Debate Thread: Scott's Hi-res Audio Test - Page 71 - AVS Forum
Forum Jump: 
 734Likes
 
Thread Tools
post #2101 of 2920 Old 07-18-2014, 09:33 AM
AVS Addicted Member
 
Chu Gai's Avatar
 
Join Date: Sep 2002
Location: NYC area
Posts: 14,964
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 317 Post(s)
Liked: 595
The president of the AES said,
Quote:
“I’ve heard some wonderful CDs, but I’ve also heard some wonderful 24/96 files,” Olive said. “I really think the difference is how well they’re recorded and mastered.”

"I've found that when you want to know the truth about someone that someone is probably the last person you should ask." - Gregory House
Chu Gai is offline  
Sponsored Links
Advertisement
 
post #2102 of 2920 Old 07-18-2014, 10:20 AM
AVS Special Member
 
RichB's Avatar
 
Join Date: Nov 1999
Location: Massachusetts
Posts: 8,889
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 328 Post(s)
Liked: 270
Quote:
Originally Posted by Chu Gai View Post
The president of the AES said,


Quote:
“I’ve heard some wonderful CDs, but I’ve also heard some wonderful 24/96 files,” Olive said. “I really think the difference is how well they’re recorded and mastered.”
Can I hear an Amen

Hey, it got religious a long time ago.

- Rich

Oppo Beta Group

Parasound A51 Black (for sale) | Oppo BDP-105D | Oppo HA-1 | Oppo PM-1 | Parasound A21 & A31 | Revel Salon, Voice, Studio | Velodyne HGS-15
RichB is offline  
post #2103 of 2920 Old 07-18-2014, 11:12 AM
GGA
Advanced Member
 
GGA's Avatar
 
Join Date: Dec 2001
Location: Topanga CA
Posts: 713
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 32 Post(s)
Liked: 14
Quote:
Originally Posted by rufus13 View Post
In the world of recording, it's pretty-well understood that the most important conversion is the first one, where a mic pre signal gets turned into a bit stream for a mixer or recorder. When the first bunch of 12 bit or 16 bit recorders became available, manufacturers included instructions to set levels as close to 0dB Full-Scale but not go over (very much very often) or there would be undesirable (horrible) distortion. This was so that they could claim a very high signal to noise ratio, compared to good analog tape recorders like a full-track mono 1/4" AG-440b at 15ip/s (70dB, no nr). .

But wait, years later there are 24 bit recorders. Even cheaper, hooray! The extra dynamic range (more than 16-19 bits of useful) provides "headroom" not unlike the ignored feature of an analog magnetic recorder, but only if you know about it. Manufacturers still tell users to set levels near zero dB, as if it was 1984. Smart users find where 0dB is and reduce the input by 20dB, making pretty clean recordings of whatever the line gives. You also don't need to use products like APHEX DOMINATOR limiter, unless you like that kind of thing.
That is very interesting. How does recording at -20dBFS improve things and what does it improve? Like you said I was always told to get near zero. I am primarily archiving my vinyl to digital. Thanks.
GGA is offline  
post #2104 of 2920 Old 07-18-2014, 12:03 PM
Senior Member
 
stereoeditor's Avatar
 
Join Date: Feb 2010
Posts: 343
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 140 Post(s)
Liked: 39
Quote:
Originally Posted by GGA View Post
How does recording at -20dBFS improve things and what does it improve? Like you said I was always told to get near zero.
Many A/D converters become less linear with signals above -2dBFS. I recommend peaking around -6dBFS, which, with most 24-bit converters actually having noise floors at the 19- or 20-bit level, means you are still recording with at least 18-bit resolution. Peaking at -20dBFS means throwing away 3 bits of resolution to no good end. If you are going to mix digitally "in the box," I think it best to have the maximum possible resolution on your stems. When you then apply attenuation in the mix, you are going to reduce the noise/quantization by the amount of attenuation. If you record with peaks at -20dBFS, you might then have to apply gain in the mix and thus reduce resolution.

At least that is what I do, to produce what Arny Krueger calls my "sonic spectaculars." :-)

Quote:
Originally Posted by GGA View Post
I am primarily archiving my vinyl to digital.
If you subsequently normalize your LP transfers, I recommend peaking no higher than -0.1dBFS, to allow for inter-sample overs if you dither down to 16 bits.

John Atkinson
Editor, Stereophile
GGA likes this.

Last edited by stereoeditor; 07-18-2014 at 12:09 PM.
stereoeditor is offline  
post #2105 of 2920 Old 07-18-2014, 12:14 PM
AVS Addicted Member
 
Chu Gai's Avatar
 
Join Date: Sep 2002
Location: NYC area
Posts: 14,964
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 317 Post(s)
Liked: 595
Did I call it, Rich?
RichB and Ratman like this.

"I've found that when you want to know the truth about someone that someone is probably the last person you should ask." - Gregory House
Chu Gai is offline  
post #2106 of 2920 Old 07-18-2014, 01:25 PM
Member
 
rufus13's Avatar
 
Join Date: Dec 2012
Location: Soviet of Portlandia
Posts: 20
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 4
Quote:
Originally Posted by GGA View Post
That is very interesting. How does recording at -20dBFS improve things and what does it improve? Like you said I was always told to get near zero. I am primarily archiving my vinyl to digital. Thanks.
When transcribing from a fixed and already peak-limited source like a vinyl record, you don't need that much headroom in the recorder. I suspect that this is one of the reasons why analog mag recorders are so loved: playback into converters is compressed and peaklimited (more at excessive levels, less at lower levels) allowing higher average level in the digital system resulting in louder=perceived-better. The "studio feel" of big reels, relay noises, PWM spooling whirrrr, and lots of meters is exotic and gives client value-satisfation-confidence (except in LA or NYC with corporate clients or sync-to-video).

Another thing I've noticed in less expensive recorders it that the meters are some combination of low-resolution and/or mis-calibrated. You can find the zero and note where clipping "turns the corner", then reduce level from that as an operating level. No dedicated TMDE required, just computer sound-card and software you already have (maybe patch cables).

If you have AES, s/p-dif, some sort of optical input, you can use your recorder as data storage for an external box that might have better analog and/or clocking. I like getting more use from older stuff. YMWV.

Some converters have integrated (switchable?) limiters to allow higher input analog levels that would otherwise occasionally exceed 0dB FS, but the price is the sound of limiting + the price of the circuit (and advertising). These converters cost more, so TANSTAAFL again.

Giving away some alleged resolution (not really, since FS is still 24 bit) to avoid clipping (ugly and impossible to "de-clip" regardless of plug-ins that promise that) is a good real-world trade when operating in the wild.

The cheapest stuff for recording is so capable, when compared to 30 years ago. I feel like I am picking at invisible pixie dust detectable only with an rfid scanner. Go record! :-)
GGA likes this.
rufus13 is offline  
post #2107 of 2920 Old 07-18-2014, 01:47 PM
Member
 
rufus13's Avatar
 
Join Date: Dec 2012
Location: Soviet of Portlandia
Posts: 20
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 4
Quote:
Originally Posted by Marc Wielage View Post
I still like the idea of high-res downloads, if only because it's a chance for musicians to make more money, for audiophiles to buy non-compressed content, and for record labels to make more money.
Brian Eno released not only uncompressed audio of some of his older material, but he shared the synchronized multitracks and invited fans to do-as-they-would with them and submit the result back. Big files took hours to get even without 56K dialup.

Don't forget people in the boonies who use narrowband internet and still use sneakernet usb-flash and dvd-r to get big media files.

It would be great if little labels made money. I must be doing something wrong. ;-)
rufus13 is offline  
post #2108 of 2920 Old 07-18-2014, 01:52 PM
Member
 
rufus13's Avatar
 
Join Date: Dec 2012
Location: Soviet of Portlandia
Posts: 20
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 4
Hey, there's a medium blue "like" on a light blue background! Cool.
rufus13 is offline  
post #2109 of 2920 Old 07-18-2014, 03:17 PM
AVS Special Member
 
RichB's Avatar
 
Join Date: Nov 1999
Location: Massachusetts
Posts: 8,889
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 328 Post(s)
Liked: 270
Quote:
Originally Posted by Chu Gai View Post
Did I call it, Rich?
Yes, I never doubted you.

- Rich

Oppo Beta Group

Parasound A51 Black (for sale) | Oppo BDP-105D | Oppo HA-1 | Oppo PM-1 | Parasound A21 & A31 | Revel Salon, Voice, Studio | Velodyne HGS-15
RichB is offline  
post #2110 of 2920 Old 07-18-2014, 10:38 PM - Thread Starter
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 18,375
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 981 Post(s)
Liked: 547
Quote:
Originally Posted by krabapple View Post
Another question is can the difference be heard without earphones?
Why? The results with headphones are sufficient to invalidate the position that these differences are inaudible. Your ear doesn't change because you are using headphones vs speakers. If the ears with headphones hear the difference, then clearly we are above threshold of inaudibility.

BTW, vast majority of compression testing is done with headphones. This includes development of international standards such as AAC audio codec.

Quote:
Originally Posted by krab
Again, let's keep sight of the forest: extravagant claim after claim, year after year (going on over a decade now), that 'hi rez' sounds obviously and distinctly better than Redbook....and not just over headphones.
The claim year after year has been that these differences are simply inaudible. So forget about folks not here, and let's stop saying things like this:

Quote:
Originally Posted by Frank Derks View Post
Above all the M&M test proved that inserting a adc and dac is transparent.
I post the results of this very test from Ethan's site and the adc and dac was not transparent to my ears.

Quote:
Originally Posted by RobertR View Post
Of course it matters. People would know that careful production of 16/44 gives them just as good sound as hires.
Except that I could hear the differences between both Scott's and Arny's files at high-res versus CD.

Quote:
Originally Posted by CharlesJ View Post
Yes, they will never accept the facts and will carry their beliefs to the grave.
So hard facts were presented in the form of my listening tests. Don't see anyone "accepting" them from our camp.

Quote:
Originally Posted by RobertR View Post
Anything that doesn't conclude that hires is audibly superior to CD is unacceptable to some people.Quote:How about you doing such an endeavor?
That's exactly what I did. I have participated in more listening tests in this thread than everyone else combined. Including the high-res vs CD.

Those are all from the first page of this thread.

Quote:
Originally Posted by antoniobiz1 View Post
Here apparently you misunderstood the whole point of the paper. If a great quality recording sounds the same in hires and through a 44.1/16 loop, what is the point of hires? Useful for the master, ok, but useless as a delivery format.
If I can tell the difference then it is not useless.

Quote:
Originally Posted by amirm View Post
So in summary, we have guidelines for a reason. Research shows the necessity of using expert listeners if we want to have high confidence results. The above paper was published in 1996 yet in 2007 we have the Boston Audio Society conducting a test and writing a paper that is devoid of important recommended practices in the industry. This is why I don't put a lot of weight on their test.
I said that in page 2 of this thread. Did folks listen? Nope. 60 pages later I demonstrated the importance of that using my listening tests. At some point, we need to stop dismissing professional experience in this regard.

Here is the response to my post above:

Quote:
Originally Posted by Frank Derks View Post
Nonsense. There is no need for trained listeners to make the test valid. That is just you opinion.
Just my opinion... right.

Quote:
Originally Posted by RobertR View Post
Meaning what? That an “industry standard” regarding listeners is the ONLY protocol that can POSSIBLY have any validity? Nonsense. The M&M study used “men and women of widely varying ages, acuities, and levels of musical and audio experience; many were audio professionals or serious students of the art.” IOW, people who are representative of the record buying public. Such people are NOT exclusively “subjects who have expertise in detecting small impairments”.

Your objection sounds an awful lot like one of the items in the subjectivist laundry list, ie "YOUR ears aren't GOOD enough to hear the difference". I find that interesting.
I hope you no longer think that everyone hears the same.

Quote:
Originally Posted by Chu Gai View Post
Certainly, it's possible that thoroughly trained listeners in a manner that Amir put forth might be able to reliably discern audible differences.
So the "might" and "possible" are certainties now.

This was all said in page 2 of this thread. Let me quote this post from Arny on page 3 and stop there:

Quote:
Originally Posted by arnyk View Post
Let's put it this way. 16 bit coding made with properly shaped dither can put all artifacts about 120 dB down referenced to 1 KHz and weighted by normal hearing @ 85 dB SPL. 80 dB suppression of those artifacts would be overkill if the goal was sonic transparency.
How possible is it that I am hearing "differences -120 db down?"

So before you ask others to change, let's see if the present vocal posters in this forum and this thread will do something different, now that they have clear counter evidence to their audio position.

Quote:
Originally Posted by krab
So, does it actually require golden ears, and possibly headphones, to hear a difference ? (and we haven't even really tested *preference* )
We are back to turning absence of data into data. I haven't gone to the supermarket in the next town over therefore, I must be afraid of going there. Illogical right? Lack of testing conditions beyond what is presented is just that: no data. You cannot conclude anything from that. Let's admit that "golden ears" do exist, that differences can be heard with headphones and these conversations become a lot more constructive.

Amir
Retired Technology Insider
Founder, Madrona Digital
"Insist on Quality Engineering"
amirm is offline  
post #2111 of 2920 Old 07-18-2014, 11:23 PM
AVS Special Member
 
NorthSky's Avatar
 
Join Date: Jan 2013
Location: Canada - West Island: Vancouver, South Direction: Go East
Posts: 4,692
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 2030 Post(s)
Liked: 593
Quote:
Originally Posted by rufus13 View Post
... I think modern movies are too loud.
Some (many), definitely. ...And others are recorded with the dialog (what counts the most) too low.

The music and movie world is first and foremost one of entertainment. ...Then business.
And, always, the sound music/movie man/woman mixing/recording engineer is the main instigator of all sound discussions related, the mastermind.
It's him/her who is @ the mixing sound control. ...He/she works with all audio stems and @ various resolutions (frequency range, bit depth, sampling rate, digital filtering, spaciousness, EQuing, tra-la-la) and using the digital/analog machines @ his/her disposition.
NorthSky is online now  
post #2112 of 2920 Old 07-18-2014, 11:31 PM
AVS Special Member
 
NorthSky's Avatar
 
Join Date: Jan 2013
Location: Canada - West Island: Vancouver, South Direction: Go East
Posts: 4,692
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 2030 Post(s)
Liked: 593
Amir, you did a heck of a job; congratulations and thank you very much. It will help me to pay more attention in the future. ...The music.
NorthSky is online now  
post #2113 of 2920 Old 07-19-2014, 01:23 AM
Member
 
rufus13's Avatar
 
Join Date: Dec 2012
Location: Soviet of Portlandia
Posts: 20
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 4
If you can hear an effect on headphones, but not your loudspeaker/room combination, it doesn't mean that the headphones are unfairly revealing. It is just the opposite: your loudspeakers are unfairly concealing.

It mean that you have not perfected your room. Most people overspend on the gear in the room and under-spend on the room. It's well known that a 4" speaker in a portable radio sounds considerably better in a recording studio after the treatment goes up.

The room inside your headphones has been carefully calibrated. If you are using $300 over-the-ear headphones, it's gonna be pretty darn good. Add a good headphone amp and it's good, really-good. A satisfying date this is not!
rufus13 is offline  
post #2114 of 2920 Old 07-19-2014, 01:38 AM
AVS Special Member
 
NorthSky's Avatar
 
Join Date: Jan 2013
Location: Canada - West Island: Vancouver, South Direction: Go East
Posts: 4,692
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 2030 Post(s)
Liked: 593
That's why headphones are used for monitoring by some sound designers.
NorthSky is online now  
post #2115 of 2920 Old 07-19-2014, 04:50 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,387
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 762 Post(s)
Liked: 1175
Quote:
Originally Posted by amirm View Post

By the way, I am traveling and the headphone I have is my Shure IEM. So now we have results across three different headphones.
Which Shure IEMs? A pair of E3s are among my favorites.

Quote:
Originally Posted by amirm View Post
Arny do you honestly not hear the difference???
Amir, given the number times I've explained my recent hearing deficiencies to you in detail, your constant reference to this fact seems to be pointless, but perhaps a little insensitive. Do you shout "Look at the crippled man" every time you see a parapalegic?

However, the number of people who have failed to obtain positive results from Ethan's comparisons seems to be large, so I am not alone. Actually, your inability to obtain negative results seems to raise a few concerns, perhaps about your laptop if not your choice of reproducers.

Also, perhaps you need to revisit your jitter-related listening tests as they were not in my recollection overly brilliant.

Quote:
Originally Posted by amirm View Post
By the way the techno clip Ethan picked is not ideally suited for this type of testing yet the audible difference is there.
I suggest that you not look gift horses in the mouth. ;-)

Positive results in ABX tests indicates that it was pretty good for its intended purpose. If you can find something better, of course you are free to put it forward.

Last edited by arnyk; 07-19-2014 at 05:16 AM.
arnyk is offline  
post #2116 of 2920 Old 07-19-2014, 05:02 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,387
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 762 Post(s)
Liked: 1175
Quote:
Originally Posted by GGA View Post
That is very interesting. How does recording at -20dBFS improve things and what does it improve? Like you said I was always told to get near zero. I am primarily archiving my vinyl to digital. Thanks.
Your point is well taken. -20 dB FS levels don't improve things in any way. However, live musical sources can be very unpredictable.

One of my rules of thumb is that peaks during a formal live performance may exceed the corresponding rehearsal by up to 10 dB.

If I do not have a relevant rehearsal and there can be post production leveling of the recording before distribution, then I try to set things up for 20 dB headroom.

However, a lot of the recordings I make cannot receive any post production processing (i.e., bring up peak levels to within 1 dB of FS). Therefore the peak level recorded is the peak level delivered, and I try to keep peaks about 4-5 dB down just in case. In this context there are no rehearsals.

BTW I agree with the idea of leaving some headroom to handle those rare but real situations where peak levels can exceed FS even though no sample exceeds FS. There is another effect where infrequently the slightly random phasing of all sources happen to line up for a split second and there is also an anomalous peak.

So this is how things tally up:

Maximum dynamic range of a live performance: 65-75 dB

Allowance for rehearsal versus formal performance: 10 dB

Reserve for putting the equipment noise floor below the venue's noise floor: >=8 dB

Total: 93 dB - still within what is available with 16 bits and uniform spectrum dither.

Last edited by arnyk; 07-19-2014 at 05:20 AM.
arnyk is offline  
post #2117 of 2920 Old 07-19-2014, 05:12 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,387
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 762 Post(s)
Liked: 1175
Quote:
Originally Posted by NorthSky View Post
That's why headphones are used for monitoring by some sound designers.
The usual rule of thumb is that headphone listening gives a very different sonic perspective from loudspeaker listening.

Overall, my listening is about 50/50 speaker versus headphones/earphones. My non-speaker listening is about 50/50 headphones versus earphones. All monitoring systems are significantly optimized using electronic equalization. I'be mentioned my use of Shure E3 earphones, but they are not the only ones I have by far. For example I have a pair of ER4s, which I loathe. I also have a collection of various headphones, with AT M50s, Superlux HD 668, Sennheiser RS170, Sony MDR 7506, etc.

Part of that difference can be an acute perception of technical flaws in the recording due to the absence of masking by the listening room and speakers.

Therefore, using headphones to hear technical flaws during recording and editing and mixing can be OK, but the final product needs to be auditioned and perhaps adjusted some more using loudspeakers before distribution.

People who use headphones frequently and also use speakers for mastering can sometimes develop a pretty good sense of how headphone listening translates to speaker listening, and get pretty close to a good finished product with headphones.

Last edited by arnyk; 07-19-2014 at 05:29 AM.
arnyk is offline  
post #2118 of 2920 Old 07-19-2014, 06:33 AM
AVS Special Member
 
Frank Derks's Avatar
 
Join Date: Jul 2002
Location: Region A,B,C
Posts: 1,897
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 103 Post(s)
Liked: 89
Quote:
Originally Posted by amirm View Post
...
I post the results of this very test from Ethan's site and the adc and dac was not transparent to my ears.

You will fail when it's not 5th generation but only one like in the M&M test.


Quote:
Originally Posted by amirm View Post
Just my opinion... right.
...
Out of context


For that particular test there is no need for trained listeners. They will fail just the same as untrained listeners.
Frank Derks is offline  
post #2119 of 2920 Old 07-19-2014, 07:30 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,387
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 762 Post(s)
Liked: 1175
Quote:
Originally Posted by Frank Derks View Post
You will fail when it's not 5th generation but only one like in the M&M test.
Good point. Ethan's test involved what was probably a far poorer DAC and ADC quality than what was used in the M&M tests.

Amir seems to be glossing over questions about the nonlinear distortion in his monitoring setup.

The distortion losses in a test like Ethan's basically add up via ordinary arithmetic over the generations. Uncorrelated noise adds up geometrically.

More details here:

http://ethanwiner.com/loop-back.htm

"In 2006 I did such a test using a $25 SoundBlaster X-Fi sound card. I never wrote an article with files for people to download and identify which file is which, but I posted the files in forums many times to let people assess the quality loss. If you're curious, below are links to an original Wave file, a single copy, plus copies after 5, 10, and 20 generations. To save server space I didn't post all the in-between files, but I still have them in case anyone wants to visit me in person to verify the test method! As you listen to these files you can hear how the clarity degrades slightly, especially after the 10th and 20th generations. Note that it's very difficult to calibrate a sound card for precisely unity gain, so all of the files linked on this page vary in volume very slightly. However, most differ by less than half a dB within each group."

There are other musical selections that can be downloaded from this page.

There is a 1 generation test file that is available. Ethan seems to be admitting that his level matching was not exact. And the Soundblaster that he uses was one of their cheapest models.

My measurements show an approximate 0.2 dB level difference in the left channel.
Attached Thumbnails
Click image for larger version

Name:	Men at work origional.png
Views:	11
Size:	13.8 KB
ID:	169482   Click image for larger version

Name:	Men at work pass5.png
Views:	8
Size:	13.7 KB
ID:	169490  

Last edited by arnyk; 07-19-2014 at 07:38 AM.
arnyk is offline  
post #2120 of 2920 Old 07-19-2014, 07:36 AM - Thread Starter
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 18,375
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 981 Post(s)
Liked: 547
Quote:
Originally Posted by Frank Derks View Post
You will fail when it's not 5th generation but only one like in the M&M test.
I already post the results of one generation from Etha's test:

Total: 19/22 (0.0%)[/COLOR]

Above I am showing my search for critical section. So when I tested the single generational loss (i.e. "most difficult") I knew what to listen for:

foo_abx 1.3.4 report
foobar2000 v1.3.2
2014/07/18 06:40:07

File A: C:\Users\Amir\Music\Ethan Soundblaster\sb20x_original.wav
File B: C:\Users\Amir\Music\Ethan Soundblaster\sb20x_pass1.wav


06:40:07 : Test started.
06:41:03 : 01/01 50.0%
06:41:16 : 02/02 25.0%
06:41:24 : 03/03 12.5%
06:41:33 : 04/04 6.3%
06:41:53 : 05/05 3.1%
06:42:02 : 06/06 1.6%
06:42:22 : 07/07 0.8%
06:42:34 : 08/08 0.4%
06:42:43 : 09/09 0.2%
06:42:56 : 10/10 0.1%
06:43:08 : 11/11 0.0%
06:43:16 : Test finished.

----------
Total: 11/11 (0.0%)


===========
I don't think I failed .

Quote:
Originally Posted by frank
For that particular test there is no need for trained listeners. They will fail just the same as untrained listeners.
Well, that is the theory but in the case of Ethan's version of the test, this trained listener did well . If you had the Meyer and Moran version, I would have been happy to take that test too. Lacking data there, we don't get to create data as I keep mentioning. Is this debating tactic so ingrained in us that we can't give it up even though it is most illogical as Spock would say?
Charles R likes this.

Amir
Retired Technology Insider
Founder, Madrona Digital
"Insist on Quality Engineering"
amirm is offline  
post #2121 of 2920 Old 07-19-2014, 07:56 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,387
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 762 Post(s)
Liked: 1175
Quote:
Originally Posted by amirm View Post
I already post the results of one generation from Etha's test:

Total: 19/22 (0.0%)[/COLOR]

Above I am showing my search for critical section. So when I tested the single generational loss (i.e. "most difficult") I knew what to listen for:

foo_abx 1.3.4 report
foobar2000 v1.3.2
2014/07/18 06:40:07

File A: C:\Users\Amir\Music\Ethan Soundblaster\sb20x_original.wav
File B: C:\Users\Amir\Music\Ethan Soundblaster\sb20x_pass1.wav


06:40:07 : Test started.
06:41:03 : 01/01 50.0%
06:41:16 : 02/02 25.0%
06:41:24 : 03/03 12.5%
06:41:33 : 04/04 6.3%
06:41:53 : 05/05 3.1%
06:42:02 : 06/06 1.6%
06:42:22 : 07/07 0.8%
06:42:34 : 08/08 0.4%
06:42:43 : 09/09 0.2%
06:42:56 : 10/10 0.1%
06:43:08 : 11/11 0.0%
06:43:16 : Test finished.

----------
Total: 11/11 (0.0%)


===========
I don't think I failed .


Well, that is the theory but in the case of Ethan's version of the test, this trained listener did well . If you had the Meyer and Moran version, I would have been happy to take that test too. Lacking data there, we don't get to create data as I keep mentioning. Is this debating tactic so ingrained in us that we can't give it up even though it is most illogical as Spock would say?
I wouldn't get too head up here.

Ethan's pass1 file has the worst matching of the ones I've tested (see attached analysis)

I've prepared a set of 1 pass matched files for download here: https://www.dropbox.com/s/9nuvojqahj...rk%20fixed.zip
See what you can do with that, Amir!
Attached Thumbnails
Click image for larger version

Name:	Men at work origional.png
Views:	17
Size:	13.8 KB
ID:	169546   Click image for larger version

Name:	Men at work pass5.png
Views:	14
Size:	13.7 KB
ID:	169554   Click image for larger version

Name:	Men at work pass1.png
Views:	15
Size:	13.6 KB
ID:	169562  
arnyk is offline  
post #2122 of 2920 Old 07-19-2014, 08:56 AM
AVS Addicted Member
 
Chu Gai's Avatar
 
Join Date: Sep 2002
Location: NYC area
Posts: 14,964
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 317 Post(s)
Liked: 595
Well, Amir, I am impressed with what you've accomplished. On Ethan's files, played through my iPad, I'll be damned if I could pick out a reliable difference. Not having the ability to time sync with rapid switching doesn't help matters.

Rather than keep beating each other up, maybe we can somehow look to advance this from a knowledge POV. IOW, it looks like Amir might be able to act as a teacher by sharing what his approach was and what it was that he listened for. There's obviously something he's keying in on that the majority is missing.

So WRT Ethan's files, Amir, would you please elaborate your approach and ultimately lead up to what time stamp(s) you used as the discriminating section that permitted you to have such great success in the ABX?
imagic and Randy Bessinger like this.

"I've found that when you want to know the truth about someone that someone is probably the last person you should ask." - Gregory House
Chu Gai is offline  
post #2123 of 2920 Old 07-19-2014, 09:42 AM
Senior Writer @ AVS
 
imagic's Avatar
 
Join Date: Dec 2005
Location: Philadelphia, PA
Posts: 6,362
Mentioned: 30 Post(s)
Tagged: 0 Thread(s)
Quoted: 1708 Post(s)
Liked: 3049
Laptop? Practice?

Well, I decided to give my laptop a try since Amir did so well using his. Lo and behold, I had little difficulty with the 16/32 key jangling test. Not quite perfect, but I suspect a bit more practice would get me up to perfect.

My laptop is a Sony Vaio PCG-41412L with the HD upgraded to a SSD. All audio enhancements are off. I used a pair of Sony MDR-1R headphones.

The results speak for themselves; I found a critical segment that revealed an audible difference. I've had some practice, which helped—just as Amir suggested. Now, I can pass an ABX test I previously failed. I'll tackle the 16/44 test next. Oh, and it was a piece of cake to pick out the differences in the 16/16 and 22/16 tests.

foo_abx 1.3.4 report
foobar2000 v1.3.3
2014/07/19 11:26:49

File A: C:\Users\mark_000\Downloads\keys jangling band resolution limited 3216 2496.wav
File B: C:\Users\mark_000\Downloads\keys jangling full band 2496.wav

11:26:49 : Test started.
11:27:29 : 00/01 100.0%
11:28:58 : 00/02 100.0%
11:29:46 : 00/03 100.0%
11:29:59 : 01/04 93.8%
11:30:06 : 01/05 96.9%
11:30:16 : 02/06 89.1%
11:30:26 : 03/07 77.3%
11:30:34 : 04/08 63.7%
11:30:45 : 05/09 50.0%
11:31:00 : 06/10 37.7%
11:31:10 : 07/11 27.4%
11:31:29 : 08/12 19.4%
11:31:41 : 09/13 13.3%
11:32:05 : 10/14 9.0%
11:32:20 : 10/15 15.1%
11:32:30 : 11/16 10.5%
11:32:41 : 12/17 7.2%
11:32:52 : 13/18 4.8%
11:33:07 : 13/19 8.4%
11:33:16 : 14/20 5.8%
11:33:28 : 15/21 3.9%
11:33:40 : 16/22 2.6%
11:33:58 : 17/23 1.7%
11:34:12 : 18/24 1.1%
11:34:25 : Test finished.

----------
Total: 18/24 (1.1%)
Charles R and amirm like this.

Find out more about Mark Henninger at www.imagicdigital.com

Last edited by imagic; 07-19-2014 at 09:47 AM.
imagic is online now  
post #2124 of 2920 Old 07-19-2014, 09:55 AM
Senior Writer @ AVS
 
imagic's Avatar
 
Join Date: Dec 2005
Location: Philadelphia, PA
Posts: 6,362
Mentioned: 30 Post(s)
Tagged: 0 Thread(s)
Quoted: 1708 Post(s)
Liked: 3049
So, on my laptop the 16/44 test was super easy. In fact I only needed to hear the sample once to make the right choice—after I honed in on a revealing transient. Now, I really am curious what happened, and why.

foo_abx 1.3.4 report
foobar2000 v1.3.3
2014/07/19 11:50:04

File A: C:\Users\mark_000\Downloads\keys jangling band resolution limited 4416 2496.wav
File B: C:\Users\mark_000\Downloads\keys jangling full band 2496.wav

11:50:04 : Test started.
11:51:13 : 00/01 100.0%
11:51:23 : 00/02 100.0%
11:51:34 : 01/03 87.5%
11:51:43 : 02/04 68.8%
11:51:50 : 03/05 50.0%
11:51:57 : 04/06 34.4%
11:52:09 : 05/07 22.7%
11:52:17 : 06/08 14.5%
11:52:24 : 07/09 9.0%
11:52:30 : 08/10 5.5%
11:52:38 : 09/11 3.3%
11:52:50 : 10/12 1.9%
11:52:57 : 11/13 1.1%
11:53:04 : 12/14 0.6%
11:53:13 : 13/15 0.4%
11:53:20 : 14/16 0.2%
11:53:27 : Test finished.

----------
Total: 14/16 (0.2%)

Find out more about Mark Henninger at www.imagicdigital.com
imagic is online now  
post #2125 of 2920 Old 07-19-2014, 10:15 AM - Thread Starter
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 18,375
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 981 Post(s)
Liked: 547
Quote:
Originally Posted by Chu Gai View Post
Well, Amir, I am impressed with what you've accomplished. On Ethan's files, played through my iPad, I'll be damned if I could pick out a reliable difference. Not having the ability to time sync with rapid switching doesn't help matters.
Thanks. Just like a sport, and let's agree this is a sport to see if you can beat the system, technique matters.

Quote:
Originally Posted by chu
Rather than keep beating each other up, maybe we can somehow look to advance this from a knowledge POV. IOW, it looks like Amir might be able to act as a teacher by sharing what his approach was and what it was that he listened for. There's obviously something he's keying in on that the majority is missing.
Thank you. There is no worse torture than having to keep listening to Ethan's files .

I will provide some hints but ultimately this is a technique that requires practice. It is like me explaining how to ride a bike. You need that "aha moment" when the bike runs without you falling over to learn what the verbal explanation really means. Mark just had one in his testing.

So here is what I do. I play A and B until I find the right classification. What is a classification? How is that high frequency note? A bit harsher? A bit stronger? Subdued? How is channel separation? Listen to how wide the sound becomes at peak. How is the decay? This one is really hard in some content like Ethan's. But it can be done. You have to ignore everything but how one of the notes goes down in amplitude.

The challenge then is finding the right segment where one of these classifications reliably works. If you have enough practice, in a few trials you can tell which one is in play or none.

What I find very useful is to NOT have to play A and B. In other words, you need to memorize what the difference is and confidently listen to X and Y. If you go back and forth to A and B, you will second guess yourself. If you feel that you can't tell the difference between X and Y, then your classification may be wrong. Or, as differences get smaller, you forget which is A and B.

Quote:
Originally Posted by chu
So WRT Ethan's files, Amir, would you please elaborate your approach and ultimately lead up to what time stamp(s) you used as the discriminating section that permitted you to have such great success in the ABX?
Courtesy of foobar forgetting the start/finish pointers, I don't have the data to give you on specific parts. The technique I used though I have explained above.

Amir
Retired Technology Insider
Founder, Madrona Digital
"Insist on Quality Engineering"
amirm is offline  
post #2126 of 2920 Old 07-19-2014, 10:21 AM - Thread Starter
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 18,375
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 981 Post(s)
Liked: 547
Quote:
Originally Posted by imagic View Post
So, on my laptop the 16/44 test was super easy. In fact I only needed to hear the sample once to make the right choice—after I honed in on a revealing transient. Now, I really am curious what happened, and why.

foo_abx 1.3.4 report
foobar2000 v1.3.3
2014/07/19 11:50:04

File A: C:\Users\mark_000\Downloads\keys jangling band resolution limited 4416 2496.wav
File B: C:\Users\mark_000\Downloads\keys jangling full band 2496.wav

11:50:04 : Test started.
11:51:13 : 00/01 100.0%
11:51:23 : 00/02 100.0%
11:51:34 : 01/03 87.5%
11:51:43 : 02/04 68.8%
11:51:50 : 03/05 50.0%
11:51:57 : 04/06 34.4%
11:52:09 : 05/07 22.7%
11:52:17 : 06/08 14.5%
11:52:24 : 07/09 9.0%
11:52:30 : 08/10 5.5%
11:52:38 : 09/11 3.3%
11:52:50 : 10/12 1.9%
11:52:57 : 11/13 1.1%
11:53:04 : 12/14 0.6%
11:53:13 : 13/15 0.4%
11:53:20 : 14/16 0.2%
11:53:27 : Test finished.

----------
Total: 14/16 (0.2%)
Well done! You are now using your ears as an instrument. You are learning to analyze and parse components of the audio stream and not get lost in the "music." The job here isn't to listen to music. It is to act as a measurement system for degradations.

You shoot video in your day job, yes? You may know then that one of the techniques for evaluating picture quality is turning off color and only looking at luminance. This is why professional monitors have that switch that turns off color. Color is deceiving in the way it makes the image "pretty." Take it out and distortions become more visible as your mind is no longer being distracted by the pretty images. Once you learn what the distortions are, then you can also see them in color images.

Amir
Retired Technology Insider
Founder, Madrona Digital
"Insist on Quality Engineering"
amirm is offline  
post #2127 of 2920 Old 07-19-2014, 10:24 AM - Thread Starter
AVS Addicted Member
 
amirm's Avatar
 
Join Date: Jan 2002
Location: Washington State
Posts: 18,375
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 981 Post(s)
Liked: 547
One note of caution guys: as you become a trained listener, you get to suffer a lot more when listening to less than perfect music. You will keep hunting for distortions whether you are in a test scenario or not. I do that routinely as I listen to radio in my car for example. While in balance I am happy to have this skill, it has its down sides. I was once in Newport Beach area shooting wildlife pictures and the rental car had XM radio. Man, I could not listen to the darn thing. The compression artifacts were a constant annoyance. The service has millions of customers so clearly this is not an issue for untrained listeners.

Amir
Retired Technology Insider
Founder, Madrona Digital
"Insist on Quality Engineering"
amirm is offline  
post #2128 of 2920 Old 07-19-2014, 10:36 AM
AVS Addicted Member
 
Chu Gai's Avatar
 
Join Date: Sep 2002
Location: NYC area
Posts: 14,964
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 317 Post(s)
Liked: 595
Do you have any plans to try this with speakers, Amir?

"I've found that when you want to know the truth about someone that someone is probably the last person you should ask." - Gregory House
Chu Gai is offline  
post #2129 of 2920 Old 07-19-2014, 10:43 AM
AVS Addicted Member
 
Chu Gai's Avatar
 
Join Date: Sep 2002
Location: NYC area
Posts: 14,964
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 317 Post(s)
Liked: 595
Oh, and I see that Ethan has some more files that I'm guessing are more challenging.

http://ethanwiner.com/loop-back.htm

"I've found that when you want to know the truth about someone that someone is probably the last person you should ask." - Gregory House
Chu Gai is offline  
post #2130 of 2920 Old 07-19-2014, 10:50 AM
AVS Addicted Member
 
arnyk's Avatar
 
Join Date: Oct 2002
Location: Grosse Pointe Woods, MI
Posts: 14,387
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 762 Post(s)
Liked: 1175
Quote:
Originally Posted by amirm View Post
One note of caution guys: as you become a trained listener, you get to suffer a lot more when listening to less than perfect music. You will keep hunting for distortions whether you are in a test scenario or not. I do that routinely as I listen to radio in my car for example. While in balance I am happy to have this skill, it has its down sides. I was once in Newport Beach area shooting wildlife pictures and the rental car had XM radio. Man, I could not listen to the darn thing. The compression artifacts were a constant annoyance. The service has millions of customers so clearly this is not an issue for untrained listeners.
While my hearing is not in the same league as yours appears to be Amir, I similarly cannot stand to listen to XM or Sirius in its current state. When they first came out they had far, far fewer different channels and thus had far more bitrate available to each of them, and it didn't sound nearly as bad to to me.
arnyk is offline  
Closed Thread Audio Theory, Setup, and Chat

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