why not the perfect reproduction?

[Gregadd]

Having been told that digital is perfect. We can measure everything. We not only have every possible material we need we indulge in overkill make.
We can't hide behind somebody else mucked up the recording. We can make our recording.

It should be easy to do with basic components and a basic microphone technique.

I throw down the challenge. Name the components. Name the recording technique. Tell me how it will sound and image just using measurements.

Please be specific and correlate the measurements to what we should hear.

 

[Bruce B]

Good-luck with that one!

 

[JackD201]

Maybe it could even win a TEC award

 

[Lee]

Maybe it could even win a TEC award

And be published in the AES Journal!

 

[Tom Danley]

We can measure the electronics and reproducers and pickups but that is only by comparing what they are told to do against what they actually do and examine the differences in various ways.
If one chose to take the ultra pure position, one could say nothing is perfect, everything has flaws.

In reality or to be realistic, one has to take into account the relative situation.
For example, at the purist level with no context, the best telescope mirror is a mountain range of atomic level irregularity.
To the eye, it is perfectly smooth, to the telescope maker it is within 1 /10 wavelength of light in the spectrum it’s made for, technical perfection.

From that approach it is not simple but possible to reduce all the errors in the electronic chain to "very small" levels.

All of this falls apart when you attach speakers, all, even the best alter the signal in many ways and then the room returns many later reflected signals arriving after the main /direct one.
Past that is the issue that how you hear is a subjective phenomena while every aspect of measurement is quantifying properties.

Lastly and a real turd in the punch bowl for realism is that there are VERY few ways to actually capture a realistic live / real stereo image.
Most recordings use panning and other acoustic tricks to place various sounds in the panorama, they are a mostly a simulation of a live event.

I don’t know if you make things or experiment, if you do and have some coins to do this, try this experiment.
If you don’t have a decent USB pre-amp, get something like am M-audio fast track pro or similar.
This allows you to record and play two channels up to 24/96K with low distortion AND has two decent mic pre-amps.
Obtain a measurement microphone (get two if you can) look at speaker building sites like Madisound or more costly ones from Earthworks, B&K, etc.
Now set up one channel and record things around your home, listen through headphones (same signal on both R and L).
If you don’t have a sound recording program Audacity is free.

Now, when you play back the one channel through one speaker, you will have a pretty good reproduction, the vast majority of the error being your speakers and the room interaction.
During nice weather, you can hear what your speakers actually sound like by themselves by listening outdoors.
If you enjoy it, there are forums and such which further open the door.

I mostly work on reproducers, but capture is a keen interest. I have devised an alternative way of capturing that image.
This is still in the R&D stage and not so exciting but give this a try with headphones.
This would be the forward facing image pair (I have a 6 channel, 360 degree plus overhead image version in parts coming along)

http://www.danleysoundlabs.com/TrainStart.wav

Please do consider recording yourself, for less than one might think you can make enjoyable recordings, experiment and unlike an open ended evaluation, if you were there when it was recorded, you really do know what it sounded like live.
Best,
Tom

 

[Gregadd]

Preety good Tom
Let's see If we measure the difference and there is none then it's perfect right(null test)? While no original is perfect a reproduction need only copy the original with all its flaws?
How can the electronics be perfect if they fall apart when asked to do thier job-drive speakers?
But we can measure everything we need to know about speakers and the room and we have the materials to correct them.
It does not have to be stereo remember you have complete freedom.

Ugh I think you misunderstood. It's your job to conduct the experiment and provide the proof. I am satisfied it can't be done.

Lastly, How can we elimiate human perception ? it's is the sine qua non.

 

[Phelonious Ponk]

Having been told that digital is perfect. We can measure everything. We not only have every possible material we need we indulge in overkill make.
We can't hide behind somebody else mucked up the recording. We can make our recording...I throw down the challenge.

You want to try that again, Greg? I haven't a clue what you just said. I was with you through "we can measure everything," and I got "I throw down the challenge," but as for the rest, I'm not at all sure what you said, much less what you want.

Tim

 

[Gregadd]

You want to try that again, Greg? I haven't a clue what you just said. I was with you through "we can measure everything," and I got "I throw down the challenge," but as for the rest, I'm not at all sure what you said, much less what you want.

Tim

Perhaps you watched the six million dollar man. You know, we have the technology, we can build it. I keep being told we have all technology and the materials we need for sound reproduction. I want to see a perfect bionic leg. I don't believe one exists. I do admit to the existence of some damn good prosthesis.

 

[Robh3606]

Tell me how it will sound and image just using measurements.

It's going to end up like every other recording. Dependant on the playback system for both how it sounds and the imaging which actually happens in your head. Certainly influenced by speaker placement though.

Rob

 

[Phelonious Ponk]

Perhaps you watched the six million dollar man. You know, we have the technology, we can build it. I keep being told we have all technology and the materials we need for sound reproduction. I want to see a perfect bionic leg. I don't believe one exists. I do admit to the existence of some damn good prosthesis.

Did someone say we have the technology to build the perfect sound system? I didn't catch that post.

I did not say, nor would I ever say,

We can measure everything. We not only have every possible material we need we indulge in overkill make.

But I have said that the typical audio amplifier is over-built. I stand by it. The typical engineer, designing an amp for a market full of passive speakers with wildly varying driver components, crossover networks and load characteristics has no choice but to over-build or put out an amplifier that will only work with a narrow range of possibilities. Even then, there are some speakers that are more difficult than most and for which a pretty narrow range of amps are a good option. You own a pair; you should understand exactly what I'm talking about.

You skipped the measurements rant this time, but give a good engineer (not me) comprehensive measurements and he can tell you if a component is going to be accurate or not and where it is going to vary from accuracy. Can he tell you what it's going to sound like? Sure. You can do that yourself. Look at the 60 cycle hump of a typical passive monitor (and too many pro actives as well). Look at a chart of the fundamental frequencies of instruments. You will be able to tell what instruments will be louder by how much when your monitors are played in an anechoic chamber or your back yard. No one can tell you what your room is going to do with it without setting it all up and measuring it there. Imaging? That's all about your room. The mix places the instruments right to left and front to back. I can pop on a good pair of headphones and tell you exactly how things should image in a perfect world, but your speakers and your room will allow the image to come through or be smeared beyond recognition. And I don't think anyone here has claimed anything else.

So, other than

We not only have every possible material we need we indulge in overkill make.

do you have a challenge for anyone here?

Tim

 

[Gregadd]

My only issue with measurements is when measurements don't correlate with what I hear or don't hear. Sorry if that's a rant.

If it serves a purpose it's not overkill. If you choose an elephant gun to kill a squirrel that's a wrong choice by the user not the manufacturer.

Great it can be done. But it has not.

Listen you can control all the variables you mentioned.

The challenge is open to anyone including you.

So far it's just theory.

 

[Phelonious Ponk]

My only issue with measurements is when measurements don't correlate with what I hear or don't hear. Sorry if that's a rant.

It's not a rant, but it's a personal issue. You hear your room. You hear the interaction between your components. I don't remember what amps you have but you may very well hear them struggling to drive your MLs at anything close to concert volume.

If it serves a purpose it's not overkill. If you choose an elephant gun to kill a squirrel that's a wrong choice by the user not the manufacturer.

You can make that argument, but the most efficient and effective solution would be to build amplifiers for speakers, or at least for speakers within a much narrower range of response than the entire universe of speakers. Instead, the market forces quality amp designers to build something robust enough to handle the worst case scenario. Is it overkill? I didn't use that word, you did. But it is "over-engineered" for most needs.

Great it can be done. But it has not.

Listen you can control all the variables you mentioned.

The challenge is open to anyone including you.

Without reference to what you're replying to here, I can't be sure how to answer, but I'll take a guess. Let's use headphones as our example, as they take the room out of the equation, have modest, headroom/amp needs, and often have FR curves because a complete lack of room gain doesn't sound right to us:

These are going to sound like what you audiophile dudes call analytical, or dry if they're driven by an accurate digital source. They're not going to be all that fatiguing or overly bright, because they start rolling off at about 5khz and are down 10db by the time you get to 10khz. Does that make the highs sound dull? No, it actually doesn't, and if you know what your'e looking at, what's in most music and what our ears can and cannot do with it, you'll know that there is so little acoustical energy in music up there that it won't make an appreciable difference, especially when you're using a digital source into ear-canal sealing headphones that will block all ambient noise leaving you capable of hearing low-level details at high FR that are down 10db at least as well as you would hear them dead flat in a room, through speakers. So why will the phones behind this chart sound dry and analytical? Because there is no appreciable lower midrange boost to simulate room gain, so what you hear, in spite of this curvy-looking chart, will sound very flat compared to most headphones, much less speakers.

As Eric Burdon and The Animals said, "Boom, boom, boom..." Not really. That big hump from the bass through the lower mids is probably a bit more than we need to simulate a bit of room gain, but not enough to make these cans boomy. It is enough, however, combined with that 7db dip at 5khz to give them a "warm" signature sound. Audiophiles like these headphones, and so do I; I own a pair. They are comforting and easy to listen to. They make good recordings sound good, great recordings sound great. A bit forward in presentation, but not overly aggressive at all. They are what y'all would call "musical." And that FR chart is a pretty decent picture of what they sound like.

Tim

 

[tony ky ma]

I don't think it is necessary to set up set of equipments to measure your room first before to buy a pair of sub add to your system, How they can sound ? sometime it depend on luck, the only thing what you looking from sub driver is watts for your room and quality level (from price). others who know, until it is done, the best way is ask friends come to help, place the sub in every possible locations and get the result to-getter , most yes will be the right direction I believe
tony ma

 

[Tom Danley]

Hi Gregadd
“Let's see If we measure the difference and there is none then it's perfect right (null test)?”
Assuming the test encompasses the important stuff then yes.

“How can the electronics be perfect if they fall apart when asked to do their job-drive speakers?”
They cannot and can only appear this way if they were tested not driving the loudspeaker and there is a difference driving one.

“But we can measure everything we need to know about speakers and the room and we have the materials to correct them.”
Actually there are two parts, mostly and NO very far from it.
While it is a negative to identify shortcomings, loudspeakers have a large number of problems / alterations relative to electronics.
I believe it was Don Davis who said that if one were to show a communications engineer a good loudspeaker’s behavior in detail but not identifying it, he would conclude it is broken.
Once one is in a room then the speakers radiation pattern at every frequency will strongly govern how transmogrified the result at listening position becomes relative to the speaker alone.

The point, so far as being “faithful” to the signal, loudspeakers are typically orders of magnitude less accurate than the electronic chain and “distort” in ways electronics can’t.
As a way of thinking about this corruption, mentally compare how headphones image compared to speakers in a room. Both use transducers which have problems but in addition the speakers usually have multiple sources and crossovers which produce an interference pattern as well as excite your room.
What is worse, one can have the appearance of “fixing” the problems and the room with DSP electronics BUT very often, this is an illusion, the result of solving the problem in one location which is not the same as a global solution. Even worse, problems caused by reflections are time issues not frequency response, trying to fix them in response (eq) can mess up time further.

“Ugh I think you misunderstood. It's your job to conduct the experiment and provide the proof. I am satisfied it can't be done.”

I think you misunderstand my job, I am an inventor who loves audio, my job it is to find ways to solve the problems that stand in the way of X, Y or Z with commercial loudspeakers (where all the problems are worse yet). To do that one usually needs to experiments and figure out the governing aspects of the problem, only then can you think about a solution. When that is possible, one still only has a solution, not a business haha.

“Lastly, How can we eliminate human perception ? it's is the sine qua non.”

We can’t eliminate it, we don’t want to if we are to produce a sonic reality, what needs to be done is re-create the acoustic conditions that allow our brain to identify the image and ultimately suspend disbelief and accept what you hear as reality. From the detection side of the chain, that is what drove the idea for the microphone thing I am working on and why (I think) it works like it does.
If you were able to hear the train start up as a realistic stereo image, I can reproduce that in say a living room with two sources albeit the image is somewhat narrower. The next version will capture a hemisphere, 360 degrees around plus overhead. If you liked it, here is another test recording, my neighbors Harley.

http://www.danleysoundlabs.com/Donny's Harley.wav

Best,
Tom Danley

 

[Gregadd]

Much discussion about why we can't do it. As I suspected.

Tom that file has excellent directional clues. I would hardly call it a perfect reproduction. I would like to hear it across a first class system. It is ironic how I, at least, am more easily convinced by sound effects than music.

 

[Ron Party]

Once one is in a room then the speakers radiation pattern at every frequency will strongly govern how transmogrified the result at listening position becomes relative to the speaker alone.

Do we have another Calvin and Hobbes fan in the room?

 

[Tom Danley]

“Tom that file has excellent directional clues. I would hardly call it a perfect reproduction. “

Thanks, here is the thing, there are several separate steps here, at each step the magic can be lost or altered.
The microphone experiment is my current approach based on "how it looks to me" to “capture” the acoustic event.

I used a nothing special recorder at 24/96 down crunched to a CD.
Reproduction starts by playing that Wave file.

“It is ironic how I, at least, am more easily convinced by sound effects than music.”

Well in the case of the recordings I made here, there are several reasons I used the things I did.
First, the “thing” is kind of cumbersome right now and has wires hanging off it and could topple. Picture something crude looking (like a project that an inventor type would kluge) and not ready for public exposure for patent reasons as well.

Second, I have observed that with music being harmonic in structure, that half of the possible harmonics a nonlinearity adds, will add to the “music” as they are like part of a chord. In recording / live there are ways to add predominantly even harmonics which makes the sound more “fat”. Odd harmonics have the familiar ick factor and do not normally ‘fit in”.

Natural sounds are rarely harmonic in structure and so adding either enough even or odd harmonics will alter the “sound of the sound”.
I think producing those kinds of things is actually harder, certainly some of the things are way way harder.

If you dare, I have a fireworks file, taken from about ¼ mile.

It won’t link as a wave file but if you go here (link) and scroll to the bottom, download the fireworks file.

http://www.danleysoundlabs.com/technical downloads.html

READ the WARNING This will initially sound very soft, because of the dynamic range the “average level” is unusually low especially in light of the loudness wars.
With headphones on, listen and relax, picture your in a back yard next to the edge of a woods, off in the distance is a highway and (unfortunately a little to the right side) the fireworks show in a small town on July 4th. I have this years fireworks but have not sent it in yet.

If you play this through your speakers afterwards, even with a subwoofer, you will run out of gas well before you will be satisfied. Such is the price of dynamic range and low frequency content.
Best,
Tom

 

[Tom Danley]

Hi Ron

“Do we have another Calvin and Hobbes fan in the room?”

Ah, well I know who they are anyway. No, I loved the idea of that word as it applied to the various possible audio distortions.
A view probably caused by too much time looking at measurements trying to associate what I see to what I hear haha.
Best,
Tom Danley

 

[fas42]

Let's use headphones as our example, as they take the room out of the equation, have modest, headroom/amp needs, and often have FR curves because a complete lack of room gain doesn't sound right to us:

I love those graphs! The phenomenal variance in frequency response above 1KHz between those two, measured with what I would presume to be reasonably accurate ways of registering what the ear picks up, would have to frighten the life out of anyone who believes that a DEQX modified world is the only one worth living in ...

And I note those two are highly regarded transducers. Thank God the ear/brain is so tolerant (where did I see that before ...) !!

Frank

 

[Phelonious Ponk]

Much discussion about why we can't do it. As I suspected.

Congratulations, Greg. You've claimed victory over a challenge to meet a set of claims made mostly by yourself. Let's review:

Having been told that digital is perfect.

Who told you that? I don't recall anyone here ever saying it. Put that challenge to someone who has made that claim.

We can measure everything.

I believe Ethan has said that we can measure everything we can hear. I know we can measure many things that we can't hear. Everything? Ask Ethan.

We not only have every possible material we need we indulge in overkill make.

This is utter gibberish, but I took a guess at what you might be talking about and answered you. Over-engineered amplifiers is a natural product of the separates market and the lack of speaker standards. I haven't a clue what "we indulge in overkill make" means.

We can't hide behind somebody else mucked up the recording.

Say what?

We can make our recording.

We're making progress; this one is a sentence of sorts. I have no idea what recording you're talking about making, but at least what you said was understandable. Then we got to this...

I throw down the challenge. Name the components. Name the recording technique. Tell me how it will sound and image just using measurements.

Please be specific and correlate the measurements to what we should hear.

I still don't know what you're talking about recording, and God knows no one has claimed you can read imaging off of a scope, but we have had some conversations here about measurement indicating what we would hear, so using headphones as an example, for reasons I articulated, I gave you the measurements and correlated those measurements to what we should hear. Evidently that's not what you were looking for because now you seem to think everyone has skirted your probing questions. Evidently you were challenging us to prove a bunch of things none of us have ever claimed.

Tim