Should this result in a different speaker than accurate production of anything else?
Hayseed Dixie being fronted by John Wheeler, a recording engineer by trade, use no eq or compression at all.
Doesn't push up the 0.1% though.
No, but I figure if you can make a speaker sound indistinguishably live then it will accurately produce anything.
Hmmm. Maybe nobody takes a felonious hand to the signal after it leaves the mic, but where you stick the mic (and everything about how that mic is made, housed, and hung) is like making a million processing decisions. Some of those decisions can be real awful.
Like "organic" pesticides that can be real poisonous to humans, many decisions are made in the course of producing "natural" recordings.
All recordings are cooked. Some more egregiously than others.
B.
Last edited:
You are asking for an effectively perfect speaker. It can be done. It is not simple.
A decision you want to make is whether to keep the stereo image intact by controlling interactions with the room. You want to know what issues are more audible so that you can make effective compromises. You want to learn to control bass in a room.
My enthusiast experience have tought me that steeper crossovers return better results. In the common examples I can think only of the Linkwitz LX monitor is an example of a well behaving first order crossover - the one spliting the midrange channel into two. All my trial and error with crossovers has lead me to the conclusion that nothing ruins speaker performance as shallow crossovers leading to bad directivity and lobing.
The second most frequent problem with crossovers is the lack of energy or the excessive energy in the crossover region. The third is impedance messing with crossover and/or attenuation. Use zobels everywhere. Open baffle is another good cure for eliminating side reflections. Wide baffle if the space is available is yet another cure for common problems.
Nothing special to teach about. You just take a bass capable cabinet and driver and construct the other two or three ways ontop of it.
Just do it!
The second most frequent problem with crossovers is the lack of energy or the excessive energy in the crossover region. The third is impedance messing with crossover and/or attenuation. Use zobels everywhere. Open baffle is another good cure for eliminating side reflections. Wide baffle if the space is available is yet another cure for common problems.
Nothing special to teach about. You just take a bass capable cabinet and driver and construct the other two or three ways ontop of it.
Just do it!
Last edited:
The hardest part is comitting to 4 ways and designing the baffle and enclosure alignment together with deciding on specific bands. Passive, active or DSP is the less labor intensive part of the building.
In the end you'll use the speakers in the esiest way through the receiver with the most convenient remote control
and hook them to the big stereo amplifier only to show off to friends with some fancy blue ray or vinyl.DSP's and active crossovers are a great design tool though. But pasive is not that worse than active when you first sort it out with a Behringer DCX2496 or miniDSP
in the end the desire of seeing and touching passive components takes the best of you.
I have had mixed results, examples where some overlap is warranted.
This is a concern, but it can be accounted for.
If the front and rear wave are opposite, wouldn't this increase diffraction around the edge?
This is going to promote 180 degree directivity. Is this going to fit your room interaction requirements?
Sometimes with lower frequencyes and smaller drivers maybe, meaning under 1 kHz and center to center distance of under 20 cm. For instance 6.5 midbass to a waveguide or smaller upper midrange. And that is only to allow for the benefits of impulse and transient response of that kind of crossover. It is still negative to have overlap. Overlaping at wavelenghts smaller than the center to center distance is just horrible.
I dont know the answer for the diffraction, but considering the directivity pattern of open baffles it seems that diffraction also cancels out. Then at higher frequency when both front and back emit in half space maybe there is double diffraction. Diffraction is said to add to distortion, but the positive effect of OB to room interaction is undoubtedly offsetting any possible added distortion. OB's are not famous as experiencing more distortion...
In wall speakers are going to emit 180° up to the point where drivers begin to narrow their emission. But that applies to any speaker. In wall takes baffle out of the eqiasion and allows for bass alignment of any sort. You could use any nice sounding low rezonance bass driver without volume constraints. But the problem with the reflections still persists.
That is why I designed a wide baffle open baffle, for the upper mid and HF it is effectivelly an in wall, and for lower midrange and around 1 kHz it has the so desired side and top null of emission.
Hellokitty, do you have a drawing of your future project?
Last edited:
Any baffle and driver type will have diffractions somewhere in passband, among other nonlinearities. That is one reasont for making multiway speakers.
Dipoles are very difficult in this respect. They have sort of three different ranges of directivity (and sideway radiation). Linkwitzlab will tell more. In short, low frequencies have no edge diffraction and radiate as "dipole model" says. Then at and a bit above on-axis nulling they radiate almost as omnipoles "rings like a bell". Thirdly, drver diameter and membrane profile start to do their effect, typically narrowing directivity of fronside radiation. But rearside is troubled with spider and voice coil/magnet system, also membrane profile is reversed and rearside directivity is just an unpredictable mess.
This is why dipole radiator should be used only for max 3 octaves and often lowpassed with a rather steep xo.
Dipoles are very difficult in this respect. They have sort of three different ranges of directivity (and sideway radiation). Linkwitzlab will tell more. In short, low frequencies have no edge diffraction and radiate as "dipole model" says. Then at and a bit above on-axis nulling they radiate almost as omnipoles "rings like a bell". Thirdly, drver diameter and membrane profile start to do their effect, typically narrowing directivity of fronside radiation. But rearside is troubled with spider and voice coil/magnet system, also membrane profile is reversed and rearside directivity is just an unpredictable mess.
This is why dipole radiator should be used only for max 3 octaves and often lowpassed with a rather steep xo.
I'm still trying to figure it out. For treble I'm probably going to use a ribbon. For midrange/woofer I'm deciding between ribbon, planar, or an array. I don't know which one produces the least distortion with the least amount of issues and this is arguably the most important section to get right. For subwoofer I'm currently looking at the BMS 18N862 with some DIY motional feedback.
At some point I'm going to build a 6+ foot tall "full" range ribbon too. At the end of the day I want to get as close to an ideal speaker as possible. I want an indistinguishably live sound. I'm beginning to think near-field set-ups are potentially superior in terms of practically achieving a close to ideal speaker, what do you think?
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.