Master Thread: Are measurements Everything or Nothing?

[dshreter]

I was actually referring to timbre, as @Inner Space described. And I still don’t see what specific measures I need in order to make valid inferences about the precision at reproducing timbre certain speaker has.

Today I have been switching from the OG LS50 To the R3 and with their similarities, a trumpet from Chris Botti sounds different in both. Often they sound quite similar, but timbre varies. What measures tell me about those variations?

“They are quite good at it” is a subjective claim.

You might actually be experiencing differences in tone instead of timbre. Neither speaker would struggle to capture the timbre of a trumpet as they both have frequency response across the entire range of a trumpet without any notches. They are tonally different, and in room response impacts the tonality further.

Speakers mostly don’t have issues reproducing timbre and this is represented in their frequency response - the spinorama is a great view. The frequency response is the specific measure.

 

[rdenney]

The job of the speaker is not to duplicate an instrument, but to duplicate a listening position in a room where the instrument is being played. I don’t want the speaker to emit “guitar”. I want the array of speakers to create the aural image of “guitar being played in a room”.

Rick “objectives are important” Denney

 

[alitomr1979]

You might actually be experiencing differences in tone instead of timbre. Neither speaker would struggle to capture the timbre of a trumpet as they both have frequency response across the entire range of a trumpet without any notches. They are tonally different, and in room response impacts the tonality further.

Speakers mostly don’t have issues reproducing timbre and this is represented in their frequency response - the spinorama is a great view. The frequency response is the specific measure.

I don’t know if I misunderstood what you are saying, but what I get is you mean that basically in any speaker capable of reproducing the frequencies in a note, instruments sound exactly the same. This can’t be, because it is not my experience. Is this truly your experience??

How can the differences be attributed to tone if the same DAC, cables, amp is receiving the same string of data, which is then converted to electric impulses and then reproduced as a group of frequencies ( a note)?

Why two speakers capable of reproducing those frequencies sound different? Do you agree they can and usually sound different?

 

[dshreter]

The job of the speaker is not to duplicate an instrument, but to duplicate a listening position in a room where the instrument is being played. I don’t want the speaker to emit “guitar”. I want the array of speakers to create the aural image of “guitar being played in a room”.

Rick “objectives are important” Denney

Within the limitations of recording techniques and playback systems as they are today, I agree.

But conceptually I would want speakers to emit guitar, and for me to be able to walk around a room and have it still sound as though a virtual guitar is playing in the same place. That information isn’t even in recordings today so the electronics and speakers aren’t designed for that goal either.

 

[dshreter]

I don’t know if I misunderstood what you are saying, but what I get is you mean that basically in any speaker capable of reproducing the frequencies in a note, instruments sound exactly the same. This can’t be, because it is not my experience. Is this truly your experience??

How can the differences be attributed to tone if the same DAC, cables, amp is receiving the same string of data, which is then converted to electric impulses and then reproduced as a group of frequencies ( a note)?

Why two speakers capable of reproducing those frequencies sound different? Do you agree they can and usually sound different?

I’m not saying they sound exactly the same. But timbre, tone, pitch all have specific meanings. If you hear yo yo ma’s cello and turn up the bass knob on your stereo the timbre hasn’t changed as far as I know. The tonality has shifted though and does sound different.

Speakers are not meant to have a timbre of their own, and what they do have is considered distortion. That’s also measurable.

 

[rdenney]

Within the limitations of recording techniques and playback systems as they are today, I agree.

But conceptually I would want speakers to emit guitar, and for me to be able to walk around a room and have it still sound as though a virtual guitar is playing in the same place. That information isn’t even in recordings today so the electronics and speakers aren’t designed for that goal either.

Not me. I want to hear a performance, not a buddy jamming in my living room. That will often require also hearing the acoustics of the performance space that may have considerable greater decay than my house.

There have been times when I wanted to duplicate an instrument sound, but I wasn’t using a recorded performance. I was using a MIDI sound generator, a keyboard controller, and often a PA speaker array. Those speakers don’t have to play back recorded performance acoustics—those are added by the room in which such a system is used, and it has to match the other instruments. That's a different set of objectives and requirements in response to a very different use case.

Rick “played enough music on stage to know that the stage kinda sucks as the place to enjoy the total performance” Denney

 

[Holmz]

I am not referring to a voltage regulator circuit.
I am referring to the rail voltages used in the voltage amp stage and current amplification stage/the unity gain stage of a discreet amplifier.
Examples of the rail voltage being off are amps that test higher or lower than their rated power output spec.

The voltage is regulated onto the rail somehow.

And secondly who knows how the amplifier is rated max sine wave, square, Etc.

Itis almost impossible to have a stable rail voltage and also have it to off buy 10-25% of the target rail voltage.

 

[Holmz]

Depends on measuring equipment. What to measure.

However, there is a difference. It is mostly a theoretical physical question. Mix time and quantum physics into the mix.Going to the extreme. But there will be differences.

Is it possible that an electron is a quanta?

 

[storing]

you mean that basically in any speaker capable of reproducing the frequencies in a note, instruments sound exactly the same. This can’t be, because it is not my experience. Is this truly your experience??

Still, that is in essence how it works. But as mentioned in other posts already: for starters there are no speakers which are exactly the same, even if they are very close they might have different dimensions, then they get put in the room which is going to affect sound a lot, there will be slight changes in placement, and so on. Some of these effects will be within audible thresholds. Still I'm fairly confident that if you take 2 speakers which measure very similar (say, 2 pairs of the same model, so as you say they will translate the frequencies i.e. input voltage at their terminals in the same sound pressure waves), put them in the same room with the same placement, then yes they are going to measure and sound (in a blind test) the same. There's no reason they should not.

 

[Holmz]

Still, that is in essence how it works. But as mentioned in other posts already: for starters there are no speakers which are exactly the same, even if they are very close they might have different dimensions, then they get put in the room which is going to affect sound a lot, there will be slight changes in placement, and so on. Some of these ffects will be within audible thresholds. Still I'm fairly confident that if you take 2 speakers which measure very similar (say, 2 pairs of the same model), put them in the same room with the same placement, then yes they are going to measure and sound (in a blind test) the same. There's no reason they should not.

Maybe define measure similar? (Which you sort of did with two of the same brand and model)

• Frequency response?
• Time domain and Impluse response?
• or something else?

In theory two different brands which were EQ’ed with Dirac should be sounding pretty similar.
And it get more similar if they have similar cross over strategies… but implementing different XO strategies in the same brand of cabinet and drivers, would not likely sound the same.

However if they measure the same in the time and frequency domain, then they should sound the same.

 

[storing]

Maybe define measure similar?

A set of measurements like Amir does with a max deviation of like 1% should cut it, I assume. Perhaps even less/simpler: suppose you'd put a microphone in a couple of positions around listening position then measure a sweep and impulse response, and shape and amplitude match for those (i.e. subtracting the waveforms ~= 0), I'd be surprised if they would be distinguishable by humans. Again: this is just a (hopefully educated) guess, inluding that 1% number, but I just don't see which differences (if any) would not be captured.

 

[ahofer]

I don’t know if I misunderstood what you are saying, but what I get is you mean that basically in any speaker capable of reproducing the frequencies in a note, instruments sound exactly the same. This can’t be, because it is not my experience. Is this truly your experience??

How can the differences be attributed to tone if the same DAC, cables, amp is receiving the same string of data, which is then converted to electric impulses and then reproduced as a group of frequencies ( a note)?

Why two speakers capable of reproducing those frequencies sound different? Do you agree they can and usually sound different?

You are trying to describe a sound characteristic that can’t be measured, but I’m afraid all you can hear in sound, including timbre, falls within the measurable domains of amplitude, phase, and frequency. The reason you hear differences is because those measurable attributes can vary at the transducer and always bounce around (or not) in your listening area, affecting the perceived FR, amplitude, and phase at the listening position. So the differences you hear lie not in some made-up immeasurable quality of sound, but in the immense complications of variations in transducer FR, directivity, and room characteristics.

Differences in timbre arise from different frequency response and amplitude - the relative volume of different harmonics and attack and decay amplitudes. So the speakers themselves might sound different based on measurable frequency response attributes of the speaker. But even if the speakers have identical (on-axis) FR, the speakers then interact with the room. If one has wider or narrower directivity at a specific frequency, the room configuration may cause a frequency to be (de)emphasized, changing the perceived timbre at the listening position even if your speakers have identical on-axis response. Which is one reason we look at directivity/off-axis response of speakers. Also, if you are comparing speakers side by side they are in different positions within the room, and the room interaction will be different, changing the measurable values at the listening position even if the speakers were identical with respect to both FR and directivity.

Which is why the philosophy here is to start with linear speakers with uniform directivity (and linearity+low distortion in everything prior to the speaker terminals, but that is now cheap and easy). Linear on-axis response, uniform directivity, and low distortion, at the speaker give you the best chance - and it’s only a chance because the room is such a big deal - of fidelity at the listening position. This is also an approach with a substantial grounding in audio science. That’s why you are at Audio Science Review.

 

[Backflash]

Speakers sound wrong, measure room, see peaks, EQ peaks, sounds good, measure room, see flat. I make a deduction measurement is everything. Monkey see, monkey do, monkey happy.

 

[sonitus mirus]

Objectivists vs. subjectivists?

I have one question for hardcore objectivists, and it is not rethorical or guiding. It is honest curiosity.

How can one measure audio quality beyond frequency response and SINAD? What I am referring to is the fact that even when instruments can reproduce a particular note, that note has a different sound, that is characteristic of each instrument. Yes, you want a speaker and components with the ability to reproduce the source content without adding anything, that is a very interesting proposition, but what about the ability to reproduce particular sounds the way instruments sound in real life? How can you measure that?

My partial objectivism (“measures are really important but never tell the whole picture”)when it comes to audio is because I am ignorant of what to actually look for/ measure.

I look forward to your thoughts.

Merry Christmas to all!!

Instruments sound different because of timbre, and this can be clearly shown in frequency response measurements. Music playback is about reproducing sounds that a microphone captured or were digitally created. These sounds are combined and processed to create an audio signal. This audio signal can be precisely measured with the right tools. It is physics. Science has a long-standing, fundamental idea about the limits of human hearing as it relates to an audio signal, and many of the challenges being made about the technical abilities of audio equipment are magnitudes beyond any rational hope of being audible.

 

[Spkrdctr]

I hate to throw a monkey wrench into this but I will anyways. The difference between speakers can be fairly large. For a quick example, everyone bashes Klipsch speakers all the time, as they usually have not listened to new ones If you really want to hear the coronets etc. in all their glory, nothing beats a horn speaker. It will put the instrument right there in your room. So, you get what you get with speakers, and the placement, room interaction and the room correction. The issue 90% of the ASR threads bump into in long threads is that the speakers are not nailed down to an easy formula as so many non-speaker things effect the sound. So, discussions go on and on about issues that are not that important (meaning very, very low impact on the sound) or we can't solve at todays or even tomorrows technology level. But it is fun to dream about future technology that will come to us in the next decades.

 

[sonitus mirus]

I hate to throw a monkey wrench into this but I will anyways. The difference between speakers can be fairly large. For a quick example, everyone bashes Klipsch speakers all the time, as they usually have not listened to new ones If you really want to hear the coronets etc. in all their glory, nothing beats a horn speaker. It will put the instrument right there in your room. So, you get what you get with speakers, and the placement, room interaction and the room correction. The issue 90% of the ASR threads bump into in long threads is that the speakers are not nailed down to an easy formula as so many non-speaker things effect the sound. So, discussions go on and on about issues that are not that important (meaning very, very low impact on the sound) or we can't solve at todays or even tomorrows technology level. But it is fun to dream about future technology that will come to us in the next decades.

This is the reason I mostly care about objective measurements with everything going to the speakers. Measurements are important for speakers as well, but I certainly agree that it is difficult to lock down as there are so many variables. A lot of the new Klipsch home theater speakers are efficient and generally capable of being EQ'd and situated within a room to get a much better overall performance for the price.

I still rely on measurements over subjective listening, even after setting up speakers. My brain will fool me and adjust to different speaker setups and rooms, but if I measure the audio with a calibrated microphone around the listening position, I'll get better results with EQ and speaker positioning than I would if I did this by ear alone. Even then, I'll still use a touch of bass or treble tweaking depending on the music.

The room is nearly always the weakest link, and many of us probably do not have the means, an ideal location, or are simply not willing to make necessary sacrifices to make any real improvements. Up to the speakers, I really don't see a whole heck of a lot of mystery that would make any noticeable audible differences that could not be deduced from reliable measurements.

 

[pablolie]

I'll repeat something I stated elsewhere:

(1) Measurements are of course meaningful
(2) The reference benchmarks that these measurements have to live up to in order to be declared perfect are IMO obsolete because
(a) The ideal speaker is still to be full range
(b) the ideal amp is still supposed to be 150W+ to drive such an ideal speaker
(3) This standard is flawed because
(a) a full range speaker is a b*tch to set up ideally *because* of the bass response - it's far easier to set up ideal bass response with a complete separate subwoofer.
(b) if you have an active subwoofer and can optimally set us frequency cut off to your active sub, you amp is only driving stuff above 60-80Hz-ish, hence power requirements are dramatically lowered
(c) room correction has to work much harder with (supposed) full range hero speakers than with a separate competent bookshelf and sub setup.

And I know this from experience. I wasted a lot of $ on setting up a full range shrine in what actually was a pretty ideal large room.

 

[Holmz]

A set of measurements like Amir does with a max deviation of like 1% should cut it, I assume. Perhaps even less/simpler: suppose you'd put a microphone in a couple of positions around listening position then measure a sweep and impulse response, and shape and amplitude match for those (i.e. subtracting the waveforms ~= 0), I'd be surprised if they would be distinguishable by humans. Again: this is just a (hopefully educated) guess, inluding that 1% number, but I just don't see which differences (if any) would not be captured.

Not really…
Amir has the radiation pattern and frequency response.
However we sort of need the impulse response or phase.

If not then 10 speakers where the drivers were all the same, would produce the same sound… changing only the XO, box dimensions and construction.

 

[Beershaun]

@hardisj interview with Sean Olive has a great presentation about the different measurable parameters and how we perceive them, and how differences in each affect the sound we hear.

 

[mdunjic]

Very interesting discussion … imo, Chapter 1 (sections “Misinformation In Audio”, “Science & Subjectivism” etc.) in Douglas Self book “Audio Power Amplifier Design” … could maybe answer many questions being argued and debated on this thread

https://nvhrbiblio.nl/biblio/boek/Self%20-%20Audio%20Power%20Amplifier%20Design.pdf

Enjoy!