• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Amir recommendation criticism

Status
Not open for further replies.

oldsysop

Senior Member
Joined
Sep 5, 2019
Messages
383
Likes
657
It seems to me that it is not a good idea for Amir to post his subjective impressions, just measurements.
But on the other hand, this is your forum and maybe it's your fun to do it.
:confused:
 

GXAlan

Major Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
3,868
Likes
5,954
It is not very subjective. Research shows otherwise.

Just to play Devil’s Advocate. Research shows the vast majority of people prefer a downward sloping response because it sounds the balanced. However that doesn’t say that no one prefers flat in-room response or that owners of multiple speaker systems might prefer to have different systems with different responses.

Example would be the speaker for background music at a cocktail party may benefit from a line source where volume control near/far is important or an omnipolar design like an MBL is preferred.

Also we have different bass preference and I don't think that it is affected by age.

Actually, Harman’s research on headphones does actually suggest that younger people prefer more bass, independent of music genre. That is why the JBL headphones tuned to the Harman data science and the AKG headphones tuned to the Harman data science sound different.

There is no reason to assume that the evolved understanding of this variability in the target curve doesn’t apply to future JBL vs. Revel approaches.
 

GXAlan

Major Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
3,868
Likes
5,954
It seems to me that it is not a good idea for Amir to post his subjective impressions, just measurements.
But on the other hand, this is your forum and maybe it's your fun to do it.
:confused:

I actually have disagreed with Amir on his subjective opinions but that just reflects room, music genre, and speaker position. That said his subjective comments are useful as he is a trained listener and is able to make comments consistently.

Combined with the science, and follow-up user/member inputs, over time we will develop a good empiric database of matching subjective to objective data.
 

KaiserSoze

Addicted to Fun and Learning
Joined
Jun 8, 2020
Messages
699
Likes
592
You know what would be very nice would be a two or three sentence summary of the pros and cons. Something like, "I don't recommend this speaker because ...", or "In spite of some minor drawbacks I recommend this speaker because ..." If people understand why Amir decided what he decided, they should be able to decide for themselves whether they agree. To my way of thinking this is only concern, i.e., the uncertainty with respect to why he decided to give a questionable speaker a thumbs up and a speaker that measures good a thumbs down.
 

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,335
Likes
6,700
The Jbl xpl90 scored 3.8 and was recommended..... I truly don't get it.

I don't see that one on the sheet. Yeah 3.8 is a little low for a $699 loudspeaker, but it's not terrible. As I said earlier, he's done so many reviews that there will no doubt be a few that don't make sense. Sort by preference score though on that page and you will see that there is a very high correlation between the preference score and whether or not the speaker gets recommended. And remember, he posts the recommendation before he sees the preference score, so the correlation is true.

The preference score isn't a hard number. I think of it more like the center point of a range of values. With a standard deviation of 0.8, that JBL could be anywhere from 3.0 to 4.6, and 4.6 is very good for a $699 loudspeaker.

Also, it's possible that the JBL is just doing something nice to the sound that the Olive score fails to recognize. The formula is pretty old now, and most of the loudspeakers measured here are actually much better than a lot of the bad loudspeakers in the sample they used. It could be that in todays market, the score holds less correlation than the .86 it did back then.

I've been critical of a few of his recommendations, but looking at that index page was eye opening for me. There really is a fairly strong correlation between the score and the recommendation.
 
OP
S

st379

Member
Joined
Jun 5, 2020
Messages
29
Likes
24
I don't see that one on the sheet. Yeah 3.8 is a little low for a $699 loudspeaker, but it's not terrible. As I said earlier, he's done so many reviews that there will no doubt be a few that don't make sense. Sort by preference score though on that page and you will see that there is a very high correlation between the preference score and whether or not the speaker gets recommended. And remember, he posts the recommendation before he sees the preference score, so the correlation is true.

The preference score isn't a hard number. I think of it more like the center point of a range of values. With a standard deviation of 0.8, that JBL could be anywhere from 3.0 to 4.6, and 4.6 is very good for a $699 loudspeaker.

Also, it's possible that the JBL is just doing something nice to the sound that the Olive score fails to recognize. The formula is pretty old now, and most of the loudspeakers measured here are actually much better than a lot of the bad loudspeakers in the sample they used. It could be that in todays market, the score holds less correlation than the .86 it did back then.

I've been critical of a few of his recommendations, but looking at that index page was eye opening for me. There really is a fairly strong correlation between the score and the recommendation.

1. There are couple of speakers missing like Kef Q350 and other speakers from last month that are not on this list.
2. I don't find strong correlation at all. Speakers that score 4.x getting recommended while others do not. Audioengine (also not on the list) scored 4.9 and was recommended.

and a 3.8 score is really low. The Q350 sound "muddy" while scoring quite high.

I don't see strong correlation between score and recommendation.
 

Snoochers

Active Member
Joined
Mar 28, 2020
Messages
187
Likes
70
What precisely about the measurements should lead me to that?

It is tricky of course, but we know some measurements are better than others and the price is known, so we could apply some metric to determine what a recommended speaker is. We already have all the scores and that, so we could use a score/price ratio or the like.
 

Juhazi

Major Contributor
Joined
Sep 15, 2018
Messages
1,717
Likes
2,897
Location
Finland
I don't care about "Olive's scoring" or "Amir's preference" Raw measurement data is what I need, all the rest is speculation and preference. Same applies to Stereophile, K&T etc. who publish valid measurements.

On the other hand I understand that most readers of ASR are not as well educated and experienced with hifi measurements and evaluation, and can't make personal judgement from raw data. It's simply up to the reader to pick up the goodies from "tests"

Amir's listening impressions from speakers are limited, he listens to a single speaker placed untypically. This reveals most faults but is not fair for different types of speakers eg. desktop/floorstander/dipole. Different radiation patterns and power response affect stereo imaging, which is important for hifi, not so much for HT or desktop.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
On the other hand I understand that most readers of ASR are not as well educated and experienced with hifi measurements and evaluation, and can't make personal judgement from raw data. It's simply up to the reader to pick up the goodies from "tests"
You just got done saying my opinion can't have any value. Why are you assuming that the "uneducated and inexperienced" membership has a prayer then? By that logic they definitely need someone to give them the recommendation.

If on the other hand, you are trying to prop yourself up as being qualified and rest of us not, let's have you pass judgement on 10 of the speakers reviewed and let's have the membership go at you. Will see how much clothes you have left when they are done. :D
 

Juhazi

Major Contributor
Joined
Sep 15, 2018
Messages
1,717
Likes
2,897
Location
Finland
Amir, you didn't get my point, nothing personal in it, sorry! Amir, Kal, Steve, John etc. just the same to me.
My personal preferences are actually deviant from most of my hifi-friends!
 

celroid

Member
Joined
Jan 15, 2020
Messages
72
Likes
44
I was thinking of a project website to have user reviews and ratings, sort of like imdb for audio equipment
 

pozz

Слава Україні
Forum Donor
Editor
Joined
May 21, 2019
Messages
4,036
Likes
6,827
They play louder hence more distortion. ....
It's the opposite. Higher sensitivity, all other things being equal, produces less distortion. It's just that for really efficient compression-based drivers horn-loading and design comes with it's own set of distortions that are hard to overcome. Compared to direct radiators, you need less cone movement and excursion to reach the same SPL.
I was thinking of a project website to have user reviews and ratings, sort of like imdb for audio equipment
Why have an abstract one-number rating/ranking system? We should be encouraging readers to learn audio concepts so they won't be overly influenced or manipulated by the bad tactics of the industry, and spend our energy to come up with better explanations and make the underlying data/research easier to access.
 

SimpleTheater

Addicted to Fun and Learning
Forum Donor
Joined
Jun 6, 2019
Messages
927
Likes
1,789
Location
Woodstock, NY
Yes indeed, an users who are very active are aware of this subjective impression follow up...but most newcomers won’t dig into 30 pages threads.
This whole thread has me thoroughly confused. @amirm did recommend the KEF R3’s. You could stop when you saw the pink panther and not have read a single word. The issue seems to be the KEF R3’s didn’t get the highest honor, the golf swinging Pink Panther, as the Revel M16’s did. And thus the problem with this site, it makes you read, unlike every other publication where you can see a simple list of the “Best of 2020”.
 

QMuse

Major Contributor
Joined
Feb 20, 2020
Messages
3,124
Likes
2,785
As stated by others, that was before he added in the room mode fix in the bass, he since has stated he likes much more if I recall.

That would be the "room mode fix" described here, right?

"So let's measure the room and see what we are working with (with KEF R3 on stand):


index.php


I have filtered this with 1/6 octave so there is not much clutter. Focusing below 200 Hz, we see our loudest and most offending peak is around 100 Hz. Let's dial in a single parametric filter by eye and see what does:"


index.php





Let's try to see what this simple filter actually did to the shown frequency response (oddly measured at 35-40dB SPL):

Capture.JPG


Yep, looks fixed.

While claiming it fixed the room modes issue you are advised to ignore the dip at 75 Hz that got deeper and the wide dip between 130 and 180Hz that also got deeper. You should also ignore peak at 240hz and especially that one at 580Hz and also dips at 310Hz and 850Hz which remained the same.

So are we supposed to believe that after applying that simple filter which effects I showed Kef R3 became a whole lot better speaker in the subjective opinion of our host? That bass is no more "boomy" and that all details are now revealed? Well sure, I mean he's a trained listener so no reason not to trust him..

At the end, let me also post an overlay between predicted in-room measurement and what Amir measured in his room after the "room modes fix". This is showing the difference between what was predicted he would be listening and what he actually was listening when reviewing the R3.

The question being raised here is can you really evaluate a speaker which measures in-room like red line but is supposed to measure like blue line? However, red line is what was actually evaluated through listening.

Capture.JPG
 
Last edited:

tjkadar

Member
Joined
Mar 20, 2020
Messages
61
Likes
160
Location
Stateburg, South Carolina
Like most things in life, you have to do your homework.

Amir first provides us with measurement data. If you're not interested in anything else, stop there. Then he provides his subjective impression. The method used to gain those impressions have been spelled out. There are even pictures somewhere on the forums of the room where he makes those speaker auditions.

Once Amir generates his data, that data is used by MZKM to generate a preference score.

Amir spends a lot of time and effort doing all of these measurements. That allows us to spend as much, or as little, time learning what those measurements mean and how Amir's subjective listening test relate to them.

For me, I feel Amir has done more than enough in as transparent a manner as possible. We are free to make our own decisions based on how we interpret, correctly or incorrectly, the information provided. We can agree, disagree, or remain ambivalent. In the end, we are responsible for our own decisions.
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,240
Likes
11,462
Location
Land O’ Lakes, FL
That would be the "room mode fix" described here, right?

"So let's measure the room and see what we are working with (with KEF R3 on stand):


index.php


I have filtered this with 1/6 octave so there is not much clutter. Focusing below 200 Hz, we see our loudest and most offending peak is around 100 Hz. Let's dial in a single parametric filter by eye and see what does:"


index.php





Let's try to see what this simple filter actually did to the shown frequency response (oddly measured at 35-40dB SPL):

View attachment 68162

Yep, looks fixed.

While claiming it fixed the room modes issue you are advised to ignore the dip at 75 Hz that got deeper and the wide dip between 130 and 180Hz that also got deeper. You should also ignore peak at 240hz and especially that one at 580Hz and also dips at 310Hz and 850Hz which remained the same.

So are we supposed to believe that after applying that simple filter which effects I showed Kef R3 became a whole lot better speaker in the subjective opinion of our host? That bass is no more "boomy" and that all details are now revealed? Well sure, I mean he's a trained listener so no reason not to trust him..

Dial back on the sarcasm a bit.

You almost never want to fix large dips with EQ, so the issue ~75Hz is going to have to stay. Now, maybe he should have used a higher-Q as to not cause the surrounding region to have a slight dip. The other areas with peaks also are surrounded by areas with dips, so that may be tricky.
 
Status
Not open for further replies.
Top Bottom