[gregorio]

Quote:

Originally Posted by milkweg /img/forum/go_quote.gif That's not something I have ever noticed and I give higher sampling rates than 44.1khz for music a big fat 0.

I give them a "big fat 0" too. But that's because I use a high quality professional ADC/DAC. Cheaper DACs are likely to have much better sounding reconstruction filters at higher sample freq, as it's difficult and expensive to create good quality steep filters which are required for 44.1k sample freq. Using higher sample freq allows the implementation of much smoother, easier and cheaper to implement filters. This is true for both ADCs and DACs. Some genres, classical music is an example, tend to use much higher recording standards and much better quality equipment, it's unlikely that mediocre ADCs have been used for recording. So, depending on what genre of music you listen to and the quality of your DAC, you may (or may not) hear higher quality at the higher sample freqs.

G

 

[Pio2001]

Quote:

Originally Posted by gregorio /img/forum/go_quote.gif Using higher sample freq allows the implementation of much smoother, easier and cheaper to implement filters.

In order not to drift off topic, just a technical point : this problem have been solved since the early days of CD players using oversampling. The DAC, fed with 44.1 kHz, oversamples it between 4x and 256x (according to the model), with a FIR filter, very easy to implement.

 

[gregorio]

Quote:

Originally Posted by Pio2001 /img/forum/go_quote.gif In order not to drift off topic, just a technical point : this problem have been solved since the early days of CD players using oversampling. The DAC, fed with 44.1 kHz, oversamples it between 4x and 256x (according to the model), with a FIR filter, very easy to implement.

This is only true up to a point. Oversampling in a DAC does away for the need for a steep reconstruction filter and can rely on a much cheaper to implement (FIR) filter at say 96kFs/s. However, a resampling DAC does not solve the problem of a poorly implemented anti-alias filter in the ADC when recording. Audio artefacts from an anti-alias filter (in the ADC) when recording are part of the mix and cannot be removed later (using oversampling or any other method).

BTW, FIR filters still have their problems if implemented in a small frequency band, so I would disagree that they are always easy to implement.

So, a 96kfs/s recording being replayed at this original sample rate can, under certain circumstances (not recorded on a great ADC or played back on a mediocre DAC), sound better. Although I don't belive there are many consumers who have the equipment or ears to percieve this potential difference, I am not willing to say that it can never make a difference, which is why I gave it only 500 points out of a million.

G

 

[JaZZ]

You (both) forget that even an ideal «reconstruction filter» is anything but ideal for transient response; the same applies for the ADC's antialiasing filtering. Which makes a higher sampling rate the technically and sonically better solution anyway.

I have decent experience with the three audio formats: CD, SACD, DVD-A. While I know that many excuse the apparently often superior sound of the high-rez formats with the extra care for (re)mastering (maybe true in single cases), that doesn't coincide with my own experiences. Provided both CD and SACD layer are made from the same master (not too hard to identify), there's always a distinct sonic difference predominantly manifesting itself in the treble, in the form of higher definition and palpability -- whereas the CD layer sounds comparably smeared and glassy. In turn the SACD layer sometimes lacks the CD layer's liquidity.

The latter is fully preserved in the DVD-A version (24/96) of an identical recording (e.g. from MDG's two double-disc editions with Shostakovitch symphonies), and it additionally shows increased transparency throughout the spectrum, plus enhanced three-dimensionality -- compared to the CD.

But the sonic advantage of the high-rez formats only pays out with corresponding masters. Below a sampling rate of 50 kHz, there's no clear benefit from SACD or DVD-A; SACD even often sounds worse than CD in those cases. Which means that the format itself isn't sonically perfect and perfectly transparent as well.

Generally all three formats sound different, although CD and DVD-A are related. If I had to choose, I'd give DVD-A the edge. Redbook CD is definitely not quite good enough to fulfill all my sonic expectations when it comes to resolution and definition. It's nevertheless good enough to be enjoyable, also given the progress represented by the current players and DACs in comparison to the early generations, and of course the familiarization with this format and its sonic characteristic, as it has become my standard format for logical reasons (add to this my personal crossfeed).
.

 

[gregorio]

Quote:

Originally Posted by JaZZ /img/forum/go_quote.gif You (both) forget that even an ideal «reconstruction filter» is anything but ideal for transient response; the same applies for the ADC's antialiasing filtering. Which makes a higher sampling rate the technically and sonically better solution anyway.

I didn't forget that view. I just didn't repeat it as it's questionable whether there is any noticable effect within the hearing spectrum. You are still going to get better sound quality with a high quality ADC at 44.1k than you are with a mediocre ADC at 96k.

"Provided both CD and SACD layer are made from the same master (not too hard to identify), there's always a distinct sonic difference predominantly manifesting itself in the treble, in the form of higher definition and palpability -- whereas the CD layer sounds comparably smeared and glassy."

So this is what is leading you to believe Hi-Rez is better? That's not entirely reasonable, as there may be a number of other factors causing this effect. For example, here's one of several possibilities; when requantising from a 24bit master to 16bit, a noise shaped dither algorithm is usually applied which adds considerable additional noise to the higher frequency range. And, the effects of this added re-distributed sound could be precisely, to the letter, the effects you have described! Top class producers and mastering engineers need years of experience and by that time usually can't hear much above 14-16kHz, so they don't notice the effect of the noise shaped algorithm. If they did notice this effect maybe they would use a different algorithm re-distributing the sound elsewhere in the audio spectrum. So the fault is not necessarily in the CD format but in how it is produced.

I'm not "having a go at you" Jazz, it's just that there is so much of this on Head-Fi. An observation is made, a conclusion is drawn and a belief system is initiated. However, with only a partial understanding of the very complex processes involved in the creation of digital audio the conclusion may not be entirely accurate. When an expert comes along to dispel the myth it's already too late, a belief system exists and it's going to take more than logic or proof to change it. The old addage springs to mind: "a little knowledge is dangerous"!

G

 

[m0ofassa]

Quote:

Originally Posted by gregorio /img/forum/go_quote.gif Interesting how high many of you put headphones, relative to the recording quality. Rather strange really as the products you are listening to have been designed for speakers and a room acoustic, not headphones. I'm not trying to insult anyone, I'm just facinated by the view and opinions of consumers, all I usually get to speak to is other professionals. Maybe if I understand better what is going on inside the heads of consumers it may influence how I mix, or at least put a warning sticker on my products!! G

recording quality is much more important when I am listening to sound effects, but I am listening to music most of the time. You can have music without a quality recording, but have a headphone and DAC (as well as the bad recording) and still have music.

 

[JaZZ]

Quote:

Originally Posted by gregorio /img/forum/go_quote.gif I didn't forget that view. I just didn't repeat it as it's questionable whether there is any noticable effect within the hearing spectrum. You are still going to get better sound quality with a high quality ADC at 44.1k than you are with a mediocre ADC at 96k.[/i]

It's not as questionable as you make it look like. Actually not at all.

Quote:

So this is what is leading you to believe Hi-Rez is better? That's not entirely reasonable, as there may be a number of other factors causing this effect. For example, here's one of several possibilities; when requantising from a 24bit master to 16bit, a noise shaped dither algorithm is usually applied which adds considerable additional noise to the higher frequency range. And, the effects of this added re-distributed sound could be precisely, to the letter, the effects you have described!

No. I know you're trying to defend your «belief system» (

) as well as you can, but that's not the effect I hear. And although my examples have been hybrid SACDs and dual-disc DVD-As -- in the interest of comparability --, the described characteristic isn't reserved to those. It's in fact a CD-inherent characteristic which annoyed me from the start of this format, although at the same time I also appreciated the inherent advantages over the vinyl disc. As to the latter, those later produced (low-rez) «digital recordings» showed similar shortcomings.

Quote:

I'm not "having a go at you" Jazz, it's just that there is so much of this on Head-Fi. An observation is made, a conclusion is drawn and a belief system is initiated. However, with only a partial understanding of the very complex processes involved in the creation of digital audio the conclusion may not be entirely accurate. When an expert comes along to dispel the myth it's already too late, a belief system exists and it's going to take more than logic or proof to change it. The old addage springs to mind: "a little knowledge is dangerous"!

Thanks nonetheless for this friendly innuendo!

I think you're just trying to put my knowledge down in favor of your arguments. I have occupied myself a lot with audio -- certainly longer than you have -- and am quite familiar with digital formats, although not in the very last technical details. And my deductions from my listening experiences are at least as logical and founded as yours -- which possibly aren't as ex-, intensive and founded as mine. You need a high-quality universal player to even begin with valid comparisons. And what's just as important: an open mind -- without any bias in the form of a belief system.
.

 

[gregorio]

Quote:

Originally Posted by JaZZ /img/forum/go_quote.gif No. I know you're trying to defend your «belief system» ( ) as well as you can, but that's not the effect I hear.

What do you mean that's not the effect you hear, I copied and pasted exactly what you said you were hearing. Either you were mistaken then or you are now?

Quote:

Originally Posted by JaZZ /img/forum/go_quote.gif It's in fact a CD-inherent characteristic which annoyed me from the start of this format

Do you know what re-distributed quantisation noise sounds like? If not, how do you know? The inherent characteristic of harshness in the higher frequencies cause by filter ringing in the early years of CDs was largely eliminated in the '90s through the use of oversampling and filtering digitally, this was improved further this century by the use of multi-layer over sampling and many of the artefacts of the early CDs can be either eliminated or reduced to virtually imperceivable levels. Again, this proves there is not a problem with the CD format itself but of how CD recordings were/are made.

Quote:

Originally Posted by JaZZ /img/forum/go_quote.gif You need a high-quality universal player to even begin with valid comparisons. And what's just as important: an open mind -- without any bias in the form of a belief system. .

No you don't. I don't have a universal player, instead, I have the equipment which actually records and mixes both standard and Hi-Rez formats, in fact, I produced my first Hi-rez recording in 1992 and have been using exclusively Hi-Rez audio ever since. I would be quite surprised if you have more experience of Hi-Rez digital audio than me.

Sorry if you feel my belief system inadequate to your high standards, you think maybe I should go back to analogue or standard-res? I've been an advocate of hi-rez digital audio before you probably even heard of it. I understand the uses of Hi-Res audio reasonably well by now and I also know it's strengths and weaknesses, where it is useful and where it is not.

G

 

[JaZZ]

Quote:

Originally Posted by gregorio /img/forum/go_quote.gif What do you mean that's not the effect you hear, I copied and pasted exactly what you said you were hearing. Either you were mistaken then or you are now? Do you know what re-distributed quantisation noise sounds like? If not, how do you know? The inherent characteristic of harshness in the higher frequencies cause by filter ringing in the early years of CDs was largely eliminated in the '90s through the use of oversampling and filtering digitally, this was improved further this century by the use of multi-layer over sampling and many of the artefacts of the early CDs can be either eliminated or reduced to virtually imperceivable levels. Again, this proves there is not a problem with the CD format itself but of how CD recordings were/are made.

See: I wasn't talking of «harshness» -- that's your interpretation. You may want to have a look at my original wording. I know what quantization noise is (although I haven't met it in person, consciously), also the merit of dithering. You were trying to make me think it's dithering which gave me the impression of «digital glare» -- a lame attempt to rebut my deductions. You may not know, but dithering is meant to make the opposite: get rid of digital artifacts.

It's not true at all that oversampling/upsampling has eliminated (nor even substantially reduced) filter ringing since the early days. Both variants (even- or odd-order oversampling) in their most common application just lead to reduced post-ringing, at the price of massive pre-ringing. The sum of the ringing is about the same -- although highly dependent on the filter slope. You only get (more or less) rid of the ringing (--> Gibbs phenomenon) with very smooth filter slopes such as Wadia's spline filter -- with a 3.5-dB drop-off at 20 kHz. That's in the nature of the thing. So digital or analogue filtering doesn't matter in this respect. There have been attempts (by EMM-Labs) to get both at the same time -- linear frequency response as well as proper impulse response -- using filter algorithms by far exceeding common filter designs (actually highly sophisticated DSPs with «intelligent» signal control).


Quote:

No you don't. I don't have a universal player, instead, I have the equipment which actually records and mixes both standard and Hi-Rez formats, in fact, I produced my first Hi-rez recording in 1992 and have been using exclusively Hi-Rez audio ever since. I would be quite surprised if you have more experience of Hi-Rez digital audio than me.

So you have indeed discovered that hi-rez sounds better? Interesting. But there's no way to make you think it could be because of the transient corruption of the low-rez format due to the sharp anti-aliasing filter? Hhmm...

I'm in the comfortable situation to have the opportunity for comparing low-rez to hi-rez any time. So you can indeed trust me.


Quote:

Sorry if you feel my belief system inadequate to your high standards, you think maybe I should go back to analogue or standard-res? I've been an advocate of hi-rez digital audio before you probably even heard of it. I understand the uses of Hi-Res audio reasonably well by now and I also know it's strengths and weaknesses, where it is useful and where it is not.

So we're in a similar situation. Why then is it so difficult to accept a possibility which is clearly within the imaginable and doesn't contradict to established electroacoustics?
.

 

[Pio2001]

Please, we are in the science forum, so let's be scientific.

The scientific investigations about this have shown that high resolution playback improves nothing, with only two exceptions : one of the 110 listeners of Detmold university scored 20/20 in an ABX test (while the 109 other failed, and himself failed with speakers). The possibility that an artifact helped him or her is equalliy probable that the perception of the sound improvement : Help! Sacd Good Or Bad? - Hydrogenaudio Forums
And of course, Oohashi's results, that are so poorly documented (since the listening test was not the goal of the experiment at all) that we don't even know where the statistical significance comes from : Inaudible High-Frequency Sounds Affect Brain Activity: Hypersonic Effect -- Oohashi et al. 83 (6): 3548 -- Journal of Neurophysiology

A recent vast experiment led by Brad Meyer and David Moran supports the hypothesis that CD resolution is completely transparent : AES E-Library: Audibility of a CD-Standard A/DA/A Loop Inserted into High-Resolution Audio Playback by Meyer, E. Brad; Moran, David R.

So if your experience says otherwise, it would be better to make it scientific (or to discuss it outside the scientific section).

First, do you think that the sonic differences are big enough for you to pass an ABX, or any other kind of randomized blind test ?
Second, would you be interested in doing so ?
If the answer is yes to both questions, we could open another topic and discuss how it could be done, given that playback levels and proper noise shaping are two of the main difficulties.

 

[Pio2001]

Message cancelled

 

[JaZZ]

Quote:

Originally Posted by Pio2001 /img/forum/go_quote.gif Please, we are in the science forum, so let's be scientific.

Is this addressed to gregorio (the first pretending to hear differences in favor of hi-rez) or me?

Quote:

So if your experience says otherwise, it would be better to make it scientific (or to discuss it outside the scientific section).

No chance. I'm as «scientific» as I like to be, and I'll certainly post in this thread/forum as long as I enjoy being here.

Quote:

First, do you think that the sonic differences are big enough for you to pass an ABX, or any other kind of randomized blind test? Second, would you be interested in doing so? If the answer is yes to both questions, we could open another topic and discuss how it could be done, given that playback levels and proper noise shaping are two of the main difficulties.

No to both questions. I'm curious about gregorio's response, though.
.

 

[milkweg]

http://theaudiocritic.com/plog/

Proven: Good Old Redbook CD Sounds the Same as the Hi-Rez Formats
Incontrovertible double-blind listening tests prove that the original 16-bit/44.1-kHz CD standard yields exactly the same two-channel sound quality as the SACD and DVD-A technologies.
In the September 2007 issue of the Journal of the Audio Engineering Society (Volume 55, Number 9), two veteran audio journalists who aren’t professional engineers, E. Brad Meyer and David R. Moran, present a breakthrough paper that contradicts all previous inputs by the engineering community. They prove beyond a shadow of a doubt, with literally hundreds of double-blind listening tests at matched levels, conducted over a period of more than a year, that the two-channel analog output of a high-end SACD/DVD-A player undergoes no audible change when passed through a 16-bit/44.1-kHz A/D/A processor. That means there’s no audible difference between the original CD standard (“Red Book”) and 24-bit/192-kHz PCM or 1-bit/2.8442-MHz DSD.

 

[Pio2001]

That was the paper that I talked about.
But remember that a negative is never a proof.

It does not "proves beyond the shadow of a doubt", it just shows that it should be considerent transparent "for all practical purposes", practical purposes meaning here audiophile critical listening.

The tests were blind tests, meaning that in sighted conditions, one can hear differences, but that this listener would probably hear the same differences if the sources were actually the same.

 

[mulveling]

With speakers, I'd say:

45% speakers
25% acoustic environment
15% recording
8% analog source
7% combined amplification electronics (phono stage, preamp, power amp)
0.1% cables

I do have nice IC cables (I like their looks and build quality), but nothing insanely expensive. Sure, a terrible recording will ruin the experience just like a terrible component or a broken cable, but I'm assuming a reasonable level of competence here - after all, we're discussing hifi, not bottom consumer grade.With a digital source, I'd put much less % emphasis on the source.