My definition of Ambience

[Scroobius]

Peter - this is very interesting stuff thanks for posting. I thought the original FM curves had been later found to be incorrect due to (possibly) measurement errors(?) Anyway it all brings back memories of that "loudness" button I used to have on my first amplifier.

It was not that long ago that I put a 20Hz (and 30 and 40 etc up to a few hundred) test tones through my ANE's and was surprised at the healthy output at 20Hz and also some surprise that it seemed slightly higher output than at 30 and 40 Hz. So now I know - I was not listening to 20Hz at all. Just shows how easy it is to get things completely wrong in this hobby.

Paul

[christoffe]

For information

http://www.audiomatica.com/wp/wp-content/uploads/2012/03/appnote_007.pdf

[PeterSt]

Joachim - Thank you. To me this shows nothing different than that picture in my yesterday's post and that indeed an official loudspeaker measurement tool (CLIO) contains an as official test for loudspeaker distortion.
In the mean time I understand less and less of it all.

All I derive from it (all) is that my standards are (way) too high.
Well, so be it.

Peter

[PeterSt]

One other thing :

What I also see from measurements like these (throughout) is fairly excessive 3rd harmonic distortion. Often higher than the 2nd. FWIW : I don't measure any 3rd here. Nothing.

With the quote of "3rd is more important to us", I (or we all) may wonder what that's really worth. To me ? nothing much exciting. But why ?

Any 3rd harmonic is about "square" distortion which in itself means that it is about the theoretically necessary infinite amount of frequency to represent that. So yes, looked at it from this angle, any 3rd will imply an infinite number of odd harmonics, knowing that in the end something won't be able to cope. Of course this is the same when any instrument puts out a more square tone, but let's say that this belongs to the instrument and only needs more sampling rate (to represent that high frequency well). And maybe superfluously : only when indeed infinite frequency is available a square can be a real square.
Anyway, when a 3rd expresses while not any square is in the music itself, you will perceive that as infite harmonics (all thus odd). Could be sibilant.

Why do I not measure even one 3rd (and thus also not beyond) ? I don't know, but it should be about cabinet reflections (no cabinet there) and otherwise room reflections (no significant room reflections there, derived from no standing waves there). In addition : I think that any reflecting wave will cause square like response (or at least it will be deformed at the reflecting surface).

Might it be important : I use continuous tones only, which -if I understand that manual properly- should only be for the worse because of driver heating up etc., such a test assuming that no longer continuous tone is playing in music, or whatever such a test anticipates upon.
Tell that the organ player.

Peter

[pedal]

Jøkleba LIVE. Its free jazz With awesome SQ. Note some old fashion synth generated looow frequency notes.

Oh yes ! So first there's the more heavy drum with its so nice spread from its peak at 48Hz to the to the left and right (same slope) which creates that ambient sound full with air and next there's indeed what must be a synth with its lowest notes at 18Hz. Notice though that from the latter you'll only perceive its sqaure based higher harmonics but in my view the perceived deep warmth again is this 18Hz.

Maybe I should tell that you can be considered to have a real subwoofer ? I mean, a sufficient amount of drivers in there, but I forgot the size and amount (per side).

Btw a message from Ciska here :
Gry won't be happy with this album. Nor does Ciska. But didn't we know that ...

Best regards to you guys,
Peter

Another wife-buster is this one:
Misha Alperin, Night (ECM)
-A jazz album recorded live at Vossa Jazz Festival, Norway.

Track "Heavy Hour": Here you will hear subsonic bass on a completely new level. The whole stage is resonating through wooden floor and everything. Certainly demo-gallery stuff! (But lock in the children first - this is terrifying music!).

[CoenP]

Hi Peter,

Thanks for sharing your ponderings on the sub frequencies!

Coen, what is this 35Hz based upon ?

Well, it's based upon me being practical on classic horn theory. For -resistive- loading down to the Fc you need at least a 1/4 of the wavelength as horn length. So that will be about 2,5 meters for 35 Hz. Even with a 1/8 space loading the mouth area will still be about the size of an AN-E. Iow these will allready be HUGE. If you don't follow the rules you are either stuck with severe peaking/notching or reduced loading (ie derate the driver and EQ whats missing). Didn't Bert have such a monster basshorn in his showroom?

Now the Orelinos are somewhat of a horn, but they are also something else. On classic theory they won't qualify neither on horn length nor on exitsurface as a horn for an 18Hz tone...
They do have a lot of surface though to move air, about the equivalent of an 18" (or two 15") for the piston. For shure they act as a waveguide given the silece behind them.
It remains a very intriguing desing that seems to work very well.

What I also see from measurements like these (throughout) is fairly excessive 3rd harmonic distortion. Often higher than the 2nd. FWIW : I don't measure any 3rd here. Nothing.

This is really odd  ! Because of the drive symmetry between forward and backward movement I would expect a perfectly executed driver to have way more 3rd H distortion than second. Usually they are in the same ballpark. If I am correct this is also the case for the infinite baffle measurements on the driver's datasheet.
Some cancellation, ok, but NO third???

regards, Coen

[PeterSt]

Hi Coen,

Forgive me, but I certainly don't know everything. In this case I have no clue why a loudspeaker (driver) would excess on the 3rd harmonic per se, while to me any 3rd harmonic only tells about "square" distortion. And "drive symmetry" would incur for distortion in the first place ? All I can see is that when a LS driver would be able behave linearly a nice sine would come out of it without distortion. But a LS driver is not linear (and can not be) so to keep it simple, the tension growing stronger at the outsides of the excursion would create a 2nd harminic distortion when the increase of tension itself goes in a linear fashion. And let's say that to some degree of excursion it will. After that anything can happen and e.g. the moving mass against the increasing tension may not go in linear fashion at all and you get unlinear "holding backs". That's 3rd (more square).

Well, at least I will admit that LS distortions are too complex for me and many exist. But I also have the rather clear idea that maybe not many understand it in the first place, because I see too many strange things written. For example, I just ran into the "fact" that doppler distortion would need two frequencies to be present in order for that distortion to be (audibly ??) present, and personal me does not believe a thing of that. Maybe two frequencies would make it better audible indeed, but the distortion itself is technically there with one frequency already. And if that is not the case then I not only don't understand much, but actually nothing.
Just an example.

It is also clear to me that intermodulation distortion is much more important than just harmonic distortion, which is just the nature of a loudspeaker driver of any kind. Here, we better use a 10-way system instead of one wide ranger and all it takes is perfect crossovers. Do those exist ? maybe. But do speaker drivers exist with 0 IM distortion ? of course not.
What is worse ?
... And again I find nothing. It all is derived to our auditary system and that we will be able to filter out distortions and blabla. Well, I don't believe one f*ck of that as long as the same persons or instances write that 10% HD is also perfectly allowed. In combination it now is about "but IM overwhelmes anyway, so why bother with HD". And in the mean time nobody knows how that IM is perceived, which of course is a tad hard to test while it needs to be about music and not test tones.

Anyway, after diving into this matter somewhat I am shocked to find so much unknowledge, to which I am not used. And then to think that a great deal of my audio life is about finding whatever devices with 0.00001% THD at nano volt noise levels. What to do this for ? Ah, because that is so much amplified.
And still all I try to achieve at that level is audible. Hmm.

I have said it so many times before : there is so, so much wrong with audio that we can improve for the rest of our lives, our grand children included.
But that is nice of course, because now we can keep on working on our hobby.

Peter

[PeterSt]

Another wife-buster is this one:
Misha Alperin, Night (ECM)
-A jazz album recorded live at Vossa Jazz Festival, Norway.

Track "Heavy Hour": Here you will hear subsonic bass on a completely new level. The whole stage is resonating through wooden floor and everything. Certainly demo-gallery stuff! (But lock in the children first - this is terrifying music!).

Hey hey ...

First off I had this album already in my Nice Stuff Gallery.
Next, I thought to play it from the beginning so possibly nobody would notice any nastyness (I regard this album almost experimental classical, might that genre exist).

Then, oh boy, exactly when this track has started 10 seconds I first got the remark from someone here "I reckon this is an advice from pedal I suppose ?". Followed by a "That's what I thought as well !". Ok ok ...

However, this time I have a kind of another remark :
With the notice that the large drum hits reached a sheer 107dBSPL and seeing that the very soft regular large-drum hits easily go beyond 20Hz ... no such "distortion" you imply is anywhere. And do notice that I exactly know what you are referring to, were it about all of the room resonating in a sort of self-mode (waiting for the building to collapse *really*).
Nothing of the sort.

So what's up here ? The 20Hz and beyond for sure is there and at levels of maybe 90dB (I didn't really check, but with the meter at 100 I saw nothing moving for those softer large-drum hits). So careful, because when 20Hz shows its second, thus 40Hz, this really is the rattling of a perceivedbly soon collapsing building.
And of course, when those 107dB hits of the loud drums (between 50-60Hz - I didn't really pay attention) go wrong then all can happen.
And the "length" of your room ? well, that for sure would incur for similar when at the wrong frequency by accident. So, 7 meters on estimate ... 344/7 = 49Hz. And I think I recently told about collapsing behaviour of a track with gliding frequency which works for my room ...

Sorry for the maybe too long post, but always interesting to see how we perceive things and how they possibly are not "real". Possibly ! ...

Peter (open for suggestions)

[PeterSt]

PS: Let me be clear : There is nothing wrong with the "effect" of the longitudinal wave virtually making the room move back and forth (if that's the effect you perceive in the first place of course) and at least for the track that works so in my own room it's a sheer gag. So this is only about "reality" and nothing about things being wrong really. So if this is perceived as e.g. wooden floor vibrating and all you can imagine, why not !

And I can tell you : such an effect is the most precious I have in my "demo" collection. It makes you dizzy and even hallucinate.
Btw, might you (or anyone) want to get it, this is Easily Embarrased - Tales of the Coin Spinner, track 02 - "Blessed Day on Distorted Shape".
I just played it to look at it through FFT and the (stepped) gliding frequency goes from 65Hz to 31Hz and is about 20dB above the normal music (sounds). Beyond that 31Hz is totally nothing. My room is 344/12 = 28,67Hz but speakers are 80cm from back wall, so 344/11.2 = 30,7Hz. Ha.
It now just depends on whether one of these steps exactly meet your room length ...

[PeterSt]

I was cleaning up my screen and found a page still there which I planned to post yesterday. Here it is after all :

http://www.neumann-kh-line.com/neumann-kh/glossary.nsf/root/F77C48111116FFBDC12578B20039968C?Open&term=THD+N

I think nothing is wrong with what is told in there. Whether the real truth is something else.

Hmm ... Since this link is about some "glossary" item for loudspeakers, I just thought to look up Intermodulation Distortion in there. Can't find a direct link, but on this page you will find a link to it :
http://www.neumann-kh-line.com/neumann-kh/home_en.nsf/root/prof-monitoring_knowledge_glossary_measurement
Pretty much the same as I said and implied yesterday.

Peter

[PeterSt]

I have an update on the "Beating" :

I have set up one of my synthesisers in order to find the merits of this "beating" and in the end I succeeded in mimicing the behavior easily;

It requires two fairly low frequencies to be close to eachother (and not too low either because the beat can't be heard then) and I found as example 65Hz and 70Hz to work very well. When these are played together, in mid air (thus not in electronics) these frequencies "beat" and the result is an estimated 4Hz of "tone". BUT :

Nothing from my theories happened, hence my theories were wrong.
What happens is that indeed the two frequencies cancel out and add (at this 4Hz frequency) but this is really all. So, the FFT will not show any 4Hz distinct frequency, and all what happens is that the 65Hz and 70Hz both disappear and get louder. The FFT just shows that.

Still what you perceive from it is a low frequency roar which sounds square like but is not. So, no harmonics appear either. Just those two (btw sine) frequencies hopping up and down and the result is a way more deep perceived tone. I think this is just deceiving but the result is "nice" so to speak anyway.

I must add that I tested this through normal PA speakers and regarding the remaining of this topic I may wonder whether there's merit in that. I mean, might these speakers not be able to produce undistorted low frequencies then maybe they don't show any of that at all. And might that be 4Hz, then of course that won't go anyway.
It is difficult to (without changing setup) to test these speakers for distortion, because the synthesiser I use for it is not able to produce clean sines to begin with. Thus not useful to test the result for THD in that speaker as well.

I will continue examining these kind of things.
Peter

[PeterSt]

For example, I just ran into the "fact" that doppler distortion would need two frequencies to be present in order for that distortion to be (audibly ??) present, and personal me does not believe a thing of that.

Well, that happens when you don't know all and just try to imagine a few things. And yes, of course I derived my imagination from that train (with whistle) coming towards you and from there it will "squeeze" the imaginary vertical waves into closer by ones - thus a higher frequency. What I did not realize though was that the speed of the train is to be regarded one frequency (a veeery low one) and that the whistle is that other frequency. So, if the train comes towards you and while being at you could drive backwards instantly, it would mimic the loudspeaker diaphragm coming towards you and moving back from you at a low frequency, while in that same diaphragm plays a higher frequency as well. What remains is my squeezing (and detracting).

Do we understand now finally ?
I hope I now do.

Anyway what it comes down to (and it always has been) is that the more the diaphragm excurts the more doppler distortion will be there. Or better put : the more excursion, the more difference between the forward moving and squeezing of the higher frequency vs the backward moving and detracting. So, the more excurting the more distortion the higher frequency receives and which is just inherent to the process. Also and to be clear : it is not so that for this doppler effect the higher frequency deforms because of the lower it rides on - it is just an in-air effect; the whistle is perfect and on a rock-steady train, but the moving train towards you will let it sound higher in frequency. And when the train has passed (moves from you again) - lower. Perform this in a faster fashion (like 100Hz) and no whistle sounds nice or pure. Is is not off key (like the train always moving to you) but it is sheer distortion because the frequency of the whistle changes 100 times per second back and forth.

Peter

[CoenP]

I find this doppler distortion rather acedemic. You have to have excursion ánd wide bandwidth to theoretically detect this effect. And how do you measure the effect of the pitch variations, how do they show in the Distortion measurement when they are a tiny modulation of the fundamentals? You may integrating the FFT over a time period of a two tone testsignal and see how wide/high the tone comes out of the speaker...

Besides that the "higher" tone's velocity also influences the pitch "lower" one if you wnt to be complete in the explanation .

Is this 4 real? Did you ever hear it and how does it sound?

Regards,  voen

[PeterSt]

Hey Coen,

I ended my previous post with something like "looks like intermodulation distortion" but I scratched that because it is more complicated. But something like that (it will express similarly).

I've been reading a good article about it and from a clear example we can derive that e.g. 28Hz with 600Hz modulated onto that will give a nice deviation of 1Hz to either side the diaphragm is moving. So, the 600Hz fluctuates between 599 and 601Hz 56 times per second (and it is no real IM because that would be continuously there while here the frequency will flutter to either side of the original (600Hz)
The math is related to the velocity of the "woofer" which by itself deviates from the excursion it expresses and further more from the speed of sound. So, because of the former it also depends on the amplitude of the lower frequency.
Notice that I could fairly easily make a simulation of the distortion figure just from the digital file, if only the woofer's excursion (for a given max reference level) is given. Of course this varies per woofer driver.

Once we know where to look for, this should show as side bands just like with IM distortion. I think it expresses in worse THD just like always, although this by itself can not be measured because of two frequencies needed to perform this testing. Also I wonder whether IM distortion figures will be valid because in my view this is not a steady distortion (but wobbles on the (28Hz from the example) carrier.

But you know what ? concensus is already that with any normal efficiency woofer speaker the distortion of this comes to 7% (just look at the FFT). Only efficient woofer horns can reach 1% and expanding the woofer surface decreases it further. All logic because the excursion gets less.

My conclusion so far (and the real knowledged may laugh) is that it is not a deformating diaphragm or failing suspension which incurs for the major (audible !) distortion, but just the simple doppler effect.
And I said "audible !" because normal harmonics like a 2nd are not so audible at all because related (but never forget the hardly audible 25 Hz which turns into a most clear 50Hz because of this). So, this is just harmonic distortion from "side bands" like IMD would show it ...

Not academic. But I can try to setup a synth sound like this where that second frequency modulates on the first lower. On/Off switchable of course. Not *so* easy to do for me, because of not being that experienced on it.
And maybe never mind this test because briefly thinking about it, any two of these two frequencies (without doing anything further) will already look exactly like modulated on eachother when looking at the digital file. So it really needs good knowledge and judgement set setup a representative test like this. And time.

Peter

[CoenP]

Thanks Peter for the elaboration.

It appears that our ear/auditory system is extremely sensitive to pitch . I found this on the net:

"The normal human ear can detect the difference between 440 Hz and 441 Hz"

So the case for audibility is there.
The two tone analogy may work for explaining the effect when there is a substantial difference in frequency but what happens for multi tones or tones that are close together. What part of the spectrum can be considered "carrier" and what "modulated". When you try to envision music through this effect it becomes a complete mess. Well maybe it is the mess we listen to in our daily lives .

Regards, Coen