Is good analog synthesis like (airplane) wing design ?

[allan_evett]

I'm an EE professor and I'd like to understand you Theo, but although I'm familiar with each of your phrases I'm not sure what your main idea is.

 

Are you saying that the nonlinearities of synth design are like nonlinearities of modeling air flow over a wing (because air isn't an idealized gas, or because of turbulence)? Or are you saying something more metaphorical?

 

Guys/gals: This isn't a problem with technical lingo. This could be a language translation problem. His grammar and sentence length aren't English-friendly.

 

This is what I've long suspected. Though I'm a fairly quick study with most things technical, I'm not at the level of an advanced engineer. So sometimes I get what Theo is saying, but other times it's a challenge - and that's compounded by translation issues. The thing I'm trying to avoid is passing judgement on intent.

 

[stoken6]

Take a listen to some of Theo's videos on Youtube (he posts as "theover2"). I just dipped into the one where he demonstrates running 48kHz and 96kHz simultaneously. Theo narrates in English which is slow and deliberate, and uses clear, simple words.

 

Theo - could you try explaining your original post in spoken English? (mp3 file on SoundCloud, youtube video, whatever). I think I would genuinely be interested in this subject, and your spoken-word delivery seems much easier to understand.

 

Cheers, Mike.

[Theo Verelst]

Still not much understanding or love for standard academic english tricks, such a shame. Maybe I've made some typo's. too...

 

Anyhow just to make the sort of technical point about sound and simulations of effects: artificial head analisys of sound sources, which indicates if a sound comes from left, right, high or low, far or near dictates sound properties that cannot easily be created with samples and DACs, however, analog synths can include accuracies of such kinds when programmed/created right.

 

A lot of speaker systems include some design around frequency graphs and a little airflow to and from the cones that make up for it's sound sources, but finite element analysis (look it up or forget it if you don't have a clue what that's for, not essential to the story) for determining the real airflow inside the box and towards the listening space using all angles can be in order for more elaborate designs. Digital synths cannot easily (or cheaply) create sound pattern elements (short ones and average based ones) with sufficient accuracy to escape standard "digital on speakers" sounds. Analog synths can easily, and for instance for bass tone "design" that's a big issue.

 

Simulations of analog synths aren't even very accurate in the sense of standard undergraduate engineering standards in the software that I know, and it is a lot harder to include various airflow in those simulations, still. So.

 

T.

 

[Gretel]

Still not much understanding or love for standard academic english tricks, such a shame. Maybe I've made some typo's. too...

 

Anyhow just to make the sort of technical point about sound and simulations of effects: artificial head analisys of sound sources, which indicates if a sound comes from left, right, high or low, far or near dictates sound properties that cannot easily be created with samples and DACs, however, analog synths can include accuracies of such kinds when programmed/created right.

 

A lot of speaker systems include some design around frequency graphs and a little airflow to and from the cones that make up for it's sound sources, but finite element analysis (look it up or forget it if you don't have a clue what that's for, not essential to the story) for determining the real airflow inside the box and towards the listening space using all angles can be in order for more elaborate designs. Digital synths cannot easily (or cheaply) create sound pattern elements (short ones and average based ones) with sufficient accuracy to escape standard "digital on speakers" sounds. Analog synths can easily, and for instance for bass tone "design" that's a big issue.

 

Simulations of analog synths aren't even very accurate in the sense of standard undergraduate engineering standards in the software that I know, and it is a lot harder to include various airflow in those simulations, still. So.

 

T.

 

You are an ... on so many levels. And you have made some "typo's", too.It gets worse when you are talking about technical terms, of which you clearly have no clue.

[Threadslayer]

.

 

Anyhow just to make the sort of technical point about sound and simulations of effects: artificial head analisys of sound sources, which indicates if a sound comes from left, right, high or low, far or near dictates sound properties that cannot easily be created with samples and DACs, however, analog synths can include accuracies of such kinds when programmed/created right.

 

A lot of speaker systems include some design around frequency graphs and a little airflow to and from the cones that make up for it's sound sources, but finite element analysis (look it up or forget it if you don't have a clue what that's for, not essential to the story) for determining the real airflow inside the box and towards the listening space using all angles can be in order for more elaborate designs. Digital synths cannot easily (or cheaply) create sound pattern elements (short ones and average based ones) with sufficient accuracy to escape standard "digital on speakers" sounds. Analog synths can easily, and for instance for bass tone "design" that's a big issue.

 

Simulations of analog synths aren't even very accurate in the sense of standard undergraduate engineering standards in the software that I know, and it is a lot harder to include various airflow in those simulations, still. So.

 

T.

Hey Theo. I think you put that quite well. Definitely a step in the right direction.

[Tusker]

Hey Theo. I think you put that quite well. Definitely a step in the right direction.

 

Agreed. Definitely a step in the right direction. Nice job on that post Theo.

 

So ... if we are to nibble on this tasty morsel ... do you have any evidence for this (a paper, or research) to support this:

 

"Anyhow just to make the sort of technical point about sound and simulations of effects: artificial head analisys of sound sources, which indicates if a sound comes from left, right, high or low, far or near dictates sound properties that cannot easily be created with samples and DACs, however, analog synths can include accuracies of such kinds when programmed/created right."

 

This is not to disagree with your observation. I've made the point in this forum that in my experience an analog sound source tends to have a coherent sonic image in sub-optimal rooms where a comparative digital sound source often gets washed out. I've wondered why.

 

Is there any data you can point me to regarding phase coherence across the frequency spectrum from these two types of sources. Or would you suggest a different way of thinking about the situation?

 

Thanks in advance.

[Gretel]

I mean seriously, this is all just b...s....

[Doerfler]

I mean seriously, this is all just b...s....

 

you have 3 negative posts in this thread. Why do you waste your time if it bothers you so much? Just put T. on ignore and let it go.

[stoken6]

OK I'm interested now. Thanks for the rephrasing Theo.

 

You seem to be making the point that psychoacoustic cues (e.g. sound source is high or low) cannot be created with samples. Why can't they?

 

And why is it possible with analogue?

 

And if you sample the output of the analogue, what do you lose?

 

(Genuine questions. I'm also interested in Jerry's observation that analogue maintains a coherent image better than digital).

 

Cheers, Mike.

[Gretel]

@dave: That's actually a very reasonable question and suggestion. The answer however is also very reasonable I think: sometimes you don't want to ignore things or people that you perceive as obviously wrong, especially if it seems that there are other people who do not (yet) have this perception.

In Theo's case I take personal offense because he claims to have an EE background which I also have. Furthermore, he claimed to believe his way of communicating would be "German". I am german. So, two factors where I feel related to him and where I think he gives EEs and Germans a bad rep.

 

Edit: Most importantly I think that Theo's posts are autogenerated somehow. To me it looks like someone is making fun of us by posting semirandom concatenations of sentences and watching our reactions. Certainly especially funny if people are actually discussing with Theo. I once read a text generated by the "random paper generator" (paper meaning scientific article, maybe you know). It read like Theo's posts. Somehow interesting and sophisticated, but you couldn't make sense of it.

I wouldn't want to ignore randomly generated scientific articles, and I don't want to ignore apparently randomly generated posts here.

 

[PianoMan51]

Lotta heat. Not much light.

[Gretel]

OK I'm interested now. Thanks for the rephrasing Theo.

 

You seem to be making the point that psychoacoustic cues (e.g. sound source is high or low) cannot be created with samples. Why can't they?

 

And why is it possible with analogue?

 

And if you sample the output of the analogue, what do you lose?

 

(Genuine questions. I'm also interested in Jerry's observation that analogue maintains a coherent image better than digital).

 

Cheers, Mike.

 

@Theo: please enlighten us. You haven't found the time to answer these clarification requests yet, but certainly we may expect a response soon, do we ?

[DulceLabs.com]

Edit: Most importantly I think that Theo's posts are autogenerated somehow. To me it looks like someone is making fun of us by posting semirandom concatenations of sentences and watching our reactions. Certainly especially funny if people are actually discussing with Theo. I once read a text generated by the "random paper generator" (paper meaning scientific article, maybe you know). It read like Theo's posts. Somehow interesting and sophisticated, but you couldn't make sense of it.

 

That is one explanation, another is "word salad".

 

[LenZoo]

@Theo: The sum of sonic to the wing of vibrations are equal to each other so it must not rotate necessarily too often. Of course, only when EE undergraduates consider a poly elastic simulation. But not me. So. Or?

 

Edit: I have it now. It is greater than 7.

[poserp]

My take on the questions posed in the thread title. I studied Aerospace engineering in college for awhile, until I found that the more math I took the worse I got at it. But I still love airplanes and airplane design.

 

Sound isn't a squiggly line, its an actual three-dimensional wave-front propagating through a space and interacting with things it runs into in the process. I think what's changed is (or, rather, at least me) sound design happens in headphones more often than loudly in real rooms. I do not think this is a digital vs. analog thing per se, it's just that the way people work with sound has changed, generally speaking, since the analog era in the 60's and 70's.

 

I'm very "guilty" of this myself as of necessity I do most of my work with headphones on, usually late at night when people are sleeping. I'd much prefer to work in a room with speakers (and have headphones too). I think of headphones as being a "microscope" and speakers as being "projectors", sound emanating from them acts differently because speakers project into an open space, which interacts with the sound, and then hits your ears. Headphones are a direct line.

 

So, I think for sound to really "fly" it needs to at least be tested in a wind tunnel (analogous to a recording studio environment), and preferably in the open air somehow. I aspire to do that, but at the moment constraints dictate otherwise. I suspect other people deal with this because I hear stuff all the time that sounds good in headphones but doesn't translate well to speakers. And that's just my music...

 

 

Edit: ...And this is why we have mastering engineers.

[Hoochie]

Edit: Most importantly I think that Theo's posts are autogenerated somehow. To me it looks like someone is making fun of us by posting semirandom concatenations of sentences and watching our reactions. Certainly especially funny if people are actually discussing with Theo. I once read a text generated by the "random paper generator" (paper meaning scientific article, maybe you know). It read like Theo's posts. Somehow interesting and sophisticated, but you couldn't make sense of it.

 

That is one explanation, another is "word salad".

 

After digesting all this I'm going to need some

Mental Floss.

 

Reruns of Baywatch or

The Love Boat

should do the trick.

 

 

John

[PianoMan51]

After digesting all this I'm going to need some

Mental Floss.

I think Mental Floss requires a big hit of Zappa.

 

Raisin' it up, waxin' it down...

[DulceLabs.com]

Reruns of Bay Babewatch

 

FTFY

[Theo Verelst]

For those such inclined and somehow understanding the engineering explanation I gave is the right one.

 

More normal language: is your speaker system creating airflows ? Would you trust your favorite digital sound designer to design a helicopter, and you take place as the first test pilot ? I don't think so.

 

Simulating electrical circuits is a main thing in the digital synths that emulate the famous analogs, but thus far a lot of those simulations aren't very accurate, and they usually promote ideas that would never make it to design something with physical properties instead of electronic ones.

 

T.

[Gretel]

First of all: My speaker does not create airflows. It makes the air vibrate, but there is no flow like a helicopter would produce, or what would be created by a turbine.

 

Then: You state that digital sound design does not relate to aerospace engineering. Okay, but that's obvious.

 

Then: You state that analog modelling is just modelling, and that modelling isn't as accurate as the real thing. Okay, but that's obvious.

 

How do you think analog sound design and airplane wing design are now potentially related? Your latest post has not at all addressed this, only that digital sound design is not perfectly capable of modelling analog sounds. What does this have to do with analog sound design being like airplane wing design?

[WillNeverPost]

You will not get a sensible answer.

[Theo Verelst]

Without going into details, an electronic circuit I consider to be a collection of electronics parts that are connected according to a certain schematic diagram, and such is a continuous response circuit, i.e. "analog", no samples.

 

A wing design leads to a wing that also is a physical object, except for most purposes it will have probably at least 2 dimensions or probably 3 dimensions as it's playground: the air surrounding it with all kinds of flows, including the main ones that cost energy to create under pressure and lift. So the air flows in, is interacted with in a physical way (no samples) and flows away on all kinds of ways, leading to forces and vibrations.

 

Simply from the airflow point of view, a speaker and it's cones take place in a mostly 3 dimensional air flow as well, though mostly there won't be much of a global in and out flow somehow. Also: no samples.

 

Now when people have a problem already with simulating electronics circuits accurately, I don't have much faith in their ability to do 3D flow simulations, which however is something done in speaker design. Maybe people don't know that, but usually there are design criteria and in modern cases, regularly computer simulations to compute all kinds of air flow components in and outside speakers, there's software for that, too.

 

The mistake circuit simulation digital persons often make has among other things to do with the presumption an electronic circuit behaves like a short lived impulse, which is generally speaking in error.

 

Don't bother if thus it still doesn't make sense: it should be clear enough to people in the relevant fields, but it's not that I need everyone to "get it", that's not so important.

 

T.

[MurMan]

The mistake circuit simulation digital persons often make has among other things to do with the presumption an electronic circuit behaves like a short lived impulse, which is generally speaking in error.

Please propose your improved circuit simulation technique to the forum.

 

And one more thing. Since "samples" are the fundamental problem in most of your posts, I assume that you will not propose a digital simulation. A purely analog simulation that uses no discrete algorithms, aka "samples".

[Theo Verelst]

"there is such a thing as samples processing distortion, and DACs aren't as perfect as you might think" is a true statement, no matter what you might have thought.

[MurMan]

Breathe slowly and focus on the question. Propose your solution.

[Theo Verelst]

Are you aware of the difference between belittling and teaching a rather dim wit b*stard ?

[MurMan]

Are you aware of the difference between belittling and teaching a rather dim wit b*stard ?

Yes, but calling you that on a public forum would be a distraction.

 

Please answer the question.

[Theo Verelst]

Are you really that dumb and self-involved ? I know the answer, both the technical scientific one, and the answer to the more general idea you think people are really interested in, you keep going on about.

[MurMan]

Theo, if your purpose is to explain answers to the forum, you will respond to questions that are posed to you without attacking the questioner.

 

On the other hand, if your purpose is to make us believe that only you have the knowledge and the answers, then this conversation is over.

[Gretel]

"Without going into details, an electronic circuit I consider to be a collection of electronics arts that are connected according to a certain schematic diagram, and such is a continuous response circuit, i.e. "analog", no samples".

 

Okay, with samples you mean time-discrete. An analog circuit is time- continuous.

 

 

"A wing design leads to a wing that also is a physical object, except for most purposes it will have probably at least 2 dimensions or probably 3 dimensions as it's playground: the air surrounding it with all kinds of flows, including the main ones that cost energy to create under pressure and lift. So the air flows in, is interacted with in a physical way (no samples) and flows away on all kinds of ways, leading to forces and vibrations.

 

Simply from the airflow point of view, a speaker and it's cones take place in a mostly 3 dimensional air flow as well, though mostly there won't be much of a global in and out flow somehow. Also: no samples."

 

Okay, wing design is more complex than analog circuit design because there are more dimensions.

 

 

"Now when people have a problem already with simulating electronics circuits accurately, I don't have much faith in their ability to do 3D flow simulations, which however is something done in speaker design. Maybe people don't know that, but usually there are design criteria and in modern cases, regularly computer simulations to compute all kinds of air flow components in and outside speakers, there's software for that, too."

 

Okay, you say if 1dim simulation doesn't work well, why would 3dim simulations work well. Of course, this does not have anything to do with the original topic.

 

"The mistake circuit simulation digital persons often make has among other things to do with the presumption an electronic circuit behaves like a short lived impulse, which is generally speaking in error."

 

I assume you think: Digital simulations of analog circuits assume that signals in analog circuits stay constant over a short period of time (e.g. the sampling interval). If my understanding is correct, you are completely wrong. The relation between digital (here: time-discrete) and analog (here: time-continuous) is assumed to be given via interpolation using the sinc-function if sampling happened at below Nyquist (which requires the signal to be bandwidth limited, which is no problem in reality (in theory) :-) ).

If this understanding of your understanding is correct, then you are neither a bot nor an expert. You are more likely simply someone with no education in electrical engineering or related fields (i.e. like a normal kid). Are you an adolescent, Theo?