To the degree that the directivity is truly constant, power response approaches the direct response. If that is flat, they both are.
Flat power response, however it occurs, is generally regarded as "overly bright" (see Toole).
Flat power response, however it occurs, is generally regarded as "overly bright" (see Toole).
MBL101 must have a flat power response, is it bright? ESP12 has flat on-axis response (except diffraction dip) - is it bright? How much power response (or 22.5-30deg) should fall at HF?
All of the WGs I've build seem to be on the bright side, especially when listening pretty much on axis. This is quite interesting because I have not tried to identify what is causing the brightness yet. My initial guess would be it's where the acoustic impedance peaks.
Then how do you know where it is? Is there an acoustic measurement, or is it just simulation of the device?
Currently, I have only simulations of acoustic impedance. It's pretty hard to get the desired directivity, the flat acoustic impedance, and clean impulse. For example, the sims I post here:
http://www.diyaudio.com/forums/multi-way/103872-geddes-waveguides-423.html#post2162214
http://www.diyaudio.com/forums/multi-way/103872-geddes-waveguides-423.html#post2162214
George The peak in your data is from the driver and not an "acoustic impedance" peak in a waveguide. The acoustic impedance of a waveguide at higher frequencies (when designed properly of course), its fairly flat. If there is a peak at all its at the lower end near what would be called "cutoff".
What's the reason for this observation? Are overly bright mixes the result of the "typical" control room acoustics which tends to absorb more high frequencies than low frequencies?
If this is true, the use of CD speakers in untreated domestic listening spaces might be the wrong approach?
Or is all of this just the result of different listening levels in control rooms vs. living rooms?
Last edited:
It's nowhere near as complex as that. Higher frequencies have relatively more power in a CD design, (as you know) compared to what most songs are mixed on/intended for. A gentle downward slope of the axial response is generally a good balance between correcting for this and not dulling down recordings that AREN'T mixed hot to balance the falling power response of cone/dome arrangements. Since you'll never have every recording work in a given system, the downward slope system is the best fit.
Aren't most studio near field monitors CD because they need to deliver the very same frequency response over a wide listening window?
Maybe it's the mastering engineer that adds even more complexity to the equation. Three very different setups:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Back when I was doing a lot of live music mixing, from time to time a band would drag in the studio engineer who had done their CD. So I would turn over the console to the studio wiz.
Without fail, their mixes were bright, bright, bright. Cut your ears off! I could see the audience cringe. My theory is that they where so used to the tonal balance and detail of nearfield monitor mixing that they were trying to recreate that sound in the big venue. It didn't work, it sounded awful. These guys were not hacks, I'd often heard their good studio work, but they just didn't understand the demands of a larger room and farther listening distance.
Somehow, I think our brains expect different listening conditions to sound different. When they don't match what we expect, they sound "wrong."
Without fail, their mixes were bright, bright, bright. Cut your ears off! I could see the audience cringe. My theory is that they where so used to the tonal balance and detail of nearfield monitor mixing that they were trying to recreate that sound in the big venue. It didn't work, it sounded awful. These guys were not hacks, I'd often heard their good studio work, but they just didn't understand the demands of a larger room and farther listening distance.
Somehow, I think our brains expect different listening conditions to sound different. When they don't match what we expect, they sound "wrong."
markus76,
Nice pics. There is a big disconnect between the real world, and the ideal one sometimes. People need to be reminded of that every once and awhile.
Nice pics. There is a big disconnect between the real world, and the ideal one sometimes. People need to be reminded of that every once and awhile.
Last edited:
Are you referring to the second pic down the right hand side? Well, VACs generates it as acoustic impedance, not sure what scale it's using. Yes, it could be from the driver, but the WG loading can change that.
I'm sorry George, but which curve is Acoustic Impedance? It would need to be in Acoustical Ohms (kg / m^2 / Sec), Specific Acoustic Impedance (kg / m^4 / Sec) or Mechanical Ohms (kg / sec.). Otherwise its not an impedance in the mechanical or acoustical domain.
From my experience as a tendency omnis or dipoles will have less
issues with "brightness" compared to monopole+WG designs, when
having comparable on axis response.
If in the Toole Article it is stated, that a flat power response will
sound overbright, i think this may partly be due to true omni and
dipoles not participating.
I am also experimenting "with some kind of waveguides" at the moment,
and i prefer off axis listening. The main radiation axis should point
to "somewhere" in the room, where reflections can be excited, that have
similar delay like the reflections present in the midrange.
Having the WG main radiation axis pointing at the listener or a (reflective)
rear wall is not a good thing IMO ...
From pure listening experience i feel, that "brightness" is not only
felt proportional to on axis response.
When the ratio of direct vs. reflected sound (in intensity and delay) changes
from the midrange to the presence and brillance region, thereby presence and
brilliance region causing much less reflected sound (in relation), a flat on axis
response will sound much too bright.
The effect can be smoothed a lot IMO , when a contribution of
- not too early - reflection is maintained in the presence to
brillance region. That contribution of reflected sound may be small
in terms of energy, but it helps to avoid midrange and presence/brilliance
getting out of balance auditively.
A more directional midrange will also help do get the balance right,
that would be the other way to go.
Balance in the end does not ask how it was achieved, does it ?
But nevertheless a slightly falling response (at listening axis and
and in power) will remain the most "good natured" tuning, making
most if not all recordings enjoyable.
Kind Regards
Last edited:
The right hand file attached in this post
http://www.diyaudio.com/forums/multi-way/103872-geddes-waveguides-423.html#post2162214
it's the second graph on the right hand colum. I probably need to go back and read what units it (VACs) actually uses, but I generally just hope to get it flat.
Sorry George, I can't read any of those, but they don't look like what I would expect an Acoustic Impedance to look like.