Why reviewers don't matter (part 2)

Kimo · « **Reply #20 on:** 3 Dec 2008, 11:11 pm »

I have heard them. I am surprised that you don't hear it with the treble. I know many others have had the same sort of odd reaction to the sound. It is almost as if I find myself tilting my head throughout the entire performance for some odd reason.

miklorsmith · « **Reply #21 on:** 3 Dec 2008, 11:14 pm »

Have you heard them in my system? Everybody that comes through my room enjoys the sound and I've never had a comment about strange treble. My audio buddies are pretty honest too.

Kimo · « **Reply #22 on:** 3 Dec 2008, 11:22 pm »

Your speakers?

miklorsmith · « **Reply #23 on:** 3 Dec 2008, 11:27 pm »

It was a rhetorical question. If I allowed myself to be swayed by your "review" of "Tripath" amps I would have missed what in my system is the total package. Now you're being the judgmental "wrong" reviewer. Touche.

Kevin Haskins · « **Reply #24 on:** 3 Dec 2008, 11:29 pm »

The high frequency response changes with loudspeaker load in many of the Class D designs. So.... change the loudspeaker and you get different subjective results because the frequency response is different.

That MAY be why you both have different perceptions.

Kimo · « **Reply #25 on:** 3 Dec 2008, 11:37 pm »

One of the most telling experiences I had was with a tube amp designer and his own gear. He put his 150 watt per channel stereo tube amp on a pair of Quad 57s in a reasonably sized room and it sounded not so great. Later, in a much larger room, a room I would have suspected too large for Quads, he used the same amp(s) but configured as mono blocks. Great sound. Then we listened to an 8 watt 300b Airtight amp. Great sound again.

We then put an Ayre AX 7 on a Bozak monster. Ugh, what happened here, flat, flat, flat. However, the stereo tube amp sounded great on the Bozaks.

Kevin Haskins · « **Reply #26 on:** 3 Dec 2008, 11:41 pm »

Here is an example. This is the International Rectifier Reference Design Iraudamp5. This is good data put out by engineers who understand how to design with the parts. This isn't a knock on the product or the company. The Tripath stuff probably showed something along these lines too.

miklorsmith · « **Reply #27 on:** 3 Dec 2008, 11:46 pm »

Excepting possible impedance considerations, a slight rising response from 10 - 20 khz could be a good thing, or not. Human hearing generally is weak at 20 khz, if a slight bump can be delivered cleanly it could be perceived as "air, sparkle". If delivered poorly it could be "bright, aggravating". Of course whether it is served well has everything to do with preamp, source, cables, and speakers.

Kimo's illustration is perfect.

Kimo · « **Reply #28 on:** 3 Dec 2008, 11:51 pm »

I certainly don't hear "bright" with tripath amps. Something askew, though. Note, I have heard the amps on single drivers most often.

Kevin Haskins · « **Reply #29 on:** 4 Dec 2008, 12:00 am »

My point is that they FR changes with impedance. The other thing to consider is that most loudspeaker designers have to pad tweeters. It isn't unusual to see a 16 ohm impedance in the top octave. It also isn't unusual to see swings in the Z at or near crossover that peak up to 30 Ohms. That changing Z will interact with the output filter, perhaps doing something that may only have a 1dB swing over a couple octaves but that becomes audible.

So...you have an amplifier that sounds different based upon loudspeaker load. I'm just trying to make sense out of why people seem to have subjective differences with many of the Class D designs.

Kimo · « **Reply #30 on:** 4 Dec 2008, 12:05 am »

Do SETs behave in a similar fashion?

miklorsmith · « **Reply #31 on:** 4 Dec 2008, 12:06 am »

No worries Kevin, I'm not arguing with you. Instead, I think your information further supports my message.

Kevin Haskins · « **Reply #32 on:** 4 Dec 2008, 12:10 am »

And just for reference.... a single driver speaker.

Here you have about almost a 30 ohm impedance at 20K. Look at the FR differences in the amplifier from 4 to 8 ohm. You get a greater magnitude the bigger the difference in Z. So you start to see relatively large differences in the amplifier FR based upon what you hook up to it.

Kevin Haskins · « **Reply #33 on:** 4 Dec 2008, 12:13 am »

Quote from: Kimo on 4 Dec 2008, 12:05 am

Do SETs behave in a similar fashion?

Yes... they interact differently on the bottom end though. The Q of the system changes based upon load. For that reason guys like Paul Joppa recommend designing the loudspeaker for the amplifier load.

Kevin Haskins · « **Reply #34 on:** 4 Dec 2008, 12:14 am »

Quote from: miklorsmith on 4 Dec 2008, 12:06 am

No worries Kevin, I'm not arguing with you. Instead, I think your information further supports my message.

I agree, there can be real subjective differences and both of you can be right!

Kimo · « **Reply #35 on:** 4 Dec 2008, 12:18 am »

Now, this is more like what I would like to see from reviewers. "This amplifier sounds this way, but consider that...."

miklorsmith · « **Reply #36 on:** 4 Dec 2008, 12:31 am »

So you want the writers to be engineers too? Good luck with that. Stick to Stereophile, they actually do have measuring equipment.

Kevin Haskins · « **Reply #37 on:** 4 Dec 2008, 12:49 am »

Quote from: miklorsmith on 4 Dec 2008, 12:31 am

So you want the writers to be engineers too? Good luck with that. Stick to Stereophile, they actually do have measuring equipment.

I don't know that all reviewers should be engineers. I think it helps to have the input from an engineer. It just provides more information to the reader. The subjective element though is something we are never going to get away from. This is by nature a subjective experience and the only thing that matters at the end of the day is the subjective experience that the equipment delivers.

Also... cheap measurement gear is like a gun in the hands of a monkey. I've seen all kinds of measurements posted on the web that are garbage. Unfortunately it takes some technical ability to understand the data and how it it taken has a large impact on it's validity. Even when taken right, it requires a number of measurements before you get meaningful data. I'm afraid that most of the data is lost on most people. Not a snub, just an observation.

Kimo · « **Reply #38 on:** 4 Dec 2008, 01:02 am »

Look, I have seen trained engineers make some of the dumbest observations possible. We would not have been blessed with products such as the Ford Pinto, etc, but for some fine engineering.

That being said, there is no reason that a reviewer should not know a little more about the technology he is reviewing than the average reader. I don't think that is asking for too much. If they aren't willing to get a basic education, they shouldn't bother with writing reviews.

It should be more than "I know what I like, and I like what I hear."

Kevin Haskins · « **Reply #39 on:** 4 Dec 2008, 01:07 am »

Quote from: Kimo on 4 Dec 2008, 01:02 am

Look, I have seen trained engineers make some of the dumbest observations possible. We would not have been blessed with products such as the Ford Pinto, etc, but for some fine engineering.

That being said, there is no reason that a reviewer should not know a little more about the technology he is reviewing than the average reader. I don't think that is asking for too much. If they aren't willing to get a basic education, they shouldn't bother with writing reviews.

It should be more than "I know what I like, and I like what I hear."

I know people with PhDs that I consider totally off the farm. I know EEs that are nuts and wouldn't listen to a thing they have to say. People are people so just because someone has a credential doesn't necessarily make them right.

In general I agree with you and I think that is largely the case. Most reviewers have tinkered around in the industry for many years. They have exposure to more equipment than I do on average. In some ways that makes them more knowledgeable about equipment in terms of personal use. Hell... I don't have fifteen different amps in my system in a given year. I don't have the time to monkey with that kind of experimentation.