I've been a live sound engineer since I was 18 and strongly disagree with this. ALL popular modern PA designs are CD on at least 1 axis (sorry Dr geddes, I know that's not true CD) ask yourself how can you get audio to sound similar in every listener position in the room without the speaker being CD? every other solution is obviously worse.
if anything, the PA world embraced CD long before studio or hifi, I imagine probably inspired by good cinema designs (which have similar goals)
Lower off axis is used to give similar loudness for audience closer to the speakers whereas higher on axis levels provides loudness for people further in the back of the room.
If they use CD in such application, they are messing things up, and I have to admit that most places do.
If they use CD in such application, they are messing things up, and I have to admit that most places do.
Last edited:
I don't like any of them 
The only one without a severe mouth reflection is N°10 but that's so poorly terminated that it's unusable.
Interestingly enough, JMLC horns (and similar) are prone to mouth reflections, as anyone can verify with Ath - it's simply not a good idea to make such long and narrow horns. It's not difficult to make it without the reflection but then it would no longer be JMLC horn and all the magic would be gone...
The only one without a severe mouth reflection is N°10 but that's so poorly terminated that it's unusable.
Interestingly enough, JMLC horns (and similar) are prone to mouth reflections, as anyone can verify with Ath - it's simply not a good idea to make such long and narrow horns. It's not difficult to make it without the reflection but then it would no longer be JMLC horn and all the magic would be gone...
Last edited:
Sure, and not only than mine. It has been designed with high loading as one of the design decisions, but completely ignoring directivity (notice there's not even a mention of this aspect in the report). Achieve both, high loading and controlled directivity, is the Holy Grail of waveguide design. No one yet has really succeded (i.e. without serious drawbacks), as far as I know. I'll take the better directivity anytime.
Last edited:
The JLMC would be much better with a wider mouth, true. But then the throat wave front becomes critical. In my previous experiments. I did not spend time to perfect that, and also the material size I was using did not allow proper termination. But a trend among the various configuration can be seen. At that time, one configuration was an OS throat with a JLMC type expansion.
See the OS-SE formula. Make the mouth smoothly rolled-back (the actual shape is not really critical, the smoothness is) and you have it.
Depending on the size of the "club" HF throw can suffer.
I did a club where some people were almost directly below the speakers (hanged 5m up) and furthest were 25m away. The system was not constant directivity type but subjectically the sound didn´t change with distance (measures +/- 2,5dB on the "floor").
It is not a heavy metal club but SPL reaches healthy figures anyway. Mains were built around BMS 15CN682´s and one sub hanged is PD 21" neo.
I did a club where some people were almost directly below the speakers (hanged 5m up) and furthest were 25m away. The system was not constant directivity type but subjectically the sound didn´t change with distance (measures +/- 2,5dB on the "floor").
It is not a heavy metal club but SPL reaches healthy figures anyway. Mains were built around BMS 15CN682´s and one sub hanged is PD 21" neo.
Just for listening, a friend and I tried a pair of speakers not too far from constant directivity with +/-30 degrees in an anechoic chamber. In an equilateral position between the speaker pointing and the listening position. The head had to be in the center for the most realistic representation, but the sweat spot was not at the expected location! it was actually close to having the speakers 120 degrees apart. Depending on the method of the twin mic recording, the sweat spot was slightly different. Chesky lab binaural recording was so realistic, yes, the person actually whispers into your ear and you can almost feel the breath!
Currently I do this, in a normal listening environment, having 0 phase and flat SPL creates a deeper soundstage and the original recorded room acoustics is more noticeable. In an anechoic chamber, minimum phase and slight deviation from a flat SPL response is slightly harsh sounding, and loses some soundstage size. The trick is that FIR filter should be based on measurement as close to the voice coil as possible. But also bear in mind that the EQ should vary based on listening loudness.
Generally, CSD will also be much cleaner with a fast decay.
Last edited: