AVS/AIX High-Resolution Audio Test: Ready, Set, Go! - Page 3 - AVS Forum
Forum Jump: 
 35Likes
Reply
 
Thread Tools
post #61 of 215 Old 07-02-2014, 09:27 AM
isa
Member
 
Join Date: Jun 2014
Posts: 35
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 29 Post(s)
Liked: 17
A bit frustrating that no one is showing their work on exactly where people are disagreeing with my work, so I'll guess: you're probably misinterpreting the gain values for "On the Street Where You Live". The HR version is the B version and has a lower Replaygain track gain adjustment, which means it's measured as louder according to EBU R128 as implemented by foobar2000.

Did I guess correctly? If not, PLEASE show your work.
isa is offline  
Sponsored Links
Advertisement
 
post #62 of 215 Old 07-02-2014, 09:28 AM - Thread Starter
AVS Special Member
 
Scott Wilkinson's Avatar
 
Join Date: Sep 2001
Location: Burbank, CA
Posts: 1,215
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 242 Post(s)
Liked: 959
Quote:
Originally Posted by isa View Post
I meant show you work with respect to your conclusion that one or more of the CDDA files measured louder than its HR version. I showed my work using foobar2000 and Replaygain that the HR versions all measured louder, but I have no idea which of the CDDA versions you found louder than the HR versions, and how you determined that.
I think he was referencing a previous post from me that said the difference in track gain is not consistent, not that he knew which is which. Of course, I know which is which, and in the screen shot of ReplayGain you posted, two of the low-res tracks are 0.2 dB louder than the corresponding high-res tracks in the Track Gain column, while in the third case, the high-res track is louder. This is quite puzzling, and we're trying to figure it out.

Scott Wilkinson
AVS Editor
Scott Wilkinson is offline  
post #63 of 215 Old 07-02-2014, 09:33 AM
isa
Member
 
Join Date: Jun 2014
Posts: 35
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 29 Post(s)
Liked: 17
The HR versions are as follows, and each are measured as louder. I'm the only person to show their work. Really appreciate everyone not showing their work but disagreeing ambiguously.

Spoiler!

Last edited by Mike Lang; 07-02-2014 at 09:49 AM.
isa is offline  
post #64 of 215 Old 07-02-2014, 09:41 AM - Thread Starter
AVS Special Member
 
Scott Wilkinson's Avatar
 
Join Date: Sep 2001
Location: Burbank, CA
Posts: 1,215
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 242 Post(s)
Liked: 959
Quote:
Originally Posted by isa View Post
A bit frustrating that no one is showing their work on exactly where people are disagreeing with my work, so I'll guess: you're probably misinterpreting the gain values for "On the Street Where You Live". The HR version is the B version and has a lower Replaygain track gain adjustment, which means it's measured as louder according to EBU R128 as implemented by foobar2000.

Did I guess correctly? If not, PLEASE show your work.
Ah ha! I was misinterpreting the Track Gain value; I thought the higher number corresponded to the louder track, but that's not it. The Track Gain is amount of gain applied to the track to equalize its level, so the track with the lower positive number is louder, because it needed less gain adjustment to get it up to 0! (In the case of the one track with negative gain, the lower negative number indicated more adjustment to get down to 0, so it was louder to begin with.) Now this makes sense, and I know what to do to correct it. Stand by for an update. Thanks isa!
Randy Bessinger likes this.

Scott Wilkinson
AVS Editor

Last edited by Scott Wilkinson; 07-02-2014 at 09:53 AM.
Scott Wilkinson is offline  
post #65 of 215 Old 07-02-2014, 09:53 AM
Advanced Member
 
andyc56's Avatar
 
Join Date: Feb 2013
Posts: 600
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 17 Post(s)
Liked: 187
Quote:
Originally Posted by Scott Wilkinson View Post
You are correct; the track gain differences are not consistent between high res and low res. That's what's so puzzling.
Not so. I just did the spectrum analysis in Audacity and compared it with the replaygain results. The high-res files all have a higher level. Kees and isa are right.

Edit: I see Scott has it figured out now.

Last edited by andyc56; 07-02-2014 at 10:01 AM.
andyc56 is offline  
post #66 of 215 Old 07-02-2014, 10:00 AM - Thread Starter
AVS Special Member
 
Scott Wilkinson's Avatar
 
Join Date: Sep 2001
Location: Burbank, CA
Posts: 1,215
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 242 Post(s)
Liked: 959
Quote:
Originally Posted by andyc56 View Post
Not so. I just did the spectrum analysis in Audacity and compared it with the replaygain results. The high-res files all have a higher level. Kees and isa are right.
You are correct; I had been misinterpreting the Track Gain values. See my previous post. I will repost the files with different A/B assignments after correcting the level mismatch.

Scott Wilkinson
AVS Editor
Scott Wilkinson is offline  
post #67 of 215 Old 07-02-2014, 10:03 AM
AVS Special Member
 
Randy Bessinger's Avatar
 
Join Date: Jun 2009
Posts: 1,279
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 24 Post(s)
Liked: 38
Quote:
Originally Posted by isa View Post
The HR versions are as follows, and each are measured as louder. I'm the only person to show their work. Really appreciate everyone not showing their work but disagreeing ambiguously.

Spoiler!
If this is correct, won't this lead to people just cheating and invalidating any results?
Randy Bessinger is offline  
post #68 of 215 Old 07-02-2014, 10:04 AM
Advanced Member
 
andyc56's Avatar
 
Join Date: Feb 2013
Posts: 600
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 17 Post(s)
Liked: 187
Quote:
Originally Posted by Scott Wilkinson View Post
You are correct; I had been misinterpreting the Track Gain values. See my previous post. I will repost the files with different A/B assignments after correcting the level mismatch.
Yes, I updated my previous post. Looks like we posted at almost the same time. Glad it's been tracked down.
andyc56 is offline  
post #69 of 215 Old 07-02-2014, 10:20 AM
isa
Member
 
Join Date: Jun 2014
Posts: 35
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 29 Post(s)
Liked: 17
Scott, glad anything I posted helped, and sorry for being impatient. Once an impatient engineer, always an impatient engineer...
isa is offline  
post #70 of 215 Old 07-02-2014, 10:29 AM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,891
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 98 Post(s)
Liked: 89
Quote:
Originally Posted by isa View Post
I meant show your work with respect to your conclusion that one or more of the CDDA files measured louder than its HR version. I showed my work using foobar2000 and Replaygain that the HR versions all measured louder, but I have no idea which of the CDDA versions you found louder than the HR versions, and how you determined that.

Forget the rounding error. Rounding happens in the last digit not the first.


I simply used auditions spectrograph to identify the hires sample files to verify my listing results.
I cross referenced with your table and found that the hires files are not consistently 0.2dB louder.
Frank Derks is offline  
post #71 of 215 Old 07-02-2014, 10:31 AM
Advanced Member
 
andyc56's Avatar
 
Join Date: Feb 2013
Posts: 600
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 17 Post(s)
Liked: 187
Just as another data point, I tried the freeware SoX for the round-trip resample, per this post. Here are the replaygain results. No manual tweaking of scale factors was required.

Attached Thumbnails
Click image for larger version

Name:	replaygain_sox.jpg
Views:	187
Size:	19.1 KB
ID:	145633  
andyc56 is offline  
post #72 of 215 Old 07-02-2014, 10:41 AM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,891
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 98 Post(s)
Liked: 89
Quote:
Originally Posted by isa View Post
A bit frustrating that no one is showing their work on exactly where people are disagreeing with my work, so I'll guess: you're probably misinterpreting the gain values for "On the Street Where You Live". The HR version is the B version and has a lower Replaygain track gain adjustment, which means it's measured as louder according to EBU R128 as implemented by foobar2000.

Did I guess correctly? If not, PLEASE show your work.


You're correct in hindsight I misinterpreted the gain values. Probably got confused by the sign.


All the hires samples are 0.2dB louder.
Frank Derks is offline  
post #73 of 215 Old 07-02-2014, 10:42 AM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,891
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 98 Post(s)
Liked: 89
Quote:
Originally Posted by andyc56 View Post
Just as another data point, I tried the freeware SoX for the round-trip resample, per this post. Here are the replaygain results. No manual tweaking of scale factors was required.

This is a rounding error.
Frank Derks is offline  
post #74 of 215 Old 07-02-2014, 10:45 AM
Advanced Member
 
andyc56's Avatar
 
Join Date: Feb 2013
Posts: 600
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 17 Post(s)
Liked: 187
Quote:
Originally Posted by Frank Derks View Post
This is a rounding error.
I think the peak value of the resampled version is higher because of the dither. Not sure if I needed to dither when going back up from 16 to 24, but I did just in case. Looks like SoX preserves signal amplitude very well in its resampling.

Last edited by andyc56; 07-02-2014 at 10:49 AM.
andyc56 is offline  
post #75 of 215 Old 07-02-2014, 11:25 AM
isa
Member
 
Join Date: Jun 2014
Posts: 35
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 29 Post(s)
Liked: 17
Thanks for sharing the SoX result, Andy. Dither should only be necessary under certain conditions when going from more bit depth to fewer bits, and should not affect peaks. I think if you reran without dither when going from 16 to 24, I think you'd get the same Replaygain results but a slightly smaller file. If you have the time, it would be interesting to see if that's what happens.
isa is offline  
post #76 of 215 Old 07-02-2014, 11:53 AM
Advanced Member
 
andyc56's Avatar
 
Join Date: Feb 2013
Posts: 600
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 17 Post(s)
Liked: 187
Quote:
Originally Posted by isa View Post
Thanks for sharing the SoX result, Andy. Dither should only be necessary under certain conditions when going from more bit depth to fewer bits, and should not affect peaks. I think if you reran without dither when going from 16 to 24, I think you'd get the same Replaygain results but a slightly smaller file. If you have the time, it would be interesting to see if that's what happens.
Okay, here is the result with dither only on the 24->16 conversion. I used WAV files when computing replaygain in foobar2000, so the files are the same size either way. Result is exactly the same for replaygain parameters.

Attached Thumbnails
Click image for larger version

Name:	replaygain_sox2.jpg
Views:	196
Size:	19.2 KB
ID:	145665  
andyc56 is offline  
post #77 of 215 Old 07-02-2014, 12:06 PM - Thread Starter
AVS Special Member
 
Scott Wilkinson's Avatar
 
Join Date: Sep 2001
Location: Burbank, CA
Posts: 1,215
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 242 Post(s)
Liked: 959
Quote:
Originally Posted by Randy Bessinger View Post
If this is correct, won't this lead to people just cheating and invalidating any results?
I'm going to reassign A and B labels to the tracks and repost them after adjusting the levels to be a closer match.

Scott Wilkinson
AVS Editor
Scott Wilkinson is offline  
post #78 of 215 Old 07-02-2014, 01:14 PM
AVS Special Member
 
Randy Bessinger's Avatar
 
Join Date: Jun 2009
Posts: 1,279
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 24 Post(s)
Liked: 38
Quote:
Originally Posted by Scott Wilkinson View Post
I'm going to reassign A and B labels to the tracks and repost them after adjusting the levels to be a closer match.
Sounds good.
Randy Bessinger is offline  
post #79 of 215 Old 07-02-2014, 01:35 PM
AVS Addicted Member
 
Mfusick's Avatar
 
Join Date: Aug 2002
Location: Western MA
Posts: 22,769
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 164 Post(s)
Liked: 818
Quote:
Originally Posted by DrMichael View Post
Dumb question but how do you know if your system has that much dynamic range?
It doesn't. Almost no system does. Typical consumer home noise floor is probably 40db so you need another 93db on top of that. 133db is beyond the capability of most all the systems around here and almost all the products made by MFG. The few exceptions would be the DIY crowd or the enlightened hardcore guys using pro audio grade stuff (think danley) that can go loud.

Some of the real serious guys have dedicated sound proofed rooms so they can bring noise floor down another 10-20db. That puts volume at about 10db over "reference" which for speakers is 105db. Reference for subs is 10db louder (115db but THX) so that means the subs actually might/should be louder.

The only people that have this full 93db of range (without significant compression or distortion) is the guys running very high end audio gear and/or DIY or pro audio gear and also have soundproofed and treated rooms to go with it. Around AVS there is some but in real life it's a needle in a haystack for normal folks.

-

"Too much is almost enough. Anything in life worth doing is worth overdoing. Moderation is for cowards."
Mfusick is online now  
post #80 of 215 Old 07-02-2014, 02:24 PM
Member
 
wgscott's Avatar
 
Join Date: Aug 2010
Posts: 91
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 31
Quote:
Originally Posted by mightyhuhn View Post
what's about that _MACOSX folder?
Resource fork files. This is an artifact of an HFS+ filesystem. For present purposes, it is garbage.
wgscott is offline  
post #81 of 215 Old 07-02-2014, 02:30 PM
Member
 
wgscott's Avatar
 
Join Date: Aug 2010
Posts: 91
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 31
Quote:
Originally Posted by Scott Wilkinson View Post
3. Very few consumer systems are capable of reproducing frequencies above 20 kHz and a dynamic range greater than 93 dB, so trying to compare a real high-res recording with a CD-quality version of the same material on such a system is pointless. There will be no acoustic difference between the two files (other than perhaps some distortion if the level of the 24/96 content exceeds 93 dB), so there is no possibility of a perceptible difference between them.
Even fewer of their owners can hear above 15kHz.

The proponents of high-resolution audio are not (in general) claiming you can detect (or need to be able to detect) sound higher in frequency than 22 kHz. Rather, they are suggesting that sampling at 44.1 kHz (truncating the signal at 22.05 kHz) produces aliasing artifacts (and perhaps Fourier truncation artifacts) within the (non-controversially) audible frequency range. Hence, if the proponents are right in this assertion, there is no requirement for specialized speakers or other equipment that reproduces ultrasonic frequencies.
wgscott is offline  
post #82 of 215 Old 07-02-2014, 02:36 PM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,891
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 98 Post(s)
Liked: 89
Quote:
Originally Posted by andyc56 View Post
Okay, here is the result with dither only on the 24->16 conversion. I used WAV files when computing replaygain in foobar2000, so the files are the same size either way. Result is exactly the same for replaygain parameters.



Imo it would be best (if Scott doesn't use SOX but sticks to the resampler he used before ) to add the gain of +0.2 dB before the down sampling without dither.
Then down sample 24>16 with dither and up-sample back to 24bit without dither.


Adding the +0.2dB gain after upsampling adds another round of dither and also amplifies the first round of dither raising the noise floor slightly.

Last edited by Frank Derks; 07-02-2014 at 02:51 PM.
Frank Derks is offline  
post #83 of 215 Old 07-02-2014, 02:40 PM
AVS Addicted Member
 
Mfusick's Avatar
 
Join Date: Aug 2002
Location: Western MA
Posts: 22,769
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 164 Post(s)
Liked: 818
Quote:
Originally Posted by wgscott View Post
Even fewer of their owners can hear above 15kHz.

The proponents of high-resolution audio are not (in general) claiming you can detect (or need to be able to detect) sound higher in frequency than 22 kHz. Rather, they are suggesting that sampling at 44.1 kHz (truncating the signal at 22.05 kHz) produces aliasing artifacts (and perhaps Fourier truncation artifacts) within the (non-controversially) audible frequency range. Hence, if the proponents are right in this assertion, there is no requirement for specialized speakers or other equipment that reproduces ultrasonic frequencies.
I can hear higher than 15k but less than 18k. I'm 36 and also have tinnitus.

But the entire concept or higher than 20khz sounds like horse manure to me. Often the speaker shoot outs the speakers that roll off like a cliff at 19khz are the ones everyone likes best. I'll save my worries for the stuff I'm more likely to hear.

-

"Too much is almost enough. Anything in life worth doing is worth overdoing. Moderation is for cowards."
Mfusick is online now  
post #84 of 215 Old 07-02-2014, 02:54 PM
Member
 
wgscott's Avatar
 
Join Date: Aug 2010
Posts: 91
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 8 Post(s)
Liked: 31
Quote:
Originally Posted by Mfusick View Post
I can hear higher than 15k but less than 18k. I'm 36 and also have tinnitus.

But the entire concept or higher than 20khz sounds like horse manure to me. Often the speaker shoot outs the speakers that roll off like a cliff at 19khz are the ones everyone likes best. I'll save my worries for the stuff I'm more likely to hear.
The audible threshold and the pain threshold converge as you approach 20 kHz, so this isn't surprising.

I agree with your assessment, but if you formulate the test such that it guarantees the desired outcome, it is not very compelling.

Put in a slightly different way: If somehow some participants could reliably identify the HD audio in a double-blind test, would that demonstrate that they can hear above 22kHz? (I think not.)
Mfusick likes this.
wgscott is offline  
post #85 of 215 Old 07-02-2014, 03:12 PM
isa
Member
 
Join Date: Jun 2014
Posts: 35
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 29 Post(s)
Liked: 17
Quote:
Originally Posted by Frank Derks View Post
Imo it would be best (if Scott doesn't use SOX but sticks to the resampler he used before ) to add the gain of +0.2 dB before the down sampling without dither.
Then down sample 24>16 with dither and up-sample back to 24bit without dither.


Adding the +0.2dB gain after upsampling adds another round of dither and also amplifies the first round of dither raising the noise floor slightly.
What? If your goal is to "trojan horse" the test, even as informal as it is, then mission accomplished if the advice of adding a 0.2 dB post-conversion bandaid is followed. If the underlying issues of gain change and sample drops are not found and fixed for the 16/44.1k versions, or if SoX isn't used to avoid these problems, then any result can and probably should be immediately dismissed by all regardless of result.

SoX is very, very good at what it does. No need to make this complicated. If the underlying issues of the original method can't be quickly found, just use SoX as Andy has demonstrated.
andyc56 likes this.
isa is offline  
post #86 of 215 Old 07-02-2014, 03:57 PM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,891
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 98 Post(s)
Liked: 89
Quote:
Originally Posted by isa View Post
What? If your goal is to "trojan horse" the test, even as informal as it is, then mission accomplished if the advice of adding a 0.2 dB post-conversion bandaid is followed. If the underlying issues of gain change and sample drops are not found and fixed for the 16/44.1k versions, or if SoX isn't used to avoid these problems, then any result can and probably should be immediately dismissed by all regardless of result.

SoX is very, very good at what it does. No need to make this complicated. If the underlying issues of the original method can't be quickly found, just use SoX as Andy has demonstrated.

I agree, SoX would be the best option to minimize the number of processing steps.
Frank Derks is offline  
post #87 of 215 Old 07-02-2014, 04:13 PM - Thread Starter
AVS Special Member
 
Scott Wilkinson's Avatar
 
Join Date: Sep 2001
Location: Burbank, CA
Posts: 1,215
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 242 Post(s)
Liked: 959
Quote:
Originally Posted by Frank Derks View Post
Imo it would be best (if Scott doesn't use SOX but sticks to the resampler he used before ) to add the gain of +0.2 dB before the down sampling without dither.
Then down sample 24>16 with dither and up-sample back to 24bit without dither.


Adding the +0.2dB gain after upsampling adds another round of dither and also amplifies the first round of dither raising the noise floor slightly.
Quote:
Originally Posted by isa View Post
What? If your goal is to "trojan horse" the test, even as informal as it is, then mission accomplished if the advice of adding a 0.2 dB post-conversion bandaid is followed. If the underlying issues of gain change and sample drops are not found and fixed for the 16/44.1k versions, or if SoX isn't used to avoid these problems, then any result can and probably should be immediately dismissed by all regardless of result.

SoX is very, very good at what it does. No need to make this complicated. If the underlying issues of the original method can't be quickly found, just use SoX as Andy has demonstrated.
I think the underlying issue has been found: Most sample-rate converters add 0.1 dB of headroom during the conversion, which lowers the overall level of the file by that much, so putting one file through two SRCs lowers the level by 0.2 dB. So the question now is, should we boost the level of the track that will undergo the double SRC by 0.2 dB before or after downsampling to 16/44.1? I agree that adding it after re-upsampling will raise the noise floor slightly. It seems to me we should increase the level of the source track before downsampling, but I wonder what you think?

BTW, we will continue to use the SRC that Mark is familiar with—Sonic Studios Sonic Process—which is a state-of-the-art SRC.

Scott Wilkinson
AVS Editor
Scott Wilkinson is offline  
post #88 of 215 Old 07-02-2014, 04:23 PM
Senior Member
 
granroth's Avatar
 
Join Date: Jan 2012
Location: Gilbert, AZ
Posts: 396
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 62 Post(s)
Liked: 102
Quote:
Originally Posted by wgscott View Post
The audible threshold and the pain threshold converge as you approach 20 kHz, so this isn't surprising.

I agree with your assessment, but if you formulate the test such that it guarantees the desired outcome, it is not very compelling.

Put in a slightly different way: If somehow some participants could reliably identify the HD audio in a double-blind test, would that demonstrate that they can hear above 22kHz? (I think not.)
There are perhaps some "pure" audiophiles that claim the ability to hear above 22kHz, but that's not a common view. The prevailing viewpoint of those who say that you can identify HD audio is that you "perceive" the high frequencies, not "hear" them. That is, that the ultrasonic frequencies are somehow (nobody has ever shown or suggested how) being picked up by our brains and incorporated into what we hear without use of our normal hearing channels.

There is precedence for something like this. Think bass. You almost surely cannot hear a 10Hz tone, but I can almost guarantee that you'd know when one was playing. That is, you are able to perceive or detect the presence of a particular frequency even though you can't hear it. And yeah, I'd expect that 10Hz tone to affect how you perceive the music or sound that is playing.

High frequencies have nowhere near the energy of low frequencies, though, so I wouldn't expect that any possible mechanism for detecting the former would be the same as the latter.

So can people perceive ultrasonic frequencies? That's precisely the point of this test! As far as I know, there are no reliable (mostly agreed upon) studies that show either one way or another whether it's possible. I'm very interested to see the results of all this.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
granroth is offline  
post #89 of 215 Old 07-02-2014, 04:49 PM
Senior Member
 
granroth's Avatar
 
Join Date: Jan 2012
Location: Gilbert, AZ
Posts: 396
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 62 Post(s)
Liked: 102
Quote:
Originally Posted by mcnarus View Post
The 0% case can just as easily be explained by transposing the two versions in his mind and simply not re-checking. That is, say he listens to A and B and notes their differences. He then goes through the five randomized samples and does notice a difference each time... but consistently presses B when he meant A and vice versa because he transposed which version was which. That's not being a moron, that's just being mildly forgetful.
If he did it once, yeah. But 5 times in a row? Come on.
You don't do if five times in a row -- you transpose A and B once and then use that transposition during the test. That's an extremely common thing to do. I was just doing hinge mortises the other day where I started out knowing that I had to set my jig for hinges but somehow early on switched that around in my head and instead set the jig for strike plates. I repeated this for all the hinges and it wasn't until I was done that I realized I had transposed the settings in my first step.

Quote:
Originally Posted by mcnarus View Post
I maintain that neither 0% nor 100% can be easily explained by either good or bad guessing, since a 3% chance is too slim to use any kind of guessing as a pat answer.
It's not a pat answer. It's a proper statistical inference. Black swans happen. Trying to rationalize them away is bad statistics. That's what you're guilty of here.

Possible, but unlikely. It's FAR more possible that he really did detect a difference of some sort consistently and just transposed his choices.
Really? What's the probability? We know the probability of bad guessing, and it's non-zero. You're just pulling a probability out of your nether reaches here.
Well, I guess we understand probability in different ways. I was always decent at math but I can't claim any high degrees specializing on that. If you have a mathematics degree, then I'd like to know what elementary error I'm making.

Because this does look elementary to me -- something you'd cover in Probability 101 in any college or advanced high school. A 3% chance means "unlikely" and "unlikely" means that even though, yes, black swans happen, any random swan is probably NOT black. If you're a betting man and the odds are only 3% that you're right, then your best bet is to fold.

Put one more way -- if I only have a 3% chance of consistently guessing a series of answers and I do consistently get the answers, then it's a fool's bet to say that I guessed -- probability theory says that I am far more likely to have known the answers than just guessed them.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
granroth is offline  
post #90 of 215 Old 07-02-2014, 05:25 PM
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 18,083
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 755 Post(s)
Liked: 434
Quote:
Originally Posted by Mfusick View Post
It doesn't. Almost no system does. Typical consumer home noise floor is probably 40db so you need another 93db on top of that.
The noise floor number is not proper. SPL meters are "perceptually blind." You can't take their one number and have it mean anything with respect to audibility. See this article I wrote on proper way to analyze room noise: http://www.madronadigital.com/Librar...amicRange.html

Quote:
Originally Posted by Mfusick
The only people that have this full 93db of range (without significant compression or distortion) is the guys running very high end audio gear and/or DIY or pro audio gear and also have soundproofed and treated rooms to go with it. Around AVS there is some but in real life it's a needle in a haystack for normal folks.
Well, normal folks would not be concerned with high resolution audio, nor be married to threads like this . That said, there is data in my article about this:



As you see the noise floor of average system surveyed in the sensitive part of our hearing is around 10 spl and best systems go well below that, below the threshold of hearing.

If you look to the left you see your "40 db" number but that only holds true for very low frequencies.

Amir
Founder,
To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.

"Insist on Quality Engineering"

amirm is offline  
Reply Audio theory, Setup and Chat

Tags
frontpage

User Tag List

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off