Digital edge, a thing of the past?

[domfjbrown]

I remember when CD first came out - I was with my dad one day in 1984 and we ended up in Laskeys - I got to hear a black frontloader - probably the Sony CDP101) while dad was mooching around (maybe he was weighing up BUYING a TV instead of renting ad infinitum ) and I thought it sounded great. Mind you, I didn't have access to a decent turntable and we didn't own one stereo item in our house until 1986! When my dad told me how much the machine cost I just said something like "that's my pocket money for the next 10 years!!!!"

Originally posted by Goomer
I have been happily living with both formats since 1985 too, and have never really understood this one versus the other argument.

I had a Trophy (think Matsui-then!) mono radio-cassette in 1985, but I do have 2 1985 vintage CD players - a Technics SLP2 (twin 16bit DACs, 8x oversampling (I think)) and a Marantz CD54 (14bit 4x oversampling, based on the Philips CD104 but with decent HDAM output stages). Both sound reasonable but then the Marantz has the top end rolled right off, and the Technics sounds phasey. Both sound like they're producing modulation noise on paino and female vocals - whether that's age or the digital electronics, I don't know.

Neither produce anywhere near as good a music as even my NAD 533 turntable; certainly on something like Little Earthquakes by Tori Amos, her Bosendorfer (?) sounds like a Bontempi organ!!! Well - ok, not that bad, but DEFINITELY synthesized - like an early PCM digital piano...

ALL LP12s sound broken by default - BECAUSE they're LP12s. :duck:

 

[michaelab]

Originally posted by technobear
Me too. I bought a Marantz CD50SE in 1989

I still have one. Bought in 1990 for £350. It was used as my transport until quite recently and is about to be re-born into my 2nd system in the office

It obviously doesn't sound anything like as good as my DAC64 but when I lent my DAC64 to a friend for a few days a month or two ago I was using it again as a CD player (the first time for a long time) and I was pleasantly surprised. It really isn't that bad at all. OK, it does have a Trichord Clock3 in there but still, pretty impressed.

About the sample frequency, I'll have to do some digging but I think it's pretty well established that you can faithfully capture any frequency upto half the sample frequency (the Nyquist frequency).

Michael.

 

[wadia-miester]

Oh, I didn't mean for this to decend into Tones is the odd one out section , Tones straight forward Honest question here, can you really not tell a difference between CD/Vinyl ? what about when you listened to Titains ?
As far as red book being a flawed medium, Vinyl prefect is it? both strenghts and weaknessies, SACD, I have a midly warm SACD 1, humm, needs a lot of work I feel The closest I've personaly heard SACD get to Vinyl is a FULL DCS firwire/Verdi V's clear audio master ref, megs cart arm blah, blah, the dcs was better in the mid band but lost out every where else, cost of each Knocking on the door of £30K, which is plain rude, personaly for me the 'most desirable' TT I've yet heard a Well Tempered 7 Keostu red cart, tricked delphini, and manley steel head, superb, streets ahead of any other TT's I've heard, however, have totaly no desire to leave red book, with the preformance I have already. prehaps it's just down to what you have a reference and judge other kit accordingly. Tone

 

[tones]

Originally posted by domfjbrown
ALL LP12s sound broken by default - BECAUSE they're LP12s. :duck:

Come out, Dom, everyone else is airing their prejudices,, so why shouldn't you?

 

[domfjbrown]

I have to admit I'm surprised (sorry dude!) that Tones can't tell the difference either - on EVERY turntable I've ever heard I can still hear subtle hf tracking noise and groove wear - even on mighty decks like Henryt's Orbe. I've not heard a Clearaudio Master Reference though

 

[tones]

Originally posted by wadia-miester
Oh, I didn't mean for this to decend into Tones is the odd one out section , Tones straight forward Honest question here, can you really not tell a difference between CD/Vinyl ? what about when you listened to Titains ?

My fault, old bean, I shouldn't have strayed into a discussion where iconoclasm wasn't wanted.

Can I not tell a difference between vinyl and CD? No, or at least nothing where I could say, "Now THAT'S definitely digital!" (Or definitely vinyl). Oh, except the vinyl surface noises give it away eventually. To me, they represent equally acceptable ways of hearing music, with CD preferred because of its convenience (I find the whole vinyl bit a bore and an inconvenience).

Titian's? Never got the chance to do an A/B comparison - I had the Handel in both CD and vinyl form with me, but there were so many other interesting things to listen to, so I didn't feel it appropriate to waste everybody's time.

 

[tones]

Originally posted by domfjbrown
I have to admit I'm surprised (sorry dude!) that Tones can't tell the difference either - on EVERY turntable I've ever heard I can still hear subtle hf tracking noise and groove wear - even on mighty decks like Henryt's Orbe. I've not heard a Clearaudio Master Reference though

Well, I meant apart from the surface noise, Dom, which gives away the vinyl sooner or later.

 

[technobear]

Originally posted by michaelab
About the sample frequency, I'll have to do some digging but I think it's pretty well established that you can faithfully capture any frequency upto half the sample frequency (the Nyquist frequency).

Well let's say we're trying to capture a tone at 22.05 kHz. Dependent on the phase relationship between the signal and the sampling clock, we might capture all the peaks and troughs or close to them and thus have a pulse train which we might be able to reassemble into a sine wave. It is equally possible that we might capture all the nulls in the signal and record *nothing at all*. So the amplitude of the captured signal will depend on the phase relationship between the signal and the sample clock at the time.

For any other high frequency, like my earlier example, we will not capture the peaks and troughs, but points along the waveform which gradually rise and fall in amplitude as time goes on. This will surely cause a tone to be produced at very low frequencies (a bit like turntable rumble I guess). We certainly won't have a pulse train which follows a smooth sine wave.

 

[michaelab]

OK, at exactly half the sample frequency it's possible that the samples will all be in the nulls and nothing will be captured so you can only reliably capture any frequency upto the Nyquist frequency. Reasonably basic explanation of all that here

If what you capture is merely the points of the sine wave then how are you to know whether the original was a sine wave or a triangular wave? Well, fourier analysis of a triangular wave at 22kHz would show that it contains components (harmonics) of higher frequencies so they have not been captured and are irrelevant. RedBook wouldn't accurately capture a triangular waveform at 22kHz - in fact at that frequency it would only capture a pure sine wave accurately but since we can't hear the difference there is no issue.

I would challenge any forum member to be able to hear any sound at 22kHz (or even 20) let alone tell the difference between a sine wave and a triangular wave (or any other shape wave) at those frequencies.

I don't believe all this stuff about ultrasonics being audible in the form of beat or difference tones either, or having some other kind of magical effect lower down the scale.

What I do believe though is that humans can hear transients very accurately and the rise of a transient attack might momentarily have a much higher frequency than 22kHz which the human ear is sensitive too. This is the only explanation I can think of why higher than RedBook resolution digital might sound better (allthough I have never heard it to do so).

Michael.

 

[julian2002]

chris,
but we're actually sampling at 44.1 khz so you'll never get that situation - hence the anything up to 1/2 the sample frequency perfectly. this 'gaps between the samples' thing rears it's ugly head every so often and it's utter shite. anything that is missed in the 'gaps' is at a frequency above the defined max frequency (in this case 22.05khz) and therefore outside the scope of the argument. most peoples hearing starts to drasticly atenuate at about 15khz so even the 'brick wall filters kill music' argument is specious and should be taken with ladles of salt. i've always found comparable vinyl and cd players to be 'different' not better much as another comparable cd player is 'different'. yes generally as you increase the quality of the components and construction vinyl gets better, but then so do cd players, the biggest difference is one of development, vinyl has had nigh on 90 years of it whilst cd is only just allowed to drink in the states, if cd players were refined for another 70 years i'm sure they'd be so far past vinyl it wouldn;t even be funny.
another thing that probably tarred cd at it's inception was the fact that a lot of early cd releases were cut from vinyl masters complete with evil riaa equalisation. this if i'm not mistaken attenuates the bass frequencies and accentuates the highs... oh just what early cd's were being accused of, funny that. i wonder if anyone has tried taking one of these supposedly awful old cd's and run it through a phono stage without the amplifier section. bet it sound just like the vinyl version does only without all that snap, crackle and pop accompanied by a bit of rumble and wow.
cheers


julian

 

[technobear]

I have to go now and I'm away for the weekend but I'd like to continue this discussion next week. If we really can't hear much above 15 kHz then why does SACD/DVD-A sound more natural and less edgy than CD? Is it perhaps just an extension of the oversampling/upsampling theory, i.e. that you can use a much more benign filter if your DAC works at a higher frequency than 44.1 kHz, and that sample rate actually has not a lot to do with it?

I'll be back. What's more I'll be back with diagrams! I would really like to know how a DAC reassembles a 21 kHz pure sine wave from a 44.1 kHz pulse train.

Have a great weekend! I intend to

 

[merlin]

You are all barking!

FWIW, I don't think CD has the resolution to accurately reproduce HF, but to explain why would only end up going round in circles so ....arsed can't be.

Just listen to some real instruments and recordings, I reckon that would do it for most, with the notable exception of Swiss gentlemen with eardrums as perforated as their most famous cheeses.

 

[michaelab]

Originally posted by merlin
You are all barking!

Ahh merlin - I see the good old "reformed smoker syndrome" is alive an kicking

From digital denizen to vinyl vulture in a few short months...

Michael.

 

[tones]

Originally posted by merlin
with the notable exception of Swiss gentlemen with eardrums as perforated as their most famous cheeses.

Eh?

 

[TonyL]

I've always questioned this and I still do. If you draw a sine wave, say ten cycles, and pretend it's a 21 kHz sine wave. Now mark on the sine wave, sample points at 44.1 kHz. Then draw as smooth a curve as you can joining the sample points. The new wave is *nothing* like the original. So can anyone explain to me just how CD supposedly manages to accurately reproduce anything at such high frequencies.

This is exactly the point I'm arguing from. The whole sampling thing is like standing on a platform and blinking really fast when a speeding train passes by ââ'¬â€œ you can see a series of short images of the train, but inevitably you do not see the whole event. The fact that the majority of serious audiophiles still find dragging a little rock through a rotating slab of plastic produces a more realistic reproduction of recorded music IMHO proves that the red book CD format is fundamentally screwed. There has to be a better way to do things, but I remain convinced CD ain't it!

As for early generation CD players - my dad has a Philips CD104 circa 1982 and it sounds a lot better than many budget and mid range players of today.

This amazes me. My first proper exposure to CD was a Philips CD150 (4xoversampling / 16 bit IIRC), a well reviewed player that someone lent me in about 1985 or so. It was truly horrible. Harsh, edgy, plasticy sound that virtually took my ears off ââ'¬â€œ my turntable was at the time a Xerxes / RB300 / AT5 running into a Naim 62 / Hicap / 140 and Kan IIs, to say it slaughtered the Philips in every way was a dramatic understatement. I also borrowed a early Sony CD player which was almost as bad. The first acceptable and affordable player was IMHO the Rotel 965BX, still a huge way behind the Xerxes, but good enough for me to buy as a second source.

Well, if the LP12 is broken, it is making a remarkably good job of pretending not to be broken! No, sorry, the EAR has made no difference. Far from murdering each other, ther LP12 and the Meridian 588 coexist rather nicely and both give enormous pleasure without any apparent tendency for one to assault the other.

Sorry Tones, I thought you had a really old Sony CD, not a Meridian ââ'¬â€œ the 588 is IIRC not a bad player at all ââ'¬â€œ the Linn should still comfortably beat it, but not to the huge extent I had in mind when I made my earlier post!

Tony.

 

[TonyL]

PS I also tried a early ex BBC ReVox 14 bit CD player - it was dreadful. If you played a quiet disc such as certain classical recordings you could clearly hear the quantisation distortion!

Early CD player really did suck!

Tony.

 

[lowrider]

Tone,

Have you tried this one, Analog Devices OP275s...

I also bought a CD player around 1985, I didn't like the Sony sound, so I bought a B&O top loader, Philips inside, it was pretty good, and gorgeous, still harsher than goodish modern designs... :MILD:

 

[MartinC]

Originally posted by TonyL
Early CD player really did suck!

Is possibly the disagreement in this thread over the performance of early CD players one of software rather than hardware? If someone heard a 1985 player in 1985 they heard it with 1985 standard digital recording (which wasn't too good ), and so could easily thought it crap. I have some CDs of that age which still sound crap on a much more modern player. On the other hand, playing a decent modern recording on a 1985 player might not sound so bad?

 

[lowrider]

Originally posted by MartinC
Is possibly the disagreement in this thread over the performance of early CD players one of software rather than hardware? If someone heard a 1985 player in 1985 they heard it with 1985 standard digital recording (which wasn't too good ), and so could easily thought it crap. I have some CDs of that age which still sound crap on a much more modern player. On the other hand, playing a decent modern recording on a 1985 player might not sound so bad?

I bought most of my CDs before 1987, and they sound better on new gear...

 

[tones]

Originally posted by TonyL
Sorry Tones, I thought you had a really old Sony CD, not a Meridian ââ'¬â€œ the 588 is IIRC not a bad player at all ââ'¬â€œ the Linn should still comfortably beat it, but not to the huge extent I had in mind when I made my earlier post!

Tony.

I guess neither of us will ever convince the other (not that we would try), so we'll beg to differ on the relative virtues of CD and vinyl. For my part, I have never heard the problems you describe with CD. Lucky me, don't you think?