Kali Audio IN-8 Studio Monitor Review

[QMuse]

It seems the 1/2 octave interference dip centered around 900Hz on the Revel accounts for more then the general broadband 1 octave dip centered between 3 and 6kHz on the Harbeth. There is also no deduction for the geneneral 6dB downward slope from 100Hz onward.

And what about Harbeth's dip at 500Hz?

Btw, Harbeth's dip between 2700 and 5000Hz is exactly in the area where our ears are most sensitive.

Sorry, I simply can't have confidence in a scoring system that gives slight advantage to Harbeth as I cannot believe blind listening test would confirm it.

 

[TimVG]

And what about Harbeth's dip at 500Hz?

Btw, Harbeth's dip between 2700 and 5000Hz is exactly in the area where our ears are most sensitive.

Sorry, I simply can't have confidence in a scoring system that gives slight advantage to Harbeth as I cannot believe blind listening test would confirm it.

They (Harman) should try that one . It seems as odd to you as it does to me, but in the end the score means little to me. The spinorama itself has most of what I'm looking for - common sense handles the rest.

 

[QMuse]

Take a look at LW comparison. IMO Harbeth not only is not slightly better than C52 but is a joke of a speaker.

 

[QMuse]

PIR comparison..

 

[QMuse]

Let's try to level match 2 PIRs to get the idea how they would sound in a listening test.

That boomy bass and wide dip at 3kHz doesn't look too primising, right? And yet they have the same scoring..

 

[QMuse]

Btw, @edechamps , @Jon AA , @TimVG - We have seen how @thewas_ did EQ correction with his speakers, can you please post your corrected in-room measurements so we have an idea how did you adjust your theory attitude to your personal listening preference?

 

[napilopez]

I have reservations on scoring formula which uses on-axis instead of LW. My head is not tightened to chair while listening as mic is to mic's stand.

I also have reservations on scoring formula which scores Harbeth slightly better than Revel C52. Don't you?

Let's have a look. Upper curves is C5s, score 7.6, Harbeth is lower curves, score 7.7:

View attachment 53163

So, they score the same. Would you say the same by looking at their curves?

Btw, from what I can see Harman guys clearly optimised C52 for LW and PIR, not for on-axis.

So I havent thoroughly followed the latest discussion in this thread, but on the point of listening window vs on-axis being used in the prference formula, I actually made a rather long post about this when I suggested to MZKM that he calculate an additional score using the listening Window as well the on-axis.

It's a long one, so I'll summarize what was perhaps my main point:

In a way, Olive's preference papers did not truly correlate individual speakers in their totality of use cases with preference. What he did was correlate specific data acquired from a group of speakers with preference.

It's a subtle but important distinction; the on-axis curve appears to have had higher correlation with preference than the listening window because it was a more accurate reflection of the direct sound given the specific study methodology.

All 70 of the speakers during the listening tests were set up on axis, toed in completely toward the listener, who were all positioned in the central same seat at 10 ft/3m away.

So it's no wonder listening window had a lower correlation in the formula. 30 degrees horizontal and 10 degrees vertical is simply too wide a span for such fixed seating at 10 feet/3m; you will rarely be more than 5 degrees off axis in any direction with such positioning during dedicated listening.

Really, on-axis should've probably been called 'direct axis' or something. The on-axis curve fares best for preference because it is a better representation of the direct sound Harman's test listeners had than the listening window.

But in the real world, we know most people don't toe in their speakers all the way. Devantier 2002 showed about 65 percent of speakers being off axis (20 degrees was most common after on axis). And coaxial speakers and a few other designs might be better represented by averaging.

I wonder if a smaller listening window - say, 5 degrees in every direction - would've yielded even higher correlations in those tests at 10 feet. We'll never know.

 

[QMuse]

So I havent thoroughly followed the latest discussion in this thread, but on the point of listening window vs on-axis being used in the prference formula, I actually made a rather long post about this when I suggested to MZKM that he calculate an additional score using the listening Window as well the on-axis.

It's a long one, so I'll summarize what was perhaps my main point:

In a way, Olive's preference papers did not truly correlate individual speakers in their totality of use cases with preference. What he did was correlate specific data acquired from a group of speakers with preference.

It's a subtle but important distinction; the on-axis curve appears to have had higher correlation with preference than the listening window because it was a more accurate reflection of the direct sound given the specific study methodology.

All 70 of the speakers during the listening tests were set up on axis, toed in completely toward the listener, who were all positioned in the central same seat at 10 ft/3m away.

So it's no wonder listening window had a lower correlation in the formula. 30 degrees horizontal and 10 degrees vertical is simply too wide a span for such fixed seating at 10 feet/3m; you will rarely be more than 5 degrees off axis in any direction with such positioning during dedicated listening.

Really, on-axis should've probably been called 'direct axis' or something. The on-axis curve fares best for preference because it is a better representation of the direct sound Harman's test listeners had than the listening window.

But in the real world, we know most people don't toe in their speakers all the way. Devantier 2002 showed about 65 percent of speakers being off axis (20 degrees was most common after on axis). And coaxial speakers and a few other designs might be better represented by averaging.

I wonder if a smaller listening window - say, 5 degrees in every direction - would've yielded even higher correlations in those tests at 10 feet. We'll never know.

Eh, conversation about this topic is spanning over at least 2 threads, pls look at this post.

 

[TimVG]

Btw, @edechamps , @Jon AA , @TimVG - We have seen how @thewas_ did EQ correction with his speakers, can you please post your corrected in-room measurements so we have an idea how did you adjust your theory attitude to your personal listening preference?

Sure. I'd have to perform some for my Revels. But here are my Genelecs 8030C from the living room before and after correcting the LW response above 1kHz.

The remaining dip is due to crossover cancellation in the vertical plane between the tweeter and woofer - EQ'ing it flat would put a boost in the LW.

Here's a measurement of one I built with 50 year old JBL parts (see avatar). You can see the on axis, LW and early (horizontal) reflections. The response below is what it looks like at the LP without EQ above 300Hz.

 

[edechamps]

All 70 of the speakers during the listening tests were set up on axis, toed in completely toward the listener, who were all positioned in the central same seat at 10 ft/3m away.

It seems like you have more information than I have. We concluded earlier that 3m is the most likely distance, but what makes you think the speakers were "toed in completely toward the listener"? What makes you think there was only a single listening seat and that the speaker being tested was always pointed towards it? It doesn't seem obvious to me that the speaker shuffler rotates to always point the speaker towards the listener - in pictures/videos of the shuffler we never see any rotation, the speaker is always pointed towards the back of the room. Maybe @amirm can clarify?

Anyway… I think if we really wanted to be rigorous about calculating scores based on the full listening window, we should not use the LW average, because, as discussed many times before, averages are misleading and can unfairly inflate the score. Instead I suspect the ideal way would be to calculate the Olive score at every angle included in the LW, and then provide the score for each angle, along with a recommendation in the lines of "best score 5.8 achieved at +10° horizontal, worst score 5.0 achieved on-axis" (which basically translates to "don't toe-in this speaker!"). That's significantly more complicated than the current calculations, but nothing that a bit of code can't crack

 

[QMuse]

Sure. I'd have to perform some for my Revels. But here are my Genelecs 8030C from the living room before and after correcting the LW response above 1kHz.

View attachment 53183

The remaining dip is due to crossover cancellation in the vertical plane between the tweeter and woofer - EQ'ing it flat would put a boost in the LW.

I don't really think it makes sense to speak about LW in the context of in-rom measurement as LW is an anechoic-only construct, when measuring in room it is always the steady state response you're measuring.
You can of course apply very short gating, but..

Can you plz post the curve with entire range?

How did you perform measurement of uncorrected response?

 

[QMuse]

Instead I suspect the ideal way would be to calculate the Olive score at every angle included in the LW, and then provide the score for each angle, along with a recommendation in the lines of "best score 5.8 achieved at +10° horizontal, worst score 5.0 achieved on-axis" (which basically translates to "don't toe-in this speaker!")

I'm not really sure such scoring system would have any practical purpose/meaning - IMO it would make speaker comparison sound like "mission impossible".

 

[TimVG]

You can use near field measurements or very nearfield measurements combined with gating where appropriate if need be to simulate an anechoic response. These are the pictures I'd saved so I can't provide much more. The Genelec one is from the with and without this type of correction as measured from the LP

 

[QMuse]

But in the real world, we know most people don't toe in their speakers all the way. Devantier 2002 showed about 65 percent of speakers being off axis (20 degrees was most common after on axis). And coaxial speakers and a few other designs might be better represented by averaging.

And this post.

 

[edechamps]

I'm not really sure such scoring system would have any practical purpose/meaning - IMO it would make speaker comparison sound like "mission impossible".

Yeah, you're probably right. Though I can see a few ways this could work. For example the score could be specified as the average of all scores along with the range of scores, e.g. "Score: 5.2 (±0.4)". The second number indicates how "consistent" the speaker is: if it's small, it means the speaker achieves consistent performance across multiple listening angles; if it's large, it means performance will vary unpredictably as the listener moves around.

As for ranking, one could either assume worst-case scenario and use the worst angle of every speaker, or use the best-case scenario, or some kind of average or median.

 

[Jon AA]

On-axis response doesn't make up a small proportion of the score....

Yes, I'm sorry, I should have looked more closely at the numbers you were stating. I saw the numbers and immediately thought of QMuse's contention that it was only responsible for 3% or so of the PIR and therefore unimportant. I was just driving home the point that the rest of the curves largely depend on it--EQ in a large bump or dip randomly in the on-axis curve and that flaw will show up in all four curves, pretty dramatically affecting the score and presumably, the way listeners would rate the speaker in a test.

I'm familiar with all you've said, but that didn't answer my question.

What it didn't do, apparently, is help your ability to accept it.

Btw, @edechamps , @Jon AA , @TimVG - We have seen how @thewas_ did EQ correction with his speakers, can you please post your corrected in-room measurements so we have an idea how did you adjust your theory attitude to your personal listening preference?

I'll be able to do some more measuring next week. But until then, here's an initial quasi-anechoic (1m 4ms gated/blended--so ignore everything below about 600 Hz) of a DIY speaker EQ'd from about 500 Hz on up:

I've tweaked the curve since then but haven't re-measured yet.

 

[QMuse]

What it didn't do, apparently, is help your ability to accept it.

It surely didn't as IMO it lacked valid arguments.

But until then, here's an initial quasi-anechoic (1m 4ms gated/blended--so ignore everything below about 600 Hz) of a DIY speaker EQ'd from about 500 Hz on up:
View attachment 53192

What you did is EQ-ing the speaker's pseudo anechoic response, not it's in-room response.

 

[TimVG]

Yes what we're doing is the best we can with modest means. Which is why Amir's measurements are extremely useful as his database grows.
It'll give us a good reference to work against (unit to unit variation left aside)

 

[QMuse]

Yes what we're doing is the best we can with modest means.

I'm not sure I understand context of this, can you please clarify?

 

[Jon AA]

It surely didn't as IMO it lacked valid arguments.

Yeah, Dr Toole and the gang at Harman, Paul Hales of QSC fame, etc, are known for not having any valid arguments.....

What you did is EQ-ing the speaker's pseudo anechoic response, not it's in-room response.

I'm quite aware of what I did. Please learn to read.