AVS Forum banner
Status
Not open for further replies.

Do Switches Effect Sonics in DBT Testing???

652 views 22 replies 4 participants last post by  Don Hoglund 
#1 ·
Recently we've had a number of members posting at various threads about some double blind testing that they've done. And as Moderator here, I commend them for they've posted tactfully without abusive terminology, something which all too often doesn't occur on the web. And I hope this positive posting continues!!!


What I'd like to know is that when you've done some of this double blind testing, what type of mechanical switches, etc. have you used to switch between speakers, cables, power cords, components, etc. And from both an objective and subjective standpoint, how have you verified that the switch mechanism may not have affected the sonics??
 
#2 ·
An example. In my system, my Theta Casablanca 2 has its own high quality digital switching mechanisms for both analog and digital audio sources. I am able to setup demos in this fashion to subjectively hear just how different sources compare sonically, from a subjective standpoint, in my system. Sometimes I do this in a blind fashion, if I have someone else listen and I am controlling the remote control - or if I have someone else over who does the remote and I listen blindly. And to me this is the ultimate demo because I am demoing just exactly as my system is setup for listening.
 
#3 ·
Interesting point Steve.

I tend to think in a DBT you are (at best) hoping for the same "interference" with the signal each way.

On a more general level, I find switches in the mains line to have a bad effect.

Here in Australia, people look at you blankly if you mention unswitched outlets, which I have found better (nicely vague on specifically how) for audio.

Or whaat about hard wiring power conditioners directly into the mains?

Do you have this Steve? Or do you still have plugs and sockets in command headquarters?:D
 
#4 ·
well you don't need a switch per se Steve. however, auditory memory being short, a switch can facilitate rapid comparisons. in your approach, or for that matter any approach, one needs to acertain that the levels that are going through both speakers are identical. is that the approach that you've investigated? That people will hear differences even when no differences exist is well documented. Perhaps it stems from our days as cave dwellers, when if a person didn't duck they were soon the meal for some feline creature? Whether a DBT is performed as some do, comparing 2 of whatever or in a slightly different permutation, i.e. a triangle test (3 samples, 2 are always the same, 1 is different...objective...select the one that's different) they've been shown to be eminently suited for the purposes of determining sensory differences. An example of this, is the person who goes in to have their eyes examined. Rapid switching is crucial in arriving at a global optimum for one's prescription.
 
#5 ·
Dave, I haven't hard wired my PS Audio Power Plant 600s directly into my dedicated wall outlets - I understand that (although it may well improve sonics as a tweaker) that its against the electrical code to do this. So therefore I wouldn't recommend this practice, either.
 
#6 ·
Quote:
Originally posted by Chu Gai
An example of this, is the person who goes in to have their eyes examined. Rapid switching is crucial in arriving at a global optimum for one's prescription.
Chu, if I have someone around to help me so I can do the test blindly, that's fine. But I don't have some always available to do this. During the past two years sometimes Don of Granite Audio is available and we've done some testing, including blind, at both my dedicated home theater room and at his Granite Audio music system. And we've been careful to level match as well, as we agree with you otherwise this can interfere with even legitimate subjective perception. But my experience also is that longer term subjective listening is also important, not just immediate A-B switching.
 
#7 ·
A more recent blind test Don of Granite Audio and I conducted, this time at his Granite Audio music system. We used his Granite Audio preamp, which Don regularly uses in his system, as the switcher. We used two of his 657 tube CD players, using his Granite Audio 470 silver interconnects and his 10 guage power cords. I had recently treated just one set of interconnects and power cords cryogenically. We listened and burned-in the cryod stuff over the course of about 3-4 days. And during that time, switching from #1 CD player with non-cryod stuff, to #2 CD player with cryod stuff, we - didn't hear any difference in sonics. As a result, I am not cryoning any of my Granite Audio power cords or interconnects, and Don isn't interested in trying it any further, either.


Gee, what went wrong?? I'm a tweaker. Per the PLACEBO EFFECT, I was supposed to hear a difference. And I didn't.


Go back to summer 2001, when Don and I demod the Great Northern Sound PASI ( connects analog between CD player and preamp/processor).

Supposedly it alleviated problems in different ground potential and improves sonics. Well, Don and I determined in my system that it actually degraded sonics just a tad, so the demo unit went back. If I recall correctly Don brought over his handmade cable switch which he believed minimally degraded sonics at most, and we used that for the purpose of this demo.

(Once again, the PLACEBO effect lost out!)
 
#8 ·
Excellent questions, Steve.


Not all switches are created equal. As an OEM I've done a lot of testing of various switches both Subjectively and Objectively. Objectively, there's a lot of differences in switches. There are several things to consider in switch quality.


1. Contact integrity. The contact resistance rating of switches can easily vary from 100 milliohms. That variable is per contact point. Things that can effect the resistance are plating material, core material, sub plating, contact surface area, contact alignment, and contact pressure.

Ever wiggle a slide or rotary switch and hear a difference in the output signal? Ever move a rotary switch back and forth between positions and hear a difference in the output signals even at the same position? I've measured rotary switches that start at 10 milliohms in Position A. Then move the switch to different positions and get a different reading each time it returns to Position A. The readings might be 10 milliohms, 2 ohms, 9 ohms, 42 ohms, etc. A switch of this quality would represent a hidden variable in a DBT test. If the switch presented a variable resistance between 10 milliohms and say 9 ohms, you would experience a voltage fluctuation. The fluctuation might even be random.


2. Contact longevity. The contacts can degrade with usage. If the switch is handling high voltage or high current and is being switched "hot", the contacts can be degraded with every use. If the contacts rub when being actuated, the plating material can wear off with time. The switch will be constantly losing quality. The switch may, over time, lose enough of it's quality to "mask" subtle sonic differences.

A super high quality switch might be rated to have a contact resistance below 10 milliohms and to maintain this rating for 25,000 or more cycles.



The "playing field" might be "level" in a DBT with all components under test being connected with the same high quality switches. But, the field will be level at a lower performance level. For example, you could make all the Olympic skaters perform on a 20' x 20' piece of ice. The field would be level for each, but none could perform at their full potential. The tester might conclude that Olympic skaters are not much better than the amateurs at the local rink.


By contrast, if low quality switches are used, the whole DBT is compromised. The field would no longer be level because the common part of the signal path would be different for each component and possibly different for the same component each time it is re-selected.
 
#9 ·
nothing precludes a DBT from being done over a period of time that can vary from seconds to minutes to whatever turns you on. just what sort of switch do you have in mind there Don that varies that much? if it suits you there's no reason contacts couldn't be made from whatever material you desire...treated with Caig or whatever. Issues of external interference, if that's a concern could always be addressed by building a little faraday cage to house the device. Switches, even old fashioned rotary ones have been around for a long time enabling scientific studies from chemistry to physics to audiology to be peformed with accuracy sufficient enough to validate and establish advances in their respective fields. To me this sounds more like setting up barriers to the effective and reproducible studies that can be conducted.
 
#10 ·
Chu, I'm interested, in this thread, in finding out what you or anyone have subjectively tried to determine the extent that switches may have affected audio quality? Have you done any testing in this regard? Or is it all objective conjecture?
 
#11 ·
And Don, as I've done some blind testing with you at your Granite Audio music system using your Granite Audio preamp, can you tell us about its switching and to what extent it does or doesn't degrade sonics vs

using say your tube CD player straight to your Granite Audio tube amps?

Is no matter how good a preamp is as a switcher, will it necessarily sound better straight from the source to the amplifiers, say in a high resolution system like yours? Of course, I appreciate you may choose not to tell all re specs re your switches in your preamp due to proprietary concerns, but whatever you can disclose, please do. Thanks.
 
#12 ·
Hi Chu,

No, I'm not "setting up barriers to the effective and reproducible studies that can be conducted". Steve was asking how switches can influence a DBT. I'm merely pointing out, with objective information, that switch quality can be a factor. You and I are in agreement that high quality switches can be used to minimize unwanted variables. I'm certainly not suggesting that experimenters "have" to use low quality switches????


Steve, when we did our AB testing with my preamp we were using a super high quality "Electroswitch" brand selector switch that is already part of the preamp. So, this AB testing did not require the addition of another switch. The switch in the preamp is sealed to keep out pollutants and keep in life-time lubricants. The contacts are silver plated and rated to maintain their integrity for a minimum of 25,000 cycles. So, the switch will stand up to say another 500 AB tests.


I support and use all manner of available testing including AB, ABC, DBT, ABX, etc. I'm just offering objective information that the test results are only as valid as the test gear and methods used.
 
#13 ·
Quote:
Originally posted by Chu Gai
--------- just what sort of switch do you have in mind there Don that varies that much? ---------------------------------- To me this sounds more like setting up barriers to the effective and reproducible studies that can be conducted.
Chu, please read the posts more carefully. Don has simply pointed out how switches can degrade sonics - that its important to be careful regarding the objective quality of the switcher. You can't just assume that any switcher you pick up does the job sonically. And certainly we have had that input at even the CRT forum here at AVS Forums in regard to some video switchers, too - some are more transparent visually than others.


I'm still interested Chu in what you have to say regarding objective quality, in what makes a good quality switcher for audio purposes? Or are you saying that any old switcher will do, it doesn't make hardly any difference?

Chu, am I misinterpreting, or does it seem from your posts that you really haven't given any thought to the quality of the audio switcher affecting your blind but subjective test results?
 
#14 ·
glad we got that cleared up! i quite agree its beneficial, expecially from the consumer's point of view, to use components that will stand up to the rigors of use.


you're quite mistaken that I haven't given it any thought Steve. designing an abx box say for evaluating speaker cables is well within the means of someone who wishes to invest a certain amount of time, effort and some money. if you wish i'd be pleased to send you a link.
 
#15 ·
Chu, you can simply give us the link here in the forum. But again, I'm asking you which ABX box you've used with your blind tests??? Or haven't you done any blind tests and are you just giving us information from what you've read???
 
#16 ·
damn, the question mark key got stuck huh? how's this one for starters Steve? http://sound.westhost.com/abx-tester.htm

also what does a switch in a preamp have to do with the topic of this thread?

I've read and studied extensively also Steve and if you wish I'll be pleased to present papers in peer reviewed journals that discuss such matters as the limitations of human hearing. I'd be most curious to hear of your testing procedures. Perhaps someday if time permits you can summarize them.

BTW, the original ABX boxes, of which a few are around have been used however regretfully those who've been tested have failed to be able to reliably differentiate between something like 12 gauge copper vs any similar gauge whatever from whomever using well made equipment...of course, reliable differentiation was able to be established with say long lengths of say 32 gauge and something like 16 gauge. but that's old news.
 
#17 ·
Quote:
Originally posted by Chu Gai
damn, the question mark key got stuck huh? [/B][/TD][/TR][/TABLE] Sorry, Chu. ...objective technical experts can discuss this.
 
#18 ·
Quote:
Originally posted by Chu Gai
what does a switch in a preamp have to do with the topic of this thread?
HELLO!!!!!??????????????????????????????? (DARN, stuck on that question mark again!!) Gee, I would think that doing a blind demo as cables or components would actually be used in my system, using my own quality preamp, would be the best test of how it would sound in my system.
 
#19 ·
Okay, I went over to the page telling about the ABX box and here's what I found.


1. The circuit is simple, effective, and basically just like my own ABX box that I built & use.


2. The parts and construction, however, appear to be low quality. The rotary switch in the picture appears to be a $1.50 part that is not sealed, is made with cheap wafer material, and has cheap brass contacts. If my observation is correct, then this is exactly the kind of switch I talked about in my earlier post that can possibly skew the test results.


3. There's a lot of wire in the pictured box. Too much in my opinion. The wires are not shielded and are close enough to each other to possibly have interference. The pictured green wires are the signal path wires and highly coupled because they are parallel, unshielded, and bunched together. Some are even wound around each other.


While the pictured box would certainly be useful and affordable; for an even better result I would recommend using the pictured circuit, minimizing the amount of wire, and using high quality parts like I have. My own shielded grounded ABX box has less than a foot of wire total, the wires are individually shielded, and none of the wires are close to each other or long enough to have much impact on the signal. The rotary switch is hermetically sealed, glass epoxy construction, silver contacts, 25,000 cycles. The volume pots are sealed ALPS brand and of high quality so that the left & right channels will track evenly over the entire range of the pot. A good quality pot will allow you to level match A & B and both the L & R channels will also be level matched to each other. The IN/OUT RCA jacks are pure oxygen-free copper with gold plating over silver, and no brass or nickel. The wires are mil-spec shielded silver and all connections are 4% silver soldered at 800 degrees.
 
#20 ·
Hi!!!!!!!!!!!!!!!!! Steve!!!!!!!!! I see that despite all these posts where you posed the question "Do Switches Effect Sonics in DBT Testing???" there has yet to be an attempt at an answer. So Don!!!!!!!!! You've stated that you've employed switches to perform DBT testing and that Steve has participated to some degree. So what sort of tests have you performed that outlines the minimum criteria needed in order to perform reliable DBT tests? Are there minimum values of say resistance, allowable variations in resistance that are necessary to better able to discern differences in say speaker cables? What were the number of trials that you performed and how did you ensure randomization and a true DBT rather than simply a Blind Test? Has the procedure been presented somewhere for comment, review, and duplication perhaps by another audiophile society like say the one in Boston? Were musical pieces employed and was there any attempt at listenter training in order to better facilitate differences? I take it the equipment used did not interact peculiarly with the speaker loads under different volume conditions like those SET amps.

If this is not the proper forum to discuss this, please accept my invitation to join myself and others on say RAO where we can enlist the participation of others such as Clark and Atkinson. I'd welcome a schematic that you can come up with and if you've got improvements on a DB remote that'd be great. Right now all I see are opinions. Using that setup, it is possible to differentiate between a 50 foot piece of 32 gauge and 12 gauge. I'd imagine its also possible to differentiate between 32 gauge and a cable of your choosing, perhaps your own.

BTW, what's your opinion on the nature of the copper used in the circuit boards for instance?


_________________________



Moderator: Chu, please reread the GUIDELINES posted at the top of this forum and try to follow them. You are going off on tangents for the most part way outside the post initiating this thread, which is contrary to the Guidelines. As stated in the Guidelines, youare welcome to start your own thread asking questions different and outside the scope of what was posted to start this thread. Thanks for understanding.
 
#21 ·
Quote:
Originally posted by Steve Bruzonsky
Recently we've had a number of members posting at various threads about some double blind testing that they've done. And as Moderator here, I commend them for they've posted tactfully without abusive terminology, something which all too often doesn't occur on the web. And I hope this positive posting continues!!!


What I'd like to know is that when you've done some of this double blind testing, what type of mechanical switches, etc. have you used to switch between speakers, cables, power cords, components, etc. And from both an objective and subjective standpoint, how have you verified that the switch mechanism may not have affected the sonics??
Here's the questions that I originally posted to start this thread. Can we please keep to them? Thanks.
 
#22 ·
#23 ·
Hi Chu,,,,,,,,,,,,,,,,,,,,,,,


Thanks for your questions. Some of your questions are already answered in this thread and most of the other questions are covered in detail in other threads on this forum. So, I'll just answer the "un-answered" questions.


All of the testing I do is for my own purposes to make the best products I can and to evaluate new products, parts, or technologies. Since the testing is for my own purposes, I follow whatever procedure suits me for the particular test. I get others, including Steve, to participate in some of the Subjective listening tests to broaden my information base. Steve has been doing more testing with me this year since my dear golden-eared friend, Don Schlicker, died in February. Sometimes I get results that are so dramatic or definitive to me that I don't bother with anyone else's participation. But, when I get test results where I can't tell the difference between A and B, or when I want to verify my results; then I get lots of additional participants. My results are "published" , if you will, in the form of my finished products. Other audiophiles evaluate my tests by auditioning my finished products and either keeping them or returning them for a refund. So, if I don't do quality tests, I get severely punished in my own wallet. ha, ha.


I'm certainly not going to publish too many details about how I do everything, because my competitors might be reading this? ha, ha.


Other than parts quality and less wire, I can't offer any improvement on the ABX box that you found and shared with the rest of us. Especially since it's the same basic circuit design I came up with 25 years ago. As I'm sure you already know, the volume pots at the A & B inputs are a super feature because they do 2 important things. They allow you to exactly level match A and B. ( I level match to within 0.1 dB.) And, they present the same constant load impedance to both A and B. So, the volume pots level the playing field in 2 ways. It's important to use quality pots so that the left & right channels are also level matched. If they're not matched, this could throw off the balance, the focus, and the sound stage effects. A low quality pot could possibly mask the subtle differences.


Several years back I built a special prototype of our Model #770 preamp that had dual circuits in it. One circuit was the reference and the other circuit was identical except for the parts. I substituted various brands & construction of capacitors & resistors in the parallel circuit. Then I could switch back & forth on the fly to instantly compare parts substitutions in the circuit.
 
Status
Not open for further replies.
You have insufficient privileges to reply here.
Top