Dr. Klaus Heinz of HEDD Audio (ex ADAM Audio) - measuring speakers, in particular speaker dynamics

[Frank Dernie]

Is that per driver matching a 10", or the 2 combined?

The two combined

 

[mitchco]

I just finished reading mitcho's article and listened to the files. My conclusion is simple: it does NOT at all apply to this topic. As I noted earlier, the sense of bass is totally gone as is deep bass reproduction. Secondly, for a binaural recording to work as far as imaging and such, your HRTF needs to be closely matched to that of the dummy head/or whatever head is used in the recording. Mine must not at all match that as all I get is the sound in the middle of my head. There is no projection even close to front of my head let alone in front of me. It is classic "headphone sound."

This setup needed to be evaluated with a neutral listener who listened to the setting live, and then compared it to the binaural recording.

Let me give you a real-life example of someone trying to show dynamic range of speaker systems. A few years ago at CEDIA conference a speaker manufacturer said they set up to differentiate themselves from others in dynamic range for movies. So they had built a surround sound setup in a mocked up theater. They then played a western movie scene with gunshots. Every time a bullet was fired, not only you heard amazing impulse power, but my pant legs would swing with the wind/sound pressure of those bursts! What was fascinating was that while the rest of the scene was loud, it was not at all uncomfortable. The average level was what you wanted the movie to be about and the peaks would do what I have heard do when listening to live gunshots.

A few years later, Harman had a demo room at CES showing their new Revel Rythem 2 subs. They played a set of drum tracks and the sensation I mentioned above went past 11! We are talking an experience you have never had with clean and super dynamic bass. Stereophile reviewed them. Here is a bit from them: https://www.stereophile.com/content/revel-ultima-rhythm2-powered-subwoofer-page-2
View attachment 27352

Sometime later we built our theater at Madrona digital and at near reference we have people who after listening once, were too scared to go back in the room! And again, this is not "loud" but super dynamic. Average levels were just fine but in movie effects, your world would come to an end just like it did in the movie.

Needless to say, nothing remotely approximating the above was in the recordings.

I have not seen binaural recordings be used to represent the above experiences in any research.

So no, you can't go there with the mitcho recordings. Neither the grand sense of soundstage in tall and large speakers is there, nor any sense of bass dynamics.

Thanks for taking a listen. The point of my article is to "compare" a small inefficient speaker versus a large high efficiency speaker while listening at reference level to hear if there is a perceived difference in size or dynamics, when both speakers share the same frequency response. I stand by my conclusion that the only real audible difference is the directivity index differences between the two speakers. They sound the same from a size and dynamics perspective, even with my expectation bias going in. That's it. Most people can hear the directivty difference on the binaural recordings who reached out to me after I wrote the article, including folks from this forum.

Of course the binaural recordings can't capture the feeling of low bass - I said that in my article and here in several posts. With 500 watts per side going into double 15" JBL cabs with rated sensitivity at 104 dB at 1 watt at 1 meter and dual 18" subs with 900 watts a side, I can "dynamically" shake my house with the intro to Spiderman into the Spiderverse at reference level where if feels like an earthquake. But who cares, cause that ain't the point.

I am waiting to see some real measurements/recordings to show/hear how one speaker, regardless of size, can be perceived to be more dynamic or "sound bigger" than another when compared side by side. Other than my comparison, I have yet to see any real scientific comparison with any results other than speculation or bravado.

 

[mitchco]

What also may be of interest is which DAC he used to drive LS50s and subs. @mitchco, can you please share that detail with us?

@PierreV has it right, it is a Lynx Hilo.

 

[Soniclife]

I am waiting to see some real measurements/recordings to show/hear how one speaker, regardless of size, can be perceived to be more dynamic or "sound bigger" than another when compared side by side.

Do I remember correctly that you are puzzled about dynamic differences between your system and the 8c (or similar) when playing back a drum recording you made?

 

[andreasmaaan]

Thanks for taking a listen. The point of my article is to "compare" a small inefficient speaker versus a large high efficiency speaker while listening at reference level to hear if there is a perceived difference in size or dynamics, when both speakers share the same frequency response. I stand by my conclusion that the only real audible difference is the directivity index differences between the two speakers. They sound the same from a size and dynamics perspective, even with my expectation bias going in. That's it. Most people can hear the directivty difference on the binaural recordings who reached out to me after I wrote the article, including folks form this forum.

Of course the binaural recordings can't capture the feeling of low bass - I said that in my article and here in several posts. With 500 watts per side going into double 15" JBL cabs with rated sensitivity at 104 dB at 1 watt at 1 meter and dual 18" subs with 900 watts a side, I can "dynamically" shake my house with the intro to Spiderman into the Spiderverse at reference level where if feels like an earthquake. But who cares, cause that ain't the point.

I am waiting to see some real measurements/recordings to show/hear how one speaker, regardless of size, can be perceived to be more dynamic or "sound bigger" than another when compared side by side. Other than my comparison, I have yet to see any real scientific comparison with any results other than speculation or bravado.

It's an interesting experiment you did, and I could certainly hear the difference in tonal balance resulting presumably from the differences in directivity, but no obvious other differences (although given the frequency response of the binaural recording is that different, I imagine it would be likely to subjectively swamp whatever other differences there might have been).

I wonder if you tried the same test at a slightly higher SPL, where the LS50 actually would have been closer to Xmax during peaks than the larger JBL? Given you used a recording with a DR of 11 or 12 and the RMS level of the direct plus reflected sound was 83dB, it's probably fair to assume that the maximum (peak) SPL @1m that either speaker was required to produce was less than 95dB.***

I would say this is probably still just within the LS50's comfort zone, based on knowledge of the driver and measurements of the speaker taken by Soundstage, which show that at 90dB the woofer is only just starting to be stressed in the lower-midrange/upper-bass.

Also, did you try comparing (for example) orchestral music with a greater dynamic range than the Mark Knopfler track you listened to? I could imagine more potential issues with an orchestral work with a DR of say 20dB, which would likely push the LS50 out of its comfort zone when played at 83dB RMS.

***This is obviously back-of-the-napkin, but in reaching this figure I'm assuming:

• the sound from the two speakers is correlated in the bass and lower midrange
• the critical distance is 1 to 1.5m
• your listening distance is 3m

all of which would mean that, anechoically, each speaker is producing about 80-83dB RMS @1m to give you a combined 83dB RMS at the listening position, or 92-95dB peak, which shouldn't be too taxing on the smaller speaker.

 

[mitchco]

Do I remember correctly that you are puzzled about dynamic differences between your system and the 8c (or similar) when playing back a drum recording you made?

Yes, I would say it is my sighted expectation bias, which drove me to try this experiment

 

[mitchco]

It's an interesting experiment you did, and I could certainly hear the difference in tonal balance resulting presumably from the differences in directivity, but no obvious other differences (although given the frequency response of the binaural recording is that different, I imagine it would be likely to subjectively swamp whatever other differences there might have been).

I wonder if you tried the same test at a slightly higher SPL, where the LS50 actually would have been closer to Xmax during peaks than the larger JBL? Given you used a recording with a DR of 11 or 12 and the RMS level of the direct plus reflected sound was 83dB, it's probably fair to assume that the maximum (peak) SPL @1m that either speaker was required to produce was less than 95dB.***

I would say this is probably still just within the LS50's comfort zone, based on knowledge of the driver and measurements of the speaker taken by Soundstage, which show that at 90dB the woofer is only just starting to be stressed in the lower-midrange/upper-bass.

Also, did you try comparing (for example) orchestral music with a greater dynamic range than the Mark Knopfler track you listened to? I could imagine more potential issues with an orchestral work with a DR of say 20dB, which would likely push the LS50 out of its comfort zone when played at 83dB RMS.

***This is obviously back-of-the-napkin, but in reaching this figure I'm assuming:

• the sound from the two speakers is correlated in the bass and lower midrange
• the critical distance is 1 to 1.5m
• your listening distance is 3m

all of which would mean that, anechoically, each speaker is producing about 80-83dB RMS @1m to give you a combined 83dB RMS at the listening position, or 92-95dB peak, which shouldn't be too taxing on the smaller speaker.

@andreasmaaan - thanks and totally agreed. The LS50 was (just) still within it's comfort zone and distortion levels. Turning it up much beyond that point one does start to hear the strain, but with the added sub(s) offloading, one can increase the level more before audible trouble, but for general reference level listening, it is not too taxing like you say.

I intend to try more experiments along these lines and use a more dynamic recording (taking suggestions I also purchased a studio grade mic pre that performs much better than the pre used on the LS 50 recordings. I am also recommending David Griesinger's approach on how to equalise headphones for accurate timbre and frontal localisation without head tracking to further enhance the binarual recording playback.

 

[andreasmaaan]

@andreasmaaan - thanks and totally agreed. The LS50 was (just) still within it's comfort zone and distortion levels. Turning it up much beyond that point one does start to hear the strain, but with the added sub(s) offloading, one can increase the level more before audible trouble, but for general reference level listening, it is not too taxing like you say.

I intend to try more experiments along these lines and use a more dynamic recording (taking suggestions I also purchased a studio grade mic pre that performs much better than the pre used on the LS 50 recordings. I am also recommending David Griesinger's approach on how to equalise headphones for accurate timbre and frontal localisation without head tracking to further enhance the binarual recording playback.

Looking forward to reading about it

 

[Kvalsvoll]

Thanks for taking a listen. The point of my article is to "compare" a small inefficient speaker versus a large high efficiency speaker while listening at reference level to hear if there is a perceived difference in size or dynamics, when both speakers share the same frequency response. I stand by my conclusion that the only real audible difference is the directivity index differences between the two speakers. They sound the same from a size and dynamics perspective, even with my expectation bias going in. That's it. Most people can hear the directivty difference on the binaural recordings who reached out to me after I wrote the article, including folks from this forum.

Of course the binaural recordings can't capture the feeling of low bass - I said that in my article and here in several posts. With 500 watts per side going into double 15" JBL cabs with rated sensitivity at 104 dB at 1 watt at 1 meter and dual 18" subs with 900 watts a side, I can "dynamically" shake my house with the intro to Spiderman into the Spiderverse at reference level where if feels like an earthquake. But who cares, cause that ain't the point.

I am waiting to see some real measurements/recordings to show/hear how one speaker, regardless of size, can be perceived to be more dynamic or "sound bigger" than another when compared side by side. Other than my comparison, I have yet to see any real scientific comparison with any results other than speculation or bravado.

I found this some time ago, and listened to the recorded files. I found there were significant differences, most of which can be attributed to differences in loudspeaker radiation pattern, but also the bass was clearly different.

I also showed how the systems are different, in comments to the original article published by @mitchco . The systems were not calibrated to give a response sufficiently similar to sound the same from lower mid and down into the bass range. This of course makes the experiment less valid for the purpose of finding how small speakers sound vs large. I did not want to pursue this any further, because it was a good experiment, given the constraints, well enough documented.

I also find it likely that the small speaker is operating at its limits in the upper bass - lower midrange, the 5" driver simply is too small. This can be calculated and estimated. You may not notice it when playing music, because the music signal will only reach those limits at brief peak transients, and experiments show that peak limiting up to 3dB is very hard to hear, especially if direct comparison is not possible.

 

[svart-hvitt]

I found this some time ago, and listened to the recorded files. I found there were significant differences, most of which can be attributed to differences in loudspeaker radiation pattern, but also the bass was clearly different.

I also showed how the systems are different, in comments to the original article published by @mitchco . The systems were not calibrated to give a response sufficiently similar to sound the same from lower mid and down into the bass range. This of course makes the experiment less valid for the purpose of finding how small speakers sound vs large. I did not want to pursue this any further, because it was a good experiment, given the constraints, well enough documented.

I also find it likely that the small speaker is operating at its limits in the upper bass - lower midrange, the 5" driver simply is too small. This can be calculated and estimated. You may not notice it when playing music, because the music signal will only reach those limits at brief peak transients, and experiments show that peak limiting up to 3dB is very hard to hear, especially if direct comparison is not possible.

@Kvalsvoll , given your experience in designing and making speakers, and as a trained listener too, how would you define a «big speaker»?

Any differences between a 2 and a 3 way system in behaving big? At what time will the woofer in a 2 way design be overworked and start to produce distortion?

 

[amirm]

I found there were significant differences, most of which can be attributed to differences in loudspeaker radiation pattern, but also the bass was clearly different.

I too found the two sounding quite different and was puzzled by the previous comments that they were close.

 

[Kvalsvoll]

@Kvalsvoll , given your experience in designing and making speakers, and as a trained listener too, how would you define a «big speaker»?

Any differences between a 2 and a 3 way system in behaving big? At what time will the woofer in a 2 way design be overworked and start to produce distortion?

Difficult to define exactly what is big - depends who you ask, big compared to what. At the extremes, it is easier - a small bookshelf or stand-mount with 5" woofer is small, a horn system with bass-system, 2x15" midbass drivers and radial horn on top is big.

Part from the physical size, there are some technical properties we expect to be different - first, capacity, then frequency range, and radiation pattern. As long as capacity limits are not compromised, differences in sound will often be mostly due to differences in radiation pattern.

Limits for a specific woofer depends on frequency range, cabinet design, properties of the driver. Operating a 5" down to 100Hz reaches capacity limits for excursion around 100dB/1m - typically, and then there is non-linearities in the magnet-coil motor system, which can be significant. This can be simulated and thus predicted, and it can be measured after the speaker is built.

I would like to mention that simply adding subwoofers to a small speaker does not make it a big speaker. They will still be lacking in capacity and transient response ("dynamics") in the important lower midrange, and the radiation pattern will still be that of a small speaker. Even a typical hifi-type speaker, with typical dome tweeter and small midrange, will have quite different properties for the larger versions, because they have larger and often several woofers arranged in an array that gives quite different radiation at lower mid frequencies, and of course more capacity.

 

[svart-hvitt]

Difficult to define exactly what is big - depends who you ask, big compared to what. At the extremes, it is easier - a small bookshelf or stand-mount with 5" woofer is small, a horn system with bass-system, 2x15" midbass drivers and radial horn on top is big.

Part from the physical size, there are some technical properties we expect to be different - first, capacity, then frequency range, and radiation pattern. As long as capacity limits are not compromised, differences in sound will often be mostly due to differences in radiation pattern.

Limits for a specific woofer depends on frequency range, cabinet design, properties of the driver. Operating a 5" down to 100Hz reaches capacity limits for excursion around 100dB/1m - typically, and then there is non-linearities in the magnet-coil motor system, which can be significant. This can be simulated and thus predicted, and it can be measured after the speaker is built.

I would like to mention that simply adding subwoofers to a small speaker does not make it a big speaker. They will still be lacking in capacity and transient response ("dynamics") in the important lower midrange, and the radiation pattern will still be that of a small speaker. Even a typical hifi-type speaker, with typical dome tweeter and small midrange, will have quite different properties for the larger versions, because they have larger and often several woofers arranged in an array that gives quite different radiation at lower mid frequencies, and of course more capacity.

Quick question: What is «lower midrange»?

 

[maty]

Why not a good 5.25" coaxial with two big subwoofers? Cut near 180 Hz?

KEF LS50 with new crossovers, by Zvu

https://www.diyaudio.com/forums/mul...factory-cabinets-simulations-post5802040.html

 

[maty]

NRC Measurements: KEF LS50 Loudspeakers

https://www.soundstagenetwork.com/i...nts-kef-ls50-loudspeakers&catid=77&Itemid=153

 

[Kvalsvoll]

Quick question: What is «lower midrange»?

From around 500-400Hz? down to.. 150-200?

It is the range right above the real bass, where most of the energy in music is. Many instruments have their fundamental here, there are important tactile sensations giving this sense of power and realism.

 

[Music1969]

I too found the two sounding quite different and was puzzled by the previous comments that they were close.

Different in sense of dynamics?

 

[Krunok]

@andreasmaaan - thanks and totally agreed. The LS50 was (just) still within it's comfort zone and distortion levels. Turning it up much beyond that point one does start to hear the strain, but with the added sub(s) offloading, one can increase the level more before audible trouble, but for general reference level listening, it is not too taxing like you say.

Maybe you can consider shifting crossover frequency between subs and LS50 a little bit upward (you set it to 70Hz, right?) to put away some strain from LS50?

 

[Krunok]

.. and the radiation pattern will still be that of a small speaker.

Can you please elaborate on this? If @mitcho was using Reference 5 "big" speaker instead of his JBLs and Reference 1 small speaker instead of LS50 he would have a scenario where identical drivers would be used from upper bass to HF. Would the radiation pattern be different in that case or it would be pretty similar?

I admit there is a problem with mid bass when adding a sub to a small speaker like LS50. It would certainly be better if a speaker like one of these center speakers is used instead of a small "bookshelf" speaker.

When you look at it this way wouldn't you say that HEDD TM80, once you rotate it, actually looks a lot like a center speaker?

 

[maty]

Talking about DSP... maybe someone is interested:

[PDF] https://www.grimmaudio.com/site/assets/files/1088/speakers.pdf