High-resolution Speaker Frequency Response & Distortion Measurements and Why We Need Them (video)

[amirm]

Recently a company rep claimed that we don't need high resolution measurements of frequency response. And that measuring speaker distortion without anechoic chamber is useless. I thought I address this with a deeper dive as to how we fundamentally measure frequency response and distortion, covering the three common methods. I then show the application of high-resolution measurements to diagnosing deficiency in design and performance of speakers (and headphones).

Sorry, I don't have a text version of this so you have to suffer through the video. You can speed it up and get through it in 20 minutes.

 

[sam_adams]

The two papers by Angelo Farina describing the swept sine technique mentioned in @amirm's video:

Simultaneous Measurement Of Impulse Response And Distortion With A Swept-Sine Technique - Angelo Farina

Advancements in Impulse Response Measurements by Sine Sweeps - Angelo Farina

 

[amirm]

Oops. I put the references in the youtube text but forgot to put them here. Thanks @sam_adams.

 

[Toni Mas]

Recently a company rep claimed that we don't need high resolution measurements of frequency response. And that measuring speaker distortion without anechoic chamber is useless. I thought I address this with a deeper dive as to how we fundamentally measure frequency response and distortion, covering the three common methods. I then show the application of high-resolution measurements to diagnosing deficiency in design and performance of speakers (and headphones).

Sorry, I don't have a text version of this so you have to suffer through the video. You can speed it up and get through it in 20 minutes.

Imho, whatever their resolution, measurements should be repeatable before all. If by slightly changing the microphone, measurement axis, loudspeaker position, etc..., the curves change too much, I consider that resolution must be taken with quite a bit of calm, and that the wisest choice is ignore this excess detail as if It were (and in part is...) a measurement artifact...

Btw, for the same reason so called precise EQ is not my cup of tea, and smooth EQ applied to smoothed measurements is what works best in my experience.

 

[amirm]

Imho, whatever their resolution, measurements should be repeatable before all.

For acoustic measurements, you have to allow tolerances. For distortion measurements, if the methodology is different, the results cannot be directly compared.

 

[617]

Moving the mic will cause the HF to droop or pick up slightly, but the basic shape of the FR will be the same, same peaks and valleys up to a very high frequency.

 

[Omar Cumming]

Excellent well worth viewing! I learned a lot.Now I know where that whoop whoop test signal comes from.

Cheers

 

[Jim Shaw]

A timely lecture, even if prompted by yet another dolt in the witchdoctor-audio-for-profit game.

These measurements are yet another place where science meets engineering, meets art, then meets the public. I am one of those who believe that measurements are meaningful, especially in the selection of materials and the design. I also believe in variabilities in quality and manufacturing. Further, I believe that the practice of measurement keeps getting better and more relevant. I don't believe in the tooth fairy.
I don't believe in letting measurements tell me what I hear. But I believe that measurements will pretty accurately predict what I'll hear. And I believe Toole when he documents that the public is surprisingly consistent in judging the quick from the dead in music playback.

Caveat music lover: Measure twice, buy once, and listen happily for a long, long time.
(Alternately, don't measure, buy again and again, and satisfaction will be elusive.)

 

[ayane]

I love that you already have a video on this, haha (and I think the juiciest part of that video is around the 3:52 mark when you discuss equivalent rectangular bandwidth). But thank you for elaborating further on this topic. The format and quality of information in your videos reminds me of "Digital Media Primer for Geeks" and "Digital Show and Tell" by xiphmont (Chris Montgomery of Xiph.org).

Keep 'em coming! =)

 

[tktran303]

For those who like to read; and skip the the parts that you find handy.

YouTube Transcript - read YouTube videos

 

[kongwee]

We are far far out of high resolution on speaker. Capture a square wave from the drivers, nobody will know it is a square wave.

 

[Sonny1]

Amir, you are an excellent teacher. Thanks for taking the time to put this together. Nicely done!

 

[amirm]

We are far far out of high resolution on speaker. Capture a square wave from the drivers, nobody will know it is a square wave.

Square wave is not an audio signal.

 

[Blumlein 88]

We are far far out of high resolution on speaker. Capture a square wave from the drivers, nobody will know it is a square wave.

From the Quad ESL-63 courtesy of Stereophile. 1 khz squarewave microphone capture.

Quad ESL-63 loudspeaker Measurements

Sidebar 2: Measurements (from June 1989, Vol.12 No.6) As I had dragged my motley collection of test equipment (footnote 1) over to Larry Archibald's listening room to carry out a set of measurements on the Mirage M1 loudspeaker he has reviewed elsewhere in this issue, I thought it might be a...

www.stereophile.com

Nice impulse result too.

942 hz squarewave from a DIY horn speaker by B. Walso (of Liberty software that wrote Diffmaker).

 

[kongwee]

Square wave is not an audio signal.

You can do a lot of music with square wave manipulation

From the Quad ESL-63 courtesy of Stereophile. 1 khz squarewave microphone capture.

Quad ESL-63 loudspeaker Measurements

Sidebar 2: Measurements (from June 1989, Vol.12 No.6) As I had dragged my motley collection of test equipment (footnote 1) over to Larry Archibald's listening room to carry out a set of measurements on the Mirage M1 loudspeaker he has reviewed elsewhere in this issue, I thought it might be a...

www.stereophile.com

Nice impulse result too.

942 hz squarewave from a DIY horn speaker by B. Walso (of Liberty software that wrote Diffmaker).

Good why we are stuck on conventional driver? It should be horn or electrostatic. We should instead those manufacturer to use them. I'm am all in for ESL, as I used ML hybrid before. Wait a minute the first graph, why in voltage not in dB? 3rd graph measure in dB?

 

[tktran303]

Oh here we go (again)

“This is a fundamental difference between commercially built speakers and DIY efforts. The enthusiast constructer can pick any design aspect and beat it to death. But if you are trying to survive in the market place you need to think about “bang for the buck”. Do your cost choices give the end user an audible benefit? Or are you creating straw men to knock down, say, (reproducing square waves) over a misguided belief that that aspects overrides others.


Try connecting and disconnecting a a 1.2-1.5V AA/A battery to a woofer.
That’s what a square wave is.
Who wants to hear a square wave, anyhow?

 

[Andysu]

i see that gr Danny speaker is still on the desk serving as paper weight . send it to me for movie soundtrack testing as single mono speaker . lol

 

[Andysu]

7:02 fixing . easy . i just place HP filter play with frequency and gain level and slot BP filter in on that other frequency that has mild raise and lower it down and be content with that .

i have small tiny dinky home no garden now you have garden how about doing same test indoors and outdoors as that sharp echo i hear in you're room and sound moves at speed of sound and pings around so many times in m/s . CB radio when i transmit my signal is going around the earth x7 times in 1 second . sort like uss enterprise at warp speed . yet the sound heard is speed of sound . does that boggle anyone's mind ?

12:14 i did test with cycle pink noise on one of the yamaha dsr70pro years ago , (i hate that term saying "years ago") i had the ch outputs into audio mixer with only single output to monitor and what i found was interesting with changing one channel by +- 1dB or adjust frequency EQ by +- 1dB or reversed polarity . i was looking at ways to improving the LCR .

that chirp birdie signal noise of fast very fast frequency sweep .

 

[Andysu]

You can do a lot of music with square wave manipulation

Good why we are stuck on conventional driver? It should be horn or electrostatic. We should instead those manufacturer to use them. I'm am all in for ESL, as I used ML hybrid before. Wait a minute the first graph, why in voltage not in dB? 3rd graph measure in dB?

sort reminds me of the Denon AVC-X8500H why i was seeing a "jagged edge" on my rigol oscilloscope could have been lots of anything's ?

 

[kongwee]

Oh here we go (again)

“This is a fundamental difference between commercially built speakers and DIY efforts. The enthusiast constructer can pick any design aspect and beat it to death. But if you are trying to survive in the market place you need to think about “bang for the buck”. Do your cost choices give the end user an audible benefit? Or are you creating straw men to knock down, say, (reproducing square waves) over a misguided belief that that aspects overrides others.


Try connecting and disconnecting a a 1.2-1.5V AA/A battery to a woofer.
That’s what a square wave is.
Who wants to hear a square wave, anyhow?

Your battery can generate impulse? like 60hz?