[Nicholars]

I must admit that I dont think the differences are as big as some people make out with amps and cd players etc. I definately noticed pretty big jump in quality from NAD315bee > NAD326bee though.... Also going from a soundcard to external DAC was quite big difference although I dont think I could ever justify spending £1000's on DAC's and amplifiers...

Tbh I doubt I will be upgrading my DAC and NAD326bee for quite a long time because past the level of what I have I don't think the electronics really justifies the price increase.... but I will definately be wanting to upgrade my speakers and maybe try some room treatment.... IMO speakers costing large amounts of money are still clearly worth the asking price, whereas amplifiers and sources mabe not so much once you have decent setup.

[Tone-uk]

I think you are on the right track and take your time is my advice. If you can find the speakers you love and they are at the right price then you are sorted imo. Speakers can last a lifetime so extra cash spent there may seem more reasonable!

Also, try some tube amps, SET's etc. You might really like them! I can never get past the feeling they are not right but, a preference is a preference.

And, look at active speakers - maybe they are the way forward? Certainly look very promising.

[phil t]

Originally Posted by Tone-uk

But isn't the objective of hi-fi (in it's true definition) to reproduce whats on the disc etc. The aim of hi fi is simple enough and if a product can't do it in the 21st century then it's a cack design and rubbish product! IMO, of course.

No, I don’t believe that to be the case. The hi fi chain starts before the data, how ever that is stored.

That data is always flawed vice the original sound. Microphones are not perfect, microphone placement can affect sound, mixing desks and mixing techniques affect sound, etc

Me, I’m trying to re-create a musical event and if the less than (technically) perfect equipment makes that “performance” more believable, then that’s where I want to be.

[Avi]

Originally Posted by Tone-uk

But isn't the objective of hi-fi (in it's true definition) to reproduce whats on the disc etc.

That may be simply impossible.

Take for example the characteristics of the low pass filter applied to CD level content with common DAC types. Numerous choices i.e. linear phase, minimum phase, attenuation curve etc must be made. There is no ideal answer and typically the design decision here force compromise in terms of artefacts. That's before the other key factor of time i.e. sample clock accuracy and the things that affect it is factored.

Here's an example of how one manufacturer approaches filter design - http://www.ayre.com/pdf/Ayre_MP_White_Paper.pdf

Similar consideration also typically apply when creating CD level content from a higher rate sample master. Here are some examples of the characteristic of some popular solutions - http://src.infinitewave.ca/

Avi

[Mark.Yudkin]

The main cause of differentiated DAC performance is a result of the distinction between theoretical perfection and actual implementation in a real chip.

There are a sequence of tutorials on ADC and DAC circuits and what's involved here. Reviewing some of these should provide a much more comprehensive and comprehensible coverage of the implementation issues than we can summarize in short posts.

[Tone-uk]

Originally Posted by phil t

No, I don’t believe that to be the case. The hi fi chain starts before the data, how ever that is stored.

That data is always flawed vice the original sound. Microphones are not perfect, microphone placement can affect sound, mixing desks and mixing techniques affect sound, etc

Me, I’m trying to re-create a musical event and if the less than (technically) perfect equipment makes that “performance” more believable, then that’s where I want to be.

Absolutely agree with all your points - especially about microphones, mixing etc. The amount of data lost from the original, live performance is massive and cannot be retrieved - no matter how much is spent at home. It's just not possible. We also have to contend with things like the loudness wars.

However, by introducing what I'd call unnecessary colourations to a system means all songs are going to be affected. Some might sound better to you but, equally, some may be ruined (ie now too warm) - unless your preference is to colour them all a certain way. I'd prefer not to.

I think where we likely disagree is how a personal preference can be used as an indication of high fidelity and if one persons preference/opinion should be assumed to be more true than another - especially when price is factored in.

[Tone-uk]

Originally Posted by Mark.Yudkin

The main cause of differentiated DAC performance is a result of the distinction between theoretical perfection and actual implementation in a real chip.

There are a sequence of tutorials on ADC and DAC circuits and what's involved here. Reviewing some of these should provide a much more comprehensive and comprehensible coverage of the implementation issues than we can summarize in short posts.

Thanks Mark - Ill have a read of that.

I would say the main cause of DAC differences being "audible" has nothing to do with implementation and everything to do with how the comparison was undertaken.

(Obviously, there are always measured differences that indicate one DAC is the better performer but, if it isn't audible who gives a monkey's?)

[Tone-uk]

Originally Posted by Avi

That may be simply impossible.

Take for example the characteristics of the low pass filter applied to CD level content with common DAC types. Numerous choices i.e. linear phase, minimum phase, attenuation curve etc must be made. There is no ideal answer and typically the design decision here force compromise in terms of artefacts. That's before the other key factor of time i.e. sample clock accuracy and the things that affect it is factored.

Here's an example of how one manufacturer approaches filter design - http://www.ayre.com/pdf/Ayre_MP_White_Paper.pdf

Similar consideration also typically apply when creating CD level content from a higher rate sample master. Here are some examples of the characteristic of some popular solutions - SRC Comparisons

Avi

Hmmm get your point. Things can never be perfect but, there is a big difference between audible transparency and absolute transparency.

This is how the manufacturers get the latest fad moving - all of a sudden there's a new problem. Like jitter. Only relatively recently has it been one of THE issues but, it's been known about for decades and solved for decades too.

Next it's EMI or something like that.

However, why not take a couple of cd player's level match and compare blind and see what the result is? Simple. Thing is we both know why reviewers/magazines and manufactueres etc don't do this. They don't like the results. And so they invest their time and money on white papers, advertising, creative writing and marketing to convince those who have no technically understanding why the product sounds so good! It's a joke!

All they seem to do is create doubt and rather than invest in progress, invest in just talking and bringing out the next fad.

[Avi]

Originally Posted by Tone-uk

Hmmm get your point. Things can never be perfect but, there is a big difference between audible transparency and absolute transparency.

Have you compared the effect of SRC characteristics ?

Originally Posted by Tone-uk

However, why not take a couple of cd player's level match and compare blind and see what the result is? Simple.

I tend to avoid reading toi much in to commercial subjective reviews or "manufacturer speak" and try to compare myself. I use blind assessment in an attempt avoid to preconception influencing perception.

Originally Posted by Tone-uk

Thing is we both know why reviewers/magazines and manufactueres etc don't do this. They don't like the results. And so they invest their time and money on white papers, advertising, creative writing and marketing to convince those who have no technically understanding why the product sounds so good! It's a joke!

I guess that's one opinion. Unfortunately it can be difficult to see the wood because of the tree's and this can lead to extreme views.

Avi

[Tone-uk]

Hi AVI

No I have not compared the effects of SRC characteristics. I will do though.

What do your blind test comparisons tell you about, say, cd players, dacs etc? Are the differences audible? If so, are they easily detectable of difficult? Also, do you think the differences are greater than that of, say, speakers and will they contribute a great deal to the sound.? How about room acoustics - do you feel the differences that may or may not be there bewtween cd players will be swamped by a room (unless the room is good).

Also, what do you think of masking? When you write about how things can be shown to be different does if take into account masking?

Finally, do you find my views extreme? I must point out that they are based on the findings of many people far more qualified than myself. People like JJ Johnston, chief scientist at DTS, Alan Shaw of Harbeth etc etc.

[Avi]

Originally Posted by Tone-uk

What do your blind test comparisons tell you about, say, cd players, dacs etc? Are the differences audible? If so, are they easily detectable of difficult?

There are lots of variables but in some cases I've found a perceivable difference between solutions and how easy or hard can depend on the solution.

Differences may be analogues to "spot the difference" pictures. Some times it's pretty obvious and sometimes it takes more time to work out what's different. A casual glance may suggest the pictures are the same when in fact they aren't.

I've found preconception is a powerful influence and I'd prefer not to fool myself if possible. The assessment I do are to satisfy my own interest not reviewing or telling others what they should or shouldn't perceive or prefer.

Originally Posted by Tone-uk

Also, do you think the differences are greater than that of, say, speakers and will they contribute a great deal to the sound.? How about room acoustics - do you feel the differences that may or may not be there bewtween cd players will be swamped by a room (unless the room is good).

Lots of variable but I think generally the speaker and room will typically have a more significant impact on the overall performance. A poor room, setup and/or speakers may mask other differences.

Originally Posted by Tone-uk

Finally, do you find my views extreme? I must point out that they are based on the findings of many people far more qualified than myself. People like JJ Johnston, chief scientist at DTS, Alan Shaw of Harbeth etc etc.

I've also read lots over the years from AES and many other sources including JJ in the audio myths workshop etc.

Don't take this the wrong way but I have no opinion or interest in your views. I'm only really interested in what I perceive in my system and attempt as much as possible to minimise factors that will influence what I "think" I hear.

Avi

[Tone-uk]

Originally Posted by Avi

There are lots of variables but in some cases I've found a perceivable difference between solutions and how easy or hard can depend on the solution.

Differences may be analogues to "spot the difference" pictures. Some times it's pretty obvious and sometimes it takes more time to work out what's different. A casual glance may suggest the pictures are the same when in fact they aren't.

I've found preconception is a powerful influence and I'd prefer not to fool myself if possible. The assessment I do are to satisfy my own interest not reviewing or telling others what they should or shouldn't perceive or prefer.



Lots of variable but I think generally the speaker and room will typically have a more significant impact on the overall performance. A poor room, setup and/or speakers may mask other differences.



I've also read lots over the years from AES and many other sources including JJ in the audio myths workshop etc.

Don't take this the wrong way but I have no opinion or interest in your views. I'm only really interested in what I perceive in my system and attempt as much as possible to minimise factors that will influence what I "think" I hear.

Avi

Thanks Avi

I loved the last paragraph! Made me smile. People tend to think their view is more true than anothers just because they say so! And people buy it - "must be true - fat bloke said"!

Thats why I rely on what science tells us rather than what a manufacturer or dealer says or indeed a forum poster. Afterall audio is a science.

[Avi]

Originally Posted by Tone-uk

Thats why I rely on what science tells us rather than what a manufacturer or dealer says or indeed a forum poster. Afterall audio is a science.

I'm very cautious of manufacturer "marketing speak" and forum posts can be challenging.

Unfortunately science doesn't always agree or reach the same conclusion either.

Avi

[Welwynnick]

Originally Posted by Tone-uk

This is how the manufacturers get the latest fad moving - all of a sudden there's a new problem. Like jitter. Only relatively recently has it been one of THE issues but, it's been known about for decades and solved for decades too.

Not that I'm aware of.

You say that you're influenced by what science tells us, so do you believe the "scientific" research paper that tells us that jitter is inaudible below 250ns? Do you believe that, instead of what people tell you?

Originally Posted by Tone-uk

Next it's EMI or something like that.

Funny you should say that. Watch this space...

[Tone-uk]

Originally Posted by Welwynnick

Not that I'm aware of.

You say that you're influenced by what science tells us, so do you believe the "scientific" research paper that tells us that jitter is inaudible below 250ns? Do you believe that, instead of what people tell you?

Funny you should say that. Watch this space...

Which research paper is that one? Oh never mind.

It's a pretty moot point now since the manufacturers have had the last few years to bring the levels down to zilch. No one can use the jitter excuse anymore and so a new baddie needs to be invented!

[Avi]

Originally Posted by Tone-uk

It's a pretty moot point now since the manufacturers have had the last few years to bring the levels down to zilch. No one can use the jitter excuse anymore and so a new baddie needs to be invented!

Wow which manufacturers achieve zilch (zero ?) jitter and does this involve employing something like asynchronous sample rate conversion ?

Avi

[Tone-uk]

Maybe zilch is too strong. Hows about tiny amounts that we cannot possibly hear and there are far far more issues to sort out first before worrying about jitter? Is that ok?

[Avi]

Originally Posted by Tone-uk

Maybe zilch is too strong. Hows about tiny amounts that we cannot possibly hear and there are far far more issues to sort out first before worrying about jitter? Is that ok?

I haven't seen any definitive science that proves what amount we cannot possibly perceive.

I have compared solutions that exhibit higher amount of correlated jitter to those that don't.

Avi

[Doomlord_uk]

so do you believe the "scientific" research paper that tells us that jitter is inaudible below 250ns?

Here's a link to the original paper. And, tbh, yes I'll take scientific (ie, properly done!) research over forum anecdotes and subjective listening tests.

[Avi]

Originally Posted by Doomlord_uk

Here's a link to the original paper. And, tbh, yes I'll take scientific (ie, properly done!) research over forum anecdotes and subjective listening tests.

I can see numbers related to simulated jitter added to a signal at the computer source but what type of measurement was done to assess actual jitter at DAC ?

Avi

[Tone-uk]

Originally Posted by Avi

I haven't seen any definitive science that proves what amount we cannot possibly perceive.

I have compared solutions that exhibit higher amount of correlated jitter to those that don't.

Avi

No, there isn't any definitive science - which is how the manufacturers got away with it and why so many bought into the idea. I remember at the time discussing it with Welwynnick and he said something along the lines of "better to be safe than sorry". I had to give up and accept that point.

However, I would have thought the manufacturers have got the levels down to even the lowest figures some papers suggested by now? I think Dunn said 20ps at 20hz or something. Maybe I'm wrong.

However, moving your head an inch is more likely to create a bigger audible effect than jitter at 20ps! With that in mind should we not all invest in head braces? (joke). I'm sure someone also figured out your pulse is more likely to be a problem!! Lol. Who can seriously say jitter at those levels is at all important - especially when the room the equipment is in has never even been looked at.

[Avi]

Originally Posted by Tone-uk

No, there isn't any definitive science - which is how the manufacturers got away with it and why so many bought into the idea. I remember at the time discussing it with Welwynnick and he said something along the lines of "better to be safe than sorry". I had to give up and accept that point.

Assess the products first hand in a properly treated listening room avoiding the influence of preconception as much as possible. If there's a preference that's reliably identifiable I guess it doesn't really matter what science or manufactures claim.

Avi

[Doomlord_uk]

Originally Posted by Avi

I can see numbers related to simulated jitter added to a signal at the computer source but what type of measurement was done to assess actual jitter at DAC ?

Avi

If you read the paper it was 'input vs output' - ie added jitter at input, listener's ability to distinguish the added jitter at 'output'. The result was that at levels below 250ns the resultant output signal distortion was inaudible - even to experienced listeners in optimal listening conditions (ie best-case scenario for motivated test subjects). Therefore what jitter was doing inside that system was beside the point. The paper answers the question of what level of jitter has to exist in a system for it to begin to be audible.

Unless someone is able to point out reasonable flaws in the methodology then I think this paper's results stand.

[Welwynnick]

Originally Posted by Doomlord_uk

Here's a link to the original paper. And, tbh, yes I'll take scientific (ie, properly done!) research over forum anecdotes and subjective listening tests.

Even when its bad science?

[Doomlord_uk]

Even when its bad science?

Is it?

[Welwynnick]

Very bad.

I think that one paper, because it HAS been easily accessible on the web for people that don't want to try too hard, has damaged a generation of audio.

Too many people assume it's scientific and credible, and aren't smart enough to ask the right questions, and see the gaping holes in its logic. Hook, line and sinker.

When listening to 24/96 material through my DAC I really cannot imagine it sounding much better although I do have recordings in 24/192, I have been pondering for some time on getting a new streamer to be able to cope with this but I only have a few is it going to be worth it?

[BlueWizard]

I'm going to offer something of a tangential perspective. In the days of its origins, CD 16b/44.1k was the peak of technology in consumer audio. It was the best they could reasonably do.

But today, I'm pretty sure 16b/44.1k DACs don't even exist. All DACs have much higher capability that is scaled down to fit what is now an antiquated standard. The cheapest crappiest NEW computer around should easily be able to handle 24b/96k internally.

Why should we move to 24b/96k as the new standard? Well... because we can.

I suspect in a few years (OK, quite a few years), as we are flying around in our jet packs, talking on our cyborg cell-phone implants, we will be discussing whether it is really necessary to take the next step up to 64b/1048K?

And the answer will be, well ... we can, so we will.

Steve/bluewizard

[Welwynnick]

Here's a link to the original paper. And, tbh, yes I'll take scientific (ie, properly done!) research over forum anecdotes and subjective listening tests.

I think that would be a mistake in this instance. I can honestly say that forum anecdotes may be more reliable than that scientific paper. The problem is that it does look scientific, but if you’re going to accept what it says, you need to be able to qualify it. That’s a bit like saying that you need to know better than the authors in order to read the paper, which is a bit unfair on most readers.

There are lots of subjective vs objective discussion threads on AVS, and I’ve been taking part in one where this paper was discussed. I said there were gaping holes in the logic and methodology of the research, and I challenged everyone to read it again and try to identify the faults. As usual, no-one got it – most people just talked about correlated vs random jitter. This paper has been discussed many times on many forums round the World, and the only people who seem to get it are Charles Hansen and AVI.

So how about it, everyone else? Read that paper again and see if you can see what they did wrong?

Nick

[Tone-uk]

I know what the problem is.

To help others and move this on a bit - think about the pc and it's jitter levels.