any suggestions for cables? - Page 12

Yes. The tests behind the JAES article were done in the mid-1970s.

The test was engineered to be more sensitive than any real world listening session. For example one can often hear frequency response variations more easily with pink noise than music. Lots of options were tried.

I didn't say that I had that problem, I said that some readers would. IME the US has the world's best collection of bass freaks. ;-)

Exactly.

You must be listening to crappy amps or the test loads were over the top.

If that data exists and it were strong then it would make an AES article with a lot of general interest.

I've seen some data from Swedish audiophile testing and they were weak on the DBT aspects of their testing. Other data from Spanish audiophile testing looks pretty good.

Speaking as an American with a strong germanic/nordic background, it disappoints me when people from the south are showing people from the north how to do things, but hey we did it right first in the US - almost 40 years ago.

I was referring to quality, not quantity. Just like in cars, something Europe does better.

Bryston had to change a few parts in the circuitry to pass the test. Think it was the 4BSST2, but I may be wrong there. Bryston listened and fixed the problem.

Emotiva got a decent review with their largest stereo amp. Good enough to be able to recommend to people after reading the article, but a bit off the best. Think they are a forum favourite, so that's good news for them.

AES can become members and read the member paper just like anyone else. Or they can go to a Swedish library and sneak read it there.

Then you are indeed referring to the Swedish tests that are meaningless because of of their poor experimental controls.

That's so easy to say, instead tell what you would like to see improved for it to make sense to you too.

They could have followed the Clark JAES paper.

BTW the failings of the Swedish amp tests were covered here some months ago.

Sure. Some of the people involved could also have bought speakers and amps instead of starting their own brands.

Doing things differently does not mean it's wrong, so do you have any specific details to say?
( I'm not going to sit down and read a 30 yo document to figure out whether there's something in it or not,
most importantly since I have tickets for the Hobbit in 40 minutes time, which is infinitely more interesting. )

 

"Great" and "crap" are subjective terms. The purpose of ABX testing is to establish if a difference can be heard between the units under test. If you can't hear a difference, then it makes no sense to say one is preferable ("great" or "crap") to the other.

 

And if you keep testing SS amps working within their design parameters for output power and clipping, and every time you test, you can find no difference, what does that tell you?

 

But even if we go along with this odd method of testing you describe... so you find and amp that is considered by all to be "great". Then when you ABX test it against any other amp SS (working within spec) and nobody can reliably hear a difference, what does that tell you?

It means that under the conditions of the test the results indicate no meaningful difference. The assumption, of course, is that the test actually reflects the real-world use of the amp(s). I used to separate out the sound and the capability for these arguments too but don't so much anymore because it's really a total package deal in those regards.

Thanks for releasing me from having to take your posts seriously...

I have seen Arny post this graph a number of times but he never provides the actual reference or the text that went with it:

He goes on to say "The way to read it is that this is the kind of matching that is required for no audible differences due to mismatching, and with a reasonable safety factor."

The paper is by David Clark, and is titled "High-Resolution Subjective Testing Using a Double-Blind Comparator." And yes, he is the same person I quoted in my article on acoustics that found in listening tests that reflections can have pleasing effects and not damaging as folks assume. So it is Ironic that he is being cited as an expert witness for another listening test when folks don't want to believe other results he has published.

Anyway the above graph is Figure 2 and this is the text in the paper:

"Fig. 2 shows the degree of level match for various bandwidths and center frequencies necessary to eliminate audible frequency response effects for most music sources. The curves are compiled from fairly limited double-blind testing of a limited number of individuals. The level used was approximately 85 dB unweighted. These curves are in general agreement with the findings of others [2], [4]. In a double-blind test, response differences greater than those allowed by the curves are likely to be responsible for audible differences. "

I have bolded some key sections. First, the testing and data was not designed to be comprehensive given its limited scope and applicability at one level (85 dB). Second, it doesn't say what is audible and what is not. It says that if you are going to run a double blind test, try to keep the variations below this level. As otherwise, it will "likely" impact the results. There certainly is no "reasonable safety factor" as my read is that it is the opposite: there is a chance there will be audible differences if you followed the graph.

Clark makes references to two other texts in [2] and [5]. They are older textbooks and unfortunately I do not have copies of them to see what they have found and fine print that went with those. If Arny has them, it would be good to know what they say because what Clark is saying is not what Arny is quoting on his behalf.

Nightlord: Can you post a brand name and nomenclature of a "straight wire with gain"? There are millions of folks in the electronics field in every discipline who are looking for this holy grail.

No, but the only one so far that hasn't been proven not to be one is a rather big Bryston 14B SST, but it has to be one made after series number 000 507 to have the "swedish fix" done.

It hasn't been proven free of coloration, just not been proven to have one and I think it has been attempted twice. But given that no other amp made it this far, it's made likely that this is about the best you can find... Due to the pricetag, chaining ten of them to try boosting any coloration to audible levels is not something anyone (but Bryston) could imagine doing.

( As you probably understand proving the non-existance of something is impossible, one can only go a very long way to try to find existance and fail in doing so and make a probability argument. )

Someone having more of a life than arguing with you is not being serious? You're quite the comedian.

The Bryston 14B SST was replaced several years ago with the 14B SST2. I do not believe the replacement amplifier has been tested by LTS. If you believe the 14B SST is the Holy Grail, you are going to have to track down a used one. We spent a 100 pages or so last year, or maybe two years ago, discussing the pros and cons of LTS's testing technique, especially their rather unique "blind testing" methodology. As you might imagine, no consensus was reached about its validity. Amirm is perhaps the only person that would have any interest in rehashing it ad nauseam again.

That's just what I said, but the great/crap comment was not regarding how they behaved relative each other, but relative to a correct reproduction. Aka in regards to the input signal.

There absolutely nothing interesting with comparing two different amps without a fixed reference. You can probably use it to sell newspapers, but that's about it.

Example:
You can test two cars against each other until your blue in the face trying to tell them apart, but if they both are red and as you all know() a real car is british racing green, it does not help very much that they two are identical, they still won't be BRG.

Nothing, there are more than one conclusion possible.

That would indicate that in everyday usage you'll probably be happy with it. I would need more info on what actually was being used as test material, some confidence in the drive in the listeners to really hear the difference if it's there, if the test was carried out enough varied power output conditions etc. If done good enough, then it would turn into a pseudo before-after test, but then the question would be why not do a real before-after on the 2nd amp instead?

It won't save you much trouble doing it your way, you still need to level match them. But if it improves the chances of you doing a better test, then by all means.

No, it requires someone willing to lend theirs for test. LTS does not have the money to buy equipment to test.

Not interested, the amps I use have been tested too so I know the nature of their flaws and put them in use in a way that I most likely don't come across their flaws in everyday use.

The Bryston is much too expensive, but sure - if I come across one below a grand one day in the future then I may very well pick it up.

As someone just post, I seem to be the only one who cares about fresh and innovative tests of amplifiers as done by LTS. You would think that there would be value in knowing if an amp colors its input. But no. Why? Because it kicks the door open ever so slightly that other audiophile views may be right, causing some to get nervous and try to dismiss it anyway they can. Here are the relevant discussions so that we don't repeat them:
Note one of the first replies from Arny: "The problem is that all of the evaluations appear to have been sighted. That makes them more like a public opinion survey than an actual audible fact." So blind testing by a respected organization is considered a public opinion survey. Arny used to go on and on about a Dolby listening tests regarding jitter. One problem: it was a sighted test! Here is that post: http://www.avsforum.com/t/1425262/are-audio-companies-all-involved-in-a-huge-conspiracy/870#post_22508179. Now this test is not blind enough for them where a sighted one was good enough since it had the outcome he wanted.

As to why I like it, I first heard about LTS and its amplifier tests back in 2009. A forum member from your neck of the woods while claiming there were no differences in such gear, let it slip inadvertently that there are tests that show differences. It took days and many posts to get him to provide these three letters: LTS. He would not say more or give any link to their web site. As you can imagine, those three letters were not enough to find the info since the web site is in Swedish. But I persevered and after a few days of searching, I found their web site. Machine translation at that time was pretty ugly affair but despite that, I could see incredible passion around objectively evaluating audio gear. The web site went on and on about how bias-prone sighted evaluation was and importance of controlled tests through blind testing. I found interviews by Ing, this time in English, where he praised the work of people like Dr. Toole and again, how important it was to be unbiased. No one reading the background of the organization would remotely come across thinking they are on the subjectivist side of the fence. They are squarely in the camp of trying to get to the bottom of "audio truth."

Fascinated, I contacted Pekka Johansson, the editor of "Music and Sound Technology" to see if I could learn more with the aim of publishing an article on their testing in the Widescreen Review magazine. Alas, after a quick and warm reply back form him, the connection got lost and that was that. Worse yet, they revamped their web site completely and none of the original editorials and articles as I described is there. And I cannot find Ing's English interview either. It is a shame as it leads to the type of mischaracterization of who they are by the members here who don't like the result of their work.

It was a pleasure reading your posts on this topic. It added some necessary "color" to my readings of the work. And bring some credit where it is due for LTS.

BTW, I read in one of Ing's forum posts (did I say I spend days reading those threads through machine translation?), he said that there have been more than one amp that has passed the transparency test. To my surprise one of them was a Tube amp! Here is the machine translation of that forum post:

"Bryston amplifier belongs to a very small group (can not remember if they are one or two more) as int has been detected in a F / E-listening.

....

One of the most transparent amplifier (thus giving low detectability in F / E-listening) that I encountered during my early experiments with F/E-listening (in the 70's) was an Audio Research, with very many tubes in!

It took many years before I encountered any transistor-equipped for amplifier as (dummy load nota bene) did the same. Though without a load so There are many transistor amplifiers that are difficult to detaktera in F/E-listening. "

I am so fascinated by this test that have made it one of my retirement projects ( ), to recreate it! Thanks again for chiming in. I won't have to reference my own posts anymore when someone challenges the test in the future .

Is the protocol for these tests available anywhere in English? The block diagram, and especially the translation, was very helpful.

Just thinking out loud, I might envision something like the following if there were no "X" as in "ABX" (just "A" and "B" per the diagram).

Let the switch box have two modes as follows, and for which the test subject can switch between "A" and "B" as needed:

1. Bypass is "A", amp + load + attenuator is "B".
2. Amp + load + attenuator is "A", bypass is "B"

The test might proceed as follows:

• For each trial, randomize whether mode 1 or 2 above is used. Allow the listener to switch back and forth between A and B as many times as possible, then choose and record whether A or B is preferred.
• Repeat as many times as required for statistical confidence.

I guess it wouldn't matter whether the bypass or the amp + load + attenuator were preferred, as long as the preference were demonstrably valid from a statistical perspective. Such a protocol isn't difficult to spell out, so I wonder why it hasn't been detailed so far. Or maybe it has, and I'm just not aware of it. If it's the latter, a link would be much appreciated.
Edited by rock_bottom - 12/28/12 at 5:33pm

amirm, you are confused (intentionally, I'll bet). Folks believe David Clark. They just don't believe you, obviously for the reasons you have displayed time and time again, even on this very thread.

Correct me if I'm wrong, aren't consumer amps designed for "everyday usage" purpose? That would be necessary if there is audible difference between Bryston 14B SST and some average consumer amp in level matched ABX set up for "everyday usage" caliber. Has there been?

Oh, you mean this: http://www.avsforum.com/t/1340051/seeking-education-about-those-ultra-expensive-interconnects/2130#post_20732920

You missed the stressed part of what I wrote, the emphasis was on _you_, not on everyday use. I expect though, that amps are made to do their job, rather than make assumptions about how they will be used.

I for one expects much more from the gear than to cope the everyday use, it's the odd days when it really has to perform that's important in my book. If I fill the room with fellow sound nerds, I'm certain we'll end up playing 20dB higher than everyday use and much more difficult and revealing material. That's what my gear is chosen for, not the normal 80-85 perhaps even 90dB normal use.

Not as far as I know. Doesn't mean it hasn't been done.

If before-after shows 'big' difference between them, and ABX doesn't, then I don't see what kind of audiophile that would be content with ABX results. It's not like you will save any money on it, the strength fom LTS testing is that you can easily find the stellar performers in the cheap ranges that will outperform many much more expensive high-end gear.
If you will be buying just anything based on the mantra that all sound the same, then it won't hurt you to pick the better before-after specimens. In your book you will be getting the same, if LTS has a point, you'll be getting something unusually good. If that isn't win-win, I don't know what is.

You actually have one win more compared to us - as the LTS results haven't had an impact on your used gear market, you'll have great bargains on those units compared to here.

For instance, I had to pay about 80% of the original retail price for 10years old NAD 208s, that's the kind of impact the LTS test/recommendation of it has had.

 

Good answers to all my points - thank you. I still disagree with the Swedish way of testing, but at least we can agree to disagree in a civilised way. I don't think there is any point in going back and forth  for dozens of pages on this as we both have entrenched views, and we each of us thinks we are right! It is an interesting topic and perhaps deserves a thread of its own (if there isn't one already).

 

I hope you enjoyed The Hobbit as much as I did (IMAX presentation) :)

Good. Fine by me, this isn't why I rejoined AVS either and I do respect differences in opinion, I just don't want others to have the wrong idea about mine due to misconceptions. So, I much rather do other things than keeping this going.

I did. Had no issues with the things I've read others complaining about, thought it was the best picture I've seen in a long time in a cinema. Not Imax nor atmos, but at least one of the largets screens in Sweden - 62 feet wide. Now begins the long wait for the next, and hopefully for an even longer Director's Cut.

 

Cool!  We're OT but just one more comment - I went expecting the 48 fps issue to be 'bad' but I ended up finding it 'good'. Amazing 3D.

More OT since the original topic "any suggestions for cables?" got lost at what post count

So the next big upgrade path in home theater PJ will be .... 48fps?
Glad I'm holding onto my Sony VW60 + Darbee Darblet until this next round of PJ's get's decided....

Try holding off on the HFR upgrade until you get UHD at the same time, would be my suggestion.

4k + 48fps?
2014 or 2015 under $5k USD?

My home theater is scheduled to open June 2017, so I aim for a bit lower than $5k.

Now... back to cables, if there's anything left to discuss in that area...
Edited by Nightlord - 12/29/12 at 6:05am

Hashed, Rehashed, old Hash, Rotten old Hash!