Time Domain measurements?

[Bjorn]

From the article below.


http://www.speech.kth.se/prod/publications/files/4108.pdf

 

[KaiserSoze]

So the answer is you can't.
Just look at ASR's measurements, they're proof that the Spinorama is manifestoy insuficient which is why you need independent driver and port measurements FR, THD, CSD, now IMD and soon cabinet.

To believe that anything audible will show in a Spinorama is pure delusion.

Anything that affects frequency response and that is audible by virtue of the effect on frequency response will manifest in measurements of frequency response. This is tautologically true. Whether you are able to identify the anomaly by visual inspection of the response curve is a somewhat different question. But if the anomaly, whatever its nature, audibly affects the frequency response, there will absolutely be evidence of it in the measured frequency response.

As for how you know whether a spike observed in the frequency response is caused by a cabinet resonance vs. something else, this is not easy in the case of panel resonances, because the affected frequencies are inherently difficult to predict. But the other kind of enclosure resonance, i.e., standing waves that set up inside the enclosure, are straightforward to predict given the interior dimensions. If you observe a pronounced spike at some specific frequency and it is so strong as to suggest a resonance of some sort, and if the corresponding wavelength is one of the wavelengths for which standing waves are predicted in accordance with the interior dimensions, there is then very high likelihood that the observed response spike is associated with the standing wave. If a given interior dimension is D, standing waves are predicted for wavelengths equal to D x 4/n where n is any odd integer, i.e., 1, 3, 5, etc. This follows from the fact that a standing wave will occur for any wavelength where the interior distance is equal to 1/4 the wavelength, or to 3/4, 5/4, 7/4 etc., of the wavelength.

 

[ctrl]

This follows from the fact that a standing wave will occur for any wavelength where the interior distance is equal to 1/4 the wavelength, or to 3/4, 5/4, 7/4 etc., of the wavelength.

You describe the wavelength for a single open tube (e.g. transmission line).
With enclosures you assume standing waves between parallel walls.

All possible types are described here:
http://www.sengpielaudio.com/StandingWaves.htm

 

[KaiserSoze]

I always kind of fancied transmission line designs, mainly for the woofers, but there are some stuff out there that use TL for small 3" fullrange drivers in a line source... (Adyton). These work really well!

Isn't a properly designed TL a way also for a woofer/mid to kill the backwave, so it doesn't backfire to the same extent through the driver?

View attachment 73424

Interesting question about what a transmission line speaker is intended to be or to accomplished. Another way to formulate the question would be to ask what theoretical speaker model does the term "transmission line" refer to. I'm not confident that there is a single answer, i.e., it seems that different people who build "transmission line" speakers do not all have exactly the same goals and expectations, or even precise agreement with respect to what it is, beyond the fact that the acoustic energy within the enclosure is expected to propagate down an interior channel where some or all of the energy will be absorbed.

One of the more prominent TL designs is the Thor kit speaker that was designed nearly two decades ago by Joseph D'Appolito, for Seas. You can still buy the drivers and plans from Seas. They have a page for kits, and it is there. D'Appolito wrote an article about the Thor design, published by AudioExpress and available on-line: http://www.seas.no/images/stories/diykits/thor/audioXpress_Thor_Review.pdf

 

[haraldo]

Thanks to all of you for this good reading in this thread

Interesting question about what a transmission line speaker is intended to be or to accomplished. Another way to formulate the question would be to ask what theoretical speaker model does the term "transmission line" refer to. I'm not confident that there is a single answer, i.e., it seems that different people who build "transmission line" speakers do not all have exactly the same goals and expectations, or even precise agreement with respect to what it is, beyond the fact that the acoustic energy within the enclosure is expected to propagate down an interior channel where some or all of the energy will be absorbed.

One of the more prominent TL designs is the Thor kit speaker that was designed nearly two decades ago by Joseph D'Appolito, for Seas. You can still buy the drivers and plans from Seas. They have a page for kits, and it is there. D'Appolito wrote an article about the Thor design, published by AudioExpress and available on-line: http://www.seas.no/images/stories/diykits/thor/audioXpress_Thor_Review.pdf

I wonder if there are as many approaches to TL designs as there are designers of TL loudspeakers. Pat McGinty of Meadowlark audio favoured a very open approach with some kind of labyrinth that he nicknamed BASSIC. His main idea with an extremely low level of internal damping is as much as possible to avoid the dynamic constraints which all that damping material gives (according to McGinty). Some of the properties of the BASSIC TL is that McGinty claims that the acoustic resistance that the drivers see is much higer below the tuning frequency of the TL, in order to better keep control on the drivers: I don't have the knowledge to understand if this is fully true or not ... The Meadowlark speakers are anyways real fun and alive and the more you raise the volume the more they lighten up. I have not experienced them as super neutral but this can also be that I just don't have good enough room for them .... working on it.

Again, Stereophile measurements of Meadowlark designs show them to be not very linear at all but I discussed this with McGinty and he claimed that his speakers were measured the wrong way, he said he could easily pinpoint measurements that would make them very neutral... I came across some Meadowlark owners around here and all of them seem to be hanging on to the speakers, even when buying new ones.
All Meadowlark speakers are time aligned and with first order acoustic slopes, claimed to be time aligned and phase linear across most of the passband.

Others like Adyton have been using TL designs for the full range line sources and some sort of acoustic filter with the subwoofers in the same transmission line as the fullrange drivers
Like the iMagic that I showed earlier in this thread. furthermore ... in an extremely narrow cabinet that as far as I know, violate all textbook rules of TL design. I have no idea how it is possible to make it work but to me it seems to work pretty well.
These are also all of them claimed to be phase coherent but I never seen measurements proving this fact

 

[fosti]

.....
One of the more prominent TL designs is the Thor kit speaker that was designed nearly two decades ago by Joseph D'Appolito, for Seas. You can still buy the drivers and plans from Seas. They have a page for kits, and it is there. D'Appolito wrote an article about the Thor design, published by AudioExpress and available on-line: http://www.seas.no/images/stories/diykits/thor/audioXpress_Thor_Review.pdf

By all D'Appolito's honors, this from his article ist not true:
"
I suggested a transmission line,


pointing out that the non-resonant behavior


of this enclosure assured that we


would hear the full capabilities of these


drivers free of “boxy” coloration.
"

 

[KaiserSoze]

Anything that affects frequency response and that is audible by virtue of the effect on frequency response will manifest in measurements of frequency response. This is tautologically true. Whether you are able to identify the anomaly by visual inspection of the response curve is a somewhat different question. But if the anomaly, whatever its nature, audibly affects the frequency response, there will absolutely be evidence of it in the measured frequency response.

As for how you know whether a spike observed in the frequency response is caused by a cabinet resonance vs. something else, this is not easy in the case of panel resonances, because the affected frequencies are inherently difficult to predict. But the other kind of enclosure resonance, i.e., standing waves that set up inside the enclosure, are straightforward to predict given the interior dimensions. If you observe a pronounced spike at some specific frequency and it is so strong as to suggest a resonance of some sort, and if the corresponding wavelength is one of the wavelengths for which standing waves are predicted in accordance with the interior dimensions, there is then very high likelihood that the observed response spike is associated with the standing wave. ... .

Edit: The particular algorithmic formula, by which the wavelengths and frequencies for standing waves that set up within an enclosure may be calculated, was not so very pertinent to the points I was making. Nevertheless it has been brought to my attention that I screwed it up, in using the 1/4 wavelength, odd integer rule. The essential fact is that a pressure antinode must exist at both hard reflective surfaces, in order for a standing wave of a particular wavelength to set up between the hard reflective surfaces. But it doesn't really matter whether it is a node or an antinode, because it must surely be the same for both hard surfaces, i.e., either a node at both surfaces or an antinode at both surfaces, and regardless of which of the two happens to be correct, 1/2 of a wavelength must match the distance between the two surfaces (or any integer multiple of 1/2 wavelength). The distance D = n x lambda/2 => lambda = D x 2/n where n is any integer, even or odd. Note, moreover, that the frequencies of "room modes" are a less straightforward calculation. The equation may be found here: https://en.wikipedia.org/wiki/Acoustic_resonance, a Wikipedia page that provides a good mathematical explanation of resonance. In the simplistic case where all three dimensions are the same (i.e., a cube), the wavelengths (for various integer values n) are reduced by the factor approximately .58 (the reciprocal of the square root of 3), compared to the corresponding wavelengths (corresponding values of n) for a single pair of hard reflective surfaces.

I wish to note that the web page that was referred by ctrl in his comment pointing out my error is a web page (sengpielaudio.com) that I have previously invested some time in and that I disdain. The problem is partly that the English translation from German was apparently done using a bad robot, but this only scratches the surface even if it happens that the translation is the primary reason that it is so difficult to make sense of it. Of course it makes perfect sense to someone who doesn't need the explanation; this is generally how it is with bad explanations. A good amount of effort was invested in the individual graphics but precious little effort was invested in the organization of the page or in the written explanations. There are content boxes with borders where the key piece of information that identifies the scenario to which the information in the box applies is located at the bottom of the box. Inside these boxes are found many run-on equations where the "=" sign occurs multiple times in a single mathematical expression, which is improper. And the explanations are just not good, well-written explanations. Another page on the same site endeavors to argue that sound power is the "cause" and that sound pressure is the "effect". This manner of differentiating cause and effect is wholly contrived; the page where this is done is gibberish. The sengpielaudio.com page says that this Wikipedia page (http://en.wikipedia.org/wiki/Standing_wave) is "Not good". This Wikipedia page provides a mathematically correct, generalized derivation of standing waves. A key feature of this derivation is the use of the sum-to-product trigonometric identity to obtain the characteristic waveform equation for standing waves. Through this analysis it is demonstrated mathematically that nodes occur at even multiples of a quarter wavelength and that antinodes occur at odd multiples of a quarter wavelength. (Pressure amplitude is always 0 at nodes, whereas at antinodes peak pressure is twice the peak amplitude of either the incident or reflected wave alone.)

Both of the Wikipedia pages to which I have given links (https://en.wikipedia.org/wiki/Acoustic_resonance and http://en.wikipedia.org/wiki/Standing_wave) are well-written, well-organized, and mathematically meaningful.

 

[KaiserSoze]

By all D'Appolito's honors, this from his article ist not true:
"
I suggested a transmission line,


pointing out that the non-resonant behavior


of this enclosure assured that we


would hear the full capabilities of these


drivers free of “boxy” coloration.
"

Could you please elaborate further?

 

[fosti]

Hello,
yes of course. Have a look at the impedance curve of the raw driver an then measured in the TL. The peaks show clearly that the TL has not a "non-resonant" behavior.

 

[KaiserSoze]

Hello,
yes of course. Have a look at the impedance curve of the raw driver an then measured in the TL. The peaks show clearly that the TL has not a "non-resonant" behavior.

In fact in the article he discusses how tapering the TL and using a lot of damping controls the ripple. To give him the benefit of the doubt, he was thinking of something meaningful and more specific, but not wanting to go into greater detail he said something that may not be correct depending on how it is interpreted. If course he was talking up the then-new Seas drivers and talking like a marketing rep. He seems to work both sides of the engineer / marketing spokesman fence with equal aplomb.

 

[Kravi4ka]

Kind of off topic for which I apologise but the Joachim Gerhard Surveyor is simply fantastic! I hope he manages to send one for testing so I can brag about it in the neighbourhood but it does sound significantly better than the Dynaudio Special 25, Special 40, Confidence one, Harbeth 40.2 Anniversary, Dutch&Dutch 8C(magnificent!), I also have more than 30 pairs of speakers built following designs from TonyG, Troels, JonMarsh and Linkwitz featuring all kinds of most expensive drivers and topologies.

It has the spooky 3D of the LX mini, it goes almost as low as the Harbeth, it has the stunning wide and deep soundstage of the 8C. The only speakers that I would trade it for is the Dynaudio Evidence Temptation that a friend has, in the following weeks I am expecting a JBL M2 set to test it against but it will take time before I can do it.

I was fortunate enough to listen to Doug Sax(RIP) working on Ray Charles' Genius Loves company and DSOTM for DSD so I have a decent idea on how things are supposed to sound and the Surveyor is all I need now. I am sure you can challenge Joachim here and there with a technical paper and a quote or two with no proof but I think he is well beyond papers, if you look at his shop he has probably hundreds of prototypes and drivers tested just for the sake of testing an idea. At some point the theoretical proof for someone with his experience and knowledge is not a challenge anymore, he makes his living on good sounding speakers and from what I have compared it to they are an extreme bargain, if given the proper conditions.

No affiliation with him, just owning 3 pairs of his speakers.

 

[Absolute]

Well, that certainly sounds intriguing! The comparison with M2 will certainly be interesting to read about, so I hope you'll share that with us.

 

[Kravi4ka]

I will for sure, it will take months before I can accommodate those beasts in the room. A meaningful comparison will be difficult as they represent quite a challenge to put in the same room though and give them optimum conditions. The Dutch and Dutch 8c was easier as it is a bit smaller. The 8C plus Weiss MAN301 was the most impressive and surprising combination in my 30 years of pursuit of musical enjoyment. mitchco's articles could have saved me literally many tens of thousands if I had read it 20 years ago as well.

If I were to start again I would simply buy the 8C and be done with it, I find it ridiculous that people spend so much on cables, fuses and magic tricks when the speakers are just placed somewhere looking good and they say they hear "improvements". Sure there are good things about a better DAC or a better amp but just moving the speakers around can give us so much more.

I try to improve my understanding of the recent advances in technology and science but I do enjoy making decision on comparing equipment as well, sometimes in a completely non scientific way yet critical of the way I do it. Somehow the Surveyor has a certain magic that the 8C has less of( and I have had review darlings that completely lack the emotion inspiring sparkle) and it will stay for the foreseeable future

 

[Andreas007]

No, nothing came.

Amir, any news regarding speakers from Mr. Gerhard?
From what I’ve read here we are supposedly ignoring one of the most competent speaker designer...

Well, still waiting...

 

[haraldo]

I will for sure, it will take months before I can accommodate those beasts in the room. A meaningful comparison will be difficult as they represent quite a challenge to put in the same room though and give them optimum conditions. The Dutch and Dutch 8c was easier as it is a bit smaller. The 8C plus Weiss MAN301 was the most impressive and surprising combination in my 30 years of pursuit of musical enjoyment. mitchco's articles could have saved me literally many tens of thousands if I had read it 20 years ago as well.

If I were to start again I would simply buy the 8C and be done with it, I find it ridiculous that people spend so much on cables, fuses and magic tricks when the speakers are just placed somewhere looking good and they say they hear "improvements". Sure there are good things about a better DAC or a better amp but just moving the speakers around can give us so much more.

I try to improve my understanding of the recent advances in technology and science but I do enjoy making decision on comparing equipment as well, sometimes in a completely non scientific way yet critical of the way I do it. Somehow the Surveyor has a certain magic that the 8C has less of( and I have had review darlings that completely lack the emotion inspiring sparkle) and it will stay for the foreseeable future

I think Joachim is one of the real great masters ... all the speakers he designed, that I auditioned, were just unreal, almost beyond belief. That was many years ago in the Audio Physic days of Gerhard; Now time has passed and drivers of today are way more capable than 10 to 15 years ago. I have committed myself to go from Norway to Brilon in Germany just to listen to what his products are capable of now. I would consider myself a complete idiot if I bought statement speakers without listening to the upscale Joachim Gerhard products. (just difficult logistics in these Corona days)

I have also no affiliation and no connection other than being great admirer of Mr. Gerhard legacy designs.
The only affiliation I can think of is my father's middle name: Gerhard

 

[Kravi4ka]

This is his Progress speaker(the almost too thin black columns in front of the Harbeths) mercilessly murdering the 40.2 Anniversary, after that session my friend bought the Beo LX which is simply on another level... The Surveyor is way better. I am really happy that I read that Stereophile review on the Sonics Anima which introduced me to his designs, recently he found a magnificent woodworker to compliment his designs.

 

[haraldo]

This is his Progress speaker(the almost too thin black columns in front of the Harbeths) mercilessly murdering the 40.2 Anniversary, after that session my friend bought the Beo LX which is simply on another level... The Surveyor is way better. I am really happy that I read that Stereophile review on the Sonics Anima which introduced me to his designs, recently he found a magnificent woodworker to compliment his designs.

Way cool

There is something very special about these slim, narrow speakers, imaging is insane

I do have the Suesskind ARA on my radar
(I think it's the best speaker Joachim Gerhard does at the moment)

 

[Joachim Gerhard]

This is high praise , many thanks. I plan to send Amir a pair of Surveyor as fast as i can. At this moment i have only one pair for review and it goes out to another person that i had promissed to do that before. One issue is that nearly all speakers we can make go out the door fast. Markus, my furniture maker also has a tough day job and we do not want to rush too fast. We are ramping up production though slowly.
I will publish a set of measurements until the end of the week though.

 

[Joachim Gerhard]

Also because this is a tread about time domain measurements i will show how i make measurements in this domain.

 

[vavan]

If I'm not mistaken surveyor utilizes purify driver and Russian viavawe twitter, probably the best ribbon ever?