Measuring for tonal characteristics

[Mivera]

Put me in the camp that says that everything in audio electronics, that we make, can be measured. Audio is simple. Its just you have to take time when you want to make in depth measurements. We can move atoms around to make initials, such as IBM, and take a picture of it, that was like 20 years ago, we can damn well measure any audio signal as it can only be frequency, phase, amplitude, etc.

Don't confuse what I said , I did not say we can measure why someone likes something, we can measure an audio signal though. There are no mysteries in measurements. Audio is kind of like American politics, you have two parties, and no party can even get off the ground, even if it is superior. In audio, you have stereo and multichannel, and nothing new that can really give us the feeling that we are in an immersive on the scene experience.

Okay describe the measurement data that will tell us the difference in tonal characteristics we should expect between the 3 tweeters I shared the spec sheets for.

 

[Thomas savage]

The way every speaker on the planet is designed, is through a combination of subjective and measurement data. Same with drivers. all of the worlds top driver manufacturers, Scan-Speak, Accuton, Seas, Vifa, Focal, Audio technology, Dynaudio etc, use extensive subjective evaluations to design their drivers. If everyone who worked for these manufacturers were completely deaf, and all they used were industry standard measurements to design their drivers/speakers, I would just imagine how they would sound.

i believe with modern computing techniques you can get close just by CAD of some kind but yes once its build you have to listen to refine. it will depend on the known factors within the design brief though. there is always a lot more to these things. i was shock at the level of known predictable events in the design process.. its not really as much of a artistic fancy special magical unknown as most audiophiles like to think.. its engineering for the most part imo.

 

[amirm]

But how is it any different than comparing DSD with PCM? At the end of the day if both measure excellent, isn't it subjective preference that matters the most?

They are different animals. Comparing two shades of red that are potentially just hair different (DSD vs PCM) is a very different matter than comparing red to purple (speaker vs speaker). The former we want to test for difference being perceptible. The latter, we know they are perceptible and want to know which is preferred.

 

[Mivera]

i believe with modern computing techniques you can get close just by CAD of some kind but yes once its build you have to listen to refine. it will depend on the known factors within the design brief though. there is always a lot more to these things.

But the designer still has to have the knowledge of the sound different materials will make on the end product. For example even if Dic designed the Vivid line of speakers without having ever listened to them, surely he had heard an aluminum dome driver before to have an idea of how it would likely sound.

 

[amirm]

Here is an example blind preference testing between Room EQ products. Each created a very different response curve and hence, tonal quality:

Focusing on the graph on the left, the dot represents the mean, and the line through it, the difference in preference. Notice how small those variations are between testers (i.e. line length) as compared between one EQ system and another (location of the blue dot). As such, preferences differences between people is immaterial with respect to which tonal response was preferred. Everyone disliked RC6 for example compared to RC1, RC2 and RC3.

 

[tomelex]

Okay describe the measurement data that will tell us the difference in tonal characteristics we should expect between the 3 tweeters I shared the spec sheets for.

That's not what I said, read the second paragraph, we can measure an audio signal, we don't intend to measure preferences!

 

[Thomas savage]

But the designer still has to have the knowledge of the sound different materials will make on the end product. For example even if Dic designed the Vivid line of speakers without having ever listened to them, surely he had heard an aluminum dome driver before to have an idea of how it would likely sound.

the drivers came before the speakers, they were inspired by his old work at B&W yes but dic also has a background in pro audio.. i think with the vivid speakers it was designing the cabinets to put the drivers in..but ultimately you would have to ask him. he does design the lot though, magnate design driver assembly cabinate.. everything, and its more or less just him i think so he has 100% control over the process.

as for the aim, its simple. the driver is to impart as little of its own character to the sound as possible. ditto for the enclosure. the end.

 

[Mivera]

They are different animals. Comparing two shades of red that are potentially just hair different (DSD vs PCM) is a very different matter than comparing red to purple (speaker vs speaker). The former we want to test for difference being perceptible. The latter, we know they are perceptible and want to know which is preferred.

What if the difference in the speaker was just the material of the dome in the tweeter? This isn't much of a difference.

In the 3 Transducer labs examples I shared, what data can I gather from the manufacturers data sheets that will provide me with the answer to what the best tweeter is for my application?

 

[Mivera]

the drivers came before the speakers, they were inspired by his old work at B&W yes but dic also has a background in pro audio.. i think with the vivid speakers it was designing the cabinets to put the drivers in..but ultimately you would have to ask him. he does design the lot though, magnate design driver assembly cabinate.. everything, and its more or less just him i think so he has 100% control over the process.

as for the aim, its simple. the driver is to impart as little of its own character to the sound as possible. ditto for the enclosure. the end.

Yes of course, many speakers are designed with the same goal in mind, but for some reason they all sound different.

 

[Mivera]

Here is an example blind preference testing between Room EQ products. Each created a very different response curve and hence, tonal quality:

Focusing on the graph on the left, the dot represents the mean, and the line through it, the difference in preference. Notice how small those variations are between testers (i.e. line length) as compared between one EQ system and another (location of the blue dot). As such, preferences differences between people is immaterial with respect to which tonal response was preferred. Everyone disliked RC6 for example compared to RC1, RC2 and RC3.

Okay how would we know what the best tweeter is for an application?

 

[Thomas savage]

Yes of course, many speakers are designed with the same goal in mind, but for some reason they all sound different.

i did ask as i was interested by the whole what came first.. like with music itself. some bands write the lyrics then the song others come up with the hook first or beat.. or whole song then words..

i asked shed loads of questions.. poor bugger

well lots of speakers are compromised by lack of engineering integrity. its just another box with a shinny tweeter and nice wood grain or we can make pinos so we can make speakers...

 

[amirm]

What if the difference if the speaker was just the material of the dome in the tweeter? This isn't much of a difference.

If listening tests show no difference in preference, then you can choose them based on other criteria (reliability, power handling, cost, looks, etc.). With statistical analysis we can still tell if a difference was heard or not.

In the 3 Transducer labs examples I shared, what data can I gather from the manufacturers data sheets that will provide me with the answer to what the best tweeter is for my application?

You can't make that decision in a vacuum in this instance. Research from listening tests tells us what the overall response of the loudspeaker should be. How to get there is up to the designer to mix and match components to get there.

 

[Mivera]

i did ask as i was interested by the whole what came first.. like with music itself. some bands write the lyrics then the song others come up with the hook first or beat.. or whole song then words..

i asked shed loads of questions.. poor bugger

Well the drivers always come first. And I highly doubt he designed them. He maybe specified the materials/design type based on his past experience building speakers, and knowledge of driver designs and materials though.

 

[Phelonious Ponk]

I think you need to define what you mean by "tone."

Tim

 

[Mivera]

If listening tests show no difference in preference, then you can choose them based on other criteria (reliability, power handling, cost, looks, etc.). With statistical analysis we can still tell if a difference was heard or not.


You can't make that decision in a vacuum in this instance. Research from listening tests tells us what the overall response of the loudspeaker should be. How to get there is up to the designer to mix and match components to get there.

There's 3 ways it could be done:

1: Buy all 3 and listen to them
2: gather subjective opinions from others who you trust and have heard them
3: Use the experience you have with drivers/materials/implementations to make an educated decision on how it will likely sound subjectively.

 

[Mivera]

I think you need to define what you mean by "tone."

Tim

What are the cone/dome materials in your current speakers?

 

[Thomas savage]

Well the drivers always come first. And I highly doubt he designed them. He maybe specified the materials/design type based on his past experience building speakers, and knowledge of driver designs and materials though.

i think he did mike lol

 

[Mivera]

i think he did mike lol

Design the drivers? Well there's different levels of designing drivers. I could design drivers too by giving Scan-Speak perimeters on exactly what I'm looking for in a driver and have them custom build it.

 

[Thomas savage]

Design the drivers? Well there's different levels of designing drivers. I could design drivers too by giving Scan-Speak perimeters on exactly what I'm looking for in a driver and have them custom build it.

http://www.vividaudio.com/drive-units.html
any doubts are best addressed by contacting them.

 

[Mivera]

http://www.vividaudio.com/drive-units.html
any doubts are best addressed by contacting them.

I seen that before, but it doesn't say who designed the drivers. From what I heard, there's only a handful of top driver engineers in the world. Most driver manufacturers hire them to design the drivers for them. a couple of the top guys were recently stolen by the Chinese to make the Chinese Wavecor, and SB acoustics line of drivers.

http://www.wavecor.com

http://www.sbacoustics.com