Member
Joined 2009
I highly doubt such systems can sound clear in a typical living room. I'm not saying that it can't be done but it will take serious commitment and knowledge. The best route to get there will be large, highly directive horns so the small/large midrange debate doesn't quite apply. I don't think anybody really advocates small midranges for clear output at 120dB.
If you want clean dynamics then you just have to run the numbers.
What is typical room anyways? My HT room is 15x25 and my family room is 30x50.
We definitely are not talking about dome designs when we talk about the best HT speaker choices. Dynamics and Directivity is the top priority when it comes to having the best "WOW" experience. Dynamics exist all the way up past 300hz.
Im also not really concerned with family rooms either, this is simply an custom HT application requirement. Too many compromises exist in a family room to start with. They can put BOSE in those rooms and people still like them
What is typical room anyways? My HT room is 15x25 and my family room is 30x50.
We definitely are not talking about dome designs when we talk about the best HT speaker choices. Dynamics and Directivity is the top priority when it comes to having the best "WOW" experience. Dynamics exist all the way up past 300hz.
Im also not really concerned with family rooms either, this is simply an custom HT application requirement. Too many compromises exist in a family room to start with. They can put BOSE in those rooms and people still like them
Member
Joined 2009
2-3dB is certainly acceptable with me. In a typical room most of the sidewall reflections will be produced by the speaker's 30-60 degrees off-axis region. If you can show fairly smooth +-3db off-axis plots up to 60 degrees I'd say you probably have an excellent speaker regardless of what drivers were used.
That's a good point! I must admit it haven't occurred to me that the audibility of crossovers may be caused by an abrupt change in the diffraction signature.
I'm less interested in discussing SPL capabilities of drivers in this thread because the tools to simulate the required parameters for a given application are readily available. I don't want to argue that most 4" drivers are designed to reach 120dB. I'm interested in close center-to-center spacing and even polar response and I'm willing to sacrifice the maximum loudness to reach that goal.
How important is the polar response? You expressed the opinion that the directivity pattern might not be as important as the baffle diffraction signature. I'd love to explore that possibility because to be honest I don't know what the best approach is. There's already an enormous "What is the ideal directivity pattern" thread because those are the most exciting new topics in audio. The ultimate goal is to map out which disturbances at what frequencies give speakers their unique signature.
I'm only interested in shaping the dispersion pattern of a speaker to make it sound more like the original source. If somebody can show that the directivity pattern does not affect perception and that there is another reason why our brains can distinguish reproduced sound from the live event I will be happy to try a different approach.
I didn't start this thread in order to push my convictions to other people because I don't know what the best way is. This discussion has solidified some of my understanding but is also raising some interesting questions that I would like to explore.
I find a smooth dispersion pattern to be VERY important.
If the tweeter has a wider dispersion than the driver it takes over from (mid or mid-woofer in a 2way) it does change the perceived timbre of the instruments playing in that region because of early reflections and this bothers me a great deal more than inaccuracies in the baffle step compensation or even a complete lack thereof.
It gets massively worse if one is moving around in the listening room.
For this reason I think waveguides are a god send as they allow for a good dispersion match.
If the tweeter has a wider dispersion than the driver it takes over from (mid or mid-woofer in a 2way) it does change the perceived timbre of the instruments playing in that region because of early reflections and this bothers me a great deal more than inaccuracies in the baffle step compensation or even a complete lack thereof.
It gets massively worse if one is moving around in the listening room.
For this reason I think waveguides are a god send as they allow for a good dispersion match.
From my Unibox simulation, the maximum linear SPL of a 4" Scanspeak driver (12MU4731T00) in a 1 litre box is 106dB at 200Hz, 114dB at 300Hz, 118dB at 400Hz (no highpass filter).
Given a system sensitivity of say 88dB, a 100W amp will only be able to reach peaks of 108dB and it will take a 1000W amp to hit the 118dB linear limit at 400Hz.
I see no issue in using a 4" midrange down to 300Hz.
Given a system sensitivity of say 88dB, a 100W amp will only be able to reach peaks of 108dB and it will take a 1000W amp to hit the 118dB linear limit at 400Hz.
I see no issue in using a 4" midrange down to 300Hz.
In a properly controlled listening test we can hear distortion/compression during the dynamic peaks more then you can hear any CTC issue but everyone has their priorities. The most important part of any design choice is simply knowing our compromises. Atleast we are lucky to understand most of the compromises and then choose which ones to minimize and which ones to accept
I deem it extremely important for Movie content and not so important for music. I actually like nearfield music listening so the idea of polars and power response isn't going to have the highest priority.
Of course in general Polar response = higher accuracy. If you want the original source then you should have a speaker that emits sound like the original source. Does anyone have the polar response of an acoustical guitar, a piano, a horn?? I can not see them emitting soundwaves like a direct radiating source (without control) or any dome tweeter at all.
I think its important to realize that different applications have different design requirements. I have read many times online that HT and Music can be covered by one design. I 100% disagree with that idea.
Last edited:
Disagree with what??
1. Anytime you have a drop off axis or your power response fails, you are not recreating the original content. Remember they create the content nearfield, they assume perfection on and off axis!!
2. Anytime you introduce driver distortion/compression and or amp clipping, you are not recreating the original content.
3. The idea that most people remotely even think they know the intent of the original content is as real as purple dragons (hehe, sorry that is my 2 year daughter's thoughts last nigh)
Our only goal really should be is to ensure clean playback and have control over our response curves and in room decay.
FWIW, I could give a crap about the original artist's intent (some are dumb about what they are trying to do). If I want to change my curve to have more bass/less bass or more/less HF, I can do that in a heart beat. Engineers are far from perfect, we need the ability to fix screw ups. I simply care about enjoying my experience. No music and movies is not the same as ART work either so I do not buy the "Would you paint/change a rembrandt" POV.
1. Anytime you have a drop off axis or your power response fails, you are not recreating the original content. Remember they create the content nearfield, they assume perfection on and off axis!!
2. Anytime you introduce driver distortion/compression and or amp clipping, you are not recreating the original content.
3. The idea that most people remotely even think they know the intent of the original content is as real as purple dragons (hehe, sorry that is my 2 year daughter's thoughts last nigh)
Our only goal really should be is to ensure clean playback and have control over our response curves and in room decay.
FWIW, I could give a crap about the original artist's intent (some are dumb about what they are trying to do). If I want to change my curve to have more bass/less bass or more/less HF, I can do that in a heart beat. Engineers are far from perfect, we need the ability to fix screw ups. I simply care about enjoying my experience. No music and movies is not the same as ART work either so I do not buy the "Would you paint/change a rembrandt" POV.
Last edited:
I wish it was a "why not". I love my ribbon designs for music. There is something just too "edgy" about high sensitivity designs when it comes to low level/near field music enjoyment but on the other end movies have never been better. I moved my ribbons into my bedroom and office.
Im still working on finding a common solution. I have TAD CDs now and I need to build a small waveguide design later.
Im still working on finding a common solution. I have TAD CDs now and I need to build a small waveguide design later.
A bit OT:
The way I see it the spatial information (ie room, radiation pattern of instruments etc) is already captured on the recording as heard from where the microphone was positioned.
At the very best we are able to recreate this with a fair degree of accuracy.
For this we need to control/minimize the effect the listening room has and omni-directional speakers do not help here.
We will never be able to recreate the impression of being anywhere other than where the mic was placed with any accuracy at all. That information is simply not available to us as it has never been recorded.
The way I see it the spatial information (ie room, radiation pattern of instruments etc) is already captured on the recording as heard from where the microphone was positioned.
At the very best we are able to recreate this with a fair degree of accuracy.
For this we need to control/minimize the effect the listening room has and omni-directional speakers do not help here.
We will never be able to recreate the impression of being anywhere other than where the mic was placed with any accuracy at all. That information is simply not available to us as it has never been recorded.
Elias,
I'm not going to engage you any further on this point in this thread both because it's almost entirely off topic to the discussion of midrange driver size, and because you are being rather disingenous with your claims about the perception of frequency response, which are not in agreement with research, and you have not provided any references of your own to back up your claims.
For anyone interested in following up the perception of direct field vs the room power response at different frequency ranges, I suggest they read the following posts by Speaker Dave, and then follow up by reading the research from Kates and Salmi that he references:
http://www.diyaudio.com/forums/mult...y-pattern-stereo-speakers-25.html#post2720861
http://www.diyaudio.com/forums/mult...y-pattern-stereo-speakers-23.html#post2716944
http://www.diyaudio.com/forums/mult...ty-pattern-stereo-speakers-6.html#post2695877
The audibility of a sudden change in baffle diffraction signature at the crossover frequency is not something that occurred to me until recently either.
Large mid / midbass with a typical wide dispersion dome tweeter is not something that works very well, I've tried the same 8" driver with a dome crossed at 3Khz before I had the ribbon tweeters, and it was ok at best, but you could definitely notice something odd at the crossover.
The large mid would be a true point source at the crossover, (if 2Khz or above) while the dome would not - it would be a point source together with an out of phase, large rectangular radiating perimeter around the edge of the baffle.
Offsetting the tweeter on the baffle might reduce the on-axis peaks and dips but does nothing to shrink the physical size of the spurious radiating perimeter nor its overall amplitude. Rounding the edges only does so much as well.
Using a wave-guide loaded tweeter is often suggested with a large mid because it more closely matches the directivity, but I wonder whether the dramatic reduction in baffle diffraction isn't just as important or even more important in aiding a seamless blend on and near on axis.
I still think the overall directivity is quite important - I generally favour fairly directional speakers, however I'm not sure that accurate matching of driver directivity at the crossover is essential, provided that both have enough directivity to minimize diffraction.
The power response only starts to dip significantly when the pattern gets quite narrow - for example there's not a lot of difference in power response when the pattern narrows from 180 degrees to 120 degrees, (or 120 degrees to 90 degrees) because only a very small part of the rooms power response comes from those very wide angles.
Diffraction in the tweeters range is certainly important. Even though my wave-guide ribbons have very little natural diffraction (to the point where even using them sitting naked on the top of the cabinet with no baffle only causes a broad 1dB diffraction dip on-axis at 6Khz) I've been playing around with various ways to absorb what diffraction there is at the edge of the baffle, and the difference is still quite noticeable.
I think a combination of some tweeter directivity to minimize the illumination of the baffle coupled with some absorption is a better approach than just trying to round the baffle edges and/or offset the tweeter.
Aren't you saying it's ok for you to spread wrong information even OT, but correcting it is not ok because it's OT
You can read Zwicker's book and learn about integration times etc. Recommended for everyone, btw. Good read.
- Elias
Member
Joined 2009
As long as the 4" midrange has high sensitivity to match the rest of the system then its fine. We can use it in a "Unity Horn"
I think, in midrange (also in bass?) that what matters is cone movement? it makes distortion. Distortion depends on driver quality of course. Last weekend I heard large 3,5-way passive speaker where lowest cross over frequency was at 80 Hz. There was mid and upper bass in 2,5-way + 2x12" crossed at 80 Hz. Mids were 4,5". Cone movement was silly large for large 3.5-way speaker though designer said that mid drivers are that good that they don't go badly to distortion. I think the most important thing in 3-way is to cut cone movement away for driver which produced sound up to 2000-3000 Hz! (and of course to achieve more balanced dispersion)
So would it be nice to play mids with 3-4" xo @ 300 Hz even if it had good sensitivity? At around 100 dB peaks cone movement is about 2 mm (@ 250-300 Hz)which is quite a lot for small true midrange. (though I used 1st order slopes in sims)
so so... 6,5" makes the noise with much less xmax
(thing that everyone knows, just wondering how important matter is to get almost all "movement" away from mid driver?)- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.