The main difference between passive direct boxes is the quality of the transformer. That shows up as less distortion and more LF energy. For most computer users even the cheapest seem to be fine. (Of course my idea of cheap is still a name brand.)
According to my experience, it is better to have set of line arrays, if you have microphones near different walls. Each pair of line arrays must be close to corresponding microphone. And of course, they must be thoroughly frequency equalized by feedback, i.e. for the complete acoustic-electric tract!
An year ago I helped on one scientific conference in the hall built for symphonic orchestra and unamplified voices in 1930'th. It was impossible to hear anything even using lapel microphones. After I brought my pair of foldable line arrays everything worked fine and nice, even with a pair of large diaphragm condenser microphones that were standing on 3 feet distance each from the speaking person.
Distributed systems may work well in long narrow theatres with walls and seats covered by sound absorbing materials. They need to contain time delay devices that carefully follow sound propagation from the stage.
An year ago I helped on one scientific conference in the hall built for symphonic orchestra and unamplified voices in 1930'th. It was impossible to hear anything even using lapel microphones. After I brought my pair of foldable line arrays everything worked fine and nice, even with a pair of large diaphragm condenser microphones that were standing on 3 feet distance each from the speaking person.
Distributed systems may work well in long narrow theatres with walls and seats covered by sound absorbing materials. They need to contain time delay devices that carefully follow sound propagation from the stage.
Wavebourn,
When you get the reverb time below 1 second just about anything works. The the issues become naturalness.
Most folks today talk at about 5 syllables per second. In the thirties public speakers spoke around 3! At 0 db S/N speech becomes adequate. At 15 there really is almost no room for improvement. So 5 syllables per second with a reverb of one second give you 12 db S/N, really quite good.
A thirties concert hall would have around 3 seconds so at the same 5 syllables/S you would have 4 db S/N in the near field.
The next issue is critical distance. With a reverb time under a second that is not an issue. With 3 seconds it probably is around 25% or less of the hall length (Reverb being an important part of the hall's design.) So you can use inverse square law to tell you when 4 db S/N drops below the reverberant field. My guess based on typical era concert halls is about 8 feet.
A line array not only gives you 3 db better feedback margins it also does not follow inverse square law.
That is why line arrays are so popular they "Fix" borderline rooms. If the reverb time goes to something silly like 9 seconds, the best fix is demolition.
When you get the reverb time below 1 second just about anything works. The the issues become naturalness.
Most folks today talk at about 5 syllables per second. In the thirties public speakers spoke around 3! At 0 db S/N speech becomes adequate. At 15 there really is almost no room for improvement. So 5 syllables per second with a reverb of one second give you 12 db S/N, really quite good.
A thirties concert hall would have around 3 seconds so at the same 5 syllables/S you would have 4 db S/N in the near field.
The next issue is critical distance. With a reverb time under a second that is not an issue. With 3 seconds it probably is around 25% or less of the hall length (Reverb being an important part of the hall's design.) So you can use inverse square law to tell you when 4 db S/N drops below the reverberant field. My guess based on typical era concert halls is about 8 feet.
A line array not only gives you 3 db better feedback margins it also does not follow inverse square law.
That is why line arrays are so popular they "Fix" borderline rooms. If the reverb time goes to something silly like 9 seconds, the best fix is demolition.
Simon;
your calculations don't include binaural selectivity effect. Actually, if direct sound coming from the speaking person is below noise floor we hear the person well. But in case of multiple point source speakers (i.e. radiating all around including floor and ceiling) and reverberation this effect does not work.
That's why line arrays, and that's why close to the sound source.
your calculations don't include binaural selectivity effect. Actually, if direct sound coming from the speaking person is below noise floor we hear the person well. But in case of multiple point source speakers (i.e. radiating all around including floor and ceiling) and reverberation this effect does not work.
That's why line arrays, and that's why close to the sound source.
Terminology difference. Reverb is part of the noise floor. Noise equal per critical band to the speech is acceptable intelligibility. A few db above the speech is where it begins to fail. Binaural and full range exclude older folks, who often are the target audience.
Simon, binaural effect means possibility to locate and filter sounds that come from certain direction by hearing apparatus. If you use single microphone, but many speakers, you screw down binaural location and filtering. Speech recognition also depends on the nature of the noise: is it incoherent noise, or reverberation.
My point was, to use one line array per microphone is the best method of speech amplification in reverberating environments. If the same sound comes from different directions it is less possible to select one of them filtering out the rest, than if direct and reflected sounds differ in timbre.
My point was, to use one line array per microphone is the best method of speech amplification in reverberating environments. If the same sound comes from different directions it is less possible to select one of them filtering out the rest, than if direct and reflected sounds differ in timbre.
Last edited:
It is not uncommon for folks to loose hearing in one ear, so when you design for the masses you can't count of binaural hearing! Same with HF response!
Line arrays are responsible for more than doubling the sales volume of pro sound! There is a reason. 🙂
Funny about the rate of speech. Gig tonight in a college gym, it's pretty bad. But we had it dialed in and those speaking didn't rush. It was very clear all the way at the back.
Tomorrow may be another story. At rehearsals they all spoke too fast for the venue. Will be interesting to hear.
Tomorrow may be another story. At rehearsals they all spoke too fast for the venue. Will be interesting to hear.
If the system is mainly speech only ( and some background music), I'd double the number of speakers. 4-5m between each is somewhat long IMO. The sepakers are probably not the cost driving factor anyway.
You'll also have to look into some kind of zoning switching to cope with the different scenarios, to avoid feedback, or to increase the feedback margin.
You'll also have to look into some kind of zoning switching to cope with the different scenarios, to avoid feedback, or to increase the feedback margin.
- Status
- Not open for further replies.