Does It have to be loud?

Another point is that I was talking about SPL at the listening position. (in my case 4.0 meter)
I now realize that some measure 107 dba at 2 feet.

@han_n

Hello han. Just letting you know that I’m aware that you are referring to sound pressure at the listening position. You clearly stated that as the first post in the thread.

However.....

In a subsequent post you made in this thread, dated 03-04-2018 8:57am, you wrote "....people admit listening at levels up to 100 or more dba at the listening position. I’m not sure how they measure, but believe me, most speakers won’t even go that loud, or if they do, they will distort tremendously...."

I took that to mean that you are now stating that most speakers cannot achieve 100 db without tremendous distortion, irrespective of the listening position. And this is simply not true, despite the Soundstage measurements submitted by shadorne as evidence.

And, I previously responded to rebut the measurements by stating that (a) the distortion is at the high frequency range only and is in fact linear at the lower frequencies, which has substantially greater impact on the ability to produce high sound pressure levels, (b) "tremendous" can be a matter of opinion, and if one believes a 1db deviation from linear is tremendous then that’s okay too, (c) live rock music at high levels also features a degree of distortion and (d) maximum speaker sound pressure level can be inferred from manufacturer specifications of sensitivity and maximum power handling capability. Now if I were to use manufacturer specification applicable to my speakers, Golden Ear Triton Reference, I could conclude that they could produce (approximate) 125db! Of course, I do realize that likely isn't possible (and certainly not practical), however, reducing that number by say 10db (to 115db) I do believe is very doable. And I’m sure there are many other speakers that can produce those levels as well.

@gdhal

The Triton Golden Ear Reference seems to have a linearity problem at quite modest volumes (90 dB). It looks like the AMT ribbon suffers from the problems that have always plagued these designs - lack of ability to produce even modestly high SPL without distortion. Infinity solved this problem by using arrays of the AMT but Triton Ref only has one.

You can see what I mean on these plots from Soundstage particularly the deviation from linearity.

http://www.soundstage.com/index.php?option=com_content&view=article&id=1770:nrc-measurements...

Of note also is the comb filtering of the AMT tweeter above 4KHz - this is due to the fact that the diaphragm is too large to produce these higher frequencies without interference and hence the comb like appearance of the frequency plot (lots of wiggles). The advantage of comb filtering is it can give a pleasing expansive effect - electric guitar players use it all the time to make guitars sound big - it makes it hard for the ear that detect the origin of the sound and will reduce pinpoint Soundstage on a stereo but conversely will make higher frequencies even bigger and expansive.

The Stereophile plots better show the comb filtering. Note that off axis is the first to be affected which is because of the different distance the sound travels from either side of the large (compared to a conventional tweeter) diaphragm of the AMT.

https://www.stereophile.com/images/118GETRfig4.jpg

@shadorne

I’m hopeful to have a more technical and comprehensive response to your 03-05-2018 12:34pm post.

However, at face value I see no issue. In fact, the non-linearity data that you are now reporting I previously disclosed.

As I look at the graphs (both 90 and 95 db "Deviation from linearity" charts), I note the left margin in both cases represents decibels, *and the resolution is in 1 db increments*! So yes, you are correct in your observation of the data, but incorrect in your interpretation of the data.

The largest deviation from linearity is not more than 3 db, and this is at roughly - and only - 3500Hz. The linearity is basically flat (less than 1 db deviation) from 50Hz all the way through roughly 2500Hz.

In any case, as I stated, I shall attempt to obtain a response beyond that of my own.

This seems to be yet another case where you are passing judgement solely on the basis of measurement data (instead of considering the sound), as you do with your DAC.

Have you heard the Triton Reference speakers?

EDIT:

By the way shadorne, as I peruse the soundstage site for measurement data of speakers other than the GET T Refs, I note there are none with perfectly flat linearity. For example, the focal sopra 2, which costs a kewl 5 grand more than the T Refs, also has linearity issues - by your criteria. http://www.soundstage.com/index.php?option=com_content&view=article&id=1445:nrc-measurements...

@shadorne 

As I promised to obtain a response beyond that of my own, the following comes courtesy of Golden Ear forum moderator.....
---------------------
The "non-linearity" he's referring to, and the associated distortion, is extremely low, inaudibly low, and insignificant compared to ALL the other benefits of the HVFR tweeter. Both the magazines that this data comes from LOVED the speakers and said nothing about this insignificant issue in their reporting of the measurements.

Isn’t the assumption that you should demo for the typical SPL that a listerner would hear were she/he attending a live show?

By that standard, MBL (which typically demos louder than anyone else I know of) doesn’t go loud enough. They often play electric blues rock at (my guess) +/- 95 DB. Very loud, but not live loud.

It may not be great for extended audition, but it does demonstrate that the volume levels in your home will likely be limited by the proud new owner’s choice, rather than the speakers’ capabilities.