Well. My experience is, that a more narrow design makes the speaker "disappear" more - if the response is still good on all axis' though. So - when going for one specific design -we always have to consider the compromises that follow along that path.
If you don't want to play party loud, then a smaller midrange of 4-5" is fine. Most of the real "power" behind the sound, comes from moving air lower in frequency.
That midwoofer actually looks fine. I would just really mate it with a waveguide tweeter of 5-6", since the midwoofer start to narrow in dispersion around 1800-2000Hz. So if you use a tweeter "freely" mounter on the front baffle, it will play in all directions at that frequency, where the midwoofer would not. And since it's an acoustical thing, then you can't correct it with a filter or EQ. So you need a wider baffle to support the tweeter - which again makes the waveguide an easier choice in this instance.
Also consider the sensitivity of the drivers. I seem to me that you aim for passive filters. So usually that is why it fits well with two woofers in parallel, to compensate for the difference in efficiency. I use an active filter, which makes it a bit easier to simply change the gain on each driver, after I've EQ'ed them flat, so they match in level.
But I have no real experience with passive filers. So I'll leave you in the hands of all the other experienced guys in that field 😊
I think Dr Geddes has talked about this before.
When there is more mass for the same motor strength, it changes the level. The cone is no longer expected to move as far, so no problem. If the response rolls off, you apply more power to the highs. Problem fixed 🙂
When there is more mass for the same motor strength, it changes the level. The cone is no longer expected to move as far, so no problem. If the response rolls off, you apply more power to the highs. Problem fixed 🙂
So is it just a harmonic distortion issue? I think that the "voices" reproduced by a tweeter or a very small driver are too "strident" and that a 6'' driver should be married with a more generous tw than a 4'' midrange which can easily match a 1'' tw.
It's my opinion.
Wait... are you mixing surface area into the equation too? Cause I thought we only talked about the same size cone, but two different masses.
Applying more power won't fix the off-axis behavior of a larger cone... so there's a limit for sure 😉
It is not just harmonic distortion. In fact, HD is not even the most important parameter in the quest for detail and clarity.
As I said, the most important parameter is the frequency response and radiation pattern.
This first speaker will lack clarity and detail compared to the second speaker
This second speaker will be percieved as much smoother, with more clarity and detail, and less listener fatigue. Same drivers, same cabinet
Distortion and other secondary effects are meaningless until the the first speaker is transformed into the second speaker by a well designed crossover. Only then does it make sense to focus on secondary effects.
No.
We are. I was referring to velocity, not Sd.
Correct.
My experience with my speakers tells that room and positioning gives most of difference, when speakers are eq'd reasonably flat.
When going to gray area I prefer 4-5" mid with LR2 xo around 300 and 3-4kHz. Naturally this makes driver choice difficult. The mid should have minimal cone breakup issues.
Difference of MT xo being LR2 or LR4 is most likely specific to driver choice. It affects distortion and directivity (mostly vertical), when response anomalies are in control.
About intermodulation distortion, I don't think it is a problem in hifi speakers with reasonable quality. Even making it show in measurements is very difficult, because of room noise, mic S/N and driver's HD are masking it. 32-tone signal is often used, but the fuzz that measurements are showing is not necessarily IMD, just regular HD. German magazines have been measuring multitone for ages with Klippel system, and they call the result "gesamtverzerrung" (common distortion)
https://www.soundandrecording.de/eq...-5-preiswerter-2-wege-nahfeldmonitor-im-test/
When going to gray area I prefer 4-5" mid with LR2 xo around 300 and 3-4kHz. Naturally this makes driver choice difficult. The mid should have minimal cone breakup issues.
Difference of MT xo being LR2 or LR4 is most likely specific to driver choice. It affects distortion and directivity (mostly vertical), when response anomalies are in control.
About intermodulation distortion, I don't think it is a problem in hifi speakers with reasonable quality. Even making it show in measurements is very difficult, because of room noise, mic S/N and driver's HD are masking it. 32-tone signal is often used, but the fuzz that measurements are showing is not necessarily IMD, just regular HD. German magazines have been measuring multitone for ages with Klippel system, and they call the result "gesamtverzerrung" (common distortion)
https://www.soundandrecording.de/eq...-5-preiswerter-2-wege-nahfeldmonitor-im-test/
Last edited:
Hi, You mean narrower baffle & (also) drivers ? equal to beamier mid upper dispersion (= better match with a tweeter) ?
Maybe it's just mostly what works best in my room. But having minimum baffle seem to make less close-by surfaces to reflect from, which can then help to reduce unwanted very early reflections (distortion) from the speaker.
Often when I hear horns to close by or wider speakers - it is like I can more easily hear the sound from the speaker. Whereas, with my now narrow speaker design - the illusion of the sound coming from the space between the speakers, seem much improved.
It could be that it's mostly related to the use of a waveguide - and not specifically a narrower baffle in general. Even though the just 5" WG for the SB26ADc that I use now - really helped to reduce a few reflections in my room, that was quite annoying before.
I would argue that my speaker could "disappear" in the soundscape - both with a 4, 5 or 6" midrange. The important part is a very smooth and even response throughout the crossover region between midrange and tweeter - on all axis'. And then maybe that the baffle is no wider than needed, for the drivers to fit. And.... rather large round-overs (smooth baffle-edge transitions) might help as well. Seems to work for Revel, D&D C8, Kii 3 among others.
Actually narrow baffle has wide radiation, so do small diameter drivers. But a waveguide makes tweeter's dispersion narrower.
The key is to make dispersion/radiation pattern smooth and not too narrow. A waveguide must match the midrange.
But, speakers positioned longitudinally in a narrow room is always problematic. Narrow dispersion speaker will sound better.
Speakers along the long wall, way out of corners will give possibility to good prescise imaging, and with wide dipersion also good 3D soundstage.
IMO "we" have the basic diversion between people who want to imagine sitting on the best seat of a concert hall, or at the mixer's desk.
Psychoacoustics guru http://www.davidgriesinger.com/
The key is to make dispersion/radiation pattern smooth and not too narrow. A waveguide must match the midrange.
But, speakers positioned longitudinally in a narrow room is always problematic. Narrow dispersion speaker will sound better.
Speakers along the long wall, way out of corners will give possibility to good prescise imaging, and with wide dipersion also good 3D soundstage.
IMO "we" have the basic diversion between people who want to imagine sitting on the best seat of a concert hall, or at the mixer's desk.
Psychoacoustics guru http://www.davidgriesinger.com/
I'll take sit at the mixer's desk for 1295, Alex! It's amazing just how different every room is when it comes to reproduction of sound, and the speaker type. Glenn.
just to back in topic 😉 .
i'd follow more test with vituixcad:
1. "remove" IEC baffle refraction response from manufacture frd response.. (as you know, SbAcoustic have measurement on IEC at 30cm)
2. Cleaned response and used on simulate baffle dimension as project draw
To export single driver measurements i'd used mic at 3Meter in tool simulator, about on listening position.. 1Meter over floor , center on horizontal plane.
3. Loaded exported directivity measure for each driver and create xover.
i share with you response , minimal phase still about in enough level fron 300 to 8000.
Next step is to merge woofer response with simulated box response.. to be and to have more details data.
If someone know how to do that.. very thanks
i'd follow more test with vituixcad:
1. "remove" IEC baffle refraction response from manufacture frd response.. (as you know, SbAcoustic have measurement on IEC at 30cm)
2. Cleaned response and used on simulate baffle dimension as project draw
To export single driver measurements i'd used mic at 3Meter in tool simulator, about on listening position.. 1Meter over floor , center on horizontal plane.
3. Loaded exported directivity measure for each driver and create xover.
i share with you response , minimal phase still about in enough level fron 300 to 8000.
Next step is to merge woofer response with simulated box response.. to be and to have more details data.
If someone know how to do that.. very thanks
