jim williams/audio upgrades

[bluebird]

The gold standard for scientific listening tests is double-blind with statistically significant results.  These are painfully difficult to execute properly and even difficult to participate in, but a number of popular myths have been disproved by those willing to invest the time and effort.

Amen,

I work for a commercial  studio that is owned by a huge media company. About 4 or 5 years ago they started an archiving program and tasked the studio with choosing the best tape decks and digital converters to use.
I built an ABX box and performed many many tests with the staff engineers, mastering engineers, and guest engineers. At the end of it all I was blown away at what people could NOT hear. What people told me they could easily hear at the beginning of the test as they clicked the A/B button, but when they turned the A/B lights out and hit the random button. Everything changed.

For instance. We used the master 1/2" tape from a well known adult contemporary artist and recorded it into pro tools at 44.1K and 192K. The music had a lot of dynamics and quiet passages and was recorded very well. We brought the 44.1K file into the 192K session with SRC. Lined up the files which were almost perfectly the same length, props to the ATR we were using, and went into the A/B box with outputs 1+2 and 3+4 of the same Digidesign 192K converter.

Out of about 6 different engineers guess who could tell the difference between the two files? Not one of them. Not only could they not tell the sample rate difference but remember there were two different paths out of the converters. If you have ever looked at all the components on one of those Digi D/A cards its seems someone should have detected some difference. Same thing happened when testing converters. They were all high end converters but no one could tell the difference between any of them! These were professionals and very talented people. There were a bunch of tests we did that finally led to having to just agree on equipment/sample rate/ procedure according to what was most peoples gut feelings were because 90% of the tests didn't reveal the clear path to take.

Now I know these were not clinical tests in a laboratory so I'm not making any official statements here but for me it was a real eye opener.
I can't give all the details or show any proof of this and wouldn't expect anyone to just take my word for it, but I suggest everyone who really wants to know whats what, build themselves a proper ABX box.

Ian

 

[critterkllr]

Kind of weird seeing a discussion go the gearslutz route on groupdiy. There will never be a universal agreement on the two separate groups of opinions here. I think everybody should keep two things in mind regardless of where you sit.

The meter reader vs. golden ear debate probably pre-dates Gearslutz and internet forums. I recall writing about it in the early '80s)

1) Science hasn't been able to explain everything yet. Keep your mind open to any possibility and understand that even an inaccurate belief can have positive benefits.

Huh? The mechanical aspects of sound is very mature technology and well understood by those skilled in the art (physics, acoustics, etc). 

The only somewhat fuzzy area (IMO) is how humans post process the audio data stream.  The brain routinely takes in massive amounts of data that it filters while discarding the vast majority. Different people have different personal filters based on experience and training, but even this has been studied for decades.

I used some basic psycho-acoustic relationships when designing noise reductions back in the '70/80s (masking, change perception time constants, etc) . Not widely discussed science, but science nonetheless. 

I'm playing devil's advocate a bit.

My statement has two parts. The first is a blanket statement to say that there is always the possibility that something isn't completely understood. Way may be 99% there, but there is a possibility that something about audio or psycho-acoustics that we don't know yet. So many of these discussions devolve into arguments and a stalemate because everybody is 100% sure they are right. For me, keeping my mind open to any possibility is the best way for me to make progress.

The other half of my statement comes from being a musician. Certain "mojo" items give people confidence and/or excitement. If that's what it takes to further their art or business, then it is a positive thing. From my experience, there is value to a placebo when it comes to performance. Of course, that doesn't mean that we can't all accept that it is in fact a placebo.

 

[JohnRoberts]

To execute a properly controlled listening test you need more than a ABX box (I remember when those first came out, I think Dave Clark was involved with developing them).

The two sources need to be matched for amplitude to a small fraction of a dB...

The frequency response needs to be matched to a fraction of a dB.

The person operating the switch needs to be blind to which source is playing so he doesn't telegraph any knowledge and influence the result.

=======

I've done several single blind tests with good results, surprises and all. The worst listening test I was subjected to was performed while I was out of town on business  , and I fear it was not adequately controlled. Millions of dollars of advertising create some expectation bias, and I was kind of blind-sided by a result I couldn't vet or defend.  No do-overs in business.  I won most such debates when I was actually present.  8)

JR

 

[bluebird]

To execute a properly controlled listening test you need more than a ABX box (I remember when those first came out, I think Dave Clark was involved with developing them).

The two sources need to be matched for amplitude to a small fraction of a dB...

The frequency response needs to be matched to a fraction of a dB.

The person operating the switch needs to be blind to which source is playing so he doesn't telegraph any knowledge and influence the result.

Yes, all those criteria were met. the test tape had calibration tones and I had a potable AP. I think overlooking proper volume matching is why people think they can tell the difference between components, sample rates, cables etc. I can easily hear a half db difference between two stereo tracks of full program material. I think most audio people can.

Do you mean the Dave Clark that use to work for Neve? If so, he is my co worker!

 

[JohnRoberts]

Do you mean the Dave Clark that use to work for Neve? If so, he is my co worker!

I don't know, the Dave Clark I am thinking of was active in a local AES chapter in detroit Mich area, back in the early '80s.

He was known in the high end audio market back then  for his work with ABX.

JR

 

[Bowie]

  Maybe I'm missing something here...  If I close my eyes and rapidly toggle the solo button a few times in my DAW, I no longer know which track is which.  I can then switch back and forth with a simple click to audition 2 separate mixes of a recording (let's say they were run through the same piece of gear, but with a different tube installed) and make a judgement of the sonic features of each track.  Then, look to see which was which and make a note.  Rinse and repeat to see if one is consistently being identified as having specific sonic properties.  Assuming the tracks were handled properly and of even dB, how is this not acceptable for me when I'm building gear, auditioning parts, making music, etc?  I don't understand why I need a neutral 3rd party in the room or a black box if my interest is finding potential differences. 

  It sometimes feels as if people are discouraged from using their ears, which is a shame.  I've seen many instances where technical advice was given based on theory amd test results, but it sounded bad in practice because the musical context of the "upgrade" wasn't properly evaluated.  It was done based on measurements.  I see this a LOT in tubes, where someone goes with specs and measurements, and doesn't listen to what's actually happening to the music.  I rely heavily on both my test equipment and my ears, so I really have no dog in this other than the truth.  Though it may exist, I still haven't seen a test that will tell me if one vocal sounds smoother, or more detailed than another as quickly and decisively as my ears do.

 

[TheJames]

  Maybe I'm missing something here...  If I close my eyes and rapidly toggle the solo button a few times in my DAW, I no longer know which track is which.  I can then switch back and forth with a simple click to audition 2 separate mixes of a recording (let's say they were run through the same piece of gear, but with a different tube installed) and make a judgement of the sonic features of each track.  Then, look to see which was which and make a note.  Rinse and repeat to see if one is consistently being identified as having specific sonic properties.  Assuming the tracks were handled properly and of even dB, how is this not acceptable for me when I'm building gear, auditioning parts, making music, etc?  I don't understand why I need a neutral 3rd party in the room or a black box if my interest is finding potential differences.  "

I do exactly this all the time.  I find it to be very helpful.  Does this sound better with X or Y?  Which do I prefer?  It's easy to test. 

The only bias is that you KNOW when the change happened (not entirely unbiased and prone to skew results).  You just have to be real honest with yourself.

One other piece of audio "quality" that I've found interesting is the exposure aspect.  It's pretty fascinating to me how the improvement a piece of gear offers isn't necessary immediately noticed on installation, but is instantly noticeable upon removal.  OR...If you've used a piece of gear for a long time, and then you "mod" it you'll more readily hear the difference.

 

[bluebird]

I don't understand why I need a neutral 3rd party in the room or a black box if my interest is finding potential differences. 
It sometimes feels as if people are discouraged from using their ears, which is a shame

Not sure if this was directed at me, but I'm definitely not discouraging you from using your ears! I 100% make my living by using my ears and making subjective decisions on what I hear. And my ABX box is not some mysterious "black box". Just some relays and control circuitry. Anyone can make one. I too have sat there with the solo button on my DAW. I have no problem with that. Although you sometime need a hardware unit to compare things like converters.

I have noticed some interesting psychoacoustic phenomena clicking around in a DAW though. Like clicking bypass on an plugin EQ with a slight bump in the high end. I'll hear the bump while watching the graphic, until I get more extreme with the boost and realize the plugin was not on the track I was listening to! My brain automatically augmented the sound I was hearing to match what I was use to seeing to a certain point. It can be hard to find the truth sometimes and it pays to be extra careful and thorough when searching for it.

 

 

[Bowie]

I don't understand why I need a neutral 3rd party in the room or a black box if my interest is finding potential differences. 
It sometimes feels as if people are discouraged from using their ears, which is a shame

Not sure if this was directed at me, but I'm definitely not discouraging you from using your ears!

No, no one in particular.  It was more in response to the general sentiment that stretched across a few different posts (I didn't pay attention to which members said what).
  In general though, this forum has the greatest harmony of any place I've been in terms of members being open minded enough to experiment, but smart enough to do some critical listening.  And, I love it for that.

 

[PRR]

> this forum has the greatest harmony

So I am wondering why we are trash-talking this guy Jim.

3 pages now.

 

[CJ]

don't want the same sound for each song on the radio, need a different mix now and then,

all about taste, and tunes!

what is the #1 of all time in Japan?  Beautiful Sunday - Daniel Boone, wtf, over?

i like the early 70's, Garden Party, Joy to the World, Come and Get It, Midnight Confession, To Late to Turn Back Now, The First Time,  I Feel the Earth Move, Question, That's Life, Fire and Rain, Snowbird, City of New Orleans, Morning Has Broken, Every Thing I Own, Summer Breeze, Tumblin Dice, Bang a Gong, Don't Pull Your Love Out, Sweet City Woman, Get Ready, In the Summer Time, I Want to Take You Higher(finally getting royalties), Spirit in the Sky,

variety, talent, tape!

what was the slew rate of that meat locker that Spector used for an echo chamber?

 

[thermionic]

They had Keith Richards on Desert Island Discs this morning. I believe the Beeb keeps an archive of all programs, stream-able indefinitely. Worth checking out.

Re: JW. Unlike in the hi-fi world (is there a more cut-throat, sociopath-infested industry?) where there's no data to prove whether an 'upgrade' works (in fact, where independently-verified data exists, many 'upgrades' actually degrade performance...), at least JW uses his AP to show his modifications improve THD / Noise etc, so power to him.

 

[JohnRoberts]

> this forum has the greatest harmony

So I am wondering why we are trash-talking this guy Jim.

3 pages now.

I don't think we were trash talking him per se, but it seems a little untoward to talk about people behind their  back. He can be engaged directly over at Gearslutz and generally answers reasonable questions.

This discussion has evolved-veered into the broader question of equipment upgrades and sound technology (I think).. 

JR

 

[Phrazemaster]

I'm certain there's something to the audio placebo effect, otherwise known as psychoacoustics. What we want to hear, or think we will hear, can affect what we perceive.

Because hearing is, after all, a function of the mind and the instrument (ears) - they are inseparable.  It's not just the ears; it's the mind interpreting the results. This means all kinda effects, including indigestion, can affect how you "hear" something since hearing is a participatory function of several systems of the mind and body. Also as someone pointed out, hearing can even be different in the morning than in the evening.

I agree the double blind test is the ultimate arbiter of "truth" in any given situation. But even just looking at the physics of the situation, it's hard not to believe that different components will yield different results. If you wanna get all EE on me, then we can discuss the transfer function of each component.

While a perfect capacitor "shouldn't" yield different results audibly, the fact jack is using different dielectrics means the space charge moves through a different medium. Water and air have different indexes of refraction, which means light travels through them differently. Why is it a stretch to believe that different dielectrics affect space charges within capacitors differently?

Ultimately, it's your gear. You get to use what components you like.

So, regardless of components, "if it sounds good, it's good."

Isn't that what's this is all about? Who cares how we arrive at our opinions. A religion this ain't.

* Please send a small donation to my non-prophet that promotes peace and unity amongst EE's and AE's. We worship at the temple of SOUND.

 

[JohnRoberts]

I'm certain there's something to the audio placebo effect, otherwise known as psychoacoustics. What we want to hear, or think we will hear, can affect what we perceive.

it's called perception bias.

Because hearing is, after all, a function of the mind and the instrument (ears) - they are inseparable.  It's not just the ears; it's the mind interpreting the results. This means all kinda effects, including indigestion, can affect how you "hear" something since hearing is a participatory function of several systems of the mind and body. Also as someone pointed out, hearing can even be different in the morning than in the evening.

I once noticed a significant difference before and after a short coffee break (at a friends recording studio). Drinking some crappy vending machine coffee (blood sugar, caffeine, maybe some second hand cigarette smoke?). I was tweaking a studio delay/flanger at the time.  I never trusted my ears the same  after that.

I agree the double blind test is the ultimate arbiter of "truth" in any given situation. But even just looking at the physics of the situation, it's hard not to believe that different components will yield different results. If you wanna get all EE on me, then we can discuss the transfer function of each component.

While a perfect capacitor "shouldn't" yield different results audibly, the fact jack is using different dielectrics means the space charge moves through a different medium. Water and air have different indexes of refraction, which means light travels through them differently. Why is it a stretch to believe that different dielectrics affect space charges within capacitors differently?

I've had this discussion before... Don't read too much into micro-electronic effects, when we have millions/billions of holes and electrons moving together it's not the individual movement, but mass or bulk movement that we experience, If a physical characteristic affects them all uniformly, it will show up in the mass measurements, otherwise it is just part of the noise floor (or less).

Ultimately, it's your gear. You get to use what components you like.

So, regardless of components, "if it sounds good, it's good."

Isn't that what's this is all about? Who cares how we arrive at our opinions. A religion this ain't.

* Please send a small donation to my non-prophet that promotes peace and unity amongst EE's and AE's. We worship at the temple of SOUND.

I consider reference path audio electronics a scientific pursuit with objective standards. Effects design is more subjective.

JR

 

[jackies]

Nice to see a good discussion with many good people sharing their educated opinions!
Even though it moved off of it's title topic into more of "upgrades viability" discussion.
Basically, the two main points are being discussed, the scientific/electronic/measurable benefits (or lack thereof) and audible/noticeable/perceivable subjective improvements.
I think one important aspect is being omitted - "hot rodding" and doing it "because we can".
So if I put big shiny wheels on my car and a stiffer suspension, let's be honest, the ride quality suffers, but I still feel good about it because the wheels were very expensive and even more so the suspension, not to mention how much I paid the mechanics to install it all. Why can't I have opamps in my mixer that cost more than the mixer? Can I use a $50 capacitor instead of $0.5 capacitor, which would most probably do the same job? You bet I can!

 

[JohnRoberts]

Nice to see a good discussion with many good people sharing their educated opinions!
Even though it moved off of it's title topic into more of "upgrades viability" discussion.
Basically, the two main points are being discussed, the scientific/electronic/measurable benefits (or lack thereof) and audible/noticeable/perceivable subjective improvements.
I think one important aspect is being omitted - "hot rodding" and doing it "because we can".
So if I put big shiny wheels on my car and a stiffer suspension, let's be honest, the ride quality suffers, but I still feel good about it because the wheels were very expensive and even more so the suspension, not to mention how much I paid the mechanics to install it all. Why can't I have opamps in my mixer that cost more than the mixer? Can I use a $50 capacitor instead of $0.5 capacitor, which would most probably do the same job? You bet I can!

Back when I was a kid, we used to have a derogatory saying about such cosmetic hot rod modifications. "If it don't go chrome it". I always found it preferable to drive a "sleeper", a car that didn't look fast, but was.

JR

 

[abbey road d enfer]

I certainly wouldn't trash-talk JW (I don't know him), but I have an issue with the whole hot-rodding religion.
It's been proved time and time again that grafting a nuclear intergalactic clock to a converter does not improve objective performance, but still, this myth is intact.
Same, it's been shown that using hyperdrive-fueled opamps to reproduce 20-20k is a waste of electrons, but still many claim unprovable benefits.
Clearly, some are making a living on misinformation. Yet it's not as bad as in HiFi. It's permeating, though; I see a number of mastering facilities that rewire their installation with directional cables.

 

[JohnRoberts]

I certainly wouldn't trash-talk JW (I don't know him), but I have an issue with the whole hot-rodding religion.
It's been proved time and time again that grafting a nuclear intergalactic clock to a converter does not improve objective performance, but still, this myth is intact.
Same, it's been shown that using hyperdrive-fueled opamps to reproduce 20-20k is a waste of electrons, but still many claim unprovable benefits.
Clearly, some are making a living on misinformation. Yet it's not as bad as in HiFi. It's permeating, though; I see a number of mastering facilities that rewire their installation with directional cables.

You can tell the difference between people who did original high performance design for a living, and people who modify other people's designs.  ;D ;D ;D

Kind of like original music and cover bands. Sometimes a cover is better than the original, if the original wasn't that good in the first place.  Most of the time covers are just different for the sake of being different, not better.

I share Abbey's disdain for pursuing the op amp du jour tweak. By the mid-late '70s commonly available op amps were faster than needed for audio path reproduction and quiet enough for most applications. Since then there have been marginal improvements in technology, with some common circuit blocks reduced to single ICs. These new circuit on a chip ICs are sure easier to use, but not necessarily better, while for a few applications the use of laser trimmed or matched precisions resistor on chip, doesn't hurt.

JR

 

[emrr]

Before you guys re-muddle the discussion, Jim by all accounts is a guy who verifies performance improvements with test gear, and has his own ear preferences.  He does as well sell a few original product designs.