The DVD/Dac thing

How about the analogue output stage... (leaving asside for a second the 1's and 0's part)

 

I think your grossly overstating your case.. How can you sit through a 90 minute movie if the sound is so bad? Afterall, speech is quite a broadband signal, and one to which the ear is very sensitive, and any shortcoming's in that frequency range would be quite obvious. Then there's the incidental music...

It also begs the question of why you chose such a bad sounding unit (which one is it?)

I asked this question because: if you heard this DVD player and it was so bad, why did you choose it? Perhaps you choose it because it wouldn't show up your CD player

Follow the links there to see why I bought the 390S:

http://www.zerogain.com/forum/showpost.php?p=48052&postcount=4
https://www.audio-forums.com/as-rediect/showthread.php?p=48433&highlight=390s#post48433

 

The Jan 2005 issue of HiFi World reviews a £120 Universal disc Pioneer DVD player and gives it a major thumbs up (except for build quality). Interestingly the technical measurement weren't so hot compared to the more expensive players they have tested. It's the DV-575A and TBH at that price it's not gone to kill the bank balance to give it a try.

 

Here are the independently measured results for the SONY DVP-NC685V:

Frequency response
20 Hz to 20 kHz +0.012, 0.37 dB

Distortion (THD+N, 1 kHz)
at 0 dBFS....................................................0.003%
at 20 dBFS...................................................0.017%

Noise level (re 20 dBFS, A-wtd)........ 75.6 dB

Would you agree that the Sony is competantly designed? If it isn't, in what way is it lacking?

 

I use the digital output downmixed to 48khz into the digital input of the eclipse....

 

Having Just Purchased a Pioneer DV 575 A, I can tell you that the sound quality is pretty dire. Even on DVD-A & SACD, it is still poor, sure there is that normal improvement, a smoother sound you get from SACD/DVD-A, but I would say the price reflects the poor sound which is really thin, with poor width. Something which the HI-FI Mags all seem to claim, that it's hard to get a well designed DVD Player that has SACD & DVD-A, to give good CD Replay. It seems the DVD Players that don't have SACD / DVD-A replay manage to produce good CD Replay. I also have a Old Toshiba SD 220e DVD Player that is pretty good at replay of CD, it can be a little harsh, but in real world terms it is streets ahead in Audio Reproduction compaired to my Pioneer DV 575A,even through it dosen't play DVD-A & SACD !!
The simple reason I purchased the Pioneer DV 575A, is so I can run it through my DAC, and thankfully it's Digital Output is pretty good, I've run it through my M Audio Superdac 96/24 & Musical Fidelity X-DAC v3, and it's sounds pretty good.
In short having SACD & DVD-A is a bonus, I now plan to have a Trichcord Clock 4 Fitted, in my eyes if you looking to use the Pioneer DV 575A as a transport only,once you have had it re-clocked it's a bargain. Of Course, if you are not brother about SACD/DVD-A replay,you always buy a Toshiba DVD Player, Im not sure of the Model, but What HI-FI Rates one on CD Replay.
Having own a Pionner 747A, I can tell you that the DV 575A is really poor, again what do you expect for £120 !!

My Equipment
PIONEER A300R PRECISION (FULL MONTY TOM EVANS MOD)
MONITOR AUDIO 20 SEC CELEBRATION
MUSICAL FIDELITY X-DAC V3
MUSICAL FIDELITY X-10 V3 TUBE STAGE
VAN DEN HUL D102 MK 111 INTERCONNECTS
QED SILVER SPEAKER CABLE BI-WIRED
PIONEER DV 575A
TOSHIBA SD 220E

 

Oedipus,

Since you consider all that is important about CD players/DACs are minor level differences between the outputs, non flat frequency response, literally and figuratively differences "in the noise"...

I would like to ask you if in your view would it be possible to detect any
differences between transports when feeding same DAC via a digital optical cable... for instance the DAC64 with the 4 sec. buffer engaged?

 

All I can summise so far is that the ATC atives will ATC preamp are so bloody good you could run an Aiwa stack through them and they would sound great

But I am still not convinced

 

also,sorry but slightly off topic..........Oedipus,the measurement thing has come up again,and I am genning up trying to learn this stuff,it seems to be important,but.................

my first learning........................flat/no flat reponse on a lab table is one thing,intregrating that flat reponse into a modern lounge is another.How do you correlate the so called accurate systems,all measured as being good,and the problems with room acoustics,that dramatically change the sopund anyway? ie its been debated ad-infinitum in here already whether anyone can actually here these minute differences in measurements,but whats not in doubt is the effect of room acoustics.

 

Hi,

Could you kindly advise as to how and why the specific measurements you have quoted are relevant measures of quality?

Simply the noise level appears appalingly high (--75db A weighted is likely much worse unweighted and at best it is 95dbh S/n Level. Hell, I getter noise levels UNWEIGTED from SE Valve Amplifiers.

Also, omited are several relvant measurements:

1) Jitter
2) Level of IMD
3) Level of "noisefloor modulation" (aka what I prefer to call noiseload test)

I have not analysed the Sony's design, so I cannot be certain, but on past experience with Sony gear, highly unlikely.

To be precise one would have to apply extensive analysis, but at an educated guess I would thinks that following are likely true:

1) Grossly bad powersupply design
2) Incompetent ground routing
3) Inadequate quality in core areas (DAC, Digital filter)
4) Analogue stage inadequate to ahndle the RF noise from the converter without IMD products folding back into the audible range (a little difficult to measure, but easy to hear)
5) grossly incomptetent implementation of the master clock and other digital circuitry, resulting in High Jitter

Just to illustrate, for "accurate" CD replay (eg 16 Bit resolution) we ABSOLUTELY REQUIRE the following easily measured qualities:

1) THD & S/N better than -96db reference to digital full scale, unweighted
2) Jitter below 100pS
3) Absence of noisefloor modulation effects in Digital Filter, DAC and analogue stage

The number of CD and DVD Players that manage the above is extremely small, though it is demonstrably not particulary difficult to achieve these (it is expensiove though, but that as they say is another story).

Ciao T

 

Hi,

A player that is competently designed will not show audible differences to the digital studio master, when playing a Disk produced from that master. Any incompetently designed player will fail this test.

And yes, it also accrues then that all competently designed players will sound absolutely identical given the same source material.

The factors are maifold.

It begins with the way CD's and DVD's are read and the error correction ivvolved. Unlike an LP which will distort or skip if ANY sufficiently large error is present, the Players use electronics to "guess" what should have been in the damaged section. Depending how the mechaniscs are designed, how well the transport manages mechanical resonances and so on - the accuracy of the datastream containing the zero's and ones will vary and more or less parts of the signal will be interpolated from adjecent samples. This is very audible.

A good illustration is to take a badly scratched CD which is not yet at the point of skipping (eg a lot of microscratches) and to rip this to Harddisk using EAC in "paranoid mode". EAC in paranoid mode will make many attempts for reading the samples correctly, unlike a CD player which will fail once and interpolate. Once you have ripped the whole disk burn the resulting file onto a recordable CD that you know to allow correct burning with low erroro rate in your writer. Compare subjectively the scratched original and the ripped and recorded copy. The copy invariably sound smuch better, because it actually has less read errors.

Okay, now lets assume we have the actual 1's and 0's of the original CD and no read errors. Home free?

Nope, if you modulate the players system clock you modulate the signal, even the rest of the conversion is perfect. The actual critical points are at the conversion to analogue signals, namely the clock at the DAC, usually the Wordclock only for multibit DAC's and both word and system clock for delta sigma DAC's and wherever sample rate conversion is involved.

Okay, so we now make sure that our jitter is low enough not to matter (How low is that? It depends on both sample frequency and wordlength, in simple terms, non-oversampled 16Bit/44.1KHz sample rate audio tolerates around 100pS jitter, if oversampling or higher sample rates are employed and if the wordlength exceeds 16 Bit it is much less). Home free?

Nope, now our signal is converted by a DAC with usually any number of imperfections and amplified by analogue circuits that are subject to any number of problems.

Sadly few designers recognise and address any of the issues, the Ostritch like "bits are bits" view is by far too prevalent.

Ciao T

 

Because frequency response, noise, distortion, loudness and to a very limited extent phase, are all that can be heard. Flat frequency response, along with low noise and distortion completely characterize a CD player as far as human hearing is concerned.

Look again. It's relative to minus 20dbFS... (Omitted from the original figures, but it is obvious from the context).

Jitter as a standalone measurement is meaningless - the same amount of jitter affects different DAC architectures in different ways. Moreover, if there is an effect of jitter it shows up in the THD+N number.

Why the concern over IMD? THD and IMD are just different manifestations of the same underlying distortion in the transfer function. If you can argue a case to show that for a CD player it is possible to have a high level of IMD, but a small level of THD, then go ahead.

Noise modulation is less than <0.5dB for the sony unit.

Well, if those "problems" don't show up in the measured peformance at the output, then they aren't really problems at all are they? Sure you can take the lid off it and do "extensive analysis", but you're chasing your tail if the output measures this well. I might also add that your extensive analysis might well be pointless given the sophistication of the chips involved, and would require 100's of thousands of dollars of test equipment (especially to measure pico-seconds of jitter) - gear that makes Audio Precision boxes look like toys I might add.

While were on the subject of test equipment, it's worth noting that simply going into these units and swapping components for "better" ones may well lead to an audibly different product, but without the (expensive) test equipment required to show a measured improvement it is quite likely that any audible difference is the result of a degradation in the units performance: at best the addition of a tone control, and at worst a significant increase in the distortion and noise performance of the unit.

Your assertion that there are effects that are "difficult to measure", but "easy to hear" holds no merit. Human hearing is quite flawed and a whole lot of research has been done on hearing using grossly distorted signals, which are easy to generate and verify (measure), yet are inaudible to the ear.

good engineering is doing for a nickle what any fool can do for a dollar. What you are asserting to be "grossly incompetant" is merely expedient design in order to reduce cost at no audible loss in performance.

Just how much better than -96dB are you expecting? BTW, it make sense to use A-weighting, so why are you objecting to it? [State your assumptions about the use of dither (and it's probability distribution) to support your case for performance beyond 96dB SNR.]

Jitter is meaningless, unless you have considerably more to say about the chips you are using, and even then, your unlikely to be able to make any bold claims about the effect of input side jitter on the analog output unless you are either the chip designer or the manufacturer.

 

Not so fast. You can't immediately jump to interpolation, that's not how it works.

CD audio data is protected by a cross-interleave Reed-Soloman block code which provides incredibly robust protection and correction to bit errors and substantial burst errors (eg. correlated errors in the region of a scratch). The net result of this redundant correcting code is the data is recovered perfectly, as though there were no bit read errors. During this process the player is not making a "guess".

If you have some data on how often the RS correcting code fails to completely cover the bit errors and therefore the number of times per second interpolation happens, then post it. Also tell how you gathered that data, or name the paper/source you are quoting.

 

Very impressive I must say Oedipus.

So why do these inexpensive DVD players sound shit then when compared to a half decent CD player? Because they do I can tell you.

So I'll leave it to you, the self proclaimed master of the scope to come up with the reason, patent the fix, and make your self millions. In the meantime, i'll listen using a decent front end if that's OK with you. But you carry on spouting this stuff - it's highly amusing to some

 

I certainly would affirm Merlins findings, cheap dvd players do indeed sound serial pants (In analogue output mode)
AS for using a benchmark dac1 with it (have 2 as you know), still a mile away from the Wadia, and using the Arcam dvd79 straight in to the SE's dac section with summer studio cable (as wire makes no difference) is chalk and cheese, you need a nos dac and dac 64 to be able to pull this difference off

 

Guys,

It must be all in your minds

 

Of course it is. Where else could it be?

 

Just mentioned it over on PFM, but it seems to be the ATC boys who claim to not hear differences in source components.

Something wrong with the speakers perhaps

 

I bet that went down well

Nevertheless, a strong possibility.

 

well,i'm totally confused (not surprise there then)

first its all source first,as in Naim philosophy,then they can't here any difference,then its actually DVD players are better than CDPs because a scientist said so.So I declare the answer to my own thread is;

bollocks to the lot of it,except at least its a thread for anyone to read at lunch time

actually,I will go further than that,I reckon ATC just couldn't be bothered to keep expensive gear in their listening room,i doubt its anymore complicated than that,and source first is still relevant,DVD players generally sound crap,however,s dedicated audio only cheapo universal player with SACDs may be the way,at least its easy to frigging understand.All the rest just seems pyscho-babble at best or just fantasy at worst.

I can't see anyone ditching 2k CDPs in favour of DVD players,but I can see some changing to reasonably priced dedicated audio only SACD players.As for the ATC/Mana versus the world ,I just think thats the immovable ball/unstoppable post thing again

still,these are atbthe end of the day quite interesting threads to read,the PFM thread in particular has been excellent imo,just what a bbs should be about.Try it,debate it,call each names,bugger off,come back,and argue again.Excellent read