Definition of Timbre Matching?

[audiofox]

I have read several threads here and elsewhere about the importance of "timbre matching" the front three speakers with each other. Can someone here define the technical aspects of a speaker's timbre and how one measures it? I am guessing that it has to do with the specific spectral response of a speaker, but I would like to have something a bit less subjective so I can deal with it as a design parameter. My primary motivation is to make a decision on which of my Dynaudio mids and woofers I should combine with each other-unfortunately, I don't have enough of any one midrange or woofer driver to use all the same ones for my setups. I do have all the same tweeters (D260s, which are VERY nice and wideband to boot), so I am hoping that if I use the same tweets for all the speakers, I will be okay and won't have to invest in a spectrum analyzer to determine the response of each speaker.

 

[Johnd]

There is not a universal defintion of timbre-match. So if you are looking for a simple Hz-kHz universal defintion, it simply does not exist. However, timbre is defined as "that distinct tonal quality" that each speaker posesses, as opposed to pitch or loudness. In short, if speakers sound the same, they are timbre-matched. Hope that helps.

 

[j_garcia]

Basically, using the same drivers (identical ideally) similar design and tuning of the cabinet, and a similarly designed crossover, should give you a good timbre match. More or less, getting mains and the center from the same line of speakers from a manufacturer. Highs tend to be immediately noticable, but I feel the midrange is also pretty easy to pick out if the timbre is not close enough.

 

[audiofox]

It sounds like I will probably be okay with the high frequencies, since I am using the same tweeter driver for the front mains and the center (actually, I am using the same tweeter for all speakers in the system). I wonder if Dynaudio has to do a lot of empirical data gathering before they know which drivers can be matched to each other-most of their lines have different woofer sizes for the mains and the center channel speakers.

 

[corey]

Perhaps "timber match" means "not flat in the same way". If you had a group of speakers for front, center, and surrond that all had an absolutly flat in room response of 15hz to 30khz, they'd be timber matched. No speakers behave this way, so the next best is to get speakers that dieviate (sp) from this in a uniform way, so that you don't notice when a sound moves from one speaker to another.

Sounds like you're designing some nice speakers. Best of luck with the project.

 

[Ben Diss]

Imagine three different sopranos, all singing the same note. On your spectrum analyser, all appear the same. To your ear, they are all different. Each has a different voice or timbre.

Speakers each have own voice. You want them to match so that an effect that moves from your right to your left continues to sound like it's coming from the same source, just a different direction.

-Ben

 

[jcPanny]

Harmonics

The reason you can differentiate from different sopranos or different instruments playing the same note is due to the harmonic signature of the voice or instrument. The amplitude of the 2nd, 3rd, 4th . . . harmonics is different for each voice or instrument. You can hear the difference and you could definitely see the difference on a spectrum analyzer.

I think that the matching tweeters is a good start for timber matching. Try to pick woofers that are similar sizes from the center and fronts if they cannot be identical.

 

[j_garcia]

Perhaps "timber match" means "not flat in the same way". If you had a group of speakers for front, center, and surrond that all had an absolutly flat in room response of 15hz to 30khz, they'd be timber matched. No speakers behave this way, so the next best is to get speakers that dieviate (sp) from this in a uniform way, so that you don't notice when a sound moves from one speaker to another.

Flat response doesn't tell you what a speaker sounds like. You can have two different speakers that are both equally flat that use very different drivers, aluminum vs paper for example, and they will sound different.

Different size drivers isn't such a big issue, as long as the drivers have a similar sonic characteristic.

 

[audiofox]

The reason you can differentiate from different sopranos or different instruments playing the same note is due to the harmonic signature of the voice or instrument. The amplitude of the 2nd, 3rd, 4th . . . harmonics is different for each voice or instrument. You can hear the difference and you could definitely see the difference on a spectrum analyzer.

I think that the matching tweeters is a good start for timber matching. Try to pick woofers that are similar sizes from the center and fronts if they cannot be identical.

Maybe I can use a noise generator to attempt to timbre match my drivers-if they have the same spectral content whe driven by the same noise source, then they will probably be close enough in "timbre" to work together.

 

[Geno]

The way I've decided if a speaker system is timbre-matched is to simply listen as I set the speaker levels from my receiver setup. When I had a set of mismatched fronts/center/surrounds, they would have a distinctly different sound as the test tones progressed from one to the other. With my newer 'matched' setup, they sound essentially identical.

 

[PENG]

Imagine three different sopranos, all singing the same note. On your spectrum analyser, all appear the same. To your ear, they are all different. Each has a different voice or timbre.

Speakers each have own voice. You want them to match so that an effect that moves from your right to your left continues to sound like it's coming from the same source, just a different direction.

-Ben

I agree with jc. A spectrum analyser that can display the full spectrum of the signal should be able to show the difference.

 

[audiofox]

I've been looking at TrueAudio's TrueRTA as a possible tool for measuring the spectral response for my drivers-the deluxe version has white and pink noise generation capabilities as well as the ability to measure the response. I also found a nice handheld scope offered by Old Colony Sound Labs that would be useful for other aspects of my project.

I also reevaluated my driver configs and discovered that I could use all the same midwoofers (the Esotec17WLQ) for the front/center/rear channel speakers for my higher end zone, and a D260/20W75 esotec config for the presence (side speakers. That should eliminate much of the timbre matching tests that I need to perform. The other zone will have the remainder of the drivers: D28/15W75 center, D28/D54/24W75 for the front mains, D260/17W75 for the rears and a single NHT 1259 sub (5.1 only for zone 2). The D260 is a better quality tweeter, but the D28 matches better with the D54 and the 24W75 in terms of sensitivity.

 

[mtrycrafts]

I've been looking at TrueAudio's TrueRTA as a possible tool for measuring the spectral response for my drivers-the deluxe version has white and pink noise generation capabilities as well as the ability to measure the response. I also found a nice handheld scope offered by Old Colony Sound Labs that would be useful for other aspects of my project.

I also reevaluated my driver configs and discovered that I could use all the same midwoofers (the Esotec17WLQ) for the front/center/rear channel speakers for my higher end zone, and a D260/20W75 esotec config for the presence (side speakers. That should eliminate much of the timbre matching tests that I need to perform. The other zone will have the remainder of the drivers: D28/15W75 center, D28/D54/24W75 for the front mains, D260/17W75 for the rears and a single NHT 1259 sub (5.1 only for zone 2). The D260 is a better quality tweeter, but the D28 matches better with the D54 and the 24W75 in terms of sensitivity.

You would need something that can do better than 1/3 octave. Even 1/6 may be lacking to show differences.

 

[audiofox]

TrueRTA level 4 can do 1/24 octave for $100.