Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[rhkrhk]

CDs are, of course, 16/44.1 - but I want to save my rips as 24/192 FLACs.
DSP is set to output sample rate at 192,000
Bit depth is set for 24-bit

Despite this, Audio path reports input file as 16/44.1   16 bit 2 channel

with desktop speakers, MC reports output as 192K/24 bit - which is fine

BUT when i use an external DAC, MC engine is not used, and the program is not upsampling - thus I want to save the files as 24/192 for use with the external DAC.

Can this be done?
thanks

[glynor]

CDs are, of course, 16/44.1 - but I want to save my rips as 24/192 FLACs.

I can't really comment on the "can" but I can comment on the "huh"?

Why would you want to do that?

If you want to upsample, for whatever reason, MC's DSP will do that for you on the fly beautifully with a nice HQ sampler.  There's no reason to alter the data on disk in any way and upsample them permanently.  It is not going to be somehow "better" if you do it that way.

[kstuart]

I can't really comment on the "can" but I can comment on the "huh"?

Why would you want to do that?

If you want to upsample, for whatever reason, MC's DSP will do that for you on the fly beautifully with a nice HQ sampler.  There's no reason to alter the data on disk in any way and upsample them permanently.  It is not going to be somehow "better" if you do it that way.

Yes, and you didn't mention the disadvantage of taking up 7 or 8 times as much hard drive space !

[glynor]

Yes, and you didn't mention the disadvantage of taking up 7 or 8 times as much hard drive space !

Right! 

[6233638]

You can't really upsample bit-depth without significantly changing how the file sounds. Most music probably has a dynamic range of less than 30dB (5-bit) and the highest fidelity classical recordings aren't going to have more than about 60dB of dynamic range. (10-bit) So there's really no benefit to 24-bit audio. (it's a different story when you're actually working in audio production)


Similarly, it's generally considered that the absolute upper limit of human hearing is about 21kHz, with most adults being considerably under 20kHz. (often 16-17kHz is the most they can hear, at a reduced level)

So 44.1kHz satisfies nyquist by fully sampling up to 22kHz - above the upper threshold of human hearing - without problems.

[rhkrhk]

Several writers questioned the reason for my wanting to store CDs as 24/192  (I understand these take up more storage space, but storage is pretty cheap).

Maybe I'm imagining it- but I think these sound better than the 16/44- mainly a broader and deeper sound stage.

My cheap little Altec desktop computer speakers seem to sound better in 24/192 than my full system does in 16/44.

When using the networked DAC, the JRiver engine is not used - so I don't get upsampling via the program, and the manufacturer recommends a digital lens/memory buffer/"NativeX" setting which does not upsample.    Seems to me that if I present the DAC with 24/96 or 192 signal I have the benefit of both upsampling and built in digital lens.

Setup is:
computer - Dell workstation, Windows 7, 64 bit,  MC 18.0106
hard wired (cat 7 cable) to router

DAC is a PSAudio PWII with Bridge, hard wired (cat 7) to router

signal -balanced- to Krell KRC preamp via crossovers to multiple amps (Krell, Conrad Johnson) powering Wilson WATT-Puppy-WHOW system.  Most cables and interconnects are Transparent Reference. Most components (other than the PS Audio DAC and the Audioquest Wild, which are new) are mid 1990s.

[InflatableMouse]

Maybe I'm imagining it

 

[kstuart]

Files that originate as 24-bit 192 khz do sound better, because the analog tapes are preserved with the highest possible clarity.

However, if you take those 24-bit 192 khz files and convert them to 16-bit 44.1 khz, you strip out all improved resolution.

Converting them back to 24-bit 192 khz does not magically restore what you have lost by going to 16/44.

Think of it this way:

Suppose you have 100 books in a big bookcase.  You need to ship the books somewhere, but they don't fit in the shipping box, so you throw out 60 of them.   If the 40 books then get shipped back to you, putting them in the big bookcase won't make the missing 60 books magically appear again.

When you convert 24-bits to 16-bits, the other 8-bits of data are thrown out - gone forever.  Same with 192khz to 44.1khz.

So, when you take a 16-bit, 44.1 khz file and change it to 24-bit, 192 khz, the new file is composed 90% of data that was never in the original.   It's a guess by the upsampling software.  If that sounds better, it is just chance.  If the soundstage is bigger, it is a fake bigger soundstage that was never in the original recording.

You can get the same thing with "spatializer software" that will give you a gigantic fake soundstage.

[eddyshere]

Files that originate as 24-bit 192 khz do sound better, because the analog tapes are preserved with the highest possible clarity.

However, if you take those 24-bit 192 khz files and convert them to 16-bit 44.1 khz, you strip out all improved resolution.

Converting them back to 24-bit 192 khz does not magically restore what you have lost by going to 16/44.

Think of it this way:

Suppose you have 100 books in a big bookcase.  You need to ship the books somewhere, but they don't fit in the shipping box, so you throw out 60 of them.   If the 40 books then get shipped back to you, putting them in the big bookcase won't make the missing 60 books magically appear again.

When you convert 24-bits to 16-bits, the other 8-bits of data are thrown out - gone forever.  Same with 192khz to 44.1khz.

Excellent example.
@rhkrhk
It would only matter if you would not rip but record the cd via audio in. Without wanting to go into audiophile voodooish talk most people agree that upsampling enhances the way they hear/perceive the direct! Playback of a file. So to have an impact (even if I doubt you'll find it day and night). You could playback the cd up sampled and record the analogue outs at 192/24. But I really doubt it will be sooooo much better.

Anyway setting aside the why, when and if's : you can achieve what you want with db poweramp converter. You'll use the resample and bit depth function. I'm not sure you can apply it on the fly while ripping but on the files once ripped.

[glynor]

Agreed.  Book analogy == excellent, kstuart.

Filed away for future use.

[Trumpetguy]

My cheap little Altec desktop computer speakers seem to sound better in 24/192 than my full system does in 16/44.

Ok, this may sound somewhat harsh, especially since I have never heard your gear, but No Way. This is your imagination.

[kstuart]

I thought of a way to simplify this.

John Coltrane song "Blue Train" HDtracks 24/192 transfer (directly from the analog tapes):    518 megabytes.

John Coltrane song "Blue Train" 1997 CD transfer at 16/44.1 of same music:    75 megabytes

If you upsample the second one to 24/192, it should be 518 megabytes, but it is 75 megabytes of information from the original master tapes, and 443 megabytes of data that is "made up" or "guessed" by the software.

So, the upsampled version is mostly fictional guesstimates, while the HDtracks 24/192 is entirely data from the original master tapes.

[6233638]

Files that originate as 24-bit 192 khz do sound better, because the analog tapes are preserved with the highest possible clarity.

However, if you take those 24-bit 192 khz files and convert them to 16-bit 44.1 khz, you strip out all improved resolution.

No human can hear audio over 22kHz - most adults cannot even hear 20kHz - their limit of hearing is more like 16-17kHz. Due to Nyquist, it is required to sample at twice that rate. In theory 40kHz would be plenty, but 44.1kHz gives you a nice bit of headroom.

24-bit audio has a dynamic range of 144dB.
16-bit audio has a dynamic range of 96dB - but that can be increased to around 120dB through the use of dithering.

Nothing is anywhere close to having 96dB of dynamic range. 24-bit is superfluous for playback, and you gain nothing from "upsampling" audio from 16-bit to 24-bit. You're literally just padding the data with zeros.


24-bit makes sense in the mastering stage, because it gives you more room to make adjustments to the audio, but it doesn't matter in a mastered track - similar to the reason why most cameras will now shoot 12/14-bit RAW images for editing, but the final output is typically an 8-bit file. You need the extra bits to make adjustments, but when your editing is done, 8-bit is sufficient for most purposes.


And for what it's worth, even if you are wanting to upsample, DACs actually perform worse at 176kHz/192kHz than they do at 88kHz/96kHz, and an argument could be made that if you are upsampling, asynchronous resampling from 44.1kHz to 96kHz is actually better than going from 44.1kHz to 88.2kHz.

[kstuart]

No human can hear audio over 22kHz - most adults cannot even hear 20kHz - their limit of hearing is more like 16-17kHz. Due to Nyquist, it is required to sample at twice that rate. In theory 40kHz would be plenty, but 44.1kHz gives you a nice bit of headroom.

24-bit audio has a dynamic range of 144dB.
16-bit audio has a dynamic range of 96dB - but that can be increased to around 120dB through the use of dithering.

Nothing is anywhere close to having 96dB of dynamic range. 24-bit is superfluous for playback.

I see these sort of posts all over the Internet.

Your post is actually similar to his upsampling idea.

You are taking the CNN or USA Today dumbed down explanation of digital audio, and then using it to debate technical points.  

The bit resolution of digital audio has to do with a lot more than just dynamic range.  If your explanation worked, then 1-bit recorders (DSD,SACD) would have far too little dynamic range.   The reality is that it is not that simple.

And, the sampling frequency affects frequencies below it, otherwise no one would use a sampling frequency higher than 18 khz.

Also, what real life experience has discovered over and over again, is that digital audio theory is nice, but it has to be implemented in physical electronics that never matches the theory...

[6233638]

Your post is actually similar to his upsampling idea.

You are taking the CNN or USA Today dumbed down explanation of digital audio, and then using it to debate technical points.  

No, I am explaining things in a simple manner because it is tedious to go into the technical details of why this is the case, especially trying to argue a point against "audiophiles" that don't understand the technical details behind a lot of their claims, with "golden ears" that can hear things beyond the limits of human perception, outside the range of what we can apparently measure (even though we can measure with far more precision than the human ear is capable of) that end up spending thousands on things that make no difference to their system other than introducing another placebo.

The bit resolution of digital audio has to do with a lot more than just dynamic range.  If your explanation worked, then 1-bit recorders (DSD,SACD) would have far too little dynamic range.   The reality is that it is not that simple.

1-bit devices operate in the megahertz frequency range, in a completely different fashion from PCM audio, and they make extensive use of noise shaping. Wikipedia has a good image to illustrate the differences in how signals are represented in PCM compared with DSD:

And, the sampling frequency affects frequencies below it, otherwise no one would use a sampling frequency higher than 18 khz.

As I said in my post, this is why 44.1kHz is used. Nyquist-Shannon Sampling Theorem says that you need to sample at twice your target rate. So 44.1kHz will handle audio up to 22kHz without any problems, and 22kHz is a lot of headroom above the threshold of human hearing - especially in an adult, as we lose the ability to hear higher frequencies with age.

[glynor]

but it has to be implemented in physical electronics that never matches the theory...

This is WAY OT, but...

Because Quantum Mechanics is both weird and, quite clearly, fundamentally wrong.

Our current understanding of Physics (both micro and macro) are limited by the "laws" we're operating under.  Our best theories, are incredibly, incredibly, incredibly close approximations of the real underlying laws of reality, but there are fundamental things we still don't understand, and...  We're as close to "right" now with Quantum Mechanics as Newton was to "right" with gravity.

It is so close it is tantalizing, but...

The issue is, that our technology is riding so close to the edge of what is "possible", that these gaps in our theory are made evident by reality itself.  That, and mass production and cost pressures (and the relentless race to the bottom) of the Consumer Electronics business world make many things around the edges just-plain-faulty-and-shoddy.  Plus, people are human and engineers are people.

In any case... I'd completely agree that The Debate Has Not Been Settled on some of this stuff.  But, I also have a firm Science-Based Life Outlook.  Show, don't Tell.  And, almost none of this matters anyhow.

I'm also convinced that if you draw a plot evaluating all of the people in the world with "Ability to appreciate music" on the Y axis, and "Amount of money spent on audio equipment" on the X axis, that the resulting plot would look something like this...

I call it the Audiophile Insanity Curve, and it is so choice to be just behind the crest of that wave.  If you have the means, of course.

[Vincent Kars]

You might have a look at SOX: http://sox.sourceforge.net/

Indeed going from 16 to 24 bits is adding 8 zero bytes, not very likely this will improve sound quality.

Upsampling can affect the sound.
Uli Brueggemann did a very nice experiment.
Upsample the original 44.1 to 176 and back to 44.1
http://www.aktives-hoeren.de/viewtopic.php?f=30&t=2242&start=44
Yes, it does have an impact.

Some DACs do change their sonic signature with the sample rate of the source.
This might be because they use a simple DSP chip to up-sample.
If you defeat this DSP by doing the up-sampling on a PC, you might get a better sound quality.
If this is the case, I do think the DAC is below par.

[gappie]

i think those anti pro audiovoodooism discussions are a bit silly. i do hear a difference using 48/24 or 44.1/16 not with random sound file, but with sounds im familiar with. do i care.. no, not at all. im a musicophile. but agreed it is not logical to want what the op wants. and when he really likes that, its easy enough. use something like reaper to upgrade your files to anything you like, dithered or not.

BUT when i use an external DAC, MC engine is not used, and the program is not upsampling - thus I want to save the files as 24/192 for use with the external DAC.

why not. what for dac are you using, when mc can play 24/192 through your dac, it can upsample it also...

 
gab

[Vincent Kars]

Hi Gab

The PS Audio Bridge is a DLNA renderer.
As far as I know, JRiver as a DLNA server can’t upsample > 48

[gappie]

so when a file is 48 it will stay 48, and when its 44.1 it will stay that way when using dnla, no way to upsample.. interesting.. i have no clue about dlna.. guess you have external dlnadacs too.  not sure why.

 
gab

[dean70]

Upsample in DSP studio to 24 bit & 88.2k or 176.4k and call it a day   Most DACs use some form of oversampling internally anyway on 44.1k sample rates. Having a brickwall filter so close to the audible spectrum is less than ideal & it is better to move this out to a higher frequency (with a smoother low pass filter).

On another note, some of the high res downloads contain a lot of noise around the 30-40khz mark (visible on a spectrum analyzer). At least it will keep the mosquitoes away 

[Vincent Kars]

not sure why.

Two devices can only communicate over the network if they use the same protocol.
If we visit this forum, it is http
Likewise for streaming AV it is often DLNA: http://en.wikipedia.org/wiki/Digital_Living_Network_Alliance

 

[gappie]

thank you vincent. interesting read, i start to get it. and look.... JRiver Media Center is also in the list.  

not understanding why 48 stays 48, 192 192 and 44.1 well etc. and the same for the bits. and why it cant be upsampeld before it leaves the door.

[Vincent Kars]

Upsample in DSP studio to 24 bit & 88.2k or 176.4k and call it a day

DSP studio only affects the audio send to a audio device.
It doesn’t work over DLNA

[Vincent Kars]

not understanding why 48 stays 48, 192 192 and 44.1 well etc. and the same for the bits. and why it cant be upsampeld before it leaves the door.

Don’t know much about the internals of DLNA
When DLNA devices discovers each other in the network they exchange their capabilities.
E.g. if the renderer doesn’t support FLAC, the server knows this and might decide to transcode it to a format the renderer does understand e.g. MP3
You will find a couple of conversion options in Media Network but sample rate looks to be capped to 48

[mwillems]

Not to pile on, but there's actually a pretty good argument that higher sampling rates for playback actually introduce distortion due to intermod in the audible range from the ultrasonic content, and the fact that most DACs and amps are far less "well-behaved" above the audio range (i.e. that 30 or 40 KHz noise mentioned above may hurt more than the mosquitos). 

Bluntly put, increasing the amount of ultrasonic content in a recording (in a playback context, as opposed to a recording or audio engineering context) is unlikely to improve sound quality, and is likely to add distortion.

A few articles and a useful thread:
http://people.xiph.org/~xiphmont/demo/neil-young.html
http://www.lavryengineering.com/pdfs/lavry-white-paper-the_optimal_sample_rate_for_quality_audio.pdf
http://www.hydrogenaudio.org/forums/index.php?showtopic=91302&st=0

Obviously these issues are hotly debated, especially in the context of what to do with files that began their life as high sample rate recordings (do you get more distortion from downsampling, or from forcing your amp to try and reproduce inaudible frequencies?).  I can also say that when I've had the opportunity to hear higher sample rate versions of identical recordings they often sound better to me, but I'm not sure how to explain that in any meaningful way.

Regardless, it seems pretty clear that upsampling just for the sake of it is just guaranteed added distortion.  If you can hear a difference, that's what you're hearing.  You may wind up liking the way that distortion sounds, but that's a separate issue entirely.

[kstuart]

mwillems wrote:
" I can also say that when I've had the opportunity to hear higher sample rate versions of identical recordings they often sound better to me, but I'm not sure how to explain that in any meaningful way. "
It is easy to explain it in a meaningful way.  Simply, we don't know the exact explanation yet.  Just as 100 years ago, doctors observed symptoms that they did not yet have an explanation for.

They sound better to you, to me, and to a number of well regarded professional studio mastering engineers, who claim there is a clear difference.  For example, one states about 24/192:
" I guess that is where we differ. I don't have to know anything about the product except how it sounds to me. If it sounds better than what I had previously, that's really all I really care about. "

It is similar to dieting.  There is a massive amount of bull in the media - mostly to keep certain industries afloat.  However, all the professional "dieters" - personal trainers - all know the real truth, because their job depends on helping their clients lose weight - such as professional actors who must lose 30 pounds in order to do a role later that year.   So, all trainers (and bodybuilders) have known for years that it is all about insulin stimulus (and resistance) - just as medical researchers have claimed.

In a similar way, I haven't met any recording engineers who have any doubts about 24/192.  Instead, a lot of people with much less musical training, can't hear a difference.  That's not surprising.

In the 19th Century, a Redwood tree was cut into pieces and re-assembled on the East Coast.   Everyone who saw it was sure it was fake - because they had never seen a tree before even close to that size.

[kstuart]

24-bit audio has a dynamic range of 144dB.
16-bit audio has a dynamic range of 96dB - but that can be increased to around 120dB through the use of dithering.

Nothing is anywhere close to having 96dB of dynamic range. 24-bit is superfluous for playback, and you gain nothing from "upsampling" audio from 16-bit to 24-bit. You're literally just padding the data with zeros.

I explained to you that dynamic range is not the only difference between 24-bits of resolution and 16-bits of resolution, but that seems to have not been absorbed at all.

So, here is a (different) professional mastering engineer who has done hundreds of major label CDs about why that is so:

" The components of a signal that are below full scale are encoded using less than the full word length.
With real music, this means a great deal of what is musically as well as sonically significant information is encoded using less than the full word length.

This means the same signal in a 24-bit system is going to be encoded with 8 bits more than in a 16-bit system.

A signal that is 40 dB down in a 16-bit system gets encoded using 10-bits. In a 24-bit system, it gets encoded using 18-bits. To my ears, 10-bits is pretty cheesy sounding. Lower level signals, encoded with even fewer bits, will sound cheesier still."

[magnust]

If I was the OP I'd feel unwelcome here  





PS:
kstuart: Good replies IMHO

[6233638]

Bluntly put, increasing the amount of ultrasonic content in a recording (in a playback context, as opposed to a recording or audio engineering context) is unlikely to improve sound quality, and is likely to add distortion.

A few articles and a useful thread:
http://people.xiph.org/~xiphmont/demo/neil-young.html
http://www.lavryengineering.com/pdfs/lavry-white-paper-the_optimal_sample_rate_for_quality_audio.pdf
http://www.hydrogenaudio.org/forums/index.php?showtopic=91302&st=0

Thank you for posting these links. I think if anyone here still has doubts, they should explain why there is no reason to go above 16/44 for playback, why 24-bit is used in the studio, and why some people think it may be beneficial to be using higher sampling rates in audio production, but if everything is working correctly, there should be no need to.

I can also say that when I've had the opportunity to hear higher sample rate versions of identical recordings they often sound better to me, but I'm not sure how to explain that in any meaningful way.

If you are aware that it's a "higher fidelity" recording, and you're trying to hear the difference, you will be listening more attentively. And I bet they do things like push the volume up a bit louder with the "high definition" audio if they have the opportunity to, as louder sounds "better" when doing comparisons like that.

That's why you need to do ABX blind testing. I can with complete consistency identify a 320kbps MP3 file in ABX testing compared with higher compression for example, so I know it's a benefit to me. I have not done enough testing to know whether lossless is worthwhile for me compared to 320kbps MP3, but I know that there is a loss there, I have the storage capacity, and I don't want to ever rip all my CDs again, so I feel like I might as well use it just in case I can hear the difference if I ever bothered to do proper testing.

But I know that I can't hear anything over 20kHz, and even outliers cannot hear over 21kHz, so I know there is no point going over 44.1kHz for audio.
Similarly, I know that 96dB is far beyond anything you will ever find in music playback - even with ReplayGain active - so 16-bit audio is plenty. And you can use a dithered 16-bit output to improve that further to at least 120dB.
However, if you have the option to output 24-bit, there's no harm in doing so. I think it's pointless to "upsample" your files to 24-bit, but if you want to use 24-bit output while you are using things like ReplayGain, go ahead.


What I do wonder, is if VideoClock should be operating at a higher internal sample rate, as it doesn't appear to be doing so right now.
I know that with ReClock it's generally recommended to be outputting 96kHz rather than 48kHz if you are resampling. But I trust that the JRiver developers know what they're doing.

I explained to you that dynamic range is not the only difference between 24-bits of resolution and 16-bits of resolution, but that seems to have not been absorbed at all.

No, you confused the fact that DSD operates at 1-bit compared with PCM audio operating at 16-bits. Dynamic range is the only difference that bit-depth makes with PCM audio.

P.S. Good luck trying to find any audio hardware that is even close to having 24-bits of dynamic range, and listening to audio at levels over 100dB without destroying your hearing.

[Hendrik]

Re VideoClock, you're free to resample to 96kHz as well before videoclock does its job, the options are all there. I'm sure the VideoClock features use as much accuracy as required, but if you request 48kHz, thats just what its going to work on.

[Alex B]

If I was the OP I'd feel unwelcome here  

I agree.

... I want to save the files as 24/192 for use with the external DAC.

Can this be done?

Welcome rhkrhk,

The answer is yes, but only by using the external encoder option. For example, SoX can do the necessary conversion on the fly. Before MC had the AIFF encoding option I posted instructions for creating AIFF files with SoX: http://yabb.jriver.com/interact/index.php?topic=67204.msg450739#msg450739. The settings for creating 24-bit / 192 kHz FLAC files are of course a bit different:

Parameters:  %IN -b 24 %OUT rate 192k
Extension:  flac

You can either convert existing files or rip directly to 24/192.

For now, you cannot use the latest Windows SoX build from Sourceforge because FLAC support is not included (http://www.hydrogenaudio.org/forums/index.php?showtopic=93850&view=findpost&p=823773). The previous build, SoX 14.4.0 seems to work fine with MC18 and MC17.


[off-topic]

I also agree with some others who have posted, that it is a bit silly to upsample and especially store the resulting huge files instead of using playback DSP.

I too was about to post a link to Monty's article at http://people.xiph.org/~xiphmont/demo/neil-young.html. It is one of the best explanations I have seen.

[/off-topic]

[JimH]

If I was the OP I'd feel unwelcome here 

I sincerely hope not.  He asked a question and he got a few answers, and a few questions.  In my opinion, the advice has been good and it's been polite.

Audiophiles come in all sizes and shapes, and some of them pass misinformation freely.  We try to be flexible and listen to these ideas, but it is important to get at the truth.

If you want a place that is truly unfriendly to such ideas, try hydrogenaudio.org.  Suggest that [name your format] sounds more [name your audiophile sound term] and you will be barraged by people demanding proof by double blind tests.  I think our forum is a lot more open minded.

[InflatableMouse]

and you will be barraged by people demanding proof by double blind tests.

And even then its not good enough and people will find fault with that and still burn you down.

[glynor]

I sincerely hope not.  He asked a question and he got a few answers, and a few questions.  In my opinion, the advice has been good and it's been polite.

Audiophiles come in all sizes and shapes, and some of them pass misinformation freely.  We try to be flexible and listen to these ideas, but it is important to get at the truth.

If you want a place that is truly unfriendly to such ideas, try hydrogenaudio.org.  Suggest that [name your format] sounds more [name your audiophile sound term] and you will be barraged by people demanding proof by double blind tests.  I think our forum is a lot more open minded.

Well said.

That's precisely how I feel.

[DoubtingThomas]

It's like religion... the non-believers want double-blind testing, and the believers don't trust logic.

[kstuart]

It's like religion... the non-believers want double-blind testing, and the believers don't trust logic.

No need for belief...

" One thing that comes up repeatedly in discussions with other music lovers and audiophiles regarding CD vs. high resolution digital formats, is the fact that most folks have no means of making a valid comparison.  Often, the high resolution version of a record is mastered at a different session, sometimes by a different engineer.  At many so-called "format shootouts", one hears level differences, EQ differences, etc., making a true comparison of the formats impossible.  Astute listeners realize these are comparisons of different masterings and not of the formats themselves.

In an effort to provide some help, Soundkeeper Recordings is providing these samples from the same album, same mastering session, etc.  A number of folks we know, who use upsampling devices to play back their CDs, have found these useful in illuminating the pluses as well as minuses of their upconversion, as they can now compare the results with a true high res version.

When comparing the files, be sure your system is not performing any sort of resampling and/or dithering to either one.  Some will by default, either upsample the 16/44 or downsample and dither the high resolution version. "

For the files go to:

http://www.soundkeeperrecordings.com/format.htm

[tgundo2003]

No, I am explaining things in a simple manner because it is tedious to go into the technical details of why this is the case, especially trying to argue a point against "audiophiles" that don't understand the technical details behind a lot of their claims, with "golden ears" that can hear things beyond the limits of human perception, outside the range of what we can apparently measure (even though we can measure with far more precision than the human ear is capable of) that end up spending thousands on things that make no difference to their system other than introducing another placebo.
..... this is why 44.1kHz is used. .....Nyquist-Shannon Sampling Theorem[/url] says that you need to sample at twice your target rate. So 44.1kHz will handle audio up to 22kHz without any problems, and 22kHz is a lot of headroom above the threshold of human hearing - especially in an adult, as we lose the ability to hear higher frequencies with age.

This thread and related topics have been a project of mine recently, looking at all the information I can find to help me formulate my own opinion further. I understand the science to a degree, and have also done extensive listening on many different types of audio equipment (I work in the industry and have access to lots of fun combinations) and too consistently I hear the benefits of the higher res recordings. I tend to keep and listen to the files in their native formats, but wouldn't chastise someone for trying other methods without listening for myself. I would also not live and die by the science alone.

Example of the science: In our shop we have 2 digital audio players, I will not name them to keep the flames at bay. Both have been reviewed on the numerous sites and magazines and both have been verified as "bit perfect" on their digital outputs. However, when testing them into the exact same DAC and playback system, just literally swapping the cable back and forth between units, playing the same exact file off the same NAS on the same network, one unit, we'll call it "Brand A", clearly sounds better than the other "Brand B". Not just to me, but everyone at the shop and the half of dozen clients I have played the demo for. Why, if they are both bit perfect, is there such an audiable difference? Likely Jitter I guess, but many people out there on the fourms are preaching that "Brand B" sounds as good as possible because it's verified as bit perfect, there could be no difference in how it sounds compared to "Brand B". Closed Minds at work...

Anyways, back to the point of my comment and why I quoted the above. I recently found this gem (http://www.youtube.com/watch?v=1CkyrDIGzOE) from the RMAF by a Speaker from ESS. It's an hour long, but there's some really good stuff in there that might open your mind a bit (If you are willing to open it). I love his lines about they "finally figured out the problem when they decided to listen to the audiophiles", and also that they (Arguably one of the premiere companies developing high performance audio devices) themselves acknowledge that even with the best instrumentation for measuring audio they cannot measure everything we have been shown to "hear".

Maybe you have all seen this video, or perhaps were there, But I thought it was appropriate to re post in this thread.

[magnust]

I sincerely hope not.  He asked a question and he got a few answers, and a few questions.  In my opinion, the advice has been good and it's been polite.

Audiophiles come in all sizes and shapes, and some of them pass misinformation freely.  We try to be flexible and listen to these ideas, but it is important to get at the truth.

If you want a place that is truly unfriendly to such ideas, try hydrogenaudio.org.  Suggest that [name your format] sounds more [name your audiophile sound term] and you will be barraged by people demanding proof by double blind tests.  I think our forum is a lot more open minded.

I totally agree that this is a very friendly place     And I know it can get really bad on many other sites.

I think the best position is that to try hard to keep an open mind. Listen to different opinions and arguments about pros and cons. To be healthily skeptical but rarely dismiss anything in a fully definite way. "That is not possible AFAIK but who knows, we just might be missing something here." And even after deciding to believe one way or another to be prepared to maybe change ones mind.

[bulldogger]

I see these sort of posts all over the Internet.

Your post is actually similar to his upsampling idea.

You are taking the CNN or USA Today dumbed down explanation of digital audio, and then using it to debate technical points.  

The bit resolution of digital audio has to do with a lot more than just dynamic range.  If your explanation worked, then 1-bit recorders (DSD,SACD) would have far too little dynamic range.   The reality is that it is not that simple.

And, the sampling frequency affects frequencies below it, otherwise no one would use a sampling frequency higher than 18 khz.

Also, what real life experience has discovered over and over again, is that digital audio theory is nice, but it has to be implemented in physical electronics that never matches the theory...

Nice post.

[BillT]

I recently found this gem (http://www.youtube.com/watch?v=1CkyrDIGzOE) from the RMAF by a Speaker from ESS. It's an hour long, but there's some really good stuff in there that might open your mind a bit (If you are willing to open it). I love his lines about they "finally figured out the problem when they decided to listen to the audiophiles", and also that they (Arguably one of the premiere companies developing high performance audio devices) themselves acknowledge that even with the best instrumentation for measuring audio they cannot measure everything we have been shown to "hear".

Thanks for that link, it's very interesting and explains a few of the defects of delta-sigma converters.

It reminds me of my BBC training many years ago - and of the days when the BBC made scientific documentaries that weren't aimed at people with the attention span of a gnat. I'll have to try his other lecture.

[pschelbert]

Hi

its clear, keep 16bit,44.1kHz CD as is.
There is anyway big doubt if anything more (24 bit, 96kHz, 192kHz, 384kHz..) would give any better sound...

see this:
http://people.xiph.org/~xiphmont/demo/neil-young.html

there is a nice ppt from D. Griesinger in above link. He is a professional, if you know Lexicon, it says all.
Its even possible to get a lot worse sound using more than 44.1kHz sampeling due to intermodulation is his conclusion.

Another link to an AES article (from Boston Audio) on above website: conclusion. More than 500 people could not tell the difference of 16bit/44.1kHz and SACD or DVDA audio quality (and these have been a lot of professional audio people!!)

Peter

[Hipper]

Five years ago when I was 54 I tested my hearing using test tones. I found I could only just hear 10kHz and could not hear 12.5kHz although it gave me a pain in the ear. Above 12.5kHz I heard nothing.

Thought I'd mention it!

And, as a thought, how then can some of the older speaker manufacturers still be satisfied with their full range speakers when they can't hear substantial portions of the frequency range!

[gappie]

Five years ago when I was 54 I tested my hearing using test tones. I found I could only just hear 10kHz and could not hear 12.5kHz although it gave me a pain in the ear. Above 12.5kHz I heard nothing.

Thought I'd mention it!

And, as a thought, how then can some of the older speaker manufacturers still be satisfied with their full range speakers when they can't hear substantial portions of the frequency range!

i guess they use their grandchildren for tests.

[kstuart]

Hi

its clear, keep 16bit,44.1kHz CD as is.
There is anyway big doubt if anything more (24 bit, 96kHz, 192kHz, 384kHz..) would give any better sound...

see this:
http://people.xiph.org/~xiphmont/demo/neil-young.html

This article is very convincingly written, but is junk science.   There are many sentences that are "assertions" - there is no substantiation, just something stated so positively that as a reader, you accept it.

Here is a recent comment by an audio professional about that article:

" I frequently receive e-mails from people who have read some anti-audiophile propaganda on the Internet that asserts 44.1kHz/16-bit audio is perfect, or that all amplifiers or cables sound the same, or other such nonsense. An excerpt from the article you mention: "Unfortunately, there is no point to distributing music in 24-bit/192kHz format. Its playback fidelity is slightly inferior to 16/44.1 or 16/48, and it takes up 6 times the space." The article appears on the surface to be technically credible but the author is naïve at best.

The resolution of a digital audio system isn't determined by the number of bits available in each sample, but rather by the number of bits being used at any given moment. For example, during very low-level passages the signal might be encoded at -80dBFS (80dB below full scale). With 16-bit quantization, a signal at -80dBFS would be encoded with three bits (eight possible quantization steps). With 24-bit quantization, that same -80dB signal would be encoded with 11 bits (2048 possible quantization steps).

The article also claims that high sampling rates offer no sonic or technical advantage over 44.1kHz because in theory 44.1kHz sampling of a 20kHz bandwidth signal is perfectly lossless. The latter part of that statement is true, but it fails to address the well-documented problem of the time-domain distortion introduced by steep filters with cutoff frequencies close to the audio band. Faster sampling rates allow filters with gentler slopes that have cutoff frequencies well away from the audio band, and thus do less damage to the signal.

Finally, anyone who is actually listened to 44.1kHz/16-bit and 176.4kHz/24-bit versions of the same recording can come to no other conclusion than hi-res digital sounds significantly better."

[kstuart]

magnust wrote:
" And even after deciding to believe one way or another to be prepared to maybe change ones mind. "

The believing one way or another IS the source of the problem.

I find that 90% of statements made by both sides are wrong - and this applies to all arguments on all subjects.

People discover that one side is correct on one little issue, and then they decide that that side is right all the time.  That's where the problems come from.

This often comes from making opinions part of one's definition of "who I am" - an unfortunate tendency.

[pschelbert]

the interesting thing about the article is, that the article references a lot of professional peoples work, who are by far no beginners, but people who work in pro audio.
For example D. Griesinger has done a lot of great work. I actually own some of Lexicons digital processors he designed. And yes they are good.

I also tried myself to see if I can hear a difference. There are some sources:
www.highresaudio.de
www.hd-klassik.de
www.qobuz.com

These websites offer testfiles, same music at different resolutions and technloogies (SACD, CD, Flac, WAV, WMA etc.). You may be surprised...

No difference from CD quality up.

I tested with streaming.
Audiointerface: RME Fireface UFX, Headphones: Stax404 with SRM-717. With speakers anyway forget it..

What makes a big difference is the recording itself. Looks that a lot of recording engineers are just damaging the quality..
Muddy sound, distorted etc. etc.

Peter

[flac.rules]

This article is very convincingly written, but is junk science.   There are many sentences that are "assertions" - there is no substantiation, just something stated so positively that as a reader, you accept it.

Here is a recent comment by an audio professional about that article:

" I frequently receive e-mails from people who have read some anti-audiophile propaganda on the Internet that asserts 44.1kHz/16-bit audio is perfect, or that all amplifiers or cables sound the same, or other such nonsense. An excerpt from the article you mention: "Unfortunately, there is no point to distributing music in 24-bit/192kHz format. Its playback fidelity is slightly inferior to 16/44.1 or 16/48, and it takes up 6 times the space." The article appears on the surface to be technically credible but the author is naïve at best.

The resolution of a digital audio system isn't determined by the number of bits available in each sample, but rather by the number of bits being used at any given moment. For example, during very low-level passages the signal might be encoded at -80dBFS (80dB below full scale). With 16-bit quantization, a signal at -80dBFS would be encoded with three bits (eight possible quantization steps). With 24-bit quantization, that same -80dB signal would be encoded with 11 bits (2048 possible quantization steps).

The article also claims that high sampling rates offer no sonic or technical advantage over 44.1kHz because in theory 44.1kHz sampling of a 20kHz bandwidth signal is perfectly lossless. The latter part of that statement is true, but it fails to address the well-documented problem of the time-domain distortion introduced by steep filters with cutoff frequencies close to the audio band. Faster sampling rates allow filters with gentler slopes that have cutoff frequencies well away from the audio band, and thus do less damage to the signal.

Finally, anyone who is actually listened to 44.1kHz/16-bit and 176.4kHz/24-bit versions of the same recording can come to no other conclusion than hi-res digital sounds significantly better."

Well, regarding sampling rate and bitdepth, professionals have donw listening tests on this, and published papers. They hardly seem to support any detectable difference in general.

And you example with "8 steps" is a bit weird, have you heard 3 bit-encoded music? It is sampled thousands of times a decond, so you don't hear the music jump between three levels, you just hear a nuts-ton of noise (if you dither it), just as predicted by the theories. The bitrate is changing the S/N.

And filters aren't really much of a problem with todays technology and a somewhat competent design.

[gappie]

i really do like your posts on this matter mister KStuart. though i always think im in some theological dispute when i read these threads.

this is sad
"anti-audiophile propaganda"
or
"audiophile propaganda"
there is logic, things we know, things we think we know, things we have no clue about. and things where later it will turn out we were so wrong..
so lets see.

enjoy the music
 
gab

[pschelbert]

an interesting thing is to look at the spectrum (spectrum analyzer) of same music at different sample and bit rates.
There looks like mirroring effects of noise around 20kHz. Really strange. Content just above 20kHz is at very low levels , below 16bits. Levels increase then up to 96kHz... whatever that ultrasound may be...?

As an above article mentioned, its like Tube-Amp sound. You may like it because of the distortion or in case of digital additionally sampling effects, but the quality actually goes belly up...

Peter