Audio Blind Testing

[fas42]

37 minutes ago, mansr said:

Maybe. Either way, there's nothing to be learned from him.

 

If at the very least people would consider that recordings have far more on them then is usually given credit for, and that improved playback can reveal this - I would be pleased if this tiny crumb was taken notice of ... .

[gmgraves]

33 minutes ago, sandyk said:

 George

 I base what I have said on the DIY gear that I use here, where I have been able to optimise various areas of the designs, so that very low level ambience cues etc. aren't lost in the low level "digital haze" or in the Analogue stages S/N and channel separation.

For me, it's mainly about further improving the SQ of my gear in order to keep enjoying high quality music reproduction as my hearing further slowly degrades due to advancing age (79 soon) and existing Industrial type hearing damage from 43 years with Telstra which resulted in an Acoustic Neuroma pressing on my right ear canal, which is sensitive to my Blood Pressure at the time.

Alex

 

Well, Alex, I have always said that I admire you as a true audiophile of the type that we all used to be back in the day. By that I mean that to be an audiophile when I was a kid, we all had to roll our own at some level. We had to either build kits from Dynaco or Heath or (in the US) Allied Radio, Chicago (Dick Smith Electronics in Oz) or we had buy components and blank chassis boxes and build our own from scratch (Popular Electronics and the Audio Amateur here in the USA carried DIY audio projects on a monthly basis). When I was a kid, we even had to make our own (totally unreliable) interconnects from raw coax and those awful tin-plated RCA plugs with no strain relief. That aspect of this hobby which used to be a major part, is mostly gone from the hobby now as people buy complete components from well established manufacturers, but I'm glad to see that there are still people like you around with the space and the ambition and enthusiasm to continue that great tradition. I spent most of my teen years bent over my garage workbench inhaling the fumes from resin-core 60/40 solder. It was all tubes in those days and whether building a 20-watt mono amp from scratch, or a kit stereo multiplex adapter for FM stereo, or a nuvistor RF amp for my Eico HFT-90 FM tuner, it was fun and I don't think the hobby has ever been as satisfying since I stopped doing that when my career as an electronics engineer began. 

In the SF Bay area, there is a yearly event called the "Burning Amp Convention" (take-off on the "Burning Man" event in Nevada). I used to attend that every year and I would marvel at the home-made gear. All of it was interesting and some down-right inspiring and even beautiful! I remember hearing the best pair of speakers I've ever heard were built by a young fellow I met at Burning Amp! They were made out of cherry wood, and they consisted of a pair of huge boxes, each containing TWO 15-inch Altec Lansing woofers and on top of each woofer box was wide, curved piece of cherry with a pair of forty-some inch long full-range ribbons and a pair of smaller ribbon tweeters, all from a company named Bohlender-Graebener mounted in tall,  wide planks and a custom low-level crossover for bi-amping. To this day, I still fondly remember how great those speakers sounded (and looked) and the guy built them from scratch, including the woodwork in his garage! I fully intended to build myself a pair of them by getting a furniture builder to make the cabinets for me, but I never did it. The cabinets would have cost me about a grand, and the big ribbons ribbons were about $600 each while the tweeters were another $100, both from a US company named Parts Express. Alas, they no longer carry either the tweeters or the 40" long ribbons and I've heard that Bohlender-Graebener went belly-up when one of the partners died. The result would have been a pair of world-class speakers for less than three thousand dollars. One couldn't buy anything that sounded as good or imaged as well for 20 times that much on the commercial market! DIY Hi-F obviously can be very rewarding.

[sandyk]

8 minutes ago, gmgraves said:

I spent most of my teen years bent over my garage workbench inhaling the fumes from resin-core 60/40 solder.

 

I spent much of my working life doing similar.

Thank Heavens Telephone Exchanges had very good air conditioning , especially near the work benches !

[STC]

2 hours ago, fas42 said:

 

That's interesting --- that backs up what you said in the previous post about perceiving a "hollow" in the middle with stereo - your brain works a litle differently from others, and highly likely mine.

 

I think you are interpreting what I said to suit your assertion. I grew up in an era where the mono sound was more common than stereo. Naturally, in mono, everything is concentrated in one channel and there wasn't any soundstage with one speaker. It is natural to perceive thinness when you listen to stereo to perceive the sound being less dense than mono. Vocals or single instruments (mono recordings) will always sound better with a single speaker.

 

This observation may not be agreeable to many as we have also learned to adapt to the stereo sound. Although they may have their own flaws, we learned to accept those flaws as natural and identify the sound as a reproduction of hifi playback. It depends on your preference.

 

When I was obsessed with hif setup, I had visitors with same interest. Now, those visitors stop coming because to them the sound is no longer hifi sound. It sounded large and often associated to the experience in IMAX cinema sound. Admittedly, most of them never heard unamplified sound in the concert hall. Now, I get new kind of visitors who like concert hall sound or musicians who listen to music from different perceptive. It all depends on one's exposure and characteristics they listen for.

[fas42]

8 minutes ago, STC said:

When I was obsessed with hif setup, I had visitors with same interest. Now, those visitors stop coming because to them the sound is no longer hifi sound. It sounded large and often associated to the experience in IMAX cinema sound. Admittedly, most of them never heard unamplified sound in the concert hall. Now, I get new kind of visitors who like concert hall sound or musicians who listen to music from different perceptive. It all depends on one's exposure and characteristics they listen for.

 

I'm with you on this - the "bigness" of the presentation may disturb some; they want to "feel in control" of what they're listening to, that it doesn't overwhelm them, emotionally. By contrast, I love this aspect of good reproduction, that it's far bigger than you are - you "live" in the energy of the music at that moment.

 

A reason why I enjoy rock recordings; no musician myself, but have spent time in the middle of a group rehearsing; the tremendous bite and drive of the instruments going right next to one is a marvelous feeling - something which is always on tap with a good system.

[STC]

16 hours ago, esldude said:

I don't think it is in the links I provided.  Those were more to show how under anechoic conditions people have a tendency to hear forward sounds in their head rather than externally if they can't move their head.  Research about such things goes back to at least the 1940s.  

 

Here is the experiment done in anechoic room using two loudspeakers determining directional localisation of the

sound in stereophony (stereo). 

 

Link

 

[STC]

On 1/14/2018 at 9:16 PM, esldude said:

So are you saying because many claim it they are managing to recreate soundfields?  

 

I have said before binaural just doesn't work for me.  This one didn't either.  Sounded much better and more dimensional over my speakers actually.  

 

I found a reference where some users were having difficulties with 3D sound via heaphones which may be relevant to you. If you are using Realtek HD Audio in your pc, try disabling that and use stereo only mode. Please let me if you hear nay difference. Thanx.

[esldude]

20 minutes ago, STC said:

 

I found a reference where some users were having difficulties with 3D sound via heaphones which may be relevant to you. If you are using Realtek HD Audio in your pc, try disabling that and use stereo only mode. Please let me if you hear nay difference. Thanx.

Thank you for the suggestion.  I avoid Realtek HD.  I mostly use recording interfaces over USB.  Typically I use a Focusrite Forte for headphone listening.  I'll either use playback software with WASAPI or ASIO output so it is bit perfect.  I don't hear what others describe in headphone listening with Binaural recordings.  

[esldude]

52 minutes ago, STC said:

 

Here is the experiment done in anechoic room using two loudspeakers determining directional localisation of the

sound in stereophony (stereo). 

 

Link

 

I'll read that later this evening.  Thank you. 

[Ralf11]

3 hours ago, gmgraves said:

He only writes in vague generalities, never with anything specific. What equipment does he use in order to get this "perfect" recreation of the sound field? How does he manage to get this great imaging, even from multi-track, multi-miked recordings that don't have any real image, just a bunch of pan-potted instruments artificially lined up from right-to-left? What is this perfect amplifier he has that's so much better than the ones everyone else here is using? He never says. What are we to glean from this lack of disclosure? 

 

I glean that he is like GUTB

[STC]

10 hours ago, fas42 said:

I'm with you on this - the "bigness" of the presentation may disturb some; they want to "feel in control" of what they're listening to, that it doesn't overwhelm them, emotionally.

 

This statement maybe true.  A few discerning audiophiles I know usually listens at low volume where they "claim" they hear the separation and low level better. Thanks.

[pkane2001]

13 hours ago, sandyk said:

 

Perhaps you may get some valuable suggestions from Dale in the Headphone area of this forum ?

 I don't always agree with him, but he does have an in-depth knowledge of many headphone amplifiers and headphones. 

 

HD650s are pretty high impedance headphones (over 300 Ohms), so perhaps they work well with my DAC output, as is, but I'll ask Dale and see what he can recommend.

 

[pkane2001]

42 minutes ago, STC said:

A few discerning audiophiles I know usually listens at low volume where they "claim" they hear the separation and low level better. 

 

Those are the few remaining ones that still have their hearing intact. The rest of us listen at much higher/dangerous levels

 

[STC]

19 minutes ago, pkane2001 said:

 

Those are the few remaining ones that still have their hearing intact. The rest of us listen at much higher/dangerous levels

 

 

IMO, an average of 75dB with peaks touching 90dB to 100dB is a comfortable level and worth having hifi system to play them at that level. This guys probably in the 60dB range as the air conditioning noise could be heard. 

 

BTW, Interestingly they also use headphones to rock. 

 

 listening level right now, a pop soundtrack ( non English).  

 

And this is Testament - Candide: Make our garden grow ( Turtle Creek Carol). 

 

 

[adamdea]

Forgive this return to the original topic, but I thought I would just throw in that in one of the early AES papers about ABX by Clark IIRC there is a reference to an experiment being done with a known amount of distortion in which it was apparently demonstrated that ABX gave more accurate results than the vaunted audiophile gold standard of long term listening. I find it disappointing that this is not more widely known although of course it maps with what is known about audio memory.

Over the years I have on several occasions pointed this out to people who insist that ABX tests somehow mask otherwise audible differences; none of them were in the lightest bit interested in the evidence. Wouldn't you know it?  

[sandyk]

1 hour ago, adamdea said:

Over the years I have on several occasions pointed this out to people who insist that ABX tests somehow mask otherwise audible differences; none of them were in the lightest bit interested in the evidence. Wouldn't you know it?

 I don't know about others, but I use both long and short term listening.

 In the majority of cases, after I make equipment modifications, I am often  able to tell in less than 30 seconds what difference there is . I then use longer term listening to confirm my initial impressions, unless I am 100% certain that the modification was not an improvement, in which case the modification is reversed, usually the next day after another quick listen..

[fas42]

1 hour ago, sandyk said:

 I don't know about others, but I use both long and short term listening.

 In the majority of cases, after I make equipment modifications, I am often  able to tell in less than 30 seconds what difference there is . I then use longer term listening to confirm my initial impressions, unless I am 100% certain that the modification was not an improvement, in which case the modification is reversed, usually the next day after another quick listen..

 

Yes, that's how it is - 30 secs is quite enough time if the modification has immediate impact on the sound. Sometimes changes require longer times to stabilise in terms of audible effects - this is normally because materials alter in their behaviours over significant periods of time. That can be a great nuisance, because if it never becomes a permanent, fixed characteristic it means always "waiting" to get the best sound - like having a TV where the colours take hours to be "right" - I would not consider an audio rig properly sorted if it took more than 5 minutes to deliver an acceptable standard.

 

People seem to have terrible trouble with the concept that distortion varies in its nature - some flavours can exist at monstrous levels, and we are quite comfortable with the listening; another manifestation can be present at minute levels, and we won't be able to stand listening to it for more than a few minutes. IOW, in audio playback distortion ... is ... not ... just ... numbers.

[esldude]

9 hours ago, adamdea said:

Forgive this return to the original topic, but I thought I would just throw in that in one of the early AES papers about ABX by Clark IIRC there is a reference to an experiment being done with a known amount of distortion in which it was apparently demonstrated that ABX gave more accurate results than the vaunted audiophile gold standard of long term listening. I find it disappointing that this is not more widely known although of course it maps with what is known about audio memory.

Over the years I have on several occasions pointed this out to people who insist that ABX tests somehow mask otherwise audible differences; none of them were in the lightest bit interested in the evidence. Wouldn't you know it?  

Thinking of this maybe:

 

http://www.aes.org/e-lib/browse.cfm?elib=5549

 

An open version of the whole article was once available, but no longer is.  

 

Some excerpts from it are here:

https://audiosciencereview.com/forum/index.php?threads/aes-paper-digest-sensitivity-and-reliability-of-abx-blind-testing.186/

 

In short, a box that generated 2.5% distortion was sent home with audiophiles for long term listening.  Some were bypasses and some distorted. Each audiophile needed to choose if the box were clean or not.  The results of this weeks long auditioning were essentially random.  Using quick switching the same group of people could identify the dirty/clean condition reliably.  Plus they could reliably identify a smaller amount of distortion when doing quick switching.   So much for long term auditioning being the gold standard.  Yes it should be better known.  Except so many don't want to know it. 

 

And just an idea of how long such things have been going on check out this NYT article from 1987. 

 

http://www.nytimes.com/1987/02/01/arts/sound-amplifiers-test-their-listeners.html

 

[STC]

Another good reference to understand biases when judging audio quality listening tests is here. Even repeating same samples several times can and often resulted in different sound quality.

[adamdea]

10 hours ago, fas42 said:

People seem to have terrible trouble with the concept that distortion varies in its nature - some flavours can exist at monstrous levels, and we are quite comfortable with the listening; another manifestation can be present at minute levels, and we won't be able to stand listening to it for more than a few minutes. IOW, in audio playback distortion ... is ... not ... just ... numbers.

I have no idea who those people may be. I'm going to blow their minds here but with numbers we can analyse more than one variable at a time, so for example you can look at whether distortion is harmonic, and if so which harmonic, as well as looking at the level. 

 

Equally with the great largely-imaginary bogeyman of jitter, we can analyse jitter frequency rather than just looking at the total integrated jitter. And this can be useful. 

 

Of course listening is not numbers but that's  a categorical distinction. It has no bearing on whether numbers can specify levels of audibility of particular spuriae, or even what is required for transparency.

 

 

[adamdea]

2 hours ago, STC said:

Another good reference to understand biases when judging audio quality listening tests is here. Even repeating same samples several times can and often resulted in different sound quality.

Thanks, very interesting article.

Picking up on your second sentence, I think that there is one sticking point which I think people do not or will not grasp: a properly functioning human being will not necessarily experience to identical auditory events in the same way. To put it another way, the experience of hearing is not purely the result of the input of various sound pressure waves, not by a long way. Thinking two things sound different (when they aren't) is not akin to a hallucination: it is the result of an integrated sensory system (and other things). IIRC one of Tom Noussaine's papers for the AES showed that there was an inherent bias towards thinking things sound different (even if they don't).

 

If one thing characterises audiophiles IMHO it is a tendency to exaggerate the importance of hifi kit on the listening experience. Hence all the spooky results along the lines, "the only thing I had changed was [the cable/support/orientation of the screws/blah blah] so the only explanation can be that it was  [the cable/support/orientation of the screws/blah blah]". No, it never is the only explanation.

 

In one sentence : you are not a sound quality measurement device.

 

 

If one grasps this point one will then understand why one can only draw conclusions about purely sonic qualities from what people think things sound like by using carefully controlled experiments.

[adamdea]

4 hours ago, esldude said:

Thinking of this maybe:

 

http://www.aes.org/e-lib/browse.cfm?elib=5549

 

An open version of the whole article was once available, but no longer is.  

 

Some excerpts from it are here:

https://audiosciencereview.com/forum/index.php?threads/aes-paper-digest-sensitivity-and-reliability-of-abx-blind-testing.186/

 

 

Yes, that's it. I was alerted by that audiosciencereview link and actually joined the AES, just to get hold of this and other papers around the time! 

[sandyk]

8 minutes ago, adamdea said:

In one sentence : you are not a sound quality measurement device.

 

 

In  one sentence : You are far from an expert in this area either !

Your average Audiophile doesn't give a damn about playing your silly DBT games. 

14 minutes ago, adamdea said:

IIRC one of Tom Noussaine's papers for the AES showed that there was an inherent bias towards thinking things sound different (even if they don't).

 

Yeah, well know-it alls like yourself also claimed that .wav files with identical checksums can't sound different.

 Yet, veteran Hi Fi Reviewer and well respected E.E. Martin Colloms from Hi Fi Critic magazine, several years ago correctly performed 6 separate sessions of your "Gold Standard" DBTs on some comparison .wav files that I supplied, and 6 out of 6 separate sessions of 8 repeats were all POSITIVE ! The differences were due mainly to the PSU area of the originating P.C.

 http://www.enjoythemusic.com/hificritic/vol6_no1/audio_networking.htm   gives a brief rundown, but several different threads in HFC forum gave quite a lot more detail about several of the various sessions.  

 

I do not intend discussing this further with yet another person who appears to have blown in just to ridicule Audiophiles and show how clever he is !

BYE !

[adamdea]

Curious isn't it that the merest reference to really basic perceptual science tends to rattle people. 

But really Sandy- you just aren't a sound quality measurement device. And you don;t have to take it from a non expert like me, just read any basic text in the field. The only issue is whether one is actually interested in the answer or not. 

As for Martin Colloms and the WAV files-  very funny. Ultimately the test of experimental credibility is repeatability. Nothing unusual there. See cold fusion.

[sandyk]

14 minutes ago, adamdea said:

As for Martin Colloms and the WAV files-  very funny. Ultimately the test of experimental credibility is repeatability. Nothing unusual there. See cold fusion.

 6 separate sessions of typically 8 repeats each , which ALL resulted in POSITIVE results showed that there were actual differences.

For people like yourself though , it wouldn't matter if there were 1 million positive results, as you will never accept them because you believe it isn't possible.

 M.C. is far from the only person to verify my results, including several prominent C.A. members and famous Recording and Mastering Engineer Barry Diament.  Use the Search facility if you doubt this, but of course you aren't interested, because you already know that Audiophiles imagine every difference they report hearing !

Read other areas of the forum, and you will also see quite a few reports by well qualified members, that even the O.S. itself can result in small audible differences with "bit perfect" files.