Has DSP turned us into audio neurotics? [rant]

[DLS79]

Amir's Klippel machine is able to accurately identify FR issues of the speaker separate from the room. If you have accurate spinorama measurements like these and the speaker has a "good" directivity curve then some careful FR tuning may be helpful. If you were just measuring the speaker in the room with a MIC you would not be able to make this determination which is what some of the "automatic DSP" claim to do.

Why do you think you need to know if the issue is caused by the speaker or the room?

 

[levimax]

Why do you think you need to know if the issue is caused by the speaker or the room?

As has been mentioned humans with 2 ears and a head and a brain do not perceive sound the same as a microphone. Humans have the ability to separate direct sound from reflections i.e. "hear through the room".

 

[MattHooper]

That's already done,Barefoot does it.
Is kind of fun but certainly not easy to do it home,this character has been done using anechoic data,etc.

I’m curious, how would Barefoot take a narrow dispersion horn loudspeaker in, say, an average 15 x 13’ room like mine, and make it sound indistinguishable from an MBL on the directional speaker in the same room?

How does it count for and re-create the dispersion characteristics and interplay with room acoustics?

 

[DLS79]

As has been mentioned humans with 2 ears and a head and a brain do not perceive sound the same as a microphone. Humans have the ability to separate direct sound from reflections i.e. "hear through the room".

You know a speaker, a mic and a little bit of math can do the exact same thing right?

 

[levimax]

You know a speaker, a mic and a little bit of math can do the exact same thing right?

No I have never seen that documented, if true then no need for a Klipple or anechoic chambers. This is also news to Dr. Toole apparently as he doesn't mention it in his writings. Can you post a link as I would be very interested.

 

[DLS79]

No I have never seen that documented, if true then no need for a Klipple or anechoic chambers. This is also news to Dr. Toole apparently as he doesn't mention it in his writings. Can you post a link as I would be very interested.

The klippel is automating the process and taking some of the variables out of the equation. The Klippel knows the distance between the speaker and the mic, and it know what angles the mic is at relative to the reference axis.

Assuming your using a decent sampling rate, and not doing anything ridiculous with regards to placement of the mic relative to the speaker, you can make some assumptions.
The first sound recorded by the mic will be the sound directly from the speaker and not a reflection. The shortest distance between to points is a strait line (note the speed of sound is constant). A reflection will arrive at the mic later, because it travels a greater distance, and it will arrive at a lower amplitude because it will have transferred energy into the surface it reflected off of. Having the speaker up on a platform, and the mic closer to the speaker than any potential reflective surfaces, just assures the above happens more consistently. It's all pretty strait forward science, and how they can tell the difference between the the speaker and the room.

Amirm normally has the klippel collects between like 500 and 1000 points of data (if memory serves). The data is then post processed, fitted, smoothed, etc before being used to generate various informative plots.

Note the above is a simplified explanation, I'm sure they are using additional techniques to help differentiate between the speaker and the room.


consumer grade DSP with 2 speakers is trying to do something like the above, but most likely they have to try and estimate distance and angles, because you have a human holding a relatively cheap mic, moving the mic around in a none precise manner, and only gathering a few dozen data points at most. If memory serves they generally only take measurements in 2 dimensions not 3 like the klippel. To and extent the simplification is ok, because they are only trying to generate corrections for a specific listening position, not model the speaker.

The main differences are the quality of the equipment, the types of data collected, amount of data collected, and the quality of the data collected.

 

[boxerfan88]

You know a speaker, a mic and a little bit of math can do the exact same thing right?

I highly doubt the possibility to make a box speaker emulate an omni speaker performance (to the ear-brain system) just using math & DSP.

 

[MattHooper]

I highly doubt the possibility to make a box speaker emulate an omni speaker performance (to the ear-brain system) just using math & DSP.

Indeed. The idea that it could would, it seems to me, go against much of what Floyd Toole and others have been telling us: you can’t take a poorly designed speaker, with various on and off axis discrepancies, and EQ it to sound like a well-designed speaker would behave. That’s why Amir (and Erin) will point out for measurements, whether one speaker with a slightly uneven frequency response can be successfully adjusted via EQ without other detriments where another could not.

 

[DLS79]

Indeed. The idea that it could would, it seems to me, go against much of what Floyd Toole and others have been telling us: you can’t take a poorly designed speaker, with various on and off axis discrepancies, and EQ it to sound like a well-designed speaker would behave. That’s why Amir (and Erin) will point out for measurements, whether one speaker with a slightly uneven frequency response can be successfully adjusted via EQ without other detriments where another could not.

What is with some of you, you have to take everything to extremes? I didn't say you could take a crappy speaker from some consumer box store and use DSP to make it perform like a Neumann KH120 MKII.

something i actually said!

In my opinion, the fundamental problem is far to many people think DSP can fix everything, like its some kind of magic.

If you have crap gear in a crap room DSP isn't going to magically make it sound perfect.

On the other hand if you have good gear with only minor deficiencies and/or a treated room with only only minor deficiencies DSP can help.

 

[MattHooper]

What is with some of you, you have to take everything to extremes? I didn't say you could take a crappy speaker from some consumer box store and use DSP to make it perform like a Neumann KH120 MKII.

something i actually said!

Sorry, but I hadn’t read every single post of yours in this thread.

When I talked about enjoying the different characteristics of loudspeakers, you reponded:

I hate to say it, but you could DSP any character you want.

Which is not an equivocal statement. It’s a blanket statement.

In talking about the character of different speaker designs, I would have, of course, been including all sorts of speaker characteristics that could not be replicated just playing with eq or room correction.

 

[DLS79]

Which is not an equivocal statement. It’s a blanket statement.

In talking about the character of different speaker designs, I would have, of course, been including all sorts of speaker characteristics that could not be replicated just playing with eq or room correction.

When you said "Unless you enjoy the character of different sound systems." I assumed you meant when people describe how various amps or speakers make the music some combination of the following Airy, Bassy, Bright, Harsh, warm, etc.

 

[MattHooper]

When you said "Unless you enjoy the character of different sound systems." I assumed you meant when people describe how various amps or speakers make the music some combination of the following Airy, Bassy, Bright, Harsh, warm, etc.

OK, cool. Just a misunderstanding we’ve clarified.

I really like the wild West aspect of high-end audio. Like going to a high-end audio show and checking out all the different systems, from something neutral to the wackiest designs. From giant horn systems, to tiny BBC Heritage monitors, to line arrays versus coaxial speakers, boutique single driver speakers To a Wilson Chronosonic, to all the varieties of electrostatic and ribbon speakers, the whole bottle of wax.
I personally don’t approach listening to loud speakers with some best practises or set of measurements in mind. Rather I see someone’s possibly unique take on loudspeakers and I think “ I really want to hear what that sounds like.” Often enough while idiosyncratic designs may not be fully successful, they can be quite compelling in their own even if limited way.

 

[Digital_Thor]

I hate to say it, but you could DSP any character you want.

There is one you can't.... dispersion character. A DSP is still nothing more than level and timing. It is ultimately the speaker construction that dictates the way sound is sent out into the room - which changes sound character a whole lot - IMO

 

[Snarfie]

There is one you can't.... dispersion character. A DSP is still nothing more than level and timing. It is ultimately the speaker construction that dictates the way sound is sent out into the room - which changes sound character a whole lot - IMO

Did measure several colume speakers same height same room an all had more or less same FR. Some had forward or backward bass ports, symmetrical load (listen 30 years to that system very tight bass) , reflex an my favorite transmission line. My impression is that after measurment an correction they more or less started to sounded the same difference was found in stereo imaging like staging depth some tighter bass some less. The reason i changed to the Vandersteen speakers which are by design build phase coherent time alignt. I expected the other speaker do the same because the DSP i'm using takes care about phase/time guess not enough.

 

[UprightCitizen]

Unless you enjoy the character of different sound systems.

Sounds cumbersome. Personally I enjoy listen to recordings, but whatever floats your boat.

 

[goat76]

Hmmmm ..... I'm not so sure about that. https://rode.com/en/about/news-info/the-beginners-guide-to-ambisonics

Jim

Other than a four cardioid capsule array built into one microphone, there is nothing new in how those capsules pick up sounds, and that works the same way as it did back in the 1950s.

 

[Sokel]

The closest we have in capture for in-rooms (not anechoic) is HATS.
But even that does not capture perception.

All we have left since,are the studies of those who dealt with it,like Dr Toole,etc.
To interpret these studies it needs formal education to the subject,been a rocket scientist for example doesn't cut it.

I think is safe to trust in it rather in our intuition,bad habits,old buys we can't get rid off,new stuff that we are not able to get,you know,the works.
Defending stuff like that has little to do with sound.

 

[Keith_W]

To interpret these studies it needs formal education to the subject,been a rocket scientist for example doesn't cut it.

Is that a subtle dig at a certain Klippel owning Youtuber with a Southern drawl who has been banned from ASR?

 

[Sokel]

Is that a subtle dig at a certain Klippel owning Youtuber with a Southern drawl who has been banned from ASR?

If you have a Klippel and know exactly how to operate it properly you let the data speak or interpreted for you

 

[Digital_Thor]

Did measure several colume speakers same height same room an all had more or less same FR. Some had forward or backward bass ports, symmetrical load (listen 30 years to that system very tight bass) , refelx an my favorite transmission line. My impression is that after measurment an correction they more or less started to sounded the same difference was found in stereo imaging like staging depth some tighter bass some less. The reason i changed to the Vandersteen speakers which are by design build phase coherent time alignt. I expected the other speaker do the same because the DSP i'm using takes care about phase/time guess not enough.

Exactly.... you point it out yourself The biggest difference between when a DSP can do wonders or not - deeply depends on the understanding of what it can - and what it can't. A complex 3D dispersion spreaded soundfield from a speaker, can never be corrected with a DSP - that's just the laws of nature - not something I cooked in my head