Respectfully, Rudolf and others, and in light of the earlier comment that all voicing is deviation from accurate, aren't ALL speaker makers' voicings a bump in the bass and treble? The west coast sound, exemplified by JBL and Altec Lansing from the days of my rock and roll youth, were all about bass and treble as I recall.
So, does it all get down to a simple truth, that people like music to have more bass and treble? It makes me wonder if there is something causing "neutral" recorded music to lack the bass and treble that is in live performance spectrums, and why that would be.
I listened to a lot of speakers at dealers recently, and when it came to the Gallo 3.5, the sense of accuracy was stunning, both reproduction of instruments, and reproduction of the space the music lived in. Imaging that no other speaker (monitors in the <5000 range) came close to. But the music was flat, uninteresting, unemotional. The speakers' abilities were exciting, but not the music. Maybe neutral accuracy is in some way not accuracy.
Jim
Last edited:
Voicing is really just the process of fine tuning the response (measured or perceived) to hit whatever target the designer desires. It doesn't require aiming for some non-flat personality, although you can if that is your desire.
My approach has aways been to get the driver blends working well, the response fairly flat and then fine tune by ear with a lot of comparison to other systems. You can always identify a few crossover components that have a broad effect and fine tune them up or down as desired.
For me the end result is generally close to flat but I feel I end up with a better balanced system through this approach, rather than just tuning by measurement.
David S
My approach has aways been to get the driver blends working well, the response fairly flat and then fine tune by ear with a lot of comparison to other systems. You can always identify a few crossover components that have a broad effect and fine tune them up or down as desired.
For me the end result is generally close to flat but I feel I end up with a better balanced system through this approach, rather than just tuning by measurement.
David S
Modern JBL Studio Monitors are ruler flat with much miore even polar response. What that means, is that the west coast sound is a relic of an era when understanding of loudspeaker design was limited at best.
Not really. As far as treble goes, ""people"" like treble to simply be present, which is a vast switchup from low-fi speakers which are the norm. Various research actually shows that the generally "preferred" treble balance tends to shelf downwards as frequency rises.
As far as synthetic contemporary music goes, yes many people like elevated bass - often on the order of 10 to 20 db - that is the classic nightclub sound. But as far as sound reproduction goes, there has been no evidence that a raised bass level correlates to preference. Great speakers are all essentially ruler flat in the upper-bass-to-lower-midrange anechoically, although not once placed into rooms.
It doesn't, though. But it's true that a lot of speakers have poor resolution. So by elevating the treble levels you can give the false illusion of "increased detail".
It's the show-room floor. People auditioning speakers don't really know what to look for. They're not thinking in the long term of accurate sound. They're listening to the speakers, not the recordings.
What grabs their attention?
For starters, it's not that they prefer more bass, but that they inherently expect "lots of bass".
Second it's not that they "expect more treble", but that they are surprised by higher treble levels and mistake it for detail. A good example is the B&W diamond series. People are told that it has a diamond tweeter, and then they listen to the speaker, which has a big peak around 4khz. Their impressions are that the "diamond" has "more detail" than any other tweeter, when really that's just "more energy". It's not accurate, but it's attention grabbing.
That doesn't mean people would prefer this speaker in the long run. But that 4khz peak (and its marketing correlation to "the resolving diamond tweeter")
So basically listening to music live is unexciting? I think you're just drawing false conclusions. An example being that what you listened to was in fact "accurate". You don't know that. There's no objective evidence of it.
And even a speaker that "seems neutral"" can be lacking due to other design decisions - underattenuated diaphrahm breakup(often the case with speakers employing shallow order crossover... like that Gallo 3.5), insufficient dynamic output (can make music sound lifeless), or a lack of spaciousness due to its room interaction. Stereo fundamentally does not have the spaciousness of multichannel music, however with the right room interaction can come very close.
So how about some impressions of a speaker whose objective measurements actually stack up? Maybe Revel Salon2, JBL LSR6332, KEF 207/2, John K's Nao Note, etc.
It's kind of ironic that that peak is caused by the FST midrange driver rather then the diamond tweeter.😀
I think a lot of brand name manufacturers do this sorta thing, don't they?
Voicing for me is done after the crossover and speaker is designed on paper and assembled. That's the point where you listen to it and make subtle changes depending a little bit on how you like it to sound or "balance".
Manufacturers probably do this, but make additional changes, like "we want a little more tightness in kick drums" so they'll poke around some crossover values and make little bumps where they think the customer's perception will lead them to make a purchase.
In the world of subject prefence speaker buying its all about expectation and excetuation of that expectation. Make a loudpeaker that is smooth and near flat (a small HF rolloff) with good polar response and power handling and then get used to that. Everything else will sound very bad after that. But take this same speaker into a typical showroom and it will sound unexciting.
To me voicing is taking different iterations of the "designed on paper (er....CAD) crossover" and comparing them. It's the stuff we're always comparing here... IE "What kind of forward lobe yields the best subjective sound" or "how good does phase tracking need to be without 60-part crossovers. things of that sort. Not really "tweaking" but rather "efficientizing" which is not a word.
Last edited:
Frequency response on axis is certainly the biggest factor in voicing - or Tonal Balance.
You can walk from room to room at an audio show and hear tonal balances all over the map. Some of that has to be the designer looking for a certain sound. Bumps and dips in certain places give speakers their character, just as they do for ladies. 🙂
Funny, no one ever mentions harmonics, which are certainly important. I find that the harmonic fingerprint of a driver has a lot to do with its sound and how well it plays with other drivers. Not all drivers play well together. Altering the FR can help that, and may be done unconsciously in the voicing stage.
You can walk from room to room at an audio show and hear tonal balances all over the map. Some of that has to be the designer looking for a certain sound. Bumps and dips in certain places give speakers their character, just as they do for ladies. 🙂
Funny, no one ever mentions harmonics, which are certainly important. I find that the harmonic fingerprint of a driver has a lot to do with its sound and how well it plays with other drivers. Not all drivers play well together. Altering the FR can help that, and may be done unconsciously in the voicing stage.
Oh yea, remember JBL and the "West Coast Sound". AKA big midrange hump so they jumped off the shelf in a showroom.
in my world its not possible to do 'voicing', the way you think of it
its not possible like that at all
you can do it by design, choice of drivers, etc
but just by adjusting the xo, no sir
you can 'use' the 'voicing' to 'find' the optimal function
and once you are 'there', thats what you can achieve, and not much else
if you do, it will only be off
it is about getting the voices right, yes absolutely 100%
and you may call it 'voicing', if you like
but thats not what it really is
its not possible like that at all
you can do it by design, choice of drivers, etc
but just by adjusting the xo, no sir
you can 'use' the 'voicing' to 'find' the optimal function
and once you are 'there', thats what you can achieve, and not much else
if you do, it will only be off
it is about getting the voices right, yes absolutely 100%
and you may call it 'voicing', if you like
but thats not what it really is
Is An Extreme Level of Better, Desirable?
This discussion is all well and good, but the sound of live performances (that we are trying to faithfully reproduce here) is mostly coming from the audio/loudspeaker system the performers brought with them. In an overwhelming number of cases, these systems are of poor quality, and are typically overdriven into high levels of distortion. The content of the recorded signal, as well, lacks dynamic range and ambiance, because the signal has been aggressively "strained" through a compressor and mikes to pickup venue acoustics are missing in the mix. So now as we improve our loudspeaker performance, the net result is to accurately reveal these short comings in their entire, cacophonic splendor. Thus, by raising the audio quality standard, we reduce the repertoire of listenable performances.
Regards,
WHG
This discussion is all well and good, but the sound of live performances (that we are trying to faithfully reproduce here) is mostly coming from the audio/loudspeaker system the performers brought with them. In an overwhelming number of cases, these systems are of poor quality, and are typically overdriven into high levels of distortion. The content of the recorded signal, as well, lacks dynamic range and ambiance, because the signal has been aggressively "strained" through a compressor and mikes to pickup venue acoustics are missing in the mix. So now as we improve our loudspeaker performance, the net result is to accurately reveal these short comings in their entire, cacophonic splendor. Thus, by raising the audio quality standard, we reduce the repertoire of listenable performances.
Regards,
WHG
Don't you think that approach kind of contradicts your pleads for adhering to the scientific school of speaker design, where flat is best? Or do you perhaps compensate slightly for some unevenness of the directivity?
I think voicing a speaker has more to do with equal-loudness compensation and specific room acoustics than with accuracy.
There's also live performances without electric amplification. This is probably what most audiophools connect with the word "live".
Anyway, if a reproduction system gets a symphonic orchestra right, it will also get Mötorhead right. If not then there's something wrong with the overall approach of sound reproduction.
Wrong!
A partial quote is a malformed basis for refutation.
The word to use here is unplugged.
The characterization of "Live" describes a recorded performance made at an out-of-studio venue for the benefit of the audience in attendance.
Considering the body of program material available, those that qualify as Live but Unplugged performances are definitely in the minority. There are a lot more artistically meritorious, plugged-in performances to choose from, albeit a lot of them range from somewhat distorted to downright nasty.
Regards,
WHG
A partial quote is a malformed basis for refutation.
The word to use here is unplugged.
The characterization of "Live" describes a recorded performance made at an out-of-studio venue for the benefit of the audience in attendance.
Considering the body of program material available, those that qualify as Live but Unplugged performances are definitely in the minority. There are a lot more artistically meritorious, plugged-in performances to choose from, albeit a lot of them range from somewhat distorted to downright nasty.
Regards,
WHG
- Status
- Not open for further replies.