Daydreaming of smooth controls...

[stonerbox]

Hi,

I was dreaming a little the other day while out in nature, with tent and backpack on my shoulders. I had come to this beautiful little stream that was running down a mountain path and along it ran an old Swedish stenmur [Sten=Stone Mur=Wall. Often around 1M / 39,4" in height and often used (long ago) as a visual line between property out in the wild].
I stopped and kneeled down to pick something up and noticed how incredibly smooth and elegantly it shelved down the high mid range and treble of the wildly flowing stream. I though to myself, I should try to design a filter in an EQ unit after this! Just try to emulate the stonewall's astonishing shelving effect the best I can from memory. Now, I know it is the rich source (the wide range of frequencies of the stream) that plays a great part in what makes the filtering sound beautiful but still the wall and it's many crevices just did something soft and warm with it's shelving, and that is what I am aiming for.
I was also tired from the long hike (and smoked a pipe while resting there) so it seemed like, in my mind, every time I kneeled down the shelving filter of the stenmur had a smooth latency to it. I thought it would be interesting to try to recreate that too, just to give the unit a further impression of smoothness when adjusting the highs. I'm thinking a latency of 30-150ms, discreet enough to almost not be noticed by the user unless he/she is really focusing on the response.

A (not pure) shelving filter (would probably adjust more than just the shelving when turning the knob) I can design but how would one go about producing latency to the filter control? I would like to avoid digital solutions if possible even though they would probably be the easier way to go.


Be sure to get out in nature sometimes, you'll never know what you'll uncover...

[PRR]

> Sten=Stone Mur=Wall.

These are often seen along urban highways, as 5m to 10m concrete walls, in an attempt to block sound from residential neighborhoods. 2,748 linear miles of noise barriers built along highways in the U.S.

They do work better if you are very close to the wall, where sound has to make a large angle to reach you.

(When I extrapolated "good" dimensions to a music school next to an expanded highway, I figured a wall a half-mile high would be a good start.... even in a project with a 400 foot cliff-cut, that was insane.)

There does not seem to be a single defensible theory. Highway planners work from simple charts and over-simplify. There's been some new work since I moved to the woods. Watch out for sites just trying to sell you something. And there is little about the effect very close to the backside of the wall.

https://blog.echobarrier.com/blog/noise-reducing-acoustic-highway-noise-walls
https://acousticstoday.org/wp-content/uploads/2016/12/Highway-Noise.pdf
https://undark.org/2017/12/27/highway-noise-barrier-science/
https://99percentinvisible.org/article/building-wall-highway-sound-barriers-evolution-noise/
https://roads-waterways.transport.nsw.gov.au/business-industry/partners-suppliers/documents/centre-for-urban-design/noise-wall-design-guideline.pdf
https://www.vejdirektoratet.dk/sites/default/files/publications/optimised_noise_barriers.pdf

These highway studies don't look at frequency response, just A-weighted broadband. Since highs are usually easier to reduce, most barriers will be high-cut, but how much?

[andy-h-h]

That all sounds very pleasant.   

Apart from effect of the wall, the absorption coefficient of all surfaces surrounding the listener should also be taken into account.  i.e.  was there something else in the area that could contribute to the effect. 

[GibsonGM]

Discussion of wave guides and audio in 3..2...1...

[stonerbox]

Maybe I should have not buried my question at the very end of the free floating babble but I was curious what you guys have to say about this.

I was also tired from the long hike (and smoked a pipe while resting there) so it seemed like, in my mind, every time I kneeled down the shelving filter of the stenmur had a smooth latency to it. I thought it would be interesting to try to recreate that too, just to give the unit a further impression of smoothness when adjusting the highs. I'm thinking a latency of 30-150ms, discreet enough to almost not be noticed by the user unless he/she is really focusing on the response.

A (not pure) shelving filter (would probably adjust more than just the shelving when turning the knob) I can design but how would one go about producing latency to the filter control? I would like to avoid digital solutions if possible even though they would probably be the easier way to go.

There does not seem to be a single defensible theory. Highway planners work from simple charts and over-simplify. There's been some new work since I moved to the woods. Watch out for sites just trying to sell you something. And there is little about the effect very close to the backside of the wall.

Yeah, the ones calling the shots on these things usually don't have a real interest for the results, just as long as nobody complains too much once it is done everything is "OK". Would be interesting too see a more detailed study of the effectiveness of these tiny (4-8m) walls along highways and railroad tracks. Along the smaller the cities, outside of mine, they usually construct a high and broad dirt hill + a high wall to fix the issue. Dirt works magic.

Thank you for the documentations I will go through it all.

Apart from effect of the wall, the absorption coefficient of all surfaces surrounding the listener should also be taken into account.  i.e.  was there something else in the area that could contribute to the effect.

I am sure a lot of phase distortions + a super dampened micro-echo was present, as there was a ridge on my side of the wall. The lows got a 4-5dB+ boost too when hunkered down close to the mur.

Code Select Expand

|                                                        |
||                                                      ||
||Ridge↓                                          Ridge↓||   
|||                                                   ||||
|||||||||||                                         ||||||
|||||||||||||||                                    |||||||
||||||||||||||||    me                           |||||||||
|||||||||||||||| ducking down                   ||||||||||
||||||||||||||||     ↓ ↓Stonewall           ||||||||||||||
||||||||||||||||||   O =    Water        ‎ ‎ |||||||||||||||||
|||||||||||||||||||  o =^^^^^^^^^^^^^^||||||||||||||||||||
------------------------ ^^^^^^^^^^^^^^--------------------

[PRR]

I saw latency and did not want to go there.

The woods of Maine have odd acoustics. Some guy even posted a YouTube. In some woods the tree trunks are dense and regularly spaced. Western hunters say they can't see the game here like they can in more open woods. I hear a quasi-random reverberation on all sounds.

No, I don't know any compact way to do audio latency in simple analog circuitry. The next step up would be a garden hose echo, "Cooper Time Cube".

I'm also sure the walk and the smoke colored your perceptions.

[Phend]

Without being there near the stream, does it make any sound at all ?

[stonerbox]

I saw latency and did not want to go there.

No, I don't know any compact way to do audio latency in simple analog circuitry. The next step up would be a garden hose echo, "Cooper Time Cube".

Hah, fair enough!

The woods of Maine have odd acoustics. Some guy even posted a YouTube. In some woods the tree trunks are dense and regularly spaced. Western hunters say they can't see the game here like they can in more open woods. I hear a quasi-random reverberation on all sounds.

I'll have to look into that! Any folk tales or superstitions around that particular area?

I'm also sure the walk and the smoke colored your perceptions.

PPR, of course it did. And I am fully determined to make others share that bent and warped reality.

[Fancy Lime]

First off, I'm jealous. I miss the days when I had the time to go solo hiking in the Great North.

Secondly, we clearly need to approach this more scientifically. All we need is reasonably accurate topographic map of the area (1m resolution should do) and some acoustic modeling (the kind they use to design opera houses). Easy peasy, massive parallel GPU setup squeazy. I am not aware how open Sweden is with their mil-grade topo maps, so you might need to create the relevant section yourself. Do you have a drone and a compact 3D laser scanner and do you know how to generate the relevant data with those? I should think that a roughly 600x600 m area around the desired location would probably suffice. If not, maybe some data exist somewhere. Sometimes the best topographical data are found in unlikely places. For example, if you want really good hiking maps for California, look at old declassified Soviet material. Beats commercially available actual hiking maps by a mile in terms of resolution and accuracy in many of the less popular destinations. But I digress. Where is our little mystery stenmur anyway? You said "mountain path" so we are not talking southern Sweden unless you and I have a different understanding of what constitutes a mountain. Also, I haven't seen a lot of stenmurar outside of settlements in Lappland. Different concept of "land property" when your raising reindeer instead of cows or sheep, I guess. And you don't mention snow cover (highly relevant for acoustics!) or that you went by ski, so I'm guessing Dalarna rather than Jämtland? Care to share coordinates? I think we might all have to go and meet there in person for a proper assessment of the situation.

Until we can agree on a date for our field trip (any suggestions on how to name this event?), I agree that a shallow low pass filter combined with a series of very short delays (the shortest ones short enough to cause comb filtering in the upper mid range) and maybe a bit of longer, slightly diffuse double echo (from the mountain sides) should get you most of the way. Smoking that same pipe again when testing the result might also help. Feasible with DSP. If you manage this with BBDs, you will instantly ascend to legend status and the bards will sing your praises until the sons of Muspell ride forth.

HTH
Andy

[stonerbox]

First off, I'm jealous. I miss the days when I had the time to go solo hiking in the Great North.

Secondly, we clearly need to approach this more scientifically. All we need is reasonably accurate topographic map of the area (1m resolution should do) and some acoustic modeling (the kind they use to design opera houses). Easy peasy, massive parallel GPU setup squeazy. I am not aware how open Sweden is with their mil-grade topo maps, so you might need to create the relevant section yourself. Do you have a drone and a compact 3D laser scanner and do you know how to generate the relevant data with those? I should think that a roughly 600x600 m area around the desired location would probably suffice. If not, maybe some data exist somewhere. Sometimes the best topographical data are found in unlikely places. For example, if you want really good hiking maps for California, look at old declassified Soviet material. Beats commercially available actual hiking maps by a mile in terms of resolution and accuracy in many of the less popular destinations. But I digress. Where is our little mystery stenmur anyway? You said "mountain path" so we are not talking southern Sweden unless you and I have a different understanding of what constitutes a mountain. Also, I haven't seen a lot of stenmurar outside of settlements in Lappland. Different concept of "land property" when your raising reindeer instead of cows or sheep, I guess. And you don't mention snow cover (highly relevant for acoustics!) or that you went by ski, so I'm guessing Dalarna rather than Jämtland? Care to share coordinates? I think we might all have to go and meet there in person for a proper assessment of the situation.

Until we can agree on a date for our field trip (any suggestions on how to name this event?), I agree that a shallow low pass filter combined with a series of very short delays (the shortest ones short enough to cause comb filtering in the upper mid range) and maybe a bit of longer, slightly diffuse double echo (from the mountain sides) should get you most of the way. Smoking that same pipe again when testing the result might also help. Feasible with DSP. If you manage this with BBDs, you will instantly ascend to legend status and the bards will sing your praises until the sons of Muspell ride forth.

HTH
Andy

Already beat you to it, headed out at midnight and captured all necessary data needed for this magical unity to materialize!

No I don't dare to disclose my location but it is south Sweden!

But wait a second.. The way you know my country, some of its farming culture and it's topographic and how you classify mountains... You.. you are one of the countrymen who come to behold and enjoy magic Elks and Allemansrätten, aren't you?!

A German!

[Fancy Lime]

Well, that sounds unnecessarily accusatory but close enough. However, would a German know that there are no elk in Sweden, only moose? And "magic" is probably the last adjective I would use for those. Have you seen one waddling through the swamp at maximum speed? It's like, what if a deer walked like a duck but it's the size of a rhinoceros on stilts and also very drunk and way the hell too fast. Sorcery, maybe. But magic?

[stonerbox]

What a blunder, yeah I meant Moose.

Greece..? Netherlands..? Denmar...?



Ok, back on topic. If I would add latency to resistance change in the digital realm what options do I have? I am not a programmer.

[amptramp]

I assume the smooth controls you want are for an amplifier, so why not add a flip-up wall in front of the speaker that you can set to an angle to suit your needs?

[stonerbox]

I dig this has almost become an full blown troll-thread. 

[Fancy Lime]

Lots of trolls in the fjäll, so that's only right, isn't it? But kidding aside, "physical modeling" using a speaker, an actual wall, maybe some foam, and one or several microphones isn't the worst idea. Not super practical live but certainly an option for the studio.

[amptramp]

I dig this has almost become an full blown troll-thread. 

I was not trying to be a troll.  Diffraction of sound around an obstacle is easy to do with mechanical effects that act like the sound wall we have beside highways in residential zones.  In electronic controls, you quite often are limited by the fact that you can get a fixed 6 db/octave or 12 db/octave or any multiples of that but you need a complicated equalizer to do what a mechanical device does very simply.  And diffraction preserves the time sequence that various frequencies come in at.  There is a delay element to diffraction.

As Fancy Lime said, it is easy enough to do in the studio but most musicians would want to have an answer they could use for live gigs.  If you had the speaker / wall combination, you would have to aim it in the same slope as the floor of the venue, so you would need an angled riser and the effect wouldn't be the same near and far from the speaker.  You could get decent results in a small venue but distance from the wall would change the sound in a larger venue.

You might be able to do some digital processing that takes frequency and delay into account but it may not be all that easy.

[Fancy Lime]

Ron's comment got me thinking: maybe there is an electro-mechanical solution that is practical live, like a spring reverb. I'm thinking maybe a very small plate reverb (30x30cm or so) with dampers between the transducer and the pickup. That's probably not going to make for a perfect or even a decent emulation of the real stenmur situation but it could be a starting point for a fun journey in that direction by varying damper material, size, placement of damper(s), transducer, and pickup, plate material, thickness, size... Very old school DIY.

Andy

[Rob Strand]

In HiFi speaker system design there's a well know method of simulating "baffle edge diffraction" by Vanderkooy.    It might be possible to do a simulation based on that.   However in the distant past  I've read-up on noise barrier design and you will find there's quite a few papers out there which provide ways of simulating noise barriers.  Many not easy to understand.    There's a lot of noise barriers out there so people have been trying to find better ways to design/engineer these things for a long time.  Some papers are scientific some are more engineering recipes to "get the job done".   You want the papers that let you estimate the frequency response.   What you will find is the fine details depend on the shape of the edge of the barrier.

[r080]

What about mounting a large brick plate on a hydraulic lift and place it between an amplifier and a microphone?

[MrStab]

Look up Analog's switched capacitor filter ICs (eg. LTC1068) and drive it with a simple CMOS 555 timer circuit with variable frequency and 50/50 duty cycle, if you'd like an analogue filter with some degree of latency.