re: Ricks Picks > > > > Leave Rick alone! Rick is cool! I enjoy Ricks posts because > > he isn't a > > total music snob unlike some of you. :( > > > .... > > > > So....in capital letters *stop dissing Rick* and respect each others > > differences! It takes Diff'rent strokes folks! I don't remember dissing him either. I quite like the fact that his list has so many things I wouldn't dream of listening to. Diversity of opinions is one of the reasons why I like this list. giluz ------------------------------ Date: Tue, 06 Mar 2001 15:23:35 -0000 From: "ian jackson" Subject: [idealcopy] WIRE in The Wire anyone noticed all the Wire mentions in March's edition of 'The Wire"?? what can this mean? (rhetorical question) ian.s.j. _________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com. ------------------------------ Date: Tue, 6 Mar 2001 17:26:08 +0200 From: "giluz" Subject: RE: [idealcopy] [OFF TOPIC] Upcoming digital audio formats > For digital audio, the sampling rate must be at least twice the highest > frequency to be reproduced. For CD,the sampling rate is 44.1kHz, > so the maximum > frequency which can be reproduced is 22.05kHz. When the digital signal is > converted into analogue, the analogue signal will contain > frequencies above > 22.05kHz. For reasons I don't claim to understand, these higher > frequencies, > although inaudible to humans will interfere with frequencies > which are audible. I think what you're referring to here are called harmonics. I don't remember exactly what they are, but I know that they actually determine the sound of the instrument, as opposed to its pitch, i.e. they determine the difference between a violin playing 440 Hz and a piano playing 440 Hz. They play the same frequency but still sound different. As far as I remember, harmonics are always a multiplication of the original frequency, meaning that they're always higher than the original. What's more, a sound is comprised of a set of harmonics, meaning that there is more than one additional harmonic sound, and each harmonic is higher than the previous. Even though they're quite weaker than the original signal and can't be distinguished as a seperate sound, they're very significant. I don't know much about the significance of harmonics that are higher than 20 kHz, but I do know that the 20k barrier is only in theory, and that most of us can hear only up to 16k (we're born with the ability to hear up to 20. As the years go by we lose that ability), so i don't know if it's that significant. > > Rephrasing my question, then: Most ordinary people, for > example, could tell > > if you played them the same song as an mp3 and later as a CD > track that the > > latter is far superior to the former. Is the fidderence between > lower and > > higher sampling rates that big? > > The difference between CD and MP3 is not just down to sampling rates. MP3 > compresses the signal so that it takes up much less bandwidth. It discards > information which it doesn't think you can hear. I know that it's completely different. I just gave it as an example of quality differences that can be distinguished by 'ordinary' people, as opposed to quality differences that can only be distinguished by engineers/musicians. My question was whether the difference between higher and lower sample rates could be distinguished by ordinary persons. giluz ------------------------------ Date: Tue, 06 Mar 2001 09:37:43 -0600 From: Michael Flaherty Subject: [idealcopy] Re: OT Television! At 03:17 AM 3/6/2001 -0800, you wrote: >I've just realised what the ATP festival is. Its the >All Tomorrows Parties thing that Wire played last >year. I thought it was a one off. The Chicago fest. is the same one Wire played at last year as well. Michael Flaherty ------------------------------ Date: Tue, 06 Mar 2001 10:43:52 EST From: PaulRabjohn@aol.com Subject: Re: [idealcopy] OT Various bits so what is the joy division list like then? p ------------------------------ Date: Tue, 06 Mar 2001 09:54:42 -0600 From: Michael Flaherty Subject: [idealcopy] Other Lists >From: =?iso-8859-1?q?Graeme=20Rowland?= >Subject: [idealcopy] Why are Beef Hearts? >If you had to choose one non-Wire related release to >listen to over & over, what would it be? Either Fripp and Eno: The Essential (all of their first album and half of their second), or White Light/White Heat--covers my two basic moods (musically) as well as any 2 releases could. >What other music related lists are Ideal Copiers >subscribed to? Elephant Talk (Fripp and KC): like Robert C., I'm more interested in Fripp than the band, and tend now to pay more attention to Fripp's web page. I use to be a very regular contributor, now I doubt most members would recognize my name. John Cale, Brian Eno, and Sonic Youth. Michael Flaherty ------------------------------ Date: Tue, 06 Mar 2001 11:03:06 EST From: PaulRabjohn@aol.com Subject: Re: [idealcopy] Q+A "........and i still don't rate xtc and i'm still amazed people bracket them with wire and JD. you're joking." Well I don't know..... they come between Wire and Yo La Tengo on my CD racks >>>>>>> well if you ever tire of colin mouldings dittys and take the ol' XTC collection to the pawn shop i could suggest plugging that gap with some wolfgang press and (an absolute essential) "germ free adolescents" by x-ray spex. i do actually own a wonderstuff album but i'll skirt around that one....p ------------------------------ Date: Tue, 6 Mar 2001 16:10:44 +0000 From: Chris.Ray@medas.co.uk Subject: [none] Other lists: The Durutti Column Would like to be in The Chameleons. The Information in this communication is confidential and may be privileged and should be treated by the recipient accordingly. If you are not the intended recipient please notify me immediately. You should not copy it or use it for any purpose nor disclose its contents to any other person. ------------------------------ Date: Tue, 6 Mar 2001 09:16:22 -0800 From: "Paul Pietromonaco" Subject: Re: [idealcopy] [OFF TOPIC] Upcoming digital audio formats > giluz wrote: > > > > And what about the higher sampling rates - are they better because of their > > larger frequency range, or because you've got more samples per time unit? > > It's not the increased sampling rate in itself that gives better sound, but the > fact that it allows the use of less aggressive filters, which have less effect > upon sound quality than those which have to be used in today's CD players. > There's two things about upcoming digital formats - an increased sampling rate (96kHz), and an increased bit count (24 bit samples). The increased sample rate lets you use the less aggressive filters, but it really doesn't have much of an audible effect. It also allows a higher frequency response (your sampling rate has to be at least twice as high as your highest recorded frequency) but we probably can't hear the extra frequencies. But - the increased bit depth, on the other hand, makes a big difference. The main explanation is that in a digital system, the distortion rises as the signal level drops. If you get down near the noise floor of a 16 bit system, say around -90 dB from digital signal level 0 dB, you're basically around 100% THD. (1 bit of data to represent the audio = 100 % distortion) Your ear can hear these distortions quite easily - our ability to detect signal to noise ratio and distortion is pretty good. So, anything that gets quiet and isn't overly compressed suffers on CD (think classical music here, or Shellac (^_^)). The 24 bit systems have a *much* lower noise floor. The three 24 bit/96 kHz discs I have have an amazing ability to reproduce the quiet spots of the sound. (These won't play in a regular CD player, by the way - you have to have a fairly modern DVD player.) These discs just stay clean "all the way down". They sound warmer, too - less fatiguing. Almost vinyl like, but without the ticks, pops, and impulse response problems. (Vinyl has a hard time with impulse signals - they tend to smear and add extra harmonics.) Or - another way to look at it: 16 bit systems have 65536 bits to represent audio. 24 bit systems have 16777216 bits to repesent audio. An analogy would be 16 bit color on a computer vs 24 bit color on a computer. (Not a perfect analogy, but it gets you in the ballpark.) The sad fact is that most people haven't heard really good audio. And, I'm not saying that to brag or anything. But, if you go to a studio, and listen to an analog 16 track master tape playing back fresh, directly-recorded sound at a tape speed of 30 inches per second, you'd be amazed at what you'd hear. By the time the songs reach the consumer, they've been compressed, EQ'd, bounced around a couple of different formats, mastered and duplicated, and basically changed quite a bit from their original format. The "audio distribution" chain is better now than it used to be - some of the original recordings from the early 70's sound terrible, while the same new remasters sound great - but it can be improved. Cheers, Paul ------------------------------ Date: Tue, 6 Mar 2001 09:45:31 -0800 From: "Paul Pietromonaco" Subject: Re: [idealcopy] [OFF TOPIC] Upcoming digital audio formats > > For digital audio, the sampling rate must be at least twice the highest > > frequency to be reproduced. For CD,the sampling rate is 44.1kHz, > > so the maximum > > frequency which can be reproduced is 22.05kHz. When the digital signal is > > converted into analogue, the analogue signal will contain > > frequencies above > > 22.05kHz. For reasons I don't claim to understand, these higher > > frequencies, > > although inaudible to humans will interfere with frequencies > > which are audible. > > I think what you're referring to here are called harmonics. Hiya! Your explanation of harmonics is correct, but that's not what we're filtering out here. (^_^) In fact the higher sampling rate systems let you capture more harmonics, which is supposed to make them sound more natural. No, what we're filtering out are "aliasing effects". Let me explain. Have you ever moved your hand fast in front of a TV screen? Or, better - spun a wheel - say like on a kid's toy? Your tv flickers at 50 or 60 Hz. Start spinning the wheel slowly, illuminated only by the tv. Everything looks good, right? Now spin it faster. As you start to approach half the frequency of the set, the wheel will appear to slow down, and eventually stop. If you keep going faster, eventually it will start to look as though it's moving backwards, slowly at first then speeding up. This is exactly the effect we get with digital audio, when you get near half the sampling frequency. You start to get ripples of audio in reverse frequency, much like your wheel (original audio) starts to look like it's going backwards (alias frequency). So, if you try to record a signal that has a higher frequency (> 22 kHz) that the sample rate (44 kHz), you'll get aliases in the audible spectrum. (And, they sound *really* weird). So, you must filter during recording. The reason we filter during playback is a little different. If you think of the sample rate as the limit of the system (bear with me here - this is a little tricky to explain) then we get a node at half the sample rate (e.g. 44 kHz makes a point at 22 kHz.). The audio we're decoding will make a "mirror image" beyond the half point of the sample rate. So, any audio we have at 21 kHz is repeated at 23 kHz. Any audio that we have at 20 kHz is repeated at 24 kHz. It's not as audible as the original aliasing, but it's noticable, and it may have effects on your amplifier chain, which is going to get audio information all the way up to 44 kHz, instead of the 22 kHz amps are generally rated for. By using 8 times oversampling, we don't gain any extra information - that was removed by the recording filter - but we do move these numbers up quite a bit. (e.g. 352 kHz with a point at 176 kHz.) So, the audio goes up to 22 kHz, then is quiet until 176 kHz where the reverse audio starts to take place - and since that is so far up the spectrum, we can filter it gently. Okay - I've got to get to work. More about digital audio later! Cheers, Paul ------------------------------ Date: Tue, 06 Mar 2001 09:42:12 -0800 From: fernando Subject: Re: [idealcopy] [OFF TOPIC] Upcoming digital audio formats At 09:16 AM 3/6/01 -0800, Paul Pietromonaco wrote: >There's two things about upcoming digital formats - an increased sampling >rate (96kHz), and an increased bit count (24 bit samples). The increased >sample rate lets you use the less aggressive filters, but it really doesn't >have much of an audible effect. It also allows a higher frequency response >(your sampling rate has to be at least twice as high as your highest >recorded frequency) but we probably can't hear the extra frequencies. > >But - the increased bit depth, on the other hand, makes a big difference. May be... it depends on the engineering... it allows for sloppy engineering to get by... I agree that the increased bandwidth should have little effects on hearing -- as a predominant factor -- but the filter designs are much nicer and that is a great help. Again, engineers can be sloppier too. >The main explanation is that in a digital system, the distortion rises as >the signal level drops. If you get down near the noise floor of a 16 bit >system, say around -90 dB from digital signal level 0 dB, you're basically >around 100% THD. (1 bit of data to represent the audio = 100 % distortion) This is not true... it is very strict what you say. First, it is attributed that 1 bit gives you about 6 dB of dynamic range... so it would seem that - -96 dB is the bottom (as you are pretty much stating), but the fact remains that with "dither", and properly shaped dither noise, you can achieve a greater SNR... and in fact, this is what happens... The big factor is that to fully use a greater SNR when listening to music... you have to be in an anechoic chamber... I doubt that many people can have that kind of dynamic range available in their home! Background noise should be louder than the smallest voltage for just 16 bits. > Your ear can hear these distortions quite easily - our ability to detect >signal to noise ratio and distortion is pretty good. So, anything that >gets quiet and isn't overly compressed suffers on CD (think classical music >here, or Shellac (^_^)). > >The 24 bit systems have a *much* lower noise floor. The three 24 bit/96 >kHz discs I have have an amazing ability to reproduce the quiet spots of >the sound. (These won't play in a regular CD player, by the way - you have >to have a fairly modern DVD player.) These discs just stay clean "all the >way down". They sound warmer, too - less fatiguing. I get the same out of a properly mastered CD. One has to be careful, and you might have, on what is attributable to the noise floor in a stereo system. Think about it... with 24 bits, you have such a small voltage resolution(*), that there is no analog hardware that can resolve that... unless it is dipped in liquid nitrogen "B^) (*) Think of peak-to-peak range of 2 volts. Then each bit represents 2/2^24 or about 12 micro volts... that is so small... who is to hear that bottom bit toggling?!?!?!?!?!? ;-) The thing that 24 brings is an ability for the hardware to retain more accuracy as it processes the music on the way to the digital to analog converter. The implementation of these filters requires that the number of bits expands... and since you have to come down back to your DAC's width... there will be some dynamic range lost. So, the thing is that, 16 bits seems good enough for 99% of the music reproduction scenarios that we will listen to... unless you would like to live at one of those chambers at Bell Labs... then you could something like 18 bits. >Or - another way to look at it: 16 bit systems have 65536 bits to >represent audio. 24 bit systems have 16777216 bits to repesent audio. An >analogy would be 16 bit color on a computer vs 24 bit color on a computer. >(Not a perfect analogy, but it gets you in the ballpark.) It is not even close in analogy... the audio and visual system work on completely different ways, so it is not a good comparison to state resolution... though I get the picture (no pun :-) Sometime more is just more... or in this case, smaller is just smaller, and this is one of the cases. >The sad fact is that most people haven't heard really good audio. And, I'm >not saying that to brag or anything. But, if you go to a studio, and >listen to an analog 16 track master tape playing back fresh, >directly-recorded sound at a tape speed of 30 inches per second, you'd be >amazed at what you'd hear. By the time the songs reach the consumer, >they've been compressed, EQ'd, bounced around a couple of different >formats, mastered and duplicated, and basically changed quite a bit from >their original format. This is true... and the other thing that happens is that the mixing has to be compromised, because people still have lousy systems, and so they have to take that into consideration. Hence the use of those dreaded and lousy (according to testimonials) of the Yamaha NS-10 speakers in mastering... those with the white cone that you see often near a soundboard. cheers! - -fernando ------------------------------ Date: Tue, 6 Mar 2001 10:39:27 -0800 From: "Paul Pietromonaco" Subject: Re: [idealcopy] [OFF TOPIC] Upcoming digital audio formats > >The main explanation is that in a digital system, the distortion rises as > >the signal level drops. If you get down near the noise floor of a 16 bit > >system, say around -90 dB from digital signal level 0 dB, you're basically > >around 100% THD. (1 bit of data to represent the audio = 100 % distortion) > > This is not true... it is very strict what you say. First, it is attributed > that 1 bit gives you about 6 dB of dynamic range... so it would seem that > -96 dB is the bottom (as you are pretty much stating), but the fact remains > that with "dither", and properly shaped dither noise, you can achieve a > greater SNR... and in fact, this is what happens... > Well, you don't really get a greater SNR ratio, since the dither increases the noise floor. According to Principles of Digital Audio by Ken C. Pohlman, pg. 69: "In addition to its beneficial contributions, dither contributes noise to a digitization system. A degradation of a few dB in the broadband noise floor results" What you get is less quantization distortion at lower signal levels and an ability to perceive information beyond the lowest quantization level. i.e. something similar to the old analog trick of "listening thru the noise floor". According to Principles of Digital Audio (pg. 68): "Our ears are quite good at resolving narrow band signals below the noise floor, because of the averaging properties of the basilar membrane. The ear behaves as a one-third octave filter with a narrow bandwidth; the quantization error, which has been given a white noise character by dither, is averaged by the ear, and the original narrow band sinewave is heard without distortion. In other words, dither changes the digitial nature of the quantization error into a white noise, and the ear may then resolve signals with levels well below one quantization level." > The big factor is that to fully use a greater SNR when listening to > music... you have to be in an anechoic chamber... I doubt that many people > can have that kind of dynamic range available in their home! Background > noise should be louder than the smallest voltage for just 16 bits. > In my opinion, that's only true if you're playing back the music at roughly the level it was recorded. I often turn quiet pieces up to catch subtle nuances - hence this becomes more of an issue. (I just have to remember to turn 'em down before they get loud again, or my neighbors complain! (^_^)) And, I keep my place fairly quiet too. Hey - I'm an audio fanatic - what can I say? (^_^) 16 bit systems - to me - sound really good when the material is fairly compressed and recorded onto the system at a loud volume level. Anything that tries to have dynamics seem to suffer. (And, I know that records don't do this well either - you have to compress the dynamic range to around 40-50 dB to master a record at all - and on pop records the dynamic range is usually around 20 dB or so. But, they just sound like they do. (^_^)) > I get the same out of a properly mastered CD. One has to be careful, and > you might have, on what is attributable to the noise floor in a stereo > system. Think about it... with 24 bits, you have such a small voltage > resolution(*), that there is no analog hardware that can resolve that... > unless it is dipped in liquid nitrogen "B^) > > (*) Think of peak-to-peak range of 2 volts. Then each bit represents 2/2^24 > or about 12 micro volts... that is so small... who is to hear that bottom > bit toggling?!?!?!?!?!? ;-) > That's my point - you can't hear the 24 bit systems. (^_^) The more I listen to 16 bit systems, then more I'm thinking you can hear the 16 bit systems at lower signal levels. (2/2^16= 3.05e-5 Volts.) To my way of thinking, a 24-bit audio system very closely approximates the voltage resolution of an analog system - i.e. you shouldn't be able to "pick out" an individual voltage step. The 96 kHz/24 bit DVDs I have sound amazing. And, I have the 44 kHz/16 bit CDs to compare them against - which sound pretty good in their own right. Plus, it's amazing how loud you can turn up the 96/24 discs. According to the back of the discs, you should get near 144 dB of Signal to Noise ratio - woo hoo! (^_^) > The thing that 24 brings is an ability for the hardware to retain more > accuracy as it processes the music on the way to the digital to analog > converter. The implementation of these filters requires that the number of > bits expands... and since you have to come down back to your DAC's width... > there will be some dynamic range lost. > That's true, but I think that the increased resolution has audible benefits beyond the "better internal numeric calculation" argument. > So, the thing is that, 16 bits seems good enough for 99% of the music > reproduction scenarios that we will listen to... unless you would like to > live at one of those chambers at Bell Labs... then you could something like > 18 bits. I respectfully disagree here. 20 bits for home systems at least. Preferably 24 bits, since we have the technology and bandwidth. (^_^) Why compromise when we don't have to? At least, that's my opinion. Cheers, Paul ------------------------------ Date: Tue, 06 Mar 2001 11:26:55 -0800 From: fernando Subject: Re: [idealcopy] [OFF TOPIC] Upcoming digital audio formats At 10:39 AM 3/6/01 -0800, Paul Pietromonaco wrote: > > >The main explanation is that in a digital system, the distortion rises >as > > >the signal level drops. If you get down near the noise floor of a 16 >bit > > >system, say around -90 dB from digital signal level 0 dB, you're >basically > > >around 100% THD. (1 bit of data to represent the audio = 100 % >distortion) > > > > This is not true... it is very strict what you say. First, it is >attributed > > that 1 bit gives you about 6 dB of dynamic range... so it would seem that > > > -96 dB is the bottom (as you are pretty much stating), but the fact >remains > > that with "dither", and properly shaped dither noise, you can achieve a > > greater SNR... and in fact, this is what happens... > > > >Well, you don't really get a greater SNR ratio, since the dither increases >the noise floor. I was being a bit slopy... thanks for the quote "B^) Yeah, you get a lower SNR, when considering the band out to 22.05 KHz. The alternate view that I had in mind is that the quantization noise is not flat anymore, but that it has been made uneven, with less noise power in the critical hearing frequencies... so... in effect, the SNR in the band of interest, say out to 13 KHz, is improved. >According to Principles of Digital Audio (pg. 68): "Our ears are quite good >at resolving narrow band signals below the noise floor, because of the >averaging properties of the basilar membrane. This is very true... even more true when you are familiar with the material! I remember a demo of low-bit rate encoders for speech, and you could not understand anything that was being said... then when listening to the clear speech, then again to the low-bit rate encoding... you can make it out. This works well for music that you are familiar with... but not for a first time listen. > The big factor is that to fully use a greater SNR when listening to > > music... you have to be in an anechoic chamber... I doubt that many >people > > can have that kind of dynamic range available in their home! Background > > noise should be louder than the smallest voltage for just 16 bits. > > > >In my opinion, that's only true if you're playing back the music at roughly >the level it was recorded. I often turn quiet pieces up to catch subtle >nuances - hence this becomes more of an issue. (I just have to remember to >turn 'em down before they get loud again, or my neighbors complain! (^_^)) >And, I keep my place fairly quiet too. Hey - I'm an audio fanatic - what >can I say? (^_^) Well, lucky you... but it seems to me that even in that case, you got to put out some serious sound pressure level... if most urban areas have a daytime background noise of 60 dB (conversation is about 50 dB????), then for the full range of sound your home stereo has to produce 150+ dB of SPL - -- that is very loud. Happy listening!!! :-) > > I get the same out of a properly mastered CD. One has to be careful, and > > you might have, on what is attributable to the noise floor in a stereo > > system. Think about it... with 24 bits, you have such a small voltage > > resolution(*), that there is no analog hardware that can resolve that... > > unless it is dipped in liquid nitrogen "B^) > > > > (*) Think of peak-to-peak range of 2 volts. Then each bit represents >2/2^24 > > or about 12 micro volts... that is so small... who is to hear that bottom > > > bit toggling?!?!?!?!?!? ;-) > > > >That's my point - you can't hear the 24 bit systems. (^_^) The more I >listen to 16 bit systems, then more I'm thinking you can hear the 16 bit >systems at lower signal levels. (2/2^16= 3.05e-5 Volts.) To my way of >thinking, a 24-bit audio system very closely approximates the voltage >resolution of an analog system - i.e. you shouldn't be able to "pick out" >an individual voltage step. The 96 kHz/24 bit DVDs I have sound amazing. I agree with this... the thing about it is how much of the 96/24 is due to the sampling at that rate, vs. the side benefits in reproduction that it brings... I am more with the latter... that if you were to "ideally" reproduce music at 96/24 vs. 44/16 you could not hear the difference... but the 96/24 allows you to build a more accurate reproduction machine. >According to the back of the discs, you should get >near 144 dB of Signal to Noise ratio - woo hoo! (^_^) Mom should be proud ;-) > > The thing that 24 brings is an ability for the hardware to retain more > > accuracy as it processes the music on the way to the digital to analog > > converter. The implementation of these filters requires that the number of > > bits expands... and since you have to come down back to your DAC's > width... > > there will be some dynamic range lost. > > > >That's true, but I think that the increased resolution has audible benefits >beyond the "better internal numeric calculation" argument. It could be to some very special ears... I would admit to that. > > So, the thing is that, 16 bits seems good enough for 99% of the music > > reproduction scenarios that we will listen to... unless you would like to > > > live at one of those chambers at Bell Labs... then you could something >like > > 18 bits. > >I respectfully disagree here. 20 bits for home systems at least. >Preferably 24 bits, since we have the technology and bandwidth. (^_^) Why >compromise when we don't have to? At least, that's my opinion. I am all for 96/24, even if it takes 10 years to get those nice low noise amplifiers up to room temperature and cheap enough... my point is more at attributing the benefits in the right places, rather than having them do 20 bits, and in a few years they move up to 24... then to pant-wetting 32. cheers! - -fernando ------------------------------ End of idealcopy-digest V4 #69 ******************************