What is your favorite house curve

[Tim Link]

I don't think that is quite all of it. If you have a "good speaker" (Flat on axis response , with smooth off axis response curve/ smooth directivity in an anechoic space) at the listening position in a room it will measure with a downward sloping FR i.e. like the Harman curve. The "curve" is not the preference, the "good speaker" is the preference and the curve is how a good speaker measures at the LP without any EQ. If you try to force a "bad speaker" to measure like a "good speaker" at the LP it may make things worse not better. The exception to this is below Schroeder frequency where room modes dominate and making EQ compensations (reducing room mode peaks, trying to fill dips is usually counter productive) is usually helpful.

I was just looking at a measurement somebody sent me of their room with their new Harbeth speakers. It looked pretty flat with a 500ms window. As I closed the window down to around 7 cycles I could see a recess around 3500Hz. I assume this is a result of widening directivity of the speaker, combined with a reduced on-axis response at the lower end of the tweeter. Is this the BBC dip? They say it does magic to vocals. Some people may prefer that sort of thing, so those kind speakers are out there for them. I built a speaker once kind of like that, with a crossover at 3500hz, and I knew the midwoofer was beaming by then. I never intended to cross it that high because of the beaming, but my ears liked it. My girlfriend liked it. That fact alone is amazing.

 

[Ron Texas]

Agreed. A microphone is not a substitute for two ears and a brain - they "hear" things very differently!

It just happens the brain is located between the two ears. Even with that strategic location, the brain has been known to not work well for everyone, LOL. Thank you for your participation and writing the book.

 

[Sokel]

I was just looking at a measurement somebody sent me of their room with their new Harbeth speakers. It looked pretty flat with a 500ms window. As I closed the window down to around 7 cycles I could see a recess around 3500Hz. I assume this is a result of widening directivity of the speaker, combined with a reduced on-axis response at the lower end of the tweeter. Is this the BBC dip? They say it does magic to vocals. Some people may prefer that sort of thing, so those kind speakers are out there for them. I built a speaker once kind of like that, with a crossover at 3500hz, and I knew the midwoofer was beaming by then. I never intended to cross it that high because of the beaming, but my ears liked it. My girlfriend liked it. That fact alone is amazing.

We often forget that the most preferred speakers had as Dr Toole says "the consequence of the nearly universal directivity discontinuity at the woofer/midrange-to-tweeter crossover"
(I quote the whole text) :

Figure 12.4 in the third edition of my book shows the evolution of a steady-state "room curve" using very highly rated loudspeakers as a guide. The population includes several cone/dome products and the cone/horn M2. The result is a tightly grouped collection of room curves, from which an average curve is easily determined. It is a gently downward tilted line with a slight depression around 2 kHz - the consequence of the nearly universal directivity discontinuity at the woofer/midrange-to-tweeter crossover. I took the liberty of removing that small dip and creating an "idealized" room curve which I attach. The small dip should not be equalized because it alters the perceptually dominant direct sound.

So nothing to worry about I think.

 

[Tim Link]

We often forget that the most preferred speakers had as Dr Toole says "the consequence of the nearly universal directivity discontinuity at the woofer/midrange-to-tweeter crossover"
(I quote the whole text) :


So nothing to worry about I think.

The Harbeth seems to be doing the opposite. The room response is smooth but the direct response has a dip. I assure you I will not worry about it either way.

 

[dlaloum]

Personally I don't get the appeal of that downward slope in the treble, it sounds flat, dull, boring, lifeless etc etc to me. Instead I've tuned my system to this curve:
View attachment 368375
Note that this is with a pair of ported bookshelves with a 4" woofer each (Ino Audio piP), I've had these for some years now but I still get impressed with their bass performance!

The "Smiley" curve was always a popular option in the days of graphic equalisers...

Salt & pepper to taste! (this would NOT rock MY boat!)

 

[dlaloum]

sound alike amps , guess depends on REW sine wave tests connected to oscilloscope does it produce harmonic frequencies ? or does the denon 8500h pure on rca outputs , not tested stomaudio thou i expect it to be flawed

Depends on the level of the harmonics as related to the main signal... if they are more than 60db below the primary signal, they will be beyond the threshold of audibility/discernability... at which point they become rather academic.

 

[007Shortz]

Since the Harman headphone target curve is indeed a target, it would be interesting if the task for a test listener would be to tune a loudspeaker in a room with the help of an equalizer so that it sounds exactly like an absolutely neutral headphone.

What would the room curve look like then?

 

[Basic Channel]

Since the Harman headphone target curve is indeed a target, it would be interesting if the task for a test listener would be to tune a loudspeaker in a room with the help of an equalizer so that it sounds exactly like an absolutely neutral headphone.

What would the room curve look like then?

Assuming I've learnt anything in the thread (i probably haven't), it could look like anything above 400hz or so as the discerning listener would hear the speaker rather than the room and speaker above that. And I suppose below 400hz would look like harman.

 

[OCA]

The exception to this is below Schroeder frequency where room modes dominate and making EQ compensations (reducing room mode peaks, trying to fill dips is usually counter productive) is usually helpful.

Your concise explanation hits the nail on the head regarding the target curve debate. I'd like to refine it slightly by replacing "usually" with "almost always" to emphasize the consistency of its helpfulness. It's undeniable that bass peaks invariably emerge in any room, and practical solutions for their absorption are limited, making EQ adjustment the most efficient and precise method.

Another crucial aspect often overlooked is the "measurement volume level," which is nearly as critical as the target curve itself. Despite a speaker maintaining a consistent tilt across most volume levels, our perception of bass levels varies with the measured volume. For example, the Harmon curve might be optimal for reducing bass peaks at lower volumes, while the less bass aggressive Dr. Toole curve could be preferable around reference level measurements. Including suggested optimal volume levels alongside these curves would be beneficial IMO.

 

[My adventures in stereo]

I prefer the Toole curve, at my usual listening levels

 

[Floyd Toole]

Since the Harman headphone target curve is indeed a target, it would be interesting if the task for a test listener would be to tune a loudspeaker in a room with the help of an equalizer so that it sounds exactly like an absolutely neutral headphone.

What would the room curve look like then?

Er-r-r, stereo recordings are made for loudspeaker reproduction. What we hear in headphones is a totally different experience directionally and spatially - not what was intended. We still find pleasure in it because the music survives well. In terms of spectral balance it is therefore no surprise that good sounding headphones deliver a sound spectrum similar to that delivered by highly-rated loudspeakers in typical domestic rooms. So, what is an "absolutely neutral headphone", especially as it cannot be judged without listening to music recorded without any standards? Only statistical averages are possible. It is a dilemma, but at least we have decades of data on what listeners think is a "neutral" loudspeaker in measurable terms. All the "Harman" curves, related to loudspeakers and headphones have been guided by double-blind subjective evaluations. In loudspeakers only the anechoic data can be regarded as resembling a "target", in that they are closely related to listener preferences - room curves are "results". In headphones there is only one curve, and even it varies with the physical interaction with individual listener bodies so, again, it is a "result" - a statistical "target" if you believe the result. Now are you confused?

 

[Floyd Toole]

I prefer the Toole curve, at my usual listening levels

What is the "Toole" curve? I'm flattered but puzzled.

 

[Basic Channel]

What is the "Toole" curve? I'm flattered but puzzled.

Don’t act all coy, you have a great figure darling and you know it

 

[OCA]

What is the "Toole" curve? I'm flattered but puzzled.

That's very interesting coming from you Dr. Toole. Both these curves are regular target curve downloads since many years all over the web:

 

[Basic Channel]

Do we hear the speaker separate from the room even when the frequency above 400hz or so is being reflected within the first wave cycle?

 

[Keith_W]

Do we hear the speaker separate from the room even when the frequency above 400hz or so is being reflected within the first wave cycle?

YES and please be quiet, a number of us are waiting here with bated breath to see what Dr. Toole says!

 

[Floyd Toole]

That's very interesting coming from you Mr. Toole. Both these curves are regular target curve downloads since many years all over the web:

View attachment 368570

And, of course the web is a reliable source of accurately reported - even properly understood - information?

The name "Harman" is liberally thrown around, but few know that the story. Most of the core research was done during my 27 years at the National Research Council of Canada -a taxpayer funded research establishment. I was headhunted to work at Harman because the research was successful. Harman allowed me to set up a research group to continue the work, and most importantly allowed us to publish the results - thereby educating the world, including Harman's competitors.

In scanning this thread it is obvious to me that many people prefer to expend energy and time debating opinions about inaccurately understood data, than to go to the original source of the information and find out where it came from and what it means. it is all in the published scientific literature and my books, but that might cost money and takes time to read. Anyone wishing to spend money wisely on audio gear is not likely to find useful guidance in opinion-dominated forums - the price may be right, but the guidance might not be.

Fortunately are some serious participants, and those who contribute more than unsubstantiated opinion, which is why I drop in from time to time for what have been some intersting discussions. Otherwise my time is currently better spent getting on with writing the 4th edition of my book, which is still selling to students, engineers, and the minority of audio enthusiasts who read beyond forums. The industry is improving. Incidentally, my books have. been translated into Chinese, and they learn well.

 

[OCA]

And, of course the web is a reliable source of accurately reported - even properly understood - information?

The name "Harman" is liberally thrown around, but few know that the story. Most of the core research was done during my 27 years at the National Research Council of Canada -a taxpayer funded research establishment. I was headhunted to work at Harman because the research was successful. Harman allowed me to set up a research group to continue the work, and most importantly allowed us to publish the results - thereby educating the world, including Harman's competitors.

In scanning this thread it is obvious to me that many people prefer to expend energy and time debating opinions about inaccurately understood data, than to go to the original source of the information and find out where it came from and what it means. it is all in the published scientific literature and my books, but that might cost money and takes time to read. Anyone wishing to spend money wisely on audio gear is not likely to find useful guidance in opinion-dominated forums - the price may be right, but the guidance might not be.

Fortunately are some serious participants, and those who contribute more than unsubstantiated opinion, which is why I drop in from time to time for what have been some intersting discussions. Otherwise my time is currently better spent getting on with writing the 4th edition of my book, which is still selling to students, engineers, and the minority of audio enthusiasts who read beyond forums. The industry is improving. Incidentally, my books have. been translated into Chinese, and they learn well.

If I understood your paper correctly, what's publicly known as Dr Toole curve (and I agree that's unfortunate and it's difficult to reach accurate information on the web especially in audio related subjects) is expected steady state response (radiated direct sound) of a speaker in a home theatre environment (small/reflective room) with a reasonably flat direct sound. What's publicly known as Harmon curve is the result of an experiment in which listeners adjusted bass and treble controls to achieve a satisfying spectral balance. I don't remember a specific volume level the test was run though and it can make all the difference in the perceived bass level which is probably coincidentally also the difference between these two curves.

 

[DWPress]

Thank you Dr. Toole for chiming in! I reference your book frequently though it may take several reads sometimes to get it into my thick head.

Not to distract from the current conversation but I found this video to have some good information:

 

[Sal1950]

And, of course the web is a reliable source of accurately reported - even properly understood - information?

Dr. Toole,
Thanks once again for stopping by our little hide-out for science oriented audiophiles.
If you have a minute I wonder if you could give me your thoughts on the below thoughts.
The gradual top end roll-off reflected in most published room curves may be fine for many - most.
But I'll be 74 and have lived an acustically hard life, LOL. What about the senior aged enthusiast (a large %)
who have already lost some level of hearing at the top?
I tend to first control the top end reverberation with room treatment.
Then aim for a flatter top end with EQ at the listening seat. ???
TIA,
Sal