"Voicing" or "tuning" a speaker after a simulation... cheating?

[fastbike1]

Re: "Voicing" or "tuning" a speaker after a simulation... cheating?

Not really. You know your room isn't helping. I understand that you answer to a higher power like many of us, but it still seems that a certain amount of room correction could be accomplished through decor.

I wish I could, though. The subjective part of the hobby is the one I dislike :( I'm not very comfortable being creative, so to speak. I'm more comfortable with tried and true solutions.

[Jeff B.]

Re: "Voicing" or "tuning" a speaker after a simulation... cheating?

I wish I could, though. The subjective part of the hobby is the one I dislike :( I'm not very comfortable being creative, so to speak. I'm more comfortable with tried and true solutions.

Back to my question above - somethimes nothing is as deceptive as a single response measurement. And many of the people who post here and believe that flat response is the goal make this mistake. I don't rely on a single measurement - I take dozens on a multitude of axes, I measure distortion, I simulate the power response,.... etc, all of this goes into the design of my crossover. Relying on a single response measurement on-axis from a Meter away can sometimes be worse than a poor indicator of a speaker's sound, it can be an altogether erroneous indicator that leads you in the wrong direction with your design choices. This is especially true if your measurements are made in a completey different environment than the one you are going to listen to them in. Over the years I have developed a process where my measurements and my modeling result in a crossover that I usually don't need to change for voicing purposes once the speaker is put together and listened to, but it took a lot of learing, and questioning, and being honest with myself, to get there.

[fjhuerta]

Re: "Voicing" or "tuning" a speaker after a simulation... cheating?

I understand that you answer to a higher power like many of us,

Post of the year. :D

I already got some sofas inside the room and they helped tremendously.

My wife said "get rid of them, there's not enough space for XXXX!".

I wish I could keep them. The room is a lot better with them inside... we use to have solid wood furniture. It's like the Caves of Altamira inside it...

[fjhuerta]

Re: "Voicing" or "tuning" a speaker after a simulation... cheating?

Back to my question above - somethimes nothing is as deceptive as a single response measurement. And many of the people who post here and believe that flat response is the goal make this mistake. I don't rely on a single measurement - I take dozens on a multitude of axes, I measure distortion, I simulate the power response,.... etc, all of this goes into the design of my crossover. Relying on a single response measurement on-axis from a Meter away can sometimes be worse than a poor indicator of a speaker's sound, it can be an altogether erroneous indicator that leads you in the wrong direction with your design choices. This is especially true if your measurements are made in a completey different environment than the one you are going to listen to them in. Over the years I have developed a process where my measurements and my modeling result in a crossover that I usually don't need to change for voicing purposes once the speaker is put together and listened to, but it took a lot of learing, and questioning, and being honest with myself, to get there.

Well, as I said, it's not that single measurement. That one proved to me there was something wrong (the in room RTA). When I measured them again with MLS I saw I had made a mistake in my original measurements...and that's when I decided to dig deeper into what was happening.

I'm barely scratching the surface on power response, off axis response, crossover topologies, etc. in my designs. I'd be very interested if you ever decide to write about your process!

[Pete Schumacher]

Re: "Voicing" or "tuning" a speaker after a simulation... cheating?

Well, as I said, it's not that single measurement. That one proved to me there was something wrong (the in room RTA). When I measured them again with MLS I saw I had made a mistake in my original measurements...and that's when I decided to dig deeper into what was happening.

I'm barely scratching the surface on power response, off axis response, crossover topologies, etc. in my designs. I'd be very interested if you ever decide to write about your process!

Sound deadening can be quite a bit less obtrusive than you might think.

A few strategically placed panels of Roxul or pressed fiberglass, with some nice, decorative cloth wrapped around them and placed on the wall, can serve dual functions of decoration and echo attenuation. And if you build them on stands, that you can move into place for a listening session, all the better.

Have a look at the red panel in the middle of the pic. The whole thing, stand and panel, weighs maybe 20 pounds. 5 or 6 of those placed in your listening room will do wonders, and you can put them away when done. Ryan even added a string of red Christmas lights to the back, so you can have a deep red glow when you turn the lights down to listen.

Very relaxing, to the room acoustics, and the eyes.

[DanP]

Re: "Voicing" or "tuning" a speaker after a simulation... cheating?

Anyway, after taking measurements with a lot of care, I found myself with a nice on-axis response, very controlled off-axis curves, and a decent enough impedance.

Yet, I still find the speakers "bright".

There's one other thing. The room it is in. It's all glass and ceramic floors, plus hardwood furniture. There's very little to no absorption and a bit of diffusion (exposed bricks). So I'd assume that any trace of harshness would be amplified quite a bit.

I think you've found your culprit. A flat speaker in that room will sound bright every time. As the other guys are saying, some room treatments are your best bet, but if that's off limits then you may have to "cheat" a bit in the crossover. Make it sound good however you have to.

Dan

[LouC]

Re: "Voicing" or "tuning" a speaker after a simulation... cheating?

Quite a provocative thread title, I think.

... Which leads me to think. I cheated. I mean, the speaker should sound fine when designed properly. It's just me and my ideas (that a big speaker should sound, well, bassy and powerful) that led me to alter it via "voicing".

I sort of think most people would think the speaker sounds great when I voiced it as flat as I could.

Does this make sense to anyone else?

Aural voicing isn't cheating, its half the process! It's the difference between science and engineering. Why trust measurements to tell the whole story? Why trust anybody else's for that matter?

At best the modeling process provides a "theory" starting point, that must be turned into a "product" by engineering. The actual environment, your taste, your source chain, your hearing all affect the final product. I can't remember designing a "flat" speaker that I could listen to for more than 10 minutes. Planars and ribbons do seem to be easier to voice flat, but how do you measure them "accurately" -- they are not point sources.

Measurements do not account for the nuances of the drivers either. I'm betting we can find a $30 Vifa that measures better than a $200 ScanSpeak. But I'll bet the ScanSpeak sounds better. It's like designing a house with square walls and expecting the wood won't be warped.

[johnnyrichards]

Re: "Voicing" or "tuning" a speaker after a simulation... cheating?

This is why I love this forum. There is nothing wrong with flat on-axis, provided off-axis doesn't create its own problems (think about the bloom above the crossover point in 4th order LR networks for example).

A dip is often used when drivers are not working in their optimal range - most cheap midwoofers have a rising 3rd order harmonic in the crossover range, suppressing the fundamental in that region can help "fix" it, call it a BBC dip if you want to, but it works equally well to combat certain driver shortcomings. I am starting to think most deliberate dips are the result of poor decisions elsewhere, not necessarily a bad thing by any stretch of the imagination - but to use Mike's "Fatbellies" as an example, I wonder if the forwardness he experiences with the tweeter in normal polarity is from the low crossover point? The dip very well could be helping to mask some distortion in that range.

On the bottom end, a high Q bump can help mask the lack of extension of smaller drivers, or help mask floor bounce. I find it amusing that many of us (myself included) will look at drivers with a high Q with suspicion and then go ahead and engineer that into our designs

Then there is the room and music preferences and personal hearing loss...

These are the threads the boys from Harman need to chime in on - they have some experience removing the eyeballs and wallets from the listening experience.

[jclin4]

Re: "Voicing" or "tuning" a speaker after a simulation... cheating?

..to use Mike's "Fatbellies" as an example, I wonder if the forwardness he experiences with the tweeter in normal polarity is from the low crossover point? The dip very well could be helping to mask some distortion in that range.

Maybe it is the 5th order harmonic? Nevertheless the Fatbellies sound great with tweeter polarity as designed and were one of my favorites at the NYDIY.

[fjhuerta]

Re: "Voicing" or "tuning" a speaker after a simulation... cheating?

Cool. I'm learning stuff! :D

I suppose it's the room. I'll be working a bit with EQ before committing to a x-over change.

[mzisserson]

Re: "Voicing" or "tuning" a speaker after a simulation... cheating?

This is why I love this forum. There is nothing wrong with flat on-axis, provided off-axis doesn't create its own problems (think about the bloom above the crossover point in 4th order LR networks for example).

A dip is often used when drivers are not working in their optimal range - most cheap midwoofers have a rising 3rd order harmonic in the crossover range, suppressing the fundamental in that region can help "fix" it, call it a BBC dip if you want to, but it works equally well to combat certain driver shortcomings. I am starting to think most deliberate dips are the result of poor decisions elsewhere, not necessarily a bad thing by any stretch of the imagination - but to use Mike's "Fatbellies" as an example, I wonder if the forwardness he experiences with the tweeter in normal polarity is from the low crossover point? The dip very well could be helping to mask some distortion in that range.

On the bottom end, a high Q bump can help mask the lack of extension of smaller drivers, or help mask floor bounce. I find it amusing that many of us (myself included) will look at drivers with a high Q with suspicion and then go ahead and engineer that into our designs

Then there is the room and music preferences and personal hearing loss...

These are the threads the boys from Harman need to chime in on - they have some experience removing the eyeballs and wallets from the listening experience.

If you have any suggestions to the poor decisions I made with the Fatbelly's, I am all ears. Furthermore distortion actually DECREASES at x-over through the tweeter's response. Simulated power response is smooth, to spite the dip in fq response. SO at what point to you quantify an arguably audible phonomea showing up on a microphone as precieveable, or not precieable to the complexity of our earS?

What's the white stuff in bird p00p? It's p00p too! Sometimes things that should not work, do. Jeff is very correct as to a single measurement not defining a speaker's sound. Heck, you look at Stereophiles measurements of ANY speaker and they are all over the place. Yet many are raved about from the $199 pioneers to the $35,000 monsters of moneymaking. Quite honestly I have found waterfalls, simulated power responce, distortion and impedance to mean more than anything. Sure, it's my flavor and everyone like's their own. But bottom line is if it sounds good, there is not question that it is a good design.

The whole point is I would hate to see anyone discourage taking whatever path works for them. Weather we like it or not, there is more than one way to skin this cat. If voicing gets anyone happy with their design, so be it. There is nothing inherently "wrong" with it.

[johnnyrichards]

Re: "Voicing" or "tuning" a speaker after a simulation... cheating?

Mike, what is the excursion of that dome tweeter?

Don't take it personally, just speculating. I think the 1K crossover was the poor choice, and no coincidence that you have a depression in that region to make it work.

[mzisserson]

Re: "Voicing" or "tuning" a speaker after a simulation... cheating?

Mike, what is the excursion of that dome tweeter?

Don't take it personally, just speculating. I think the 1K crossover was the poor choice, and no coincidence that you have a depression in that region to make it work.

None taken. The Fs is 475Hz and 1.6mm is listed as the excursion. .3mm is "linear". As I said the weird thing is the distortion measured very low either way. .. Mabye jclin is onto something that 5th order does come way up from 600 to 1K. The brickwall on the Fatbelly's tweeter is 800Hz, crossed at -3dB at 1200 with a 2nd order slope to 800.

5th order... weird. What the heck caused 5th order? I never pay much attention beyond 3rd unless something is heinously wrong.

[johnnyrichards]

Re: "Voicing" or "tuning" a speaker after a simulation... cheating?

Some 90db distortion sweeps would be interesting to see.

[GranteedEV]

Re: "Voicing" or "tuning" a speaker after a simulation... cheating?

It's not cheating, but I do sincerely believe the need to voice is highlighting some fundamental design flaw, somewhere. For example, those drivers would have very wide dispersion... that might not be the best idea.