All good and correct points.
All supporting the value of the proposal.
We indeed all suffer with trying to find the answers to all of these questions for ourselves before even starting on the design journey. But with a comparable reference point we can improve the efficiency of making better choices.
So you measure your hearing and find you have a null at 1.7k. OK. Now in the future when you design speakers that others build, you can tell the community that they may want to play with this band and adjust to taste.
We may find that if we had a reference design, that their is a correlation between age and subjective tuning options. Once we have this data, we can tell people in advance that if they are over 55 (for instance) there is a 70% chance they will prefer crossover type '3B'.
You are misunderstanding me. This is primarily a TOOL for designers. This tool can of course be used as your main speaker, but is designed to gather data so you can place yourself in the bell curve of opinions on many topics.
Think about it this way.....
You are deciding between 10 possible speaker designs each designed by a different person that suit your physical requirements like speaker size, extension and aesthetics.
What if I could tell you FOR CERTAIN that when listening to music on the reference system:
-8 of the designers had preferences for setting type 6
-1 designer liked type 2 best and 3 was close, but type 6 was god awful to him
-1 designer liked setting 1
SETTING 6:... 'Padding the tweeter down by 2db , moving the crossover up by 2khz, and changing to a 3rd order slope combined with modification of cabinet to shape 3'
when you listen to setting 6 IT WAS GOD AWFUL to your ears, and setting 2 and 3 were MUCH more like it.
Would that not be valuable to you when choosing which design to make?
Now imagine we collected all the measured and subjective data from all the people who had this reference speaker. Then we placed it on various graphs to look for trends.
1) This would lead to insight into WHERE designers should be placing their efforts when creating popular designs based on WHAT WORKS, not what measures well for most people. I mean, Imagine that 95% of 1000 people under age 35 preferred setting 6, but the top 25 designers were over 50!!!
2) Those who are outliers in specific areas will know they are outliers, and so can also now be designed for.
Or, it can be used in an entirely different way......
'This year, the DIYAudio group is investigating the audibility of wave guide profiles.
To participate in this test, you will need to buy the 'horn profile pack' from DIYsoundgroup. 100 sets have been produced, and are available to purchase now. Please set your equipment up per the standards attached and fill out the attached form. Once you have submitted your results, you can access the data to see how your results compaired to other members based on:
-Age
-Sex
-Country of origin
etc....'
Last edited:
I know for instance that my current project (JMLC 350hz horn and TD10M mid has a pattern control difference between the mid and tweeter.
That's something I have traded off in order to get lower distortion, higher output and an aesthetic that I can live with. But I'm guessing that this is worth the trade off against the center to center distance being much larger than the 1/2 wavelength that is commonly suggest is minimum for driver integration. But the truth is that I have nothing else to go on.
With a reference point, I would have a much better chance of making the right choices.
That's something I have traded off in order to get lower distortion, higher output and an aesthetic that I can live with. But I'm guessing that this is worth the trade off against the center to center distance being much larger than the 1/2 wavelength that is commonly suggest is minimum for driver integration. But the truth is that I have nothing else to go on.
With a reference point, I would have a much better chance of making the right choices.
I'd rather just go on and measure to guide me. As that part really worked for me thus far. I'm not building for anyone but me and find that a pretty good reference does exist ,just look at the published work of firms like JBL.
They have the money to do the needed tests in a controlled environment. This does not automatically mean I have to agree with every thing they publish, but I have to say that it has been fun to find out that their reference material is an excellent guide to start with.
I'm just not that interested in what the masses want to hear. I'm interested in what I want, somehow that turned out pretty close to what JBL found too. This is a good enough check-up for me. Yes, my ears have been tested to
.
Just for giggles, look up the several full range drivers we tested subjectively over on the other side of the forum:
A Subjective Blind Comparison of 3in to 5in Full Range Drivers
A Subjective Blind Comparison of 3in to 5in drivers - Round 2
A Subjective Blind Comparison of 2in to 4in drivers - Round 3
A Subjective Blind Comparison of 2in to 4in drivers - Round 4
A Subjective Blind Comparison of 2in to 3.5in drivers - Round 5
Subjective Blind ABX Test of EnABLed FF85WK - Round 6
This originally started after a debate about good and bad sounding full range drivers. It is a flawed test of course as you'd need a pair of headphones or speakers with probably an unknown frequency response to listen to this.
But there were quite some mixed results in these tests, with lots of people choosing something that I thought was pretty unlistenable.
Would I put my trust in those people's hands and let that guide me? I won't. I rather go for measuring my results and compare this to published data, not strictly for guidance, just as a sanity check. As said, I'm out to please me, not the general public. Comparing results to known references only functions to keep myself on track.
They have the money to do the needed tests in a controlled environment. This does not automatically mean I have to agree with every thing they publish, but I have to say that it has been fun to find out that their reference material is an excellent guide to start with.
I'm just not that interested in what the masses want to hear. I'm interested in what I want, somehow that turned out pretty close to what JBL found too. This is a good enough check-up for me. Yes, my ears have been tested to
.Just for giggles, look up the several full range drivers we tested subjectively over on the other side of the forum:
A Subjective Blind Comparison of 3in to 5in Full Range Drivers
A Subjective Blind Comparison of 3in to 5in drivers - Round 2
A Subjective Blind Comparison of 2in to 4in drivers - Round 3
A Subjective Blind Comparison of 2in to 4in drivers - Round 4
A Subjective Blind Comparison of 2in to 3.5in drivers - Round 5
Subjective Blind ABX Test of EnABLed FF85WK - Round 6
This originally started after a debate about good and bad sounding full range drivers. It is a flawed test of course as you'd need a pair of headphones or speakers with probably an unknown frequency response to listen to this.
But there were quite some mixed results in these tests, with lots of people choosing something that I thought was pretty unlistenable.
Would I put my trust in those people's hands and let that guide me? I won't. I rather go for measuring my results and compare this to published data, not strictly for guidance, just as a sanity check. As said, I'm out to please me, not the general public. Comparing results to known references only functions to keep myself on track.
Last edited:
A larger distance between drivers sure has it's own drawbacks but it doesn't mean you can't get great results at a pre-determined listening distance. The rule of thumb is indeed a quarter wave distance between drivers, if you understand the why of that, you can deviate and still get excellent results, as long as you understand what you need to do and why. Measurements can guide you there. A reference speaker would mean ****, as it would interact differently with your surroundings.
Just don't forget how your speaker interacts with the room, and measure this. Adjust the room to integrate the speakers within that room.
Don't stop after close up measurements of your speakers, measure them where you actually listen to them. Learn how that changes your perception.
(as it definitely will change that perception)
It does not stop at the speakers if you want great sound in your room.
Learning to read the measurements will help, as there's more to it than frequency response only. Learning to listen helps too. But overall, trust that microphone to guide you.
Last edited:
Scottjoplin is right the thread has morphed into a discussion of design philosophy/approach/
data gathering rather than whether or not driver distortion matters.
seems like the majority would rather quibble and argue fine points and philosophy/religion rather then come to a consensus as to what is truly relevant and how it constructs the bigger picture.
does low driver distortion matter?
data gathering rather than whether or not driver distortion matters.
seems like the majority would rather quibble and argue fine points and philosophy/religion rather then come to a consensus as to what is truly relevant and how it constructs the bigger picture.
does low driver distortion matter?
Yes, sure matters to those of us who are enthusiasts for the clean sound of ESLs.
Attached is a normalized-to-fundamental THD curve for my speakers which has THD across most of the most audible band around -50dB (.3%).... and that includes passing trucks, mic, laptop, DSP, USB, SPDIF, and my old amps*.
REW sometimes complains that my THD results are so low they are falling into the bottom mud level and it just doesn't want to display them (REW users will know what I am talking about here). Ah, well....
B.
*not to mention the Behringer DCX2496 DSP, constantly maligned at this forum
Attachments
Last edited:
It matters to me too, as an array lover. But I don't worry about it once it's down far enough.
What do I consider far enough? The objective was below 1% from 200 Hz and up, I'm comfortably there.
As measured at the listening position, probably not the best example I can come up with, I was just testing a new amplifier.
This also includes traffic noise, DSP etc, -3 dB at 17 Hz and ~17 kHz. This graph is also normalised to fundamental.
No crossovers, just many drivers.
What do I consider far enough? The objective was below 1% from 200 Hz and up, I'm comfortably there.
As measured at the listening position, probably not the best example I can come up with, I was just testing a new amplifier.
This also includes traffic noise, DSP etc, -3 dB at 17 Hz and ~17 kHz. This graph is also normalised to fundamental.
No crossovers, just many drivers.
Last edited:
Hey Gedlee,
Sy touches on a lot of the biases that apply to audio in his Testing, One, Two Three article
View attachment LA Vol 2 Yaniger.pdf
Non-Auditory Cues; Placebo Effect; My Wife (or the need for Double Blind); Timing; Level ...
"
Our brains and sense organs, as a system, are limited in what they can do and how they can efficiently process the sort of data that kept our ancestors from being eaten or trampled.
...
Remember: Trust your ears, but don’t trust your lying brain!
"
"
References:
(1) Randi, James. 1982. “Flim-Flam” Prometheus Books.
(2) Proctor, Philip; Austin, Phil; Bergman, Peter and Ossman, Dave. 1970. “Don’t Crush That Dwarf,
Hand Me the Pliers,” Columbia C 30102.
(3) Toole, Floyd. 2008. “Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and
Rooms,” Elsevier.
(4) Martin, Geoff. 2011. “Introduction to Sound Recording,” available through www.tonmeister.ca; Lipshitz,
Stanley and Vanderkooy, J. 1981. “The Great Debate: Subjective Evaluation”, J. Audio Eng. Soc.,
Vol. 29 (July/Aug.), pp. 482-491.
(5) Rosenthal, Robert. 1998. “Covert Communication in Classrooms, Clinics, and Courtrooms,” Eye on
Psi Chi. Vol. 3, No. 1, pp. 18-22.
(6) Lipshitz, Stanley. 1984. “The Digital Challenge: A Report” BAS Speaker Aug.-Sept., available at
Boston Audio Society - ABX Testing article.
(7) Clark, David. 1982. “High-Resolution Subjective Testing Using a Double-Blind Comparator”, J. Audio
Eng. Soc., Vol. 30 (May), pp. 330-338.
(8) See, for example, Grant, Doug and Wurcer, Scott. “Avoiding Passive Component Pitfalls,” Analog Devices
Application Note AN-348.
"
Cheers,
Jeff
Last edited:
As discussed earlier and in detail in this thread, EQ can be used to flatten the FR of drivers due to resonances and hence reduce distortion.
EQ is used to compensate for room response to a degree, does this reduce room distortion, does it have a negative effect on speaker distortion or do we have to balance the pros and cons?
Then there is EQ for personal preference, I presume this should be used in moderation to avoid significant distortion?
@lbstyling, do you know what Dr. Geddes wife does for a living?
I see the above idea as seriously flawed, unless you'd also include a listening room to go with the set of reference speakers. Even though you mentioned the room (only once), this part is one of the most important pieces of the puzzle.
I see no use in a reference set of speakers without defining the way it should be setup too. And that's just not going to work i.m.h.o.
Wesayso - correct, and if lbstyling grants us , oh say $100,000 we'll get started on the project right away!
And the room absolutely has to be a part of the evaluation since how a loudspeaker interfaces with the room is not a small factor.
It was car audio - it emulated the poor sound of a typical car radio of the day.
Having "magic" in audio is a real necessity for many people. When you take it away they loose interest.
A reference is not going to happen. There are so many compromises that have to be made wehn one designs a loudspeaker, that even the best loudspeakers are highly flawed in some way. One can take a collection of very good loudspeakers, and they sound quite different, yet are all equally valid.
dave
I don't think that this is true. Your link is a "recommendation" - there are no "standards" for studio work and the ones that I have witnessed were not done at 85 dB. I was in a studio recently where the band was playing at > 100 dB and being recorded. The mixing room was not quite that loud, but certainly > 85 dB.