Brines Acoustics

My Blog

In general,

I hate blogs and bloggers. The average blog is a mindless exposition of trivia by the self aggrandizing. So why do I launch into such cyberspace? Certain topics come up on the various forums that I visit that hit my pressure points. I respond to some of them with varying degrees of agreement. I find myself 1) repeating the same stuff over and over and 2) I get tired of defending my opinion against other opinions. Facts are facts and that is the end of the discussion. Opinions are fleeting interpretations of facts and personal preferences that (last least should) change over time. Having different opinions about the same subject is a good and healthy thing. What follow is a collection on my current opinions on a number of audio topics. You are invited to comment on any of this at the Brines Technical Forum. I am completely willing to change my opinion given a compelling argument. However, keep on topic, forgo the profanity and don't flame me. Your message can be deleted in a heartbeat. Any well thought out and supportable comments will remain and probably get a response. Bring it on!

Blind Testing

Yet another mindless debate about blind testing occured on Audio Asylum. I decided to write down my thoughts on the subject.

The audio hobby is a wonderful thing. Individuals seem to gravitate to sets of believes that approach religous ferver. Caables -- tubes -- magic pebbles -- high dollar anything. Then "they" will report that the latest tweek make a HUGE difference. "They" know these things and will not be diswaded. Then comes blind testing. "They" will not accept blind testing as a valid scientific method, and "they" will not acknowege the results. The problem with blind testing, you see is that it produces the wrong answer too many times for "them". After all, "they" know that there particular religion is the only correct one.

For those who do now know, blind testing in the audio world is a listening test between to different setups where the listener does not know which setup is being heard. This is often abriviatied as ABX. The ultimate test is the double blind test is which the operator does not know which setup is being heard, abreviated DBX. This is done by some kind of switching system that makes the selection randomly and without human intervention. Double blind is hard to do and beyond the reach of most of us. Single blind is good enough if the operator is honest that the number of tests is large enough. Check with your favorite statistition for a definition of "enough".

One misconception about ABX is that it profides an answer of good/bad. It does not. ABX only profides an answer that A is or is not different from B. The operator can either switch from one setup to the other or can do nothing. The subject is then asked to say if a difference cam be detected. Then, after a number of switches/non-switches, the number of correct responses is tabulated and if the subject has guessed right a statistically significant number of times, then the is a DIFFERENCE between the two setups. This is an objective result. Personal preference is not a part of the test. Only the difference between the setups. The subject is not asked which setup is better. That would be a subjective result and totally misleading. Anyone that can hear should be able to tell the difference between a single-ended triode amp and a Class D amp, but to tell me that I should PREFER the SET or a T-amp (which I don't, but that is irrelevant)is nonsense.

So why is DBX and ABX so hated by the purists? Because ABX invalidates their position. The classic case is speaker cables. Time and again ABX/DBX testing has shown that 14-18ga zip cord sounds no different from botique cables, provided of course that the botique cable isn't wildly non-linear. Since "they" know at teflon insulation or cryogenic treatment or whatever makes a better sounding cable, it must be the ABX test that is wrong. Anyone who vehemently states that ABX testing is in valid is simply stating that their opinions are sacrosanct and is not willing to be infuenced by fact.

How should ABX testing be handled by the individual audiophile? When you get a new piece of equipment, particularly something that is going to at best make a subtle difference, is to make sure the new thing does indeed make a difference. The problem here is that if you just bought a $1000 set of speakers, they darn will better make a difference. And, if you just put them in and give a listen, they WILL sound better. This is the plecebo effect. What is supossed to make a difference will! What you need to do is perform a simple ABX. Have a friend switch or not switch the new and old cables a few times as see if there is a difference. Only if you successfully determine tha ther IS a difference should you try to determine which is better.

Bob 10/13/10

Cables

More blood has been let over this topic than any other I've seen. You've seen the posts. "When I installed ABC cables, The vale was lifted. Great dynamics. The highs became higher, the bass lower. Imaging and sound stage became alive." Yea, right. Besides some of this being mutually exclusive, it is never supported by measurement or proper A/B testing. Generally, what the guy is saying is "I just spend kilobucks on a set of cable and there darn well better be a big difference." First of all, let be say that yes, I can hear the difference between cables. It is much easier to hear differences between interconnects than speaker cables, but still the differences can be heard. Sometimes. Given the right circumstances. Just to demonstrate that I am not a complete troglodyte, I use silver in Teflon IC's between my DAC and amp because I like the sound better than Radio Shack IC's. But let's keep the discussion to speaker cables.

If you are going to compare cables, you have to set up reasonable test conditions. The "right" way to do it is a double blind test (DBT). Someone/something switches cables without knowing what cable will be used for any given test. You don't know what the switcher is doing. This way, you have no idea which cable is in use for any given test. After a number of tests, you will have a statistically significant set of results. You either chose one cable over the other or you didn't. The problem with DBT cable tests is that it pretty much requires a computer driven switching devise that is beyond the means of most audiophiles. So, you fall back to single blind testing. You have an assistant do the switching based on a random number table -- switch, don't switch. As long as the assistant's actions cannot be detected, this will work. Doing your own switching is risky. How many of us can independently agree that those new $1000 cables sound just the same, or maybe worse than lamp cord?

As much blood has been spilled over DBT as the qualities of the cables tested. The problem with DBT is that it doesn't always give the "correct" results. Folks go into DBT with expectation as to the results and are too often presented with the contrary answer. "Well," they say, "it must be the test that is in error. Everyone knows that cable A is superior to cable B." Audio is a very personal thing. We spend a lot of time and money tweaking our systems to get the sound we like. It is hard to accept the concept of "different" vs "Good/Bad". Now on to the meat of this discourse.

What are speaker cables all about? Basically, they are lumped parameter components. They have resistance, inductance and capacitance. That's about it. The ideal cable will have zero resistance, inductance and capacitance. Resistance is pretty much determined by the gauge of the wire. Inductance and capacitance are determined by the geometry of the cable. The catch is that if you lower one, you increase the other. Inductance is a bad thing because it rolls off the high frequency response. Capacitance aught to be a bad thing because it rolls off the low frequencies, but to get enough capacitance to actually make a high pass filter is beyond what you can get from a cable. Therefore, boutique cable tend to trade vanishingly low inductance for high capacitance. This is generally OK, but in the extreme, amplifiers that are sensitive to reactive loads can go into oscillation. Not a good thing. There are a bunch of other things that are proported to affect the sonic qualities of cables -- extreme low oxygen copper, silver, cryogenics, insolation material -- but I have yet to read a convincing argument that these thing affect the signal at audio frequencies. These things may affect the signal at radio frequencies and higher, but at audio frequencies? Consider this: Dirty old contractor grade CAT5e cable, ordinary copper in PVC isolation and jacket, passes a sufficiently square wave to reliably transmit data at 350MHz. Now how much distortion is generated at 20kHz? Show me!

The standard by which I judge all cables is 16ga zip cord from Walmart -- the stuff they sell in the Automotive Audio section. Whack off a couple of lengths of this stuff, apply you standard terminations and give it a listen. What you you will hear is a pretty clean, uncolored cable. You may like your favorite cable better, and that's fine. But, understand that you are using your cable as a tone control. I find this concept interesting to discuss with those who won't allow any filtering between the amp and speakers because they don't want anything to corrupt the music!

What do I use? Often, Walmart zip cord. Sometimes single pairs of CAT5. My "show" cables are two pair star wound CAT5, primarily because it isn't stock. There is a lot of elitism at shows. Among other things, you MUST use a tube amp, which I do. Here is a list of cables I find to be good, as well as cheap. No particular order:

Bob 5/5/09

Imaging

One of the most frequent questions I get is "How do they image?" I sigh and launch into my standard reply. But this is what I really think:

Imaging is basically whatever the recording engineer wants it to be. That includes any other sound stage artifact like width and depth. 99% of all non-classical music is done in a studio with one microphone for each performer, often two -- one for voice and another for the"acoustical" instrument. Too much classical music is also multi-mic'd -- several microphones scattered throughout the orchestra or again one microphone per performer in chamber music. Multi-mic'ing reduces a performance to a number of isolated tracks with no relationship to each other. The recording engineer then plays with the tracks to produce the recording we ultimately hear. Volume is adjusted to change the balance between tracks, the track is panned left/right to establish position (or not, which can lead to some bizarre effects), the phasing is adjusted to give depth, and sometimes the phasing is screwed with to make the sound stage wider than speakers. Also, remember that the different tracks are often (usually?) laid down at different times and sometimes in different studios. Consider a performer doing a duet with himself.

Some of the engineering tricks I hate:

If you want to know what imaging is really all about, you need to go to live concerts. Rock concerts and most pop concerts have huge speaker arrays all over the venue, including piping into the house speakers. The sound comes from EVERYWHERE. Zero imaging. The same goes for most stage performances. The band is multi-mic'd and the performers use wireless mics. Again the house speakers are usually involved and there is zero imaging. Classical orchestra (usually, but not always -- the Metropolitan Opera mic's the singers), chamber concerts and acoustic jazz are not mic'd. So how's the imaging? Sort of, particularly in those expensive seat in the middle of the main floor because there are so many reflections. ( I find that the first row of the balcony are the best seats because of minimal reflections.) The bottom line is that imaging in live concerts is at best vague and at worst non-existent. But, for some reason, we demand pin-point imaging from studio recordings!!!!

Now, given the above, "How do they image?" Well, here is my standard answer:

My speakers image like mad! What screws up imaging is early reflections. The worst are reflections off of the side walls. You will want your speakers as far away from the side walls as possible. For this reason, I always place the speakers on the LONG wall of a rectangular room. The other destroyer of imaging is defraction off of the cabinet edges. Narrow cabinets image better than wide cabinets because edge defraction occurs at higher frequency and is thus less damaging.

Multi-driver speakers with dome tweeters are the hardest to get to image well. The tweeter is be operating at near 180° beam width at the edge defraction frequencies. To get good imaging, the designer has to by attention to the baffle surface and edge shaping. It can be done, but it is not easy. By contrast, a single driver speaker is beaming rather strongly at the important defraction frequencies, so little energy gets to the cabinet edges. Actually, the edge treatment on my standard speakers with the oval trim panel is pretty good. The trim panel plus the rounded edges of the cabinet give an effective edge radius of 1 1/2". On those speakers that use a tweeter, I chose a horn loaded tweeter that has a controlled beam width, so again little energy reaches the cabinet edge. Note that my speakers with grills and raised edges image almost as well because of the driver beaming.

"And that's the way it is"

Bob 7/23/09