Tekton M-Lore Speaker Review

[KSTR]

Looking at how the NFS mechanism work, I may have been mistaken. It looks to be moving the microphone up and down without changing axis (I was assuming it tilted as it moved around somehow). So it actually did capture the on-axis measurement of the woofer as well if I understand it correctly. I was assuming the microphone tilted as it moved around the speaker. So you actually don't need to remeasure. It's a matter of just selecting what is reference axis from the existing measurement/data?

There is a small error coming the directivity of the measurement mic at very high frequencies (it loses its omnidirectional properties) but that is usually not significant as it affects the upper HF response only at very large vertical off-axis angles. In a perfect world the Klippel NFS would tilt the mic so that it always points directly to the tweeter, mechanically. Post-correction (when the angle and the directivity is known, the impact can be factored out mathematically) would be simpler, though.

 

[VintageFlanker]

My understanding is that Amir has given Eric plenty of chances to explain what was wrong here, so I am not saying Amir did anything wrong.

The issue, @sigbergaudio, whoever would be "right" or "wrong" on this matter, is that data has been contested in the first place. And as far as I'm concerned, I cannot see how data could be contested without other data. That is just basic logic for me.

Feel also kinda sad for @Eric Alexander at this point. The damage will probably be permanent and yes, we're beating a dead horse. Yet, I have to remind that Erin has been the first victim of all this mess, and that should never been ignored.

You could either apologize to @amirm and Erin Hardisson and go on with your life and company. People tend to love when other people admit having made a mistake

This.

I don't understand why there are still speaker designers without the Klippel NFS

I do. That means 100k$/€ + a dedicaced (yet not treated) room + the needed expertise for using and interpret it.

 

[markus]

As you guessed later, Klippel NFS fully captures the response of the speaker in 3-D space. The reference axis is just a presentation value. All the data is captured independent of it.

I just went through and recompuated the reference axis being 5 inches higher. Review is updated with more detail. But here is the key results:

And this is exactly what I said to Eric. He simply said he disagrees but provided no comparison like I am showing above. Or explanation of why that would make a difference in far field listening. Or how a customer was supposed to sit to have their ear constantly be at that precise angle.

Above is why I don't jump when the manufacturer says this and that is wrong. The responsibility is theirs to prove their point, not just claim it. Otherwise, it winds up making me do a bunch of busy work per above.

I am not guessing at above. As I have mentioned, I have made this adjustment many times with similar results. Unless you are moving way higher or lower, a few inches doesn't make a difference.

Thanks for doing this Amir. Now we can scratch "botched measurements" from the list of accusations.

 

[sigbergaudio]

Not Amir but yes, that's correct. I don't understand why there are still speaker designers without the Klippel NFS. Do they all have anechoic chambers linear down to 20Hz or do they drag out their speakers to measure them in a real free field environment?

Well, it is pretty expensive. I can only talk for myself, and I happen to live 20 minutes from Seas. They have an anechoic chamber that we use. It is sort of the opposite of an NFS in the sense that the microphone is stationary, and the speaker is placed on a rotating pidestal/stand, and it is connected to Klippel measuring software. So there is more manual work involved than with an NFS, but you can get comparable accuracy.

We also do use free field / outdoors measurements, which works fairly well too.

 

[markus]

I do. That means 100k$/€ + a decided (yet not treated) room + the needed expertise for using and interpret it.

And that shows whether you're really serious, professional and knowledgeable as a loudspeaker manufacturer or just a crazy enough DIY guy.

P.S. My lawyer is waiting for your lawyer to have a 20 minute "chat" about my comment above

 

[sigbergaudio]

And that shows if you're really serious and professional as a loudspeaker manufacturer or just a crazy enough DIY guy.

I don't think that is fair. You don't need 100,000USD NFS to create good speakers. Many manufacturers (in the business of speakers or other things) started out literally in their garage, moving on to become very successful. I'm sure many were viewed as crazy to begin with.

 

[amirm]

Well, it is pretty expensive. I can only talk for myself, and I happen to live 20 minutes from Seas.

US distributor of Klippel will do CEA-2034 scan for about $2000. There is no excuse for any company bigger than one person to not pay for this service.

 

[markus]

I don't think that is fair. You don't need 100,000USD NFS to create good speakers. Many manufacturers (in the business of speakers or other things) started out literally in their garage, moving on to become very successful. I'm sure many were viewed as crazy to begin with.

There are plenty of examples like the Dutch & Dutch which started as a project at diyaudio.com. Sure you can produce great loudspeakers and concepts with knowledge and workarounds but I'm talking about loudspeaker design and manufacturing as a business that needs to sustain yourself and all the people involved running the business. When you take everything into account 100k is not that much. Not everyone has SEAS around the corner. You can consider yourself very lucky.

 

[solderdude]

There is a small error coming the directivity of the measurement mic at very high frequencies (it loses its omnidirectional properties) but that is usually not significant as it affects the upper HF response only at very large vertical off-axis angles. In a perfect world the Klippel NFS would tilt the mic so that it always points directly to the tweeter, mechanically. Post-correction (when the angle and the directivity is known, the impact can be factored out mathematically) would be simpler, though.

While I agree about the mic angle I don't think that is a problem here, in fact it is the opposite in this particular case.

With the Klippel the mic is raised and lowered in horizontal position. The angle is calculated (distance and height) so for a spherical dome measurement you would be correct.
However, the reasoning was that the mic should be at the height of the woofer (and horizontal mic) which is exactly what Amir's measurement is. You actually would not want a spherical measurement in this particular case.

I am not guessing at above. As I have mentioned, I have made this adjustment many times with similar results. Unless you are moving way higher or lower, a few inches doesn't make a difference.

About the timing between tweeter and mid/woofer (step response). I think the delay time (XO filter phase dependent as well) difference Eric states is wrong could really only be 'perfect' on a certain distance and height. The question is whether or not that plot will change much when measured on axis of the woofer.

I reckon @amirm could also show the difference in step response between the 2 scans at different heights or is that not possible ?

 

[Snickers-is]

I think this debate about the NFS also should include if this is the type of measurement you as a manufacturer wants as your tool to make a great speaker. I would say no to that, both because it takes so long to measure a speaker, and because I typically look for different data than what the NFS shows. A bit the same as when I develop a driver, I use simulations that give me far more information than a Klippel large signal analyzer would. But the Klippel analyzer is useful to see that the driver works as intended, or that the coil is centered properly.

 

[Robbo99999]

As you guessed later, Klippel NFS fully captures the response of the speaker in 3-D space. The reference axis is just a presentation value. All the data is captured independent of it.

I just went through and recompuated the reference axis being 5 inches higher. Review is updated with more detail. But here is the key results:

And this is exactly what I said to Eric. He simply said he disagrees but provided no comparison like I am showing above. Or explanation of why that would make a difference in far field listening. Or how a customer was supposed to sit to have their ear constantly be at that precise angle.

Above is why I don't jump when the manufacturer says this and that is wrong. The responsibility is theirs to prove their point, not just claim it. Otherwise, it winds up making me do a bunch of busy work per above.

I am not guessing at above. As I have mentioned, I have made this adjustment many times with similar results. Unless you are moving way higher or lower, a few inches doesn't make a difference.

This seems rock solid and would settle the matter in my eyes. As long as the 5 inch difference is indeed the difference in alignment height between woofer & tweeter, and as long as there are no possible loopholes/problems with "resetting" the alignment position after the measurement has been done. He mentioned the feet were not included in your measurement in his video, but I don't know if that's the case or not.

 

[Newman]

@Eric Alexander, …Would you be willing to send a pair of your highly acclaimed speakers to Amir for further review? If you feel that Amir is bias since he already reviewed one of your speakers, perhaps send a pair to Erin from Erin's Audio Corner, since he has never reviewed any of your speakers (at least not to my knowledge)?

Whoa! So you started this?

 

[restorer-john]

I think this debate about the NFS also should include if this is the type of measurement you as a manufacturer wants as your tool to make a great speaker. I would say no to that, both because it takes so long to measure a speaker, and because I typically look for different data than what the NFS shows. A bit the same as when I develop a driver, I use simulations that give me far more information than a Klippel large signal analyzer would. But the Klippel analyzer is useful to see that the driver works as intended, or that the coil is centered properly.

There is no no "debate" about the Klippel NFS's efficacy. Don't attempt to intimate there is.

And a Klippel analyser is not required to determine the concentricity of a voice coil is it? A Dayton DATS V3 can do that for $129.98 (USD)...

Throwing shade isn't going to happen.

 

[AscendDF]

.... I happen to live 20 minutes from Seas. They have an anechoic chamber that we use. It is sort of the opposite of an NFS in the sense that the microphone is stationary, and the speaker is placed on a rotating pidestal/stand, and it is connected to Klippel measuring software. So there is more manual work involved than with an NFS, but you can get comparable accuracy.

We also do use free field / outdoors measurements, which works fairly well too.

I called in a favor with SEAS a while ago and sent them a speaker to measure in their chamber to compare with our Klippel NFS measurements of the exact same speaker. Accuracy was not only comparable, but shockingly close all the way down to the low frequency limitations of their chamber, where the accuracy of the NFS is superior. As I recall, the measurements were within about +/- 0.5 dB of each other, possibly even tighter but I don't have the time to look this up right now.

I won't comment on Eric's handling of this situation, but it would be in his best interest to understand and learn that the resolution of the NFS is far greater than any measurement device he has ever used. Because of this, it is highly likely he does not even recognize what the true response of his speakers actually are. That stated, he must accept the FACT that the NFS measurements do show precisely how these speakers actually measure, and they really aren't too bad, considering the resolution and accuracy of the device.

 

[thewas]

I called in a favor with SEAS a while ago and sent them a speaker to measure in their chamber to compare with our Klippel NFS measurements of the exact same speaker. Accuracy was not only comparable, but shockingly close all the way down to the low frequency limitations of their chamber, where the accuracy of the NFS is superior. As I recall, the measurements were within about +/- 0.5 dB of each other, possibly even tighter but I don't have the time to look this up right now.

Exactly, many great measuring loudspeaker companies like Genelec, Neumann, KEF, Revel etc don't have or use a NFS and their measurements still match perfectly, even hobbyists at home can make very close measurements as shown here.

 

[AscendDF]

I think this debate about the NFS also should include if this is the type of measurement you as a manufacturer wants as your tool to make a great speaker. I would say no to that, both because it takes so long to measure a speaker, and because I typically look for different data than what the NFS shows. A bit the same as when I develop a driver, I use simulations that give me far more information than a Klippel large signal analyzer would. But the Klippel analyzer is useful to see that the driver works as intended, or that the coil is centered properly.

I think you might be confusing the Klippel NFS with one of the many other Klippel measurement devices.

 

[sigbergaudio]

US distributor of Klippel will do CEA-2034 scan for about $2000. There is no excuse for any company bigger than one person to not pay for this service.

Sure, I commented on not OWNING a Klippel NFS.

 

[markus]

Above is why I don't jump when the manufacturer says this and that is wrong. The responsibility is theirs to prove their point, not just claim it. Otherwise, it winds up making me do a bunch of busy work per above.

Being sceptic is at the core of the scientific mind. Sceptic about what we know, evidence, what is presented as facts. Especially if it is our own knowledge, findings and understanding about what is real. And yes, it's frigging laborious and tedious and no fun!

 

[sigbergaudio]

I called in a favor with SEAS a while ago and sent them a speaker to measure in their chamber to compare with our Klippel NFS measurements of the exact same speaker. Accuracy was not only comparable, but shockingly close all the way down to the low frequency limitations of their chamber, where the accuracy of the NFS is superior. As I recall, the measurements were within about +/- 0.5 dB of each other, possibly even tighter but I don't have the time to look this up right now.

Interesting and good to know! No reason to expect great differences between our own published measurements and something done by Erin or Amir then (if someone were to send in speakers).

 

[sigbergaudio]

Accuracy was not only comparable, but shockingly close all the way down to the low frequency limitations of their chamber, where the accuracy of the NFS is superior.

If you are bored at some point it would be really interesting to see the deviations in the low frequency at Seas compared to the NFS, then I could theoretically correct for that.