I think you will find that to come at all close to the experience of hearing a tenor or soprano at close range you will need vastly more acoustic output than the Orion (or any other speaker with typical sensitivity) can deliver, even with the most powerful amplifier it can tolerate.
You will need to add 10 dB of sensitivity and/or move the decimal one to the right in terms of watts of drive available, to come close.
You will need to add 10 dB of sensitivity and/or move the decimal one to the right in terms of watts of drive available, to come close.
That's disappointing. Now I wish I had had a decibel meter with me, I would love to know what the sound volumes were.
I did not experience the slightest discomfort or hearing pain; far from it. I certainly did not experience any temporary deafness afterward; if anything, there was a heightened acuity of hearing...but that was, I think, because I had become sensitized to the approximately 2800 Hz "ring" in a well-trained, talented voice.
I do find headphones irritating, even supposedly very good ones with a perfectly decent amp; I wish I knew why or how to fix it.
But what you've said makes me think even more strongly about the 1,000 watt amp I've been thinking of making. Not to play at 1,000 watts, of course, but to have the massive amounts of headroom.
Then, of course, the speakers would have to be able to handle the transients without breaking up or literally catching fire. (I did "smoke" an array of speakers once, driving them from a 900 volt mu-stage, pentode over triode, having goofed and forgotten to put two 500 volt electrolytics in series between the output and the speakers....the single 500 volt 'lytic promptly blew up, sending a jet of steam to the ceiling, and shorted, so the close to half an amp at eight hundred or so volts made all the voice coils simultaneously so red hot the paper cones started smoking!.....the array was about 150 identical small speakers all in series, so the output stage saw an impedance of about 1.2 Kilohms. It had possibilities, that, but the speakers were cheap and nowhere near full-rangers...the same idea with 150 really good full-rangers all in series, so there's no need for an OPT, would probably work fabulously, if one could afford it.)
I was planning on using a woofer in the dipole capable of a 1,000 watts, with 100 dB sensitivity (a 21" monster), but the problem is the full-range driver. I could make a vertical line array with four 6.5" FR's to at least get to several hundred watts of capability, but there goes the point source concept (although the Beethoven certainly is not pretending to be a "point-source" speaker...but perhaps with OB design it's not so important).
In the designs at Common Sense Audio with FR's it's all about point-source, but, as you say, I cannot get the "full immersion" effect I want, probably, from a single FR per channel. My budget limits me to things like Audio Nirvana's, so I have to figure out a compromise that will work.
Along the lines of hedonism rather than accuracy, I wonder what would happen if I made a graphic equalizer, just for the frequencies from 2700 to 3000 Hz, and tuned the system to emphasize the ring of the singer's voice? I am sure what I experienced in that room was total immersion in the approx. 2800 to 3000 Hz overtone in the operatic voice.
As Duke Ellington said (this will be like waving a red flag in front of some members): "If it sounds good, it is good."
Best, Charlie
I did not experience the slightest discomfort or hearing pain; far from it. I certainly did not experience any temporary deafness afterward; if anything, there was a heightened acuity of hearing...but that was, I think, because I had become sensitized to the approximately 2800 Hz "ring" in a well-trained, talented voice.
I do find headphones irritating, even supposedly very good ones with a perfectly decent amp; I wish I knew why or how to fix it.
But what you've said makes me think even more strongly about the 1,000 watt amp I've been thinking of making. Not to play at 1,000 watts, of course, but to have the massive amounts of headroom.
Then, of course, the speakers would have to be able to handle the transients without breaking up or literally catching fire. (I did "smoke" an array of speakers once, driving them from a 900 volt mu-stage, pentode over triode, having goofed and forgotten to put two 500 volt electrolytics in series between the output and the speakers....the single 500 volt 'lytic promptly blew up, sending a jet of steam to the ceiling, and shorted, so the close to half an amp at eight hundred or so volts made all the voice coils simultaneously so red hot the paper cones started smoking!.....the array was about 150 identical small speakers all in series, so the output stage saw an impedance of about 1.2 Kilohms. It had possibilities, that, but the speakers were cheap and nowhere near full-rangers...the same idea with 150 really good full-rangers all in series, so there's no need for an OPT, would probably work fabulously, if one could afford it.)
I was planning on using a woofer in the dipole capable of a 1,000 watts, with 100 dB sensitivity (a 21" monster), but the problem is the full-range driver. I could make a vertical line array with four 6.5" FR's to at least get to several hundred watts of capability, but there goes the point source concept (although the Beethoven certainly is not pretending to be a "point-source" speaker...but perhaps with OB design it's not so important).
In the designs at Common Sense Audio with FR's it's all about point-source, but, as you say, I cannot get the "full immersion" effect I want, probably, from a single FR per channel. My budget limits me to things like Audio Nirvana's, so I have to figure out a compromise that will work.
Along the lines of hedonism rather than accuracy, I wonder what would happen if I made a graphic equalizer, just for the frequencies from 2700 to 3000 Hz, and tuned the system to emphasize the ring of the singer's voice? I am sure what I experienced in that room was total immersion in the approx. 2800 to 3000 Hz overtone in the operatic voice.
As Duke Ellington said (this will be like waving a red flag in front of some members): "If it sounds good, it is good."
Best, Charlie
I was once in a very live room, a classroom of mine. The walls were cinderblock and there was very little damping - hard desks and chalkboards... but the ceiling may have been acoustic tile; I'm not sure. Anyway, we had a soprano perform, and even near the back of the room (approximately 6-8 meters away from her) the SPL was terribly, uncomfortably high.
EDIT: It hurt my ears.
EDIT2: If a system is nominally 90dB/1W@1m...
at 110dB at 1 m, you need 100W
at 120dB at 1 m, you need 1000W (neglecting power compression effects)
at 120dB at 2 m, you need 4000W
at 120dB at 3 m, you need ~10000W
Putting the system in a room will decrease the quantities somewhat, but the math is pretty clear. But, if we start with a system that's 100dB/1W@1m...
at 110dB at 1 m, you need 10W
at 120dB at 1 m, you need 100W
at 120dB at 2 m, you need 400W
EDIT3: The drivers need to have sufficient excursion within their respective passbands to make this power-limited SPL possible
EDIT: It hurt my ears.
EDIT2: If a system is nominally 90dB/1W@1m...
at 110dB at 1 m, you need 100W
at 120dB at 1 m, you need 1000W (neglecting power compression effects)
at 120dB at 2 m, you need 4000W
at 120dB at 3 m, you need ~10000W
Putting the system in a room will decrease the quantities somewhat, but the math is pretty clear. But, if we start with a system that's 100dB/1W@1m...
at 110dB at 1 m, you need 10W
at 120dB at 1 m, you need 100W
at 120dB at 2 m, you need 400W
EDIT3: The drivers need to have sufficient excursion within their respective passbands to make this power-limited SPL possible
Perhaps I was in a lucky spot. I was off to the side, the singer was singing toward the side wall, not right at me. And the room was full of rug, paintings, a piano, furniture...undoubtedly not as bright.
The soprano you heard might have been a bit on the shrill side, too. I have heard sopranos who were overly bright - or sang with a bit of a tense voice, and their tone did make me uncomfortable. But none of the ones I heard in McClosky's studio. Not even the Wagnerian soprano who literally made the windows shake.
Anybody who has been to a Messiah has, I'm sure, noticed the "light-house" effect of a soloist who sweeps the room while singing; as the focus of their voice sweeps over you and then passes, it's rather like a spot-light rotating, or, more aptly, a train going by.
454Casull, if four 6.5" FR's were driven simultaneously, in phase of course, in a vertical line array, with 96 dB sensitivity (1W/1M), at 2 or 3 meters out and the array curved slightly so they more or less merge into one source, how would one calculate what the equivalent sensitivity would be?
Best, Charlie
The soprano you heard might have been a bit on the shrill side, too. I have heard sopranos who were overly bright - or sang with a bit of a tense voice, and their tone did make me uncomfortable. But none of the ones I heard in McClosky's studio. Not even the Wagnerian soprano who literally made the windows shake.
Anybody who has been to a Messiah has, I'm sure, noticed the "light-house" effect of a soloist who sweeps the room while singing; as the focus of their voice sweeps over you and then passes, it's rather like a spot-light rotating, or, more aptly, a train going by.
454Casull, if four 6.5" FR's were driven simultaneously, in phase of course, in a vertical line array, with 96 dB sensitivity (1W/1M), at 2 or 3 meters out and the array curved slightly so they more or less merge into one source, how would one calculate what the equivalent sensitivity would be?
Best, Charlie
I heard a system just the other day which underlined for me the inadequacy of much I have listened to of late in terms of mid range dynamics capability.
It was able to reproduce the sound of a 12 voice choir I had recorded where the voices were about 6 feet from the single stereo microphone and the singing style was in the Bulgarian tradition - slightly harsh. When listening at a distance of 5 feet from the speakers (as I do when mixing) I was able - just - to reach approximately real levels without the obvious distortion my current monitor delivers when attempting the same.
The key design characteristic of the speaker was the mid range driver - that it had one, and that it was called upon to cover only a modest range - 650Hz to 3.3kHz. It was a soft dome, and the speaker was internally tri-amplified with a line level crossover (24 dB per octave slopes).
The speaker was the Klein + Hummel 0300 and it made me wonder about the divergence of home and "pro" speakers, in that I wondered how a system could be configured to match the performance I heard for a lower retail price.
I don't think it could be done with any components I am aware of, so I'm buying these, with the 0800 sub.
I imagine the Summas would be better still, but would be overkill for my needs and listening distance, plus the size of these and weight (30lb each, including amps) is a bonus as I can take them to certain remote recording sites.
It was able to reproduce the sound of a 12 voice choir I had recorded where the voices were about 6 feet from the single stereo microphone and the singing style was in the Bulgarian tradition - slightly harsh. When listening at a distance of 5 feet from the speakers (as I do when mixing) I was able - just - to reach approximately real levels without the obvious distortion my current monitor delivers when attempting the same.
The key design characteristic of the speaker was the mid range driver - that it had one, and that it was called upon to cover only a modest range - 650Hz to 3.3kHz. It was a soft dome, and the speaker was internally tri-amplified with a line level crossover (24 dB per octave slopes).
The speaker was the Klein + Hummel 0300 and it made me wonder about the divergence of home and "pro" speakers, in that I wondered how a system could be configured to match the performance I heard for a lower retail price.
I don't think it could be done with any components I am aware of, so I'm buying these, with the 0800 sub.
I imagine the Summas would be better still, but would be overkill for my needs and listening distance, plus the size of these and weight (30lb each, including amps) is a bonus as I can take them to certain remote recording sites.
To the people wanting to use this thread to talk about Orion cloning, I do apologize for the off topic stuff. The Summa talk that's bothering you lately was the Orion talk three years ago. I keep thinking everything that could possibly be said about Orions has been by now but I know it's not for me to decide. Many people were inspired by them and have built not quite clones, but similar speakers. Do search the archives and read Linkwitz and John K's sites cover to cover because I think you'll find them much more concentrated sources of the information you seek then yet another orion thread in the middle of all this noise.
radianceaudio said:
Yes the "box" dimensions are available, but not the curves. In regards to diffraction I'm not sure if you've explore this study?
http://www.linkwitzlab.com/diffraction.htm
Basically SL concluded that edge diffraction is overrated.
Russell Dawkins said:
Not at all. The most powerful singers put out 100-110dB at 1 meter which means about 90-100dB at 3 meters. I don't know how much gain the performance hall would give but say 10dB+ and we are back at 100-110dB.. and that is LOUD. I do not enjoy a soprano going all out at that distance, 5-6 meter minimum.
The Orion's midrange units are among the lowest distortion drivers on the market and used in pairs (which effectively is the case for a solo-singer placed in the middle of the soundstage) you get 6dB extra output and the listening room add some as well.
The 100-110dB output from a soprano is not in the bassrange (in the bass range s singer has much less output) and one W22 driver can put out 136dB at 400Hz in a small box/baffle. Reduce a couple of dB for the open baffle.
The sensitivity is 90dB/2.83V/1m so 100W-300W amp power would be enough.
If we are talking choir or a jazz drumset we may run into trouble though, but that depends on material and recording distance.
/Peter
poptart said:
I know I said I'll butt out, but I just saw this and wanted to reply. Fair point, and I was guaging the relevance of my post from the title and initiating post more than the recent discussion. Sorry if this has clouded matters and I will look elsewhere to put my updates/comment re: my little project.
Cheers,
Dan
I have read (again) the article on diffraction - by some considered very evil, SL says in the real world the effect is not that great. And I have also discovered another point of why he prefers dipoles - baffle step. Baffle step compensation adds bass to the off-axis sound and changes the power response balance, right? As far as I understand, the rear and front diffraction cancels out at certain frequencies.
Is baffle step compensation used in Summas or one compensates for this with recommended placement?
Is baffle step compensation used in Summas or one compensates for this with recommended placement?
I wonder what else is not relevant for good music. This guy seem to take it to the extreme, he reckons best way to design speakers is first to format c: 
http://lampizator.eu/NIRVANA/nirvana.html
http://lampizator.eu/NIRVANA/nirvana.html
pelanj said:
As with much of the discussion on SL's pages it is necessary to think about what is said and interpret it. What is stated about baffle step is true, but does it apply? If the speaker is in the free field, away from all boundaries, then it is correct. However, most speakers are placed in a room, with woofers close to the floor or ground plane. If we consider only the presence of a ground plane, and assume the GP is acoustically opaque, then there is a transition back to 2Pi radiation at low frequency and the radiated power for a conventional speaker and a dipole would look something like this:
An externally hosted image should be here but it was not working when we last tested it.
Of course, if the GP is not acoustically transparent then some amount of low frequency power will pass through the GP, not to mention the floor, etc. Thus it can be argued that a dipole midrange should actually be coupled to a monopole woofer if the woofer is close to the floor. Or that the dipole woofer axis should be aligned differently than the midrange dipole axis.
I've done it both ways. The NaO Mini combines a dipole mid with monopole woofer and the results are very good. The NaO II combines a dipole mid with a quasi cardioid woofer, again with very good results. (The dipole and cardioid have the same on axis directivity index. Both radiate 4.8dB less than a monopole in the free field.)
The biggest problem with baffle step and conventional speakers is that there will at least be a region in the lower midrange where there is 4Pi radiation and baffle step comp applied here will result in added power in this frequency band, as shown in the figure above.
The second thing is that I have read a lot about some of you wanting to build dipole but not at the cost of the Orion or NaO II. If you want to DIY but don’t have a good idea where to start you may want to consider my ABC Dipole design guide.
pelanj said:
Baffle Step is a theoretical concept - an approximation at best. When I do the design of the crossover I use a full set of polars for both drivers measured in a free field with the drivers installed in the actual enclosure to be used. Thus any "baffle step" is actually in the data and will automatically be "accounted for". This method is the most accurate since the baffle step is only an approximation of what will actually happen in a given enclosure. The actual measured effect is exact.
Another thing that may not be obvious is that the bump that can be seen off axis in the frontal hemipshere as a midrange driver handles over to a tweeter may not be all that problematic in a dipole.
Looking at the phenomena in isolation one may think that it is a problem that needs to be solved, however I'm not so sure that anything needs to be done at all. A waveguide in this situaution may casue worse over all response than leaving the tweeter as is on a more or less flat baffle.
Obviously I'm thinking of the added midrange contribution from the back of the dipole midrange and how that to some degree balances the power response. I haven't done any measurements on this but maybe some day.
Unless all of the "backwards" radiated is absorbed this energy will sum with the frontal radiating energy and mix in the room. What likely is important though is to make sure that the first reflection points of the side walls are absorbed, diffused or "deflected" in order to avoid strong early contribution of the "jump" from mid to tweeter.
I know SL's take on this is to add a 2:nd tweeter at the back of the dipole however i prefer to use some serious absorbtion on the wall behind the speakers to reduce excess room reverberation. Still there will be some energy left from the back of the dipole.
/Peter
Looking at the phenomena in isolation one may think that it is a problem that needs to be solved, however I'm not so sure that anything needs to be done at all. A waveguide in this situaution may casue worse over all response than leaving the tweeter as is on a more or less flat baffle.
Obviously I'm thinking of the added midrange contribution from the back of the dipole midrange and how that to some degree balances the power response. I haven't done any measurements on this but maybe some day.
Unless all of the "backwards" radiated is absorbed this energy will sum with the frontal radiating energy and mix in the room. What likely is important though is to make sure that the first reflection points of the side walls are absorbed, diffused or "deflected" in order to avoid strong early contribution of the "jump" from mid to tweeter.
I know SL's take on this is to add a 2:nd tweeter at the back of the dipole however i prefer to use some serious absorbtion on the wall behind the speakers to reduce excess room reverberation. Still there will be some energy left from the back of the dipole.
/Peter
Well yes, its a "real" phenomina, but the equations used to predict it or compensate for it are approximations. If you use real measured data in the crossover design then there is no need to "account" for it or "compensate" for it.
"I don't know a single serious speeker designer that does not measure their designs."
They just don't share that data I guess.
"I don't know a single serious speeker designer that does not measure their designs."
They just don't share that data I guess.
John K.,
thanks for clarification, now I have put all together and it makes sense for me. Just to be sure - how would the corresponding on axis FR response look like for the power response charts? I am afraid I cannot imagine it from the power response - if it makes sense (or is possible) to equalize for both power and on-axis response?
thanks for clarification, now I have put all together and it makes sense for me. Just to be sure - how would the corresponding on axis FR response look like for the power response charts? I am afraid I cannot imagine it from the power response - if it makes sense (or is possible) to equalize for both power and on-axis response?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.