or Connect
Head-Fi.org › Forums › Equipment Forums › Sound Science › I Don't Understand You Subjective Guys
New Posts  All Forums:Forum Nav:

I Don't Understand You Subjective Guys - Page 13

post #181 of 861
Quote:
Originally Posted by stv014 View Post

purrin's listening test is now also discussed at the comments section of the "ODAC May Update" article of a certain blog. He is invited to an ODAC blind testing challenge run by an independent third party, similarly to the not yet taken $500 one for the O2.

Do you have the details of how this testing challenge will be run?  Without a properly conducted MSA and testing methodologies the results may be useless.  Merely citing that it's run by an independent third party doesn't mean it will be properly conducted.  Tip: The test is  useless without first establishing at what precision and accuracy the test subject is capable of detecting. Something I have yet to see from any audio "challenge".

 

I design experiments for clinical trials for a living and would be happy to critique the methodology so we can actually feel comfortable that the results of this test prove a thing.


Edited by TWIFOSP - 7/26/12 at 8:51am
post #182 of 861
Quote:
Originally Posted by TWIFOSP View Post

Do you have the details of how this testing challenge will be run?

 

I do not know the exact details, but you can ask for information at the above mentioned blog article, or "O2 Summary" (August 2011). Blind testing details can also be discussed with "RocketScientist" at the diyAudio forums.


Edited by stv014 - 7/26/12 at 9:17am
post #183 of 861

TWIFOSP we would all be in your debt if you could check nwavguy/rocketscientist's methodology. I agree that it would be foolish for purrin to take a listening test without first knowing how good his/her hearing is!

 

popcorn.gif

post #184 of 861
Quote:
Originally Posted by TWIFOSP View Post

 

External measurement systems do not measure pressure differences though.  They measure specific frequency at a given point in time and plot those across time. Ok so what is a measurement of frequency?  Pressure waves.  But they are measuring the signal and not the differences in pressure.  The difference is important when you look at what waves do to each other.  Sound waves aren't some clean signal response like is illustrated by electronics.  There is a mess of collisions and pressure drops.

 

Returning to the example of two DACs, it would make sense to measure the (voltage) output of the DACs, not the "pressure output" of the headphones/speakers. The latter is possible though of course, it's called a "microphone" wink.gif

 

The frequency spectrum is constructed from a fourier transform of the signal, not by measuring one frequency at a time.


Edited by joeyjojo - 7/26/12 at 9:35am
post #185 of 861

Sometimes, subjectivity and objectivity can go along... Just an example:

 

As the subject goes over high end DACs, let's subjectively evaluate how your DAC manages higher sample rates... with an objective test signal :

 

Go to http://www.audiocheck.net/testtones_highdefinitionaudio.php and head to the "High Definition Frequency Sweep" section; download the 192 kHz sweep.

 

"These sweeps start from 1 Hz up to the Nyquist frequency (half the sample rate) in a 30-second time span (linear time sweep).The test starts by playing frequencies that are in your hearing range, and then they will steadily increase. As the sweep enters ultrasonic range, the sound will fade away. Listen carefully; the rest of the file should remain perfectly silent."

post #186 of 861
Quote:
Originally Posted by joeyjojo View Post

TWIFOSP we would all be in your debt if you could check nwavguy/rocketscientist's methodology. I agree that it would be foolish for purrin to take a listening test without first knowing how good his/her hearing is!

 

popcorn.gif

If you can point me in a direction where I can see the testing methodology, I would be happy to.  There is a real quick check that anyone can make to determine if their methodology is flawed.  Did they conduct a measurement systems analysis prior to attempting a double blind test.  If the answer is no, throw the results in the trash.

post #187 of 861
Quote:
Originally Posted by Xaborus View Post

Why would anyone buy a $1000 DAC when the $150 ODAC performs just as good in blind testing as a DAC1?

 

Sometimes i think most of you guys are just buying an expensive placebo effect.. just like $1000 Cables.

 

Edit: Please note- I'm not trolling. Explanation of my viewpoints on the second page.

 

Quote:
Originally Posted by Xaborus View Post


NOT TROLLING. JUST SCIENCE. 192K is harmful. "

192kHz digital music files offer no benefits. They're not quite neutral either; practical fidelity is slightly worse. The ultrasonics are a liability during playback.

Neither audio transducers nor power amplifiers are free of distortion, and distortion tends to increase rapidly at the lowest and highest frequencies. If the same transducer reproduces ultrasonics along with audible content, any nonlinearity will shift some of the ultrasonic content down into the audible range as an uncontrolled spray of intermodulation distortion products covering the entire audible spectrum. Nonlinearity in a power amplifier will produce the same effect. The effect is very slight, but listening tests have confirmed that both effects can be audible."-- http://people.xiph.org/~xiphmont/demo/neil-young.html

 

"It's worth mentioning briefly that the ear's S/N ratio is smaller than its absolute dynamic range. Within a given critical band, typical S/N is estimated to only be about 30dB. Relative S/N does not reach the full dynamic range even when considering widely spaced bands. This assures that linear 16 bit PCM offers higher resolution than is actually required.

It is also worth mentioning that increasing the bit depth of the audio representation from 16 to 24 bits does not increase the perceptible resolution or 'fineness' of the audio. It only increases the dynamic range, the range between the softest possible and the loudest possible sound, by lowering the noise floor. However, a 16-bit noise floor is already below what we can hear"

 

Digital is digital. USB = Optical = Coax. This argument is only valid if you cant physically use usb as an input, and only then and there i admit, you have a valid argument.

 

 

Now i will give a single case in point to subjectivists, and it is a very valid point. Placebo effect is very, very real. And if you believe something will make your audio sound better, it simply will sound better. It doesn't matter if its a $1000 bag of rocks taped to your speakers, it will work if you truely believe it does.

 

But i personally believe that scientifically proven low-distortion methods of music playback reguardless of price is the best way to obtaining the highest audio quality. My placebo effect is not my wallet, but scientific charts and graphs.

 

While I agree that there are some wild claims out there, I don't see how your arguments regarding 192kbps, digital transfer mediums, and bit-widths invalidate subjective experience at all. In order to determine all these parameters, subjective experiences and observations had to be collected in the first place. Models regarding hearing thresholds, loudness levels, audibility thresholds, etc. are/were sanity checked with collective experiences of real people (unless papers about it flat out lie.) All these objective models have limitations, and will hopefully be improved as our understanding of ourselves improves. Measurements do not substitute our real world perception, they complement it.

 

Regarding the design of the audio systems out there, there is a huge collection of topologies built under a different set of requirements, goals, and assumptions. Saying that two "good" but different topologies will sound the same based on our limited knowledge is flawed. In the selection of one topology to the next, there are usually trade-offs and unknowns. I know this is the case in the design of communication systems, and I don't see how audio systems differ in this regard.

 

I believe a consumer is better served using all tools available to him/her in his/her selection of a product, and that includes BOTH subjective and objective instruments.


Edited by ultrabike - 7/26/12 at 11:46am
post #188 of 861
Quote:
Originally Posted by estreeter View Post

Since I seem to be on your blocked user list, it wont matter that I was right, I would ask others to go over the last 4 14 pages of this thread and tell me it didnt belong in SS from the outset. The centipede vs snake vid does ameliorate some of my pain, tho, and responses from purrin and rhythmdevils were gold. I'm just wondering how these guys can soar with eagles when surrounded by turkeys.  rolleyes.gif

This thread should have been deleted or moved to the general discussion page. It's a little insulting that sound science has become a recycle bin since those subforum regulars are left to polish a turd. Although I join you in commending the few brave souls who took the time to spit out an honest few paragraphs on the topic, I don't think that your posts were all that productive. Then again it's not like any of my comments really helped, so I'm sorry for the childish hypocrisy of the whole blocked thing. I'm going back to my vow to leave that list empty.
post #189 of 861
Quote:
Originally Posted by TWIFOSP View Post

You are mistaken.  Signal measurements measure one frequency response at a time.  There is not a single method of analysis that measures the aggregate of every frequency in a musical segment and is capable of displaying it in a meaningful fashion.  

 

I think you are the one who is mistaken.  As pointed out already this is what Fourier series and  transform is all about!  In its very basic form Fourier series theorem states that a continuous time-varying signal can be representated by a combination of sine waves of different frequencies and amplitudes and is a proven mathematical theorem.  So even a complex musical signal if you measure it as a function of time vs "pressure" (ie recording it from a microphone and plot its "pressure" as in amplitude in terms of voltage as joeyjojo suggested), you then apply the Fourier transform to this signal and you get the complex waveform as a representation of all the frequency it is made up of as a mathematical series (which is the Fourier series), plot this series on a graph you get an even better visual representation instead of just a string of numbers - a plot of every frequency in the signal against its amplitude.  This is trivial signal processing 101 stuff which is applied in electronics and digital communications every day and is highly accurate because we build our current fast digital communications network on this very theory and application.  We've had methods to do these "frequency analysis" things for a very long time - heck this stuff goes back to the humble FM radio.


Edited by nanaholic - 7/26/12 at 12:36pm
post #190 of 861

I've reassessed the veracity of the arguments made in the last few pages of this thread, run the numbers again, and come to the conclusion that I prefer the vid where there giant centipede eats a mouse. 

post #191 of 861
Quote:
Originally Posted by nanaholic View Post

 

I think you are the one who is mistaken.  As pointed out already this is what Fourier series and  transform is all about!  In its very basic form Fourier series theorem states that a continuous time-varying signal can be representated by a combination of sine waves of different frequencies and amplitudes and is a proven mathematical theorem.  So even a complex musical signal if you measure it as a function of time vs "pressure" (ie recording it from a microphone and plot its "pressure" as in amplitude in terms of voltage as joeyjojo suggested), you then apply the Fourier transform to this signal and you get the complex waveform as a representation of all the frequency it is made up of as a mathematical series (which is the Fourier series), plot this series on a graph you get an even better visual representation instead of just a string of numbers - a plot of every frequency in the signal against its amplitude.  This is trivial signal processing 101 stuff which is applied in electronics and digital communications every day and is highly accurate because we build our current fast digital communications network on this very theory and application.  We've had methods to do these "frequency analysis" things for a very long time - heck this stuff goes back to the humble FM radio.

 

Fourier series uses a convergent approximation to plot a harmonic.  You can only compare fourier series when A and B are of the same signal type, Periodic, Semi, ect.  Music is a combination of harmonic signal types and turns out can not be compared with a fourier transform.  
 

This article basically sums it up:  Music signals can not be accurately compared.  Convergence is a bitch.

 

http://en.wikipedia.org/wiki/Time%E2%80%93frequency_analysis_for_music_signals

post #192 of 861
Quote:
Originally Posted by TWIFOSP View Post

You are mistaken.  Signal measurements measure one frequency response at a time.  There is not a single method of analysis that measures the aggregate of every frequency in a musical segment and is capable of displaying it in a meaningful fashion.  

Have you looked? There have been many studies on the sound fields of musical instruments which account dispersion patterns, room reflections, harmonics, etc. Then there are many studies on sound reproduction that weigh omnidirectional and point sources in how they are able to approximate the original signal (& more importantly, whether a human being can tell the difference).

There's a difference between hunting down a measurement limit in pure mathematics and accounting for a margin of error in practical applications.
post #193 of 861
Quote:
Originally Posted by anetode View Post


Have you looked? There have been many studies on the sound fields of musical instruments which account dispersion patterns, room reflections, harmonics, etc. Then there are many studies on sound reproduction that weigh omnidirectional and point sources in how they are able to approximate the original signal (& more importantly, whether a human being can tell the difference).
There's a difference between hunting down a measurement limit in pure mathematics and accounting for a margin of error in practical applications.

 

My post was phrased poorly, but the keyword that I meant is method of analysis.  Sure we can plot signals a million accurate ways and put them into formulas.  But to my knowledge there is no method of accurately analyzing them because they all form approximation of harmonics.  I hope that we all understand the difference between a result plot and an analysis.

post #194 of 861
Quote:
Originally Posted by TWIFOSP View Post

My post was phrased poorly, but the keyword that I meant is method of analysis.  Sure we can plot signals a million accurate ways and put them into formulas.  But to my knowledge there is no method of accurately analyzing them because they all form approximation of harmonics.  I hope that we all understand the difference between a result plot and an analysis.

Perhaps I'm not catching on to the specific type of analysis you're referring to. Studies of instruments allowed for physical modelling synthesis, which dates back to the seventies and allows for some impressive Yamaha synths. Then there's speech recognition and speech synthesis software that necessitates the sort of analysis I think you're referring to.
post #195 of 861

as a gedankenexperiment a recorded music performance can be "looped", therefore rendered "perfectly periodic" - thus the Fourier Transform can "perfectly" represent the time series

 

they are mathematical "Duals" the complete Fourier amplitude, phase plots contains every bit (pun) of information that the time series does

 

lots of Sophomoric things have been written about digital audio - look deeper and you often find engineers do know what they are doing

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › I Don't Understand You Subjective Guys