Vinyl and seperates

[mtrycrafts]

you may want to

When did vinyl have 3 channels on it? or 2, for that matter? Discrete?
It was not amps or pre that had the limitations as you can line up any number in a row, but software.

A 3 chan analog recording would reproduce music better than an encoded software enhanced recording.


Actually, what is needed is three of those speakers with the two ch and 3 ch ability to be switched and behing acoustic curtins so you are not biased, then we can see which setup is more accurate, more realistic, etc.
Your proposal with different speakers will tell not much other than speaker differences. But, even that can be tested though, DBT, of course.

2 ch is history. It cannot capture nor reproduce what a multi channel can.
But, if you like the 2 ch, no problem.

You have 2 ears, in a live performance the performer is playing in front of you, everything else you hear is caused by the enviorment surrounding you. Movies, i agree surround and multi chan is the way to go, but to reproduce music it dosnt work.

read this about vinyl, long threads:

http://groups-beta.google.com/group/rec.audio.high-end/browse_frm/thread/b5b3d74ebf5195d9/653de88007541118?dq=&hl=en&lr=&ie=UTF-8&prev=/groups?hl=en&group=rec.audio.high-end#653de88007541118


And this


http://groups-beta.google.com/group/rec.audio.high-end/browse_frm/thread/759d781c924238c7/2fa1f85a2a7c985b?dq=&hl=en&lr=&ie=UTF-8&prev=/groups?hl=en&group=rec.audio.high-end#2fa1f85a2a7c985b

 

[mtrycrafts]

Do you mean the waveform created after that stylus goes around the vinyl for the very first time? Or do you mean the new waveform created each time the friction of the stylus changes the surface of the vinyl ever so slightly…and then overtime…what have you then? Must be perfection!


I agree the music is corrupted every time the album is played. In that aspect digital is truly superior. If music was digitized at an adequate rate with GOOD digitizers, it would be a different story. Digitizer technology has grown by leaps and bounds since the 80's (when it was implemented in home audio) lets get with the program.

Music is digitaized at an adequate rate as it is, better than vinyl is.

Read the whole thread

http://groups-beta.google.com/group/rec.audio.high-end/browse_frm/thread/b5b3d74ebf5195d9/653de88007541118?dq=&hl=en&lr=&ie=UTF-8&prev=/groups?hl=en&group=rec.audio.high-end#653de88007541118


And this

http://groups-beta.google.com/group/rec.audio.high-end/browse_frm/thread/759d781c924238c7/2fa1f85a2a7c985b?dq=&hl=en&lr=&ie=UTF-8&prev=/groups?hl=en&group=rec.audio.high-end#2fa1f85a2a7c985b

 

[Buckeyefan 1]

Music is analog.

But the signal to your brain is not. It is in packets of digital signals. However, so is the music from a CD at the speaker terminals. And, very accurate to the original master recorded, a far cry for vinyls.

I thought the cochlea reproduced sound to the brain electromagnetically through nerves. 0's and 1's ??? Hmmmm. I missed that class.

If you get really bored tonight...

www.beyonddiscovery.org/content/view.page.asp?I=257

 

[Rotarhead69]

Well, im not going to sound brilliant or anything, but I'll put my 2 cents in:

Ive heard a/b comparisions of the same recordings on vinyl and cd back to back and the vinyl always seem more involving - always!

Now mind you i love my surround sound and SACD's, but there is somthing about 2ch analog that comes off more pure.

Lets put aside all worries of extracting every ounce of info from some particular source, nice holy grail, but not all together the real issue in my opinion. Think about it, we are discussing what sounds better an analog or a digital source, but of what? I LIVE event! I've only heard 1 or 2 systems in my whole life that can come close to that (one an analog one a digital)! So extracting the last bit of detail is only part of the equation.

I cant explain why the vinyl sounds better, but after all the years and CD's I've collected, I've begun to experiment with vinyl myself. I've found an old Pioneer Turntable, go it fixed up with ebay and 50 bucks and found some albums at GoodWill for $6, and you know what, It sound damn good concidering. My Vandersteens don't seem to have a problem with the vinyl.

I my uneducated guess is that digital is all interpolated. Time is counted by a digital clock. The computer has to say "this note first, this note second and so on. . . " So time and space are more removed for the actual analog wave form it came from. Analog, while not as exacting as its digital brethren has a more direct connection to the music.

I guess what Im saying is, digital has to figure out the music between the 0's and 1's where as analog doesn't, its an infinant source of information.

As long as you enjoy it - who cares.

 

[Dan]

I think that things may often be a bit simpler than dissecting the pros and cons of digital encoding and decoding processes. Many good quality vinyl recordings were made very carefully with ideal settings and talented engineers and performers. Those days are long gone. Many recordings including some classical ones are recorded nowadays with such a limited dynamic range that they sound horrible. This issue has been discussed extensively on this site elsewhere. It is a marketing decision. Further, the decline in the talent level especially in popular music has led to extensive multitracking and processing to cover performers lack of ability. I first noticed it with Madonna 20years ago and now it is ubiquitous. I don't care if you have two or twenty speakers just try and tell me where these pop singers are standing on the stage when there are five layers of multitrack!

Further much postprocessing is now considered standard. Unfortunately almost all of it degrades the sound. For a contrast, try a Chessky or Mapleshade recording that has no postprocessing except A to D conversion. The soundstage is superb as well as the dynamic range and the frequncy range is 20 to well beyond 20K.

As for vinyl vs. digital, I think the above recordings beat any vinyl I have ever heard. Some of the best vinyl I have heard (especially Henry Kaiser's Metallanguage label) is excellent and better than most CD material I have heard. However I have a lot of vinyl that is unlistenable on my new system because it exposes all the flaws of the recording techniques of the early sixties rock music. They were intended for AM car radios not $15K stereos. Think early Beatles and all sixties Motown.

I guess what I'm saying is that the CONVERSION of sound onto the medium be it digital or analog is FAR MORE IMPORTANT than what that medium is. Don't underestimate the importance of the recording engineer. Therefore I am in no hurry to replace my vinyl with CDs or even SACD. I have tried it and there is no significant difference except in ease of use.

As far as two vs 5 channel I favor two on my setup. I think it depends on the speakers imaging ablility, their placement vis avis the seating and the room dynamics which are not totally reflected in room eq. My speakers image superbly (Vandersteen 3a Sig) so that I don't need a center channel even for movies. I have a decent but not great room and the speakers are decently but not optimally sited due to practical contraints. I have never heard any other speakers image as well. Therefore when some prefer 5 channel over two, that may be quite valid on their system. Their center channel may be better than mine (Vandersteen VCC-1 sig).

 

[MacManNM]

Why digital is harsh and bad

No need fror further discussions before you look into this further since 16 bits give a 96dB dynamic range, a number of orders above vinyl, not to mention all the other junk problkems with vinyl, like distortion, frequency response issues, and on.

I apologize, my math was in fact wrong. 16 bits of dynamic range is 48 db (65536 counts, log) . But since we are talking voltage you double it 96 db. However, the effective bit rate can't be 16, its simply not possible. Lets be optimistic and assume it really is 14 effective bits (16384 log) gives you 42 db, or 84 for audio. So bottom line we are looking at 84 db of true dynamic range. A good turntable and Vinyl recordings have been measured to have 80 db of dynamic range. Average ambient noise level in a quiet room is about 18db. Taking that into account, the dynamic range is not an issue.

Lets now get into the important part, sampling. A standard cd is sampled at 44.1 KHz. At the limit of human hearing 20KHz, you would have 2.205 data points. Can you make a sine wave with 2.205 data points? Even at 10 KHz, thats 4.41 data points, better but still poor. This is the reason digital music sounds harsh. As you may know the sharper the rise time the more frequency content in the waveform. With this under sampling of music there is MORE harmonic distortion in CD's and DVD's. Most are rated at 1KHz, there you have 44.1 data points, a very reasonable amount of data to reconstruct a signal, hence low THD ratings. On paper they look far superior, but in reality they are inferior. So, here we are, back to the start. Digital needs to fill in all of those gaps, all of the data that is not there has to be "made up". Thats what your processors and sound fields do, fill in the empty data gaps. That is the basic science of it in a nutshell. I would be happy to explain this further if you wish.

 

[zumbo]

That's it! I'm busting out my dingaling. On 8-track. By Chuck Berry. It's gonna jam.

 

[Frzdrdhppy]

That's what it's all about! That warm fuzzy feeling...lifted to another place and time. How each of us gets there is different, but as long as we're there that's all that really matters.

Right on! And then ya might want to grab some munchies and put on some music!

---Frz

 

[WmAx]

However, the effective bit rate can't be 16, its simply not possible. Lets be optimistic and assume it really is 14 effective bits (16384 log) gives you 42 db, or 84 for audio.

Plenty of modern devices achieve 94-96dB range in 16 bit mode today. 3rd party measurements abound, need not look very long.

A good turntable and Vinyl recordings have been measured to have 80 db of dynamic range. Average ambient noise level in a quiet room is about 18db. Taking that into account, the dynamic range is not an issue.

A very good turntable, with extremely clean record, that in turn has a very good pressing and with very little wear--this is not a realistic scenario except in the most unusual circumstances.

Can you make a sine wave with 2.205 data points? Even at 10 KHz, thats 4.41 data points, better but still poor. This is the reason digital music sounds harsh.

While it may seem that way at first glance, all that is needed according to Nyquist theory to perfectly reproduce a signal is rate of twice the signal rate to be stored. The anti-alias filter of the CD player, after the D-A convertor, is what converts the 'stepped' waveform into a perfect sine wave. The time domain of the wave is correct, also, though this seems impossible at first thought. But the time and frequency and amplitude domain are restored. The amplitude points vs. sample points allow for this to be [1]accomplished.

As for the assertion of sounding 'harsh' -- according to whom? A credible perceptual study or audiophile heresay?

With this under sampling of music there is MORE harmonic distortion in CD's and DVD's.

Please provide credible data to prove this claim. If you are interested in harmonic distortion levels, why not discuss how phonograph reproductions end up with THD into the single digits in the best case [2]scenarios. Why not factually correlate what level of THD is in fact [2]audible within music program according to proper studies.

On paper they look far superior, but in reality they are inferior.

Do you care to support these assertions with facts?

-Chris

[1]
Sampling Theory For Digital Audio
Lavry, Dan
http://lavryengineering.com/documents/Sampling_Theory.pdf


[2]
Just Detectable Distortion Levels
James Moire, F.I.E.E.
Wireless World, Feb. 1981, Pages 32-34 and 38

 

[MacManNM]

Plenty of modern devices achieve 94-96dB range in 16 bit mode today. 3rd party measurements abound, need not look very long.

I would love for you to show me one 16 bit digitizer that has 16 effective bits.


A very good turntable, with extremely clean record, that in turn has a very good pressing and with very little wear--this is not a realistic scenario except in the most unusual circumstances.

So, it is possible.

While it may seem that way at first glance, all that is needed according to Nyquist theory to perfectly reproduce a signal is rate of twice the signal rate to be stored. The anti-alias filter of the CD player, after the D-A convertor, is what converts the 'stepped' waveform into a perfect sine wave. The time domain of the wave is correct, also, though this seems impossible at first thought. But the time domain is restored because the points of data sampling are relative when compared to the amplitude values, providing for complete restoration of the waveform.

Audio signals do have fast, large transients. All significant phase shifts in a waveform are not perfectly reproduced. IE, @ 18KHz, if the content has a 180 deg phase shift, that effectively doubles the frequency, this causes a duplicate "reflection" on the other side of nyquist. The anti-alias filter does not have the ability to remove all of these transients without significantly deteriorating the signal at and around nyqiust. Therefore the sample rate is not adequate. The anti-alias filter of the CD player tries remove these transitions, it sees them as errors because the frequency is out of band, when in actuality it is actually part of the music. How can a filter smooth out the steps? All it is doing is integrating the waveform. IF you take a complex waveform (that has freq content higher than nyquist) and digitize it 10 times, depending on the start time of the digitization you will have 10 different answers.

As for the assertion of sounding 'harsh' -- according to whom? A credible perceptual study or audiophile heresay?

I believe people have been saying that since digital music came out. Just like the Tube VS Transistor argument. Tubes are slower, they dont reproduce everything as well as transistors do. Hence you get an integration of the signal. These digitizer errors, because of the sample rate, cause other fast sharp transients in the waveform that can be heard, some call it harsh

Please provide credible data to prove this claim. If you are interested in harmonic distortion levels, why not discuss how phonograph reproductions end up with THD into the single digits in the best case [1]scenarios. Why not factually correlate what level of THD is in fact [1]audible within music program according to proper studies.

Again a 1KHz sinewave. Put a complex signal in near and above nyquist, you will see tons of distortion. I would love to make the measurements, unfortunately this is not something you can perform with a cw source, setting up to take the measurements properly would take a great deal of time and effort. Any help would be appreciated. I have all of the equipment necessary.

Do you care to support these assertions with facts?

 

[WmAx]

I would love for you to show me one 16 bit digitizer that has 16 effective bits.

A common Creative Audigy 2 series PCI soundcard, achieves 94-95dB SNR with no problem in 16 bit recording and playback modes. Search for reviews and 3rd party measurments. This is not exactly a cutting edge product, either.

Audio signals do have fast, large transients. All significant phase shifts in a waveform are not perfectly reproduced. IE, @ 18KHz, if the content has a 180 deg phase shift, that effectively doubles the frequency, this causes a duplicate "reflection" on the other side of nyquist.

And when such transient signal has data that is beyond what is relevant, audibly, the ultrasonic data has no use and may be discarded.

As for phase response, the audibilty of high frequency phase shift as would be induced by a low-pass filter required for RBCD was considered and tested in both the established JAES standard [1] and by other parties [2] and found to be unimportant for purposes of audibility. Your issues with time domain signal distortion are also covered in the Dan Lavry paper to which I referred in the last post.

The anti-alias filter does not have the ability to remove all of these transients without significantly deteriorating the signal at and around nyqiust. Therefore the sample rate is not adequate.

The phase distortion in this band is unimportant. [1] [2]

The anti-alias filter of the CD player tries remove these transitions, it sees them as errors because the frequency is out of band, when in actuality it is actually part of the music. How can a filter smooth out the steps?

The 'steps' are ultransonic content, that is not a product of any recorded music, but a byproduct of the digital to analog conversion. All that is required to remove such is a sufficient low-pass filter. Refer to the Dan Lavry paper, if this is not clear.

All it is doing is integrating the waveform. IF you take a complex waveform (that has freq content higher than nyquist) and digitize it 10 times, depending on the start time of the digitization you will have 10 different answers.

How so? Any proper audio digitazation system has anti-alias filter before sampling the analog data. The system will only record the frequencies within the band supported by the sample rate as a consequence, thus avoiding the aliasing artifacts that you imply will exist.

I believe people have been saying that since digital music came out. Just like the Tube VS Transistor argument. Tubes are slower, they dont reproduce everything as well as transistors do. Hence you get an integration of the signal. These digitizer errors, because of the sample rate, cause other fast sharp transients in the waveform that can be heard, some call it harsh

People have also been saying that ghosts exist, or alien abductions. Why should I(or you) believe them in lack of substantial evidence? Where is the proof of audible artifacts you claim exist with any digital reproduction system?

Again a 1KHz sinewave. Put a complex signal in near and above nyquist, you will see tons of distortion.

Sure, if the device does not have the required anti-alliasing filters, it will produce all sorts of spurious high frequency artifacts. But this is not how the majority of equipment is produced.

I would love to make the measurements, unfortunately this is not something you can perform with a cw source, setting up to take the measurements properly would take a great deal of time and effort. Any help would be appreciated. I have all of the equipment necessary.

Since this is supposedly a simple matter, then just refer to a credible study of another demonstrating these artifacts to be an issue, and thus audibly effecting a properly operating digital reproduction chain as confirmed with coorelated perceptual testing.

-Chris


[1]
Which Bandwidth Is Necessary for Optimal Sound Transmission?
G. PLENGE, H. JAKUBOWSKI, AND P. SCHONE
JAES, Volume 28 Number 3 pp. 114-119; March 1980

[2]
Perception of Phase Distortion in Anti-Alias Filters
Preis, D.; Bloom, P. J.
AES Preprint: 2008

 

[MacManNM]

A common Creative Audigy 2 series PCI soundcard, achieves 94-95dB SNR with no problem in 16 bit recording and playback modes. Search for reviews and 3rd party measurments. This is not exactly a cutting edge product, either.

First off, if its 95-94db s/n, its not 16 effective bits. 2nd, I doubt those are true measurements, I work with digitizers every day, advertised s/n is never on the mark. A National Instruments NI6115 advertises 16 bit performance, yet the advertised s/n is around 85db, actual s/n that I have measured is more like 80db

And when such transient signal has data that is beyond what is relevant, audibly, the ultrasonic data has no use and may be discarded.

Is this discarded before digitizing?

As for phase response, the audibilty of high frequency phase shift as would be induced by a low-pass filter required for RBCD was considered and tested in both the established JAES standard [1] and by other parties [2] and found to be unimportant for purposes of audibility. Your issues with time domain signal distortion are also covered in the Dan Lavry paper to which I referred in the last post.

You are good at quoting papers, but it's quite obvious you don’t understand how the Nyquist theory works, or for that matter how a digitizer functions. There is no analog filter that has a steep enough cutoff to block everything below nyquist freq, and maintain the integrity of the upper band of the audio signal (when sampled so close to intended freq), therefore the analog upper band of the analog spectrum suffers. This is why the digitizer rate must be increased. If the filter is not sufficient, then the digitizer will see content above nyquist. The latter is what happens. When the digitizer sees these out of band frequencies, it reproduces factions of them in its nyquist band. Common to all digitizers.



The phase distortion in this band is unimportant. [1] [2]


The 'steps' are ultransonic content, that is not a product of any recorded music, but a byproduct of the digital to analog conversion. All that is required to remove such is a sufficient low-pass filter. Refer to the Dan Lavry paper, if this is not clear.

See above. Ive looked at the output of several cd players, you can see the steps on a decent DSO, great filters eh?

How so? Any proper audio digitazation system has anti-alias filter before sampling the analog data. The system will only record the frequencies within the band supported by the sample rate as a consequence, thus avoiding the aliasing artifacts that you imply will exist.

That do exist.

People have also been saying that ghosts exist, or alien abductions. Why should I(or you) believe them in lack of substantial evidence? Where is the proof of audible artifacts you claim exist with any digital reproduction system?
please!

Sure, if the device does not have the required anti-alliasing filters, it will produce all sorts of spurious high frequency artifacts. But this is not how the majority of equipment is produced.

I'm glad to see you agree with me, now lets see some data that shows the cutoff of these filters, they must be 100db/octave! If they aren't perfect, then, there is either more, or less, content than there is supposed to be. So I guess every decent cd/dvd player has a low pass filter that cuts off at 20KHz exactly, and has 4 or 5 orders of magnitude in attenuation!



Since this is supposedly a simple matter, then just refer to a credible study of another demonstrating these artifacts to be an issue, and thus audibly effecting a properly operating digital reproduction chain as confirmed with coorelated perceptual testing.


You sure put a lot of responsibility on one filter that is supposed to save your system. But im sure that every one that comes off the line in china is exactly perfect. Must have some really good quality control over there.

 

[gene]

This is an old argument usually fought by people into esoteric tube amps and compression horn loaded speakers. While there is some validity to the arguments, properly executed redbook is nothing short of stellar and has unfortunately gotten a bad rap b/c of poor quality recordings, or poor room acoustics and/or loudspeaker integration allowing for a more revealing system to sound "bright".

I encourage checkout out our articles on Audio Formats and Compression.

The bottom line is CD has the ability to be both worse and better than LP equivalents. Usually the latter is a result of too much compression and/or poor xfering at the studio.

 

[WmAx]

First off, if its 95-94db s/n, its not 16 effective bits. 2nd, I doubt those are true measurements,

Yes, while not true 16 bit, it is 1-2dB shy. To make a point of this is being trivial.

Those are not factory specified -- they are 3rd party review measurements from several sources -- available with a quick google search if you were interested.

You are good at quoting papers, but it's quite obvious you don’t understand how the Nyquist theory works, or for that matter how a digitizer functions. There is no analog filter that has a steep enough cutoff to block everything below nyquist freq, and maintain the integrity of the upper band of the audio signal (when sampled so close to intended freq), therefore the analog upper band of the analog spectrum suffers. This is why the digitizer rate must be increased. If the filter is not sufficient, then the digitizer will see content above nyquist. The latter is what happens. When the digitizer sees these out of band frequencies, it reproduces factions of them in its nyquist band. Common to all digitizers.

I never claimed that absolutely every bit of aliasing was relieved. But it is of such small magnitude after proper filtering, that it is not shown to be of any consequence audibly in properly designed systems. Why don't you try to coorelate with audiblity, rather than just say there is an effect? I can drop a pebble on a concrete floor, and it will vibrate the concrete 50 feet away, but will you feel it vibrate the floor 50 feet away?

See above. Ive looked at the output of several cd players, you can see the steps on a decent DSO, great filters eh?

At what magnitude? The wave form should be essentially perfect looking, unless of course you are talking about artifacts at very low levels/magnitudes(relevance to audibility?) or are purposely testing an output that has extremely poor performance due to poor design(Audio Note comes to mind) or defect.

That do exist.

Perhaps. But I would consider such to be poorly designed or defective.

I'm glad to see you agree with me, now lets see some data that shows the cutoff of these filters, they must be 100db/octave! If they aren't perfect, then, there is either more, or less, content than there is supposed to be. So I guess every decent cd/dvd player has a low pass filter that cuts off at 20KHz exactly, and has 4 or 5 orders of magnitude in attenuation!

When did I claim 'technically' perfect filters? I am careful to qualify my statements in the context of audibility.

The presence of a distortion is not sufficient to establish audibility. You must coorelate it with audibility, through reference to proper perceptual studies, or by performing and publishing such research yourself.

-Chris

 

[MacManNM]

Audibility, which you refer to so much, is a subjective term.

Yes, while not true 16 bit, it is 1-2dB shy. To make a point of this is being trivial.

not true, i looked at the data from several sources that claimed 94.1 db dynamic range (measured), when you look at the data, the interpretation of the data is flawed. The measurement should be made using RMS noise, and max amplitude of the test signal, according to my calculations it was more like 75 db. Also the card cuts off sharply at 18KHz. That is more reasonable for a 44.1KHz sample rate. So, it is in fact not trivial.


I never claimed that absolutely every bit of aliasing was relieved. But it is of such small magnitude after proper filtering, that it is not shown to be of any consequence audibly in properly designed systems. Why don't you try to coorelate with audiblity, rather than just say there is an effect? I can drop a pebble on a concrete floor, and it will vibrate the concrete 50 feet away, but will you feel it vibrate the floor 50 feet away?

It depends on the pebble, the room resonance, the concrete. I've worked with seismometers that would go crazy from that. You can't define audibility by some "experts" terms, obviously there is an effect that some people can hear, or this wouldnt be a big controversy in the audio world.



At what magnitude? The wave form should be essentially perfect looking, unless of course you are talking about artifacts at very low levels/magnitudes(relevance to audibility?) or are purposely testing an output that has extremely poor performance due to poor design(Audio Note comes to mind) or defect.


Usually bit noise is the biggest thing here, on playback usually you will have 1-2 bits of noise. Now, the big thing is, while small in amplitude, these rapid high bandwith transitions can have great effects on your entire system. Tesla and resonance come into play, this is speculation based on 20 years of RF source and diagnostic design.

Perhaps. But I would consider such to be poorly designed or defective.

I agree.


When did I claim 'technically' perfect filters? I am careful to qualify my statements in the context of audibility.

And they're not, as can be seen by your example of a 16 bit digitizer, that rolls off at 18KHz. With a rolloff that sharp, at that frequency, there will be much less aliasing.

The presence of a distortion is not sufficient to establish audibility. You must coorelate it with audibility, through reference to proper perceptual studies, or by performing and publishing such research yourself.

That is in your opinion, and again audibility is a subjective term.

We come to a few solutions here, make the cutoff earlier in the audio spectrum, as most mfgrs. do, or take the extremely expensive route and build really good filters. The best I believe, is to sample the music twice, first oversample it, at say 60KHz, then digitaly filter it, and then resample at 44.1KHz.

 

[mtrycrafts]

I apologize, my math was in fact wrong. 16 bits of dynamic range is 48 db (65536 counts, log) . But since we are talking voltage you double it 96 db. However, the effective bit rate can't be 16, its simply not possible. Lets be optimistic and assume it really is 14 effective bits (16384 log) gives you 42 db, or 84 for audio. So bottom line we are looking at 84 db of true dynamic range. A good turntable and Vinyl recordings have been measured to have 80 db of dynamic range. Average ambient noise level in a quiet room is about 18db. Taking that into account, the dynamic range is not an issue.

Lets now get into the important part, sampling. A standard cd is sampled at 44.1 KHz. At the limit of human hearing 20KHz, you would have 2.205 data points. Can you make a sine wave with 2.205 data points? Even at 10 KHz, thats 4.41 data points, better but still poor. This is the reason digital music sounds harsh. As you may know the sharper the rise time the more frequency content in the waveform. With this under sampling of music there is MORE harmonic distortion in CD's and DVD's. Most are rated at 1KHz, there you have 44.1 data points, a very reasonable amount of data to reconstruct a signal, hence low THD ratings. On paper they look far superior, but in reality they are inferior. So, here we are, back to the start. Digital needs to fill in all of those gaps, all of the data that is not there has to be "made up". Thats what your processors and sound fields do, fill in the empty data gaps. That is the basic science of it in a nutshell. I would be happy to explain this further if you wish.

Let's assume nothing. It is 96dB, period. Vinyl is in the basement in all respects.

You have 18dB noise level in your room? Really? Recording studios have difficulty getting there, let alone a home. Please.

You need to read some literature on this. J. Stewart of Meridian has something to say about vinyl, and some others.

You also need to check into the sampling theorem, no? Better than to assume things not in evidence and speculate instead. CDs are not harsh, but you may think as you wish.

CD doesn't cut off at 18kHz, by a long shot. You just need to check a few CD specs. Can you hear 20kHz??? I didn't think so. Can you hear 18kHz?

And we are nowhere near where we were. You may be; that is fine.

 

[WmAx]

Audibility, which you refer to so much, is a subjective term.

Audiblity:adj.: heard or capable of being heard

What is subjective about that? A properly executed DBT will establlish audibility.

not true, i looked at the data from several sources that claimed 94.1 db dynamic range (measured), when you look at the data, the interpretation of the data is flawed. The measurement should be made using RMS noise, and max amplitude of the test signal, according to my calculations it was more like 75 db. Also the card cuts off sharply at 18KHz. That is more reasonable for a 44.1KHz sample rate. So, it is in fact not trivial.

I am not sure which source you used, but the common review method is to use RMAA software, using either a reference machine for measurement or loopback testing(though loopback testing gives worse results than reality due to two conversions on the same grade hardware). RMAA works by measure of dynamic range by stimultion of a constant signal vs the calculated RMS noise floor. It also measures pure RMS noise level(though this obviously is subect to inaccuracy if a noise gate is present). You can inquire or critisize how RMAA specifically operates at www.rightmark.org if you have a problem with it. The software was designed, and ensured to work properly by comparison by the author to an AP analyzer system. If you have an issue with the way it calculates, it would useful for you to post at the forum at that address --- the author will respond.

It depends on the pebble, the room resonance, the concrete. I've worked with seismometers that would go crazy from that.

Seismometers are not the issue. I specified 'feeling' the vibration -- not measuring such with sensative measurement device(s). The point is that just because you can measure something, does not mean it is of consequence in so far as human sensory threshold is concerned.

You can't define audibility by some "experts" terms, obviously there is an effect that some people can hear, or this wouldnt be a big controversy in the audio world.

There is no accounting for what some group of people have decided to have faith in without evidence. As my afformentioned example groups of faith believers in the previous post.

Usually bit noise is the biggest thing here, on playback usually you will have 1-2 bits of noise. Now, the big thing is, while small in amplitude, these rapid high bandwith transitions can have great effects on your entire system. Tesla and resonance come into play, this is speculation based on 20 years of RF source and diagnostic design.

You continue to fail to coorelate with audiblity.

Wmax writes: The presence of a distortion is not sufficient to establish audibility. You must coorelate it with audibility, through reference to proper perceptual studies, or by performing and publishing such research yourself.

That is in your opinion, and again audibility is a subjective term.

That is scientific protocol.

-Chris

 

[Nick250]

maybe so, with your fancy tube audio seperates and white-gold divers watches, but in the end, it all comes down to one thing: I have the bigger penis.

That is just what I was thinking, this has turned into one big **** waving contest. Starting with the initial inflammatory post I might add. One could argue this topic ad infinitum but the bottom line is that audio is completely subjective and depends on what floats your boat. Everything else is just some much BS. And if you guys have not figured that out yet then get with the program! End of story.

 

[MacManNM]

Audiblity:adj.: heard or capable of being heard

What is subjective about that? A properly executed DBT will establlish audibility.

Audiblity:adj.: heard or capable of being heard, By whom? My mother cant hear 20KHz. Can you? Can I? Audibility is NOT a scientific term, it needs quantified, to something, dbm, spl, some calibratable NIST standard, not just what some guy can hear.

I am not sure which source you used, but the common review method is to use RMAA software, using either a reference machine for measurement or loopback testing(though loopback testing gives worse results than reality due to two conversions on the same grade hardware). RMAA works by measure of dynamic range by stimultion of a constant signal vs the calculated RMS noise floor. It also measures pure RMS noise level(though this obviously is subect to inaccuracy if a noise gate is present). You can inquire or critisize how RMAA specifically operates at www.rightmark.org if you have a problem with it. The software was designed, and ensured to work properly by comparison by the author to an AP analyzer system. If you have an issue with the way it calculates, it would useful for you to post at the forum at that address --- the author will respond.

What test equipment did they use? Tektronix, HP, Agilent, Acurus, R&S? Or some crappy sound card, not even known to the real scientific community. Are the standards calibrated, what are the error bars in the measurement? All standard information provided in true scientific data.
www.vosssci.com This is standard lab data acquisition software, used by USAF, USN, Los Alamos national labs, Sandia national labs. Im sure the piece of code and hardware used to analyze a $300 sound card is the same as the govt uses to perform multi billion dollar research.



Seismometers are not the issue. I specified 'feeling' the vibration -- not measuring such with sensative measurement device(s). The point is that just because you can measure something, does not mean it is of consequence in so far as human sensory threshold is concerned.

True, but every person has their own threshold



You continue to fail to coorelate with audiblity.
That is scientific protocol.

You have no idea what scientific measurements, or protocol really are, again, audibility is not a scientific term, it cannot be quantified! How can you correlate to something that has no scientific definition? There is not a Physics term that gives true measurable parameters to audibility. If there are lets see them.

 

[WmAx]

Audiblity:adj.: heard or capable of being heard, By whom? My mother cant hear 20KHz. Can you? Can I? Audibility is NOT a scientific term, it needs quantified, to something, dbm, spl, some calibratable NIST standard, not just what some guy can hear.

The threshold of audibility is traditionally determined by a test group of trained listening subjects, as is common in the established perceptual research. In a statistically significant sample group of highly trained listeners, a baseline of what is audible is able to be accurately established. It might be that a genetic anomoly could cause some isolated individual to hear 30kHz, for example, but that is so far off the mark of known human samples that is not important.

What test equipment did they use? Tektronix, HP, Agilent, Acurus, R&S? Or some crappy sound card, not even known to the real scientific community. Are the standards calibrated, what are the error bars in the measurement? All standard information provided in true scientific data.

I already specified the device used for comparitive analysis(Audio Precison Analyzer) by the software author. If you want to ask or critisize the RMAA software and it's reliability -- go to the forum for RMAA found at www.rightmark.org

You have no idea what scientific measurements, or protocol really are, again, audibility is not a scientific term, it cannot be quantified! How can you correlate to something that has no scientific definition? There is not a Physics term that gives true measurable parameters to audibility. If there are lets see them.

You certainly seem accusatory, for a person that has so far substantiated his factually made audibility claims with exactly 0 references as of this point. If you can not, then it is pointless to continue this conversation, as you will only be speculating.

-Chris