XXHighEnd

I was forwarded a link to an interesting article a few days back regarding the virtues (or not) of high resolution recording, that I thought most here might enjoy reading.

http://people.xiph.org/~xiphmont/demo/neil-young.html (http://people.xiph.org/~xiphmont/demo/neil-young.html)

Now, I happen to agree with most of the ideas put forward, but I'm sure there will be a few here in disagreement  :grin:

Enjoy the read .. I did!

Cheers,

Russ

Yes it is interesting.
Not much mumbo jumbo in that article.

Still I have a feeling that age is quite important regarding these issues.

Lots of people really enjoy hirez music and hirez dacs. Still just as many prefer 16bit/18bit dacs (tda1541, ad1865 etc..)
Strange isn't it?

Interesting article.  The one area that really lost me was the argument about bandwidth limiting.  I'd rather not make the format a crutch for bad amplifier design.

Hi Calibrator - a very good read thanks for posting. It certainly supports what I hear with my system - some of the very best recordings I have were recorded in the 50/60's and are recorded at 16/44.

Although I have some exceptional modern recordings they tend to be produced by specialist recording engineers and are mostly 16/44. I have some superb 24/96 recordings also but not (I think) because they are 24/96 more because the recording engineer was superb.

Paul

What a stupid blog post. It reminds my about those who state that "there are no audible differences between cables, because there are no measurable differences to speak of". Or who claims that "there cannot be differences between software players or computers if both are bit-perfect, etc etc bla bla"

For this purpose, ABX sucks. Rapid repetions in an ABX test is doomed to end in a nul result. ABX is close to useless for finetasting in sound. You have to listen for longer periode to various music without stress.

He also advocates oversampling. The NOS1 DAC is all about avoiding oversampling.

Several recording engineers has stated many times that when recording from the same source, higher bit/sampling (than redbook) sounds better.

Even ripping your LPs sounds better in 24/96 than 16/44.

Unfortunately articles like this creates more confusion.

Of course, the difference is not that big, and some CDs sound spectacular too. And some old recordings sound bad whatever format. But that doesnt change the fact that a modern well recorded album sounds better in hi-rez.
-The sound is more "fluent", microdynamics are more "real" and there are more nuances even in the bass. It is the tiny aspects of sound reproduction which improves.

(Sorry for faulty spelling. I wrote it on my phone sitting on the loo. A fitting tribute to the blog post reference).

:yahoo: :rofl:

pedal - feel free to elaborate further.

When we know about all the "false" marketing tricks from some resellers of HD music - and all they do is to introduce "noise" in one form or another - how can this be good?

If your ears tell you otherwise it is a good thing for you.
I have analyzed some HD material (using izotope RX, audiocity etc.). I can tell you, not all is good.
I think it is a very good article. Maybe everything is not 100% correct but it makes you think doesn't it?

Well, what the article did - or merely postings about it elsewhere - is that I FINALLY seem to have found a reason why the best natively recorded HiRes can sound worse than the to 16/44.1 decimated same recording.

So as we all know, I keep on being persistent that HiRes doesn't sound better at all, and this with the NOS1 behaving electrically 100% the same as when it's fed with Redbook.
That this is subjective to some extend (like pedal being out of line with this (or me !) is clear too of course, but at least it is me trying to squeeze the best out of Redbook. Ehm, you all do of course, but I can use different means (and you can only use mine for it :scratching:).

It is too early perhaps, but I have now found a sheer technical reason how native HiRes can sound worse.
But, it needs the NOS1 and Arc Prediction Upsampling to be in that technical realm. I mean, with any normal DAC this difference can hardly be there (or actually not at all), which should imply that on that normal DAC HiRes sound better than Redbook.
That now Redbook sounds relatively miserable is something else ...

Btw, to some extend, just using Arc Prediction with a normal DAC will also let HiRes *not* sound better. But here YMMV very much, and I myself can't even judge it (lacking such a DAC).

Eh, more later ?

I think so too. And this is to such a degree that I am not much able to "attack" it right from that origine.
That I can write pages full of it is another matter.

The point is : there is much more going on, and the article is not about that. So it is about the in-parallel other things going on which makes the article "wrong". But within itself I see it as quite consistent ...

Peter

Is it? Well yes as Peter rightly avoids using the word "oversampling" because of its interpolation connotations. But of course NOS1 does increase the input data stream to one with a much higher sample rate and for exactly the same reasons that oversampling was intended to do. Unless of course any of us listen at 44.1 which I certainly don't.


Certainly agree with that but I would add "if at all". And then I would ask why are my very best sounding recordings 16/44?

With the advantage of a qwert keyboard and after a good night sleep, I can deepen my view on this so called “article”:

I found 2 interesting parts:

1. The tech talk about IM distortion. Which is an audible factor for owners of faulty constructed amplifiers. –Stable amplifiers with low distortion and bandwidth up to, and beyond 100kHz is easily achievable in regular AB transistor amplifiers these days, if the designer wants to. The “faulty” ones in this regard are to be found among tube amplifiers and class D amplifiers. Owners of such should aware, ha-ha. But pls don’t deny high resolution music formats just because you have a low tech amplifier…  :P

2. The physical mechanics of our ear and the reference to the Fletcher and Munson curve is also interesting read.

------------

The rest of the blog post by "Monty" I find to be malicious and highly biased propaganda.

Mr Monty happens to be Christopher Montgomery, who has spend the last 15 years developing and advocating Vorbis. A lossy music compression format competing in the marked with such as MP3 and Apples AAC. In fact, the blog post is a direct attack on Apple’s iTunes. So much for his credibility.

Monty is a low-rez guy. If you read carefully all his blog post, and the sources he lists, they are mainly from the “objectivist” camp living in a simplified world.
-They don’t trust individual listening experiences. Subjective listening only becomes objective when run through a ABX process with multiple repetitions of the same piece of music.
With this kind of mentality and ABX method, they find no difference between 24/96 and 16/44. Also there are numerous ABX tests showing no difference between lossy MP3 and CD. So there you go: Lossy MP3 is as good as 24/96! Say no more.

Mr Monty is a malicious propagandist because he omits known facts, which I am sure he knows very well. This man has spent 10 years programming and writing algorithms for his lossy music format. For sure he knows that our hearing is not only a physical matter (the outer mechanical part of our ear). Just as important is the psychoacoustic part. How our brain analyzes and percepts. In this field the scientists are only scratching the surface. There is so much going on in our brain, processing sounds, etc.

His analogy to the eye is quite misguiding and irrelevant. Especially when it comes to performing comparisons. Comparing A and B with sight is easy, because you can watch (let’s say) 2 photographs side by side, in real time.
But you cannot compare 2 pieces of music simultainsly. Both A and B must be heard one at a time and “stored” in your memory. Then compared from memory. It’s a completely different story, which Mr Monty elegantly avoids to mention.

Also he simplifies the Nyquist–Shannon sampling theorem. Mainly because increased bit depth is not only a matter of wider dynamic range, but also a more accurate calculating of the tiny amplitudes. Shortly put, for those into fishing with nets: More bits create a bigger net, but also a net with smaller masks. It captures more fish (bigger net) but also smaller fish (smaller masks). 

And what about square waves? Many music instruments are full of both DC- and square wave signals. That’s why we hear them as “real”. And that is a large reason why we hear them as “reproduces” on records. Because neither the format nor the hardware has enough bandwidth. –This is also a gross omission in Mr Montys piece of propaganda.     

There are numerous statements from well respected music engineers that higher sampling rates and bit depths improves the sound quality. And this is of course when they record and compare from the same source. 16/44 is not the threshold of our hearing, as Mr Monty falsely advocates. To paraphrase Mr Churchill I would say Red Book is not the end, but merely the begging of our sonic journey!

I have done quite some research in this matter. If anybody is interested I can post some links from recording engineers, discussing the limitations of 16/44.

-------------------

Much of the first generation of hi-rez music was made shortly after year 2000, when DVD-A was launched. Most of the effort went into the formats 5.1 surround sound virtues. They dusted of the mastertapes of the old top selling albums and made some very good multichannel releases. But they didn’t pay same attention to the stereo versions on the same disk. Some of these are cr*ppy sounding with heavy compression, which at that time had developed into ia decease among mastering engineers. :censored:

I guess some hi-fi enthusiasts have developed skepticism towards hi-rez, due to bad experience from such 1st. generation releases. But you cannot judge a technical format, based on wrongly usage. That is quite obvious. Let’s be patient, and let the format develop.

I have several good hi-rez recordings and they sound wonderful. More alive and more “natural” than Red book which in comparison has a slight “artificial” signature. A kind of subtle “dryness”, even if the recording is very soave, soaked in a glorious “tubey” sound signature.  (It’s hard to explain, sorry).

PS: My amplifiers have flat response from 0 to almost 1MHz, with near immeasurable distortion. In fact my playback chain is entirely DC-coupled, from NOS1 DAC to my loudspeakers, including the DC-coupled active crossovers. 

Okay, at bringing forward those thoughts about the article, I can't go much against that.
I must admit that I stopped reading when the MP3 stuff started.

Peter

I think we must be careful with accusations which may seem to allow to spring from logic, but where we still can't guarantee it is so.

Let's take one example : PeterSt. What does he do ? explicitly advocate 16/44.1. Okay, not because it "is so", but because there is a good reason to make it that. This, against that handful of Hires, which in the mean time PeterSt judges as s*cking all over. He really does you know.

So you see ? this is just an example, which hopefully (!!!) proves that we must be careful.

Btw, some months back I proposed MP3 as an explicit filtering means. Not that it worked out, but at least I had some theories about it.
This, in the context of how bad *I* can find that article.
But honestly, I have no judgement. It is too well written to go against it right away. Remember, in my view ...

Peter

Relax, Peter - your work cannot be compared with Monty’s work. You are aiming for the sky; he is aiming for the ground. With XXHE/NOS1 DAC you are trying to create the best possible sound quality in the world for the relatively few connoisseurs of high end. Mr Monty has created and advocates Vorbis, a lossy format for low-rez streaming audio in the mass marked.

The arguments in Mr Montys blog post circles around “how bad can things be done without audible degrading”. He is advocating mediocrity. You are pushing for perfection.

Basically all his blog post is a contradiction to your work. Here are some examples:
Red Book CD is good enough as it is today, applied with standard oversampling technology.
Music signals below 20 and above 20k has no relevance for people.
Sound differences don’t exist unless they are proven in ABX.
ABX proves that people doesn’t discern between CD and hi-rez.
ABX proves that people doesn’t discern between MP3 and CD.
Etc, etc

-------------

I think Mr Monty wrote his blog piece badly on purpose. Applying a very black/white perspective he has managed to provoke the high end community. His article is linked everywhere and made a big impact. He is probably having a good laugh now.

I dislike people who are pissing on my passion for highest fidelity sound reproduction. That’s why I wrote my previous post, pointing out Mr Montys malicious propaganda.

Hey pedal,

Again, when you put it forward like that, I can only agree.
Still I think I "behave" similar.

... And as you know by know the whole article can be p*ssed on from all imagineable directions. Also by me.
But it is a complete indirect excercise and the reasoning about it is always for ... the better. Well, hopefully.

As you have seen I am now in the direction of "almost being able" to prove why Hires isn't better after all. It may fail, especially with you around as the exception of someone who really finds that Hires *is* the better sounding. So, including all like a good resolving system, XXHighEnd and a NOS1.
This should make me wonder, right ?

But what so stupidly remains is that all so much depends on the contents of that Hires itself. You know too that you judged something like good, while I did not plus some graphs should prove my right.
But merely, it is not about proving one's right. It is about IMproving, and for me it intrigues that Hires should do that while for me it does not at all. Or not enough.

The lucky thing is that Redbook can be so good.
I personally think this is way more important than trying to get Hires right. It isn't there anyway, so to speak.

In any event ... respectfully ...  :)
Peter

I am pragmatic about these format matters. I have about 3000 CDs ripped to hard disk. It’s my main music source. Hi-rez is only a supplement, with perhaps 30 titles sounding outstanding. (But the number is increasing). So my CD ratio is 100:1.

What counts in the end is the sound quality, whatever format. Phasure has taken 16/44 playback to such a high level, it is almost unbelievable. If the marked knew its performance, the NOS1 DAC would sell in thousands.

I am eagerly waiting your new discoveries. (Every night when I go to bed, I fantasize about “those cymbals brought more forward, without sounding too brash or harsh” which you announced for the next XXHE version).