mpm@hpfcms.UUCP (mpm) (03/04/85)
I've been shopping around for a good CD player for several months
now. Being in love with technology I, of course, want 4x oversampling
and digital filtering. (I guess that means 14-bit DACs, n'est pas?)
Being rich in spirit and poorer in cash (just bought a house), I, of
course, want low cost. (My target: under $500 if possible; I'm look-
ing for discounts.)
Anyway I've seen three players that I know qualify: the Revox
(overwhelming front panel), the Meridian (lacking the program features
I want - i.e. ability to REPEAT a TRACK), and the Nakamichi (the BIG
one). Except for the Meridian, they are all well beyond what I want
to spend.
Well, my saga continues: This weekend I saw a review of the new
Carver CD player in Audio. I think I'm in love! It has all the above
goodies PLUS one of those nifty special processing circuits that Car-
ver is well known (but is he loved) for. AND, that circuit - the
Digital Time Lens - is switchable, so if you don't like it, don't use
it. The best news - price is (suggested) $649.
Well naturally I would like to see, touch, hear, etc. one. Does
anyone know anything about this unit? Any comments on Carver's repu-
tation for reliability, quality of construction, etc.? Does ANYONE
discount Carver equipment? Is the suggested price the same as the
ACTUAL price? (Anybody seen one in a store yet?) What is the player's
REAL immunity to shock and disk defects?
I wouldn't be surprised if someone wants to jump in with basenote
drift on this one. Well, I would be glad to hear what you have to say
on: Carver, other "fancy" players, effectiveness of "technological
whizzies", pricing, and (of course) my ignorance of technical issues.
Anything else, use your own judgement.
-- Mike "time to get back into audio equipment" McCarthy
{ihnp4, hplabs} hpfcla!mpm
P.S. Being rather naive about the underlying technology I would like
some feedback on this:
From previous discussions I have guessed that CD players incor-
porate 4x oversampling and (usually?) digitial signal processing use
the underlying Phillips mechanism. I think that Phillips uses 14-bit
converters (D-A). Thus, apparently, there are no players that offer
16-bit DACs (do you need them) AND 4x oversampling.
Frankly I don't think this conjecture makes much sense. But with
a degree in CS, most of the technical discussions so far have blown
right by me. So please help straighten me out.shauns@vice.UUCP (Shaun Simpkins) (03/08/85)
> {ihnp4, hplabs} hpfcla!mpm > > P.S. Being rather naive about the underlying technology I would like > some feedback on this: > > From previous discussions I have guessed that CD players incor- > porate 4x oversampling and (usually?) digitial signal processing use > the underlying Phillips mechanism. I think that Phillips uses 14-bit > converters (D-A). Thus, apparently, there are no players that offer > 16-bit DACs (do you need them) AND 4x oversampling. > > Frankly I don't think this conjecture makes much sense. But with > a degree in CS, most of the technical discussions so far have blown > right by me. So please help straighten me out. > -- Mike "time to get back into audio equipment" McCarthy Yes, Mike, there are 16 bit 4x oversampling CD players; they just use one 16 bit DAC per channel. I posted something about this to the net a few months ago. The upshot is that data words are presented to the DAC once every 5.6us with 4x oversampling. If you peruse your catalogs for cheapo 16-bit DACs you will find that they settle (i.e., the output converges to within a least significant bit of the correct value) in about 4-5us. Thus, only one data channel can be presented to a 16-bit DAC at a time. DAC sharing (i.e., time sharing one DAC between channels with an analog sample/hold for each channel after the DAC) effectively doubles the data rate. Solution? Go to 14 bit DACs that settle faster if you want to use only one, or make do with 2x oversampling if you still want 16 bit conversion. This is what Yamaha did. I suspect that Philips made a silk purse out of a sow's ear with their 14bit 4x scheme \- at the time that chip set came out cheap 5us 16bit DACs weren't even on the drawing board but 14 bits was within the reach of existing IC processes. Burr-Brown and Harris didn't report on their monolithic 16-bit 5us DACs until 1983. Companies that make 16 bit 4x players are: Kyocera, Nakamichi, NEC. I'm not sure about the Carver - probably dual DAC. 16 bit 2x: Yamaha, possibly the new Sony lines - CDP302 et al, TEAC. 14 bit 4x: Philips, Meridian, and licensees. This list is not by any means exhaustive. Check out your dealer and the October issue of Audio for a more complete listing. BTW, gramophone just reviewed the CD-X1 by Yamaha. They almost tripped over their tongue praising it. Unusual for that usually kind but critical rag. The wandering squash, -- Shaun Simpkins uucp: {ucbvax,decvax,chico,pur-ee,cbosg,ihnss}!teklabs!tekcad!vice!shauns CSnet: shauns@tek ARPAnet:shauns.tek@rand-relay
rfg@hound.UUCP (R.GRANTGES) (03/10/85)
[] Before you plunge completely, you might want to look at the umpteenth generation SONY CDP-102. I glimpsed a review of it in that new magazine (Digital Audio??). List is $450. It has 2x oversampling and apparently 16 bit decoding and quotes its phase linearity (I recall). It interested me, anyhow. I have the CDP-101 which everyone likes to act so superior to. Nevertheless, tin ear that I am, (that's Sn-ear, by the way (why is it that everyone hates puns except their author?)) I have been nothing other than pleased for about a year now. I have not had even one incident of mistracking (knock on wood) and while some of my approx 45 cds are much less super than others, they are all readily distinguished from cassettes, records, etc by an inherent clarity that stamps them as something special. -- "It's the thought, if any, that counts!" Dick Grantges hound!rfg
kunz@hplsle.UUCP (kunz) (03/12/85)
(Re: why do 4x oversampling?) vice!shauns says.... > ... DAC sharing (i.e., > channel after the DAC) effectively doubles the data rate. Solution? Go to > 14 bit DACs that settle faster if you want to use only one, or make do with > 2x oversampling if you still want 16 bit conversion. This is what Yamaha did. > I suspect that Philips made a silk purse out of a sow's ear with their 14bit > 4x scheme \- at the time that chip set came out cheap 5us 16bit DACs weren't > even on the drawing board but 14 bits was within the reach of existing IC > processes. ... > The wandering squash, > -- > Shaun Simpkins > > uucp: {ucbvax,decvax,chico,pur-ee,cbosg,ihnss}!teklabs!tekcad!vice!shauns > CSnet: shauns@tek > ARPAnet:shauns.tek@rand-relay > The real reason to do 4x oversampling is not to accomodate the lack of technology in DACs, but rather, to ease the design requirements on the analog filter that follows the DAC. If one were to use 1x sampling (44.1 KHz), the analog filter would need to have a very sharp cutoff around 20 KHz which would result in tremendous phase shift (but can we hear it!) and potential ripple in the pass band (say... over 0.1 dB, but can we hear it?). With the cutoff for the filter out at 88 KHz (approx), the phase shift at 20 KHz can be substantially reduced and the passband ripple likewise reduced. Components in the frequency domain between 20 Khz and 88 KHz are reduced by the digital filter before the DACs. There is also the whole topic of aliasing, which 4x oversampling makes easier to accomodate for. (I.E. eleminate) 4x oversampling is accomplished by padding the one real sample with 3 zero samples (BTW - done in the digital filter) and producing a sample stream at the output of the digital filter at a rate of 176.4 KHz. This sampled data only needs to be converted with 14 bit precision because the 'extra' bits come from the oversampled samples. (It's a difficult concept to write about but can be shown graphically very easily) The resultant output still has 96 dB dynamic range even though the DAC is a 14 bit one. Using 4x oversampling and 16 bit DACs is overkill since there is no information available in the LSB's (unless one wants to manufacture it as noise somewhere in the system). I own some of Carver's gizmos and they do seem to work as claimed. The holographic pre-amp is a real winner, especially with small satellite speakers (I have M&Ks with a subwoofer). Good luck in your quest for the 'perfect' CD! Bob Kunz {hplabs!hp-pcd, fluke, teltone}!hplsla!kunz Hewlett-Packard Lake Stevens Instrument Division (206) 335-2135
sullivan@harvard.ARPA (John Sullivan) (03/15/85)
> > Using 4x oversampling and 16 bit DACs is overkill since there is no > information available in the LSB's (unless one wants to manufacture it as > noise somewhere in the system). You are right about the main reason for oversampling being to reduce the demands on the analog filter. But you still should use 16 bit DAC's. Suppose your input signal is 17,18,17,18,17,18,... Whether your extra samples inserted before the digital averaging are 0's or repeats of the last sample makes no difference. You still can get out a signal 17,17,17,17,18,18,18,18,17,17,17,17,... from this process, UNLESS you drop the lower two bits, in which case it will look flat. John M. Sullivan sullivan@harvard
mpm@hpfcms.UUCP (mpm) (03/19/85)
I found another article (somewhere, I don't remember where) about
the Carver player. Apparently it uses 2x oversampling (not 4x), but
has 16-bit DACs.
I think Sherwood also uses 2x oversampling. Why stop there?
-- Mike "in the interests of audio education" McCarthy
(hpfcla!hpfcms!) mpm
P.S. After two years on the market, there is still a lot of confusion
and misunderstanding of this technology. I now look with skepticism on
advertising claims (even statements from trade journalists) about CD
players.
Watch out for the following:
- misrepresentation of sampling rate
Multiplexing a single DAC admittedly gives a real sampling
rate of 88.2 KHz for both channels, but the standard 44.1
KHz rate for each channel. This is NOT the same thing as
2x oversampling. Obvious to us perhaps, but what about
the typical consumer?
- confusion between error correction and digital filtering
As far as I know, ALL players have ECC. (The Sony/Phillips
CD standard calls for the redundant info on the media. I
guess there'e no guarantee a player will use it. [Maybe
that's what was wrong with the old Sears deck.]) Only a
few (more expensive) players have nifty digital filtering -
often in conjunction with oversampling.
Add these to other areas of confusion: the difference between
mastering and digital recording, frequency response versus sampling
rate, indexing, tracks versus bands, etc. (This is getting complicated.)kunz@hplsle.UUCP (kunz) (03/20/85)
> Suppose your input signal is 17,18,17,18,17,18,... > Whether your extra samples inserted before the digital averaging are > 0's or repeats of the last sample makes no difference. You still can > get out a signal 17,17,17,17,18,18,18,18,17,17,17,17,... from this process, > UNLESS you drop the lower two bits, in which case it will look flat. It seems that we hit the topic of dithering... Now you've forced me to learn something new (again -- I knew it in school). Yes, I'll agree that dithering noise in a sampled system can be a real problem and probably can be eliminated by 16 bit DACs (for CDs) and then 'tossing' the LSB's. This is what you mean, isn't it? I wonder though, is that signal down at -90 dB a concern. I suppose for the purists it is! :-) This is a good discussion, I'd like to keep up the technical discussion of CDs and related audio stuff in this category. It keeps me up on things. -------------------------------------------------------------------------- Bob Kunz {ihnp4!hplabs!hp-pcd, uw-beaver!fluke}!hplsla!kunz Hewlett-Packard Lake Stevens Instrument Division 8600 Soper Hill Rd Everett, Washington 98205-1298 (206) 335-2135
mpm@hpfcms.UUCP (mpm) (03/20/85)
Is anybody out there?
By the way: does anybody have opinions, info, stories about
Kyocera? They have a CD player that looks like it would satisfy
me (after bankruptcy that is - it's $1600.00). Any locals know
where to go for an audition?
-- Mike "I'm not paranoid, just a little worried." McCarthy
{ ihnp4 | hplabs }!hpfcla!mpmshauns@vice.UUCP (Shaun Simpkins) (03/20/85)
> > (Re: why do 4x oversampling?) > > > ... DAC sharing (i.e., > > channel after the DAC) effectively doubles the data rate. Solution? Go to > > 14 bit DACs that settle faster if you want to use only one, or make do with > > 2x oversampling if you still want 16 bit conversion. This is what Yamaha did. > > I suspect that Philips made a silk purse out of a sow's ear with their 14bit > > 4x scheme \- at the time that chip set came out cheap 5us 16bit DACs weren't > > even on the drawing board but 14 bits was within the reach of existing IC > > processes. ... > > > > The real reason to do 4x oversampling is not to accomodate the lack of > technology in DACs, but rather, to ease the design requirements on the > analog filter that follows the DAC. > ^^^^Undoubtedly. But my point was: given that you wish to oversample, how can it be done the cheapest? There are certain technology limitations. Quote Philips: "The conversion of the 16 bit words into a analog signal is performed ... by a 14 bit digital-to-analog converter available as an integrated circuit and capable of operating at the high sampling rate of 176.4 kHz." > > 4x oversampling is accomplished by padding the one real sample with 3 zero > samples (BTW - done in the digital filter) and producing a sample stream at > the output of the digital filter at a rate of 176.4 KHz. This sampled data > only needs to be converted with 14 bit precision because the 'extra' bits > come from the oversampled samples. (It's a difficult concept to write > about but can be shown graphically very easily) The resultant output still > has 96 dB dynamic range even though the DAC is a 14 bit one. > > Using 4x oversampling and 16 bit DACs is overkill since there is no > information available in the LSB's (unless one wants to manufacture it as > noise somewhere in the system). > > Bob Kunz > Continuing the quote from Philips: "Partly because of the fourfold oversampling and partly because of the feedback of the rounding-off errors in antiphase, rounding off to 14 bits does not result in a higher noise contribution in the audio band. This remains at the magnitude corresponding to a 16 bit quantization ... so that even though there is a 14 bit digital-to-analog converter it is still possible to think in terms of a 16 bit conversion system." Strike me dead if I'm wrong here, but the resolution of the Philips system is NOT the same as the 16bit systems - it's 14bits. It's just the S/N that's the same. Let's follow a signal through the Philips system. At the beginning, the 2 LSBs are truncated and applied to the digital filter. The filter interpolates 3 new data points between samples, `recovering' the lost 2 bits if 16 were used at the output, which they are not. Thus the base S/N of the Philips system is 84dB. Since the final bandwidth of the system is 1/4 the effective bandwidth of the filter output(i.e., the maximum bandwidth of an input signal sampled at the filter clock frequency) we gain 6dB in S/N. The roundoff feedback further averages the quantization error to return to a 96dB S/N. But what if we kept the 2 bits that we threw away at the filter output and ran them into a sixteen bit DAC? Assuming a noise free input (< 1 input LSB), we'd get an 17-bit S/N ratio from the action of the reconstruction filter! But-if we put in a slow staircase signal, the output would only move in 16 bit increments, just very stable 16 bit increments - with a touch of rounding at the edges. Similarly, the Philips system moves in 14bit increments, just very stable 14 bit increments. I would say that there's information in those last 2 bits - as long as the input noise is less than an LSB - even at 4x oversampling. Indeed, if we were able to find an 18-bit DAC that could run at 176 kHz (and a 16-bit ADC with a S/N of more than 108dB) we could use a 16-bit input digital filter and get an output S/N of 19+ bits! 114dB! Anyone care to clear up this matter? The wandering squash, -- Shaun Simpkins uucp: {ucbvax,decvax,chico,pur-ee,cbosg,ihnss}!teklabs!tekcad!vice!shauns CSnet: shauns@tek ARPAnet:shauns.tek@rand-relay
bruceh@tekig.UUCP (Bruce Harrington) (03/20/85)
> > > > Using 4x oversampling and 16 bit DACs is overkill since there is no > > information available in the LSB's (unless one wants to manufacture it as > > noise somewhere in the system). > > You are right about the main reason for oversampling being to reduce the > demands on the analog filter. But you still should use 16 bit DAC's. > Suppose your input signal is 17,18,17,18,17,18,... > Whether your extra samples inserted before the digital averaging are > 0's or repeats of the last sample makes no difference. You still can > get out a signal 17,17,17,17,18,18,18,18,17,17,17,17,... from this process, > UNLESS you drop the lower two bits, in which case it will look flat. > > John M. Sullivan > sullivan@harvard *** REPLACE THIS MESSAGE WITH YOUR LINE *** I'm not sure this is exactly what is going on in a 4x 14 bit oversampling system, but we are currently using the following system in an instrument to get 14 bit resolution from a 12 bit DAC. 4 "time slices" of a 12 bit DAC are integrated over time to get voltages between the steps on the DAC. DAC Output 2 LSBs | T1 | T2 | T3 | T4 | -------------------------------- 11 /--------------\____/ The low and high levels here are output voltage level N, and 10 /---------\_________/ output voltage level N + 1, respectively, where N is the 01 /----\______________/ 12 bit result of dividing the intended 14 bit value by 4. 00 \___________________/ Bruce Harrington Tektronix, Beaverton
agn@cmu-cs-k.ARPA (Andreas Nowatzyk) (03/23/85)
I hope cancel worked, otherwise please excuse my empty post. This was
caused by our job-security program for UNIX system maintainer: they
keep on moving EMACS to strange locations...
On the subject on 4X 14bit vs. nX 16bit (n=1,2,4):
4x oversampling with 14bit is theoretically equivalent to 16 bit *if*
the 14 bit DAC has 16 bit precision (that is, if it is equivalent to
a 16 bit one that has its 2 lsb's tied to some constant). However, this
is not the case. See the IEEE Journal on Solid State Circuits, Vol SC-14,
pp. 552-556 for an interesting paper on the Philips 14bit DAC. At 25C,
it approaches 15bit precision, but over the entire temperature range it
is just a good 14bit DAC. They also claim a S/N of 'only' 90db while
16bit could approach 96db. At the time of the design (arround 1979), it
was not feasable to integrate a 16bit DAC without expensive laser trimming.
So part of the reason for 14bit is plain cost reduction.
On the subject of ringing:
Ringing is a characteristic of a particular filter design and has little
to do with the components used ('capacitors'). A given n-pol filter
design has a certain ammount of ringing, no matter of its realization:
analog or digital. Digital filters however can be designed to have
better performance by the use of a large number of poles. The Philips
design has something like 64 poles while the one in Sony's D5 has
9 (if I understand the circuit correctly). The number of poles in
an analog filter is limited by the number of discrete parts involved
and by their precision. The number of poles in a digital design is
limited only by the number of gates you can afford.
Cheers, Andreas Usenet: ...!seismo!cmu-cs-k!agn
Arpa: Andreas.Nowatzyk@cmu-cs-k.arpa