|
Transcript
Pat Brown: Sure. Played guitar in high school - studied Electrical Engineering Technology at University of Louisville - ran a retail music store for nine years before going into sound. I started working with Syn-Aud-Con in the early 90s as Don Davis' assistant. Weogo Reed: On the live-audio board, there has been much discussion of optimum bass. Two questions: for smaller venues, are horn-loaded subs out, or are acoustic suspension (like Bag End) the best way to go? Second, what about having odd numbers of subs, like center, left and right? Pat: That's a good question, but one that we couldn't cover in a day. I tend to break it down into section. What is the LF array doing? What is the room doing? The latter is as important as the former. yam4000vca: As teaching is an art I feel, do you think that you can actually teach a person to mix music or do you have to be satisfied to teach how the gear works and how to get the best out of it? And hope for the best? Pat: You can teach someone what the knobs do, but you can't teach them how to listen. I make this clear in all of our seminars. Mixing is both art and science. The science can make one a better artist, but it is not a substitute for talent. Moderator: Pat, since you travel and teach internationally, do you find there are problems that exist everywhere? Pat: Definitely. The big ones are "acoustic gain" and "blowing loudspeakers”. Both of these lend themselves to scientific solutions. Harry: Could you expand on those problems a bit? And the solutions? Pat: I'm a real believer in working within the physical limitations of the system. "Acoustic gain" refers to gain-before-feedback. There is a potential gain that is established by component placement that cannot be exceeded. One needs to know how to manipulate that point. Most systems that are in feedback are trying to violate this limit. Most blown loudspeakers result from failure to consider the crest factor of the program material. That is what determines the actual power delivered to the loudspeaker and therefore what it must dissipate. We spend considerable time on both in our seminars. Moderator: For the acoustic gain problem, what can we do besides "keeping the mics behind the speakers"? Pat: First on my list is keeping the mic close to the source. As long as the source is louder at the mic that the loudspeaker, the system will be stable. C.K.: Pat, for those of us who don't know much about Syn-Aud-Con, please describe a typical attendee and session. Pat: We do a two-day seminar on System Optimization. Gain structure, grounding, equalization, alignment, etc. Next is a three-day seminar on system design. This one deals more with room acoustics and loudspeaker placement. Vic Brown: Can you give us a "typical" profile of seminar participants in the last several years? Pat: We get a lot of contractors, consultants, and these days a lot of church sound personnel. We also have a lot of military attendees. They have more systems out there than anyone. Many manufacturers send us their techs and engineers to get a feel for how people are using their gear. Moderator: Can you define "crest factor"? Pat: Crest factor can be thought of as the "area under the curve" of a waveform. It's the RMS voltage, as opposed to the peak. You can't let it get too high (excessive compression, clipping, etc.) Dave Howden: What do you consider the proper way to set compressors/limiters in a live sound reinforcement system for system protection? Pat: You have to be careful when you say "protection" and "compression." The loudspeaker prefers a high crest factor signal (less heat). Compressors and limiters reduce the crest factor. This makes it louder, but generates more heat. I prefer compression as an artistic tool, and limiting for managing amplifier headroom. Too much of either will get you into trouble. The best way to protect the loudspeaker is to monitor the Leq of the system and keep it under control. Tucci: Better differentiate the different “L” levels for us. Pat: Tools like TEF and CLIO allow continuous Leq measurements. Most shows escalate the level as the night goes on. It's a subtle increase but enough to burn things up by the end of the night. Level measurements will show this happening. Ken: Can you define Leq for us old slow folks? Pat: The Leq is the "equivalent level" that a complex signal will integrate to. It is a good measure of the actual energy being radiated by the loudspeakers. Mike S: Pertaining to acoustic measurement systems… I am fairly confident with SIM, SMAART and just beginning on the wonders of TEF. What do you anticipate as the next big step in measurement systems? Pat: I think the next big step is to get everyone using the currently available systems. They are amazing at their current state of development. I'm not sure we need a lot more information. We need to learn how to interpret what we can currently measure. Phase is a good example. Easy to measure. Hard to interpret. Ken: What about assistants to help interpret? Pat: That scares me a bit. I wouldn't want a computer interpreting my lab results. I'd prefer a gray-haired doctor :>) Most of the measurements are not too difficult to interpret given the proper background. Moderator: Do you have a preferred method of tuning a system? Pat: I use several. Number one is making sure that loudspeaker placement is optimal. You can't fix a bad placement electronically. I prefer to look initially at the direct field of the loudspeaker. That's what carries the most information. I also like to consider independently of the room for mid and high frequencies. Trooper Hales: That’s right, one man’s reverberation is another man’s reflections. Pat Brown: This is where TEF, SMAART, etc. comes in. The time window is a wonderful tool. Roger Wade: For someone needing to get started with SMAART, would you suggest one or both of the courses at NSCA? Pat: I'm not sure what they are offering. I have taught them in the past and was disappointed in the amount of time to cover such a large topic. My best advice is to get a copy of SMAART and measure everything that you can get your hands on. Ken: True Story - many more can measure than can understand what they are seeing. Pat: The biggest benefit of any measurement tool is what you can learn from it. Whatever system you get, work through the tutorials step by step. Do it every day. Also, a lot of study in basic acoustics. Tucci: Time window choice will dictate the frequency resolution. That’s physics I suppose. How do the differing systems differ as far as S/N ratio of the measurements go? Pat: TEF comes out on top for S/N. Specifically TDS (time-delay spectrometry). You can measure almost under any conditions with TDS. SMAART yields good S/N - probably good enough for most applications. But when they want to drive trucks through the room TDS or MLS have some benefits. Dave Howden: Many younger engineers seem to be slave to the software and PC's. The ears have to be trained before the eyes are IMO. Pat: Agree with Dave. The PCs are great, but the best analyzer you'll ever own is the one mounted to your shoulders. Ken Regular: How do you teach people to correlate what they measure to what they hear? Pat: You just have to calibrate it. That's where the analyzers come in. Tim Mc: It seems to me that the hardest part of learning SMAART or any other analysis tool is correlating what you see with what you hear. Pat: I think that you can teach yourself that. This is one area where SMAART is great. The real-time transfer function with music as a program source is a great teaching tool. Move knobs. Listen. Watch. Ask why? Also, spend a lot of time learning about wavelengths vs. frequency. They’re two different ways of looking at the same thing. C.K.: Can you elaborate on that? Pat: I worked in this business for years before I ever considered the wavelength of the sound. Sound has a physical size and it's this size that determines how it behaves. Almost all things in acoustics are wavelength dependent. It's also the foundation of designing arrays. Once you get a grip on wavelength, then follow-up with superposition. Any relevant text should have some info. Moderator: How does that affect your approach to equalization? Pat: I spend most of my equalization time figuring out what to equalize and what to leave alone. A properly placed loudspeaker with a decent transfer function should not need a lot of EQ. Mostly some fixes for low-freq coupling with boundaries around the loudspeaker, etc. Moderator: Can you give us a primer on superposition? Pat: Superposition is combining waves according to their phase offset. Most know the two extreme cases. In-phase you get a 6 dB add. 180 degrees out and you get a complete cancellation. There are an infinite number of in-between states. If each wave comes from a different loudspeaker at a different position, superposition is what determines the combined response. Weogo Reed: Speaking of wavelengths, please clarify for subs in small rooms: do horn-loaded subs really not have enough space to develop full LF waves? Pat: If you couldn't have LF waves in a small space, then headphones couldn't work. The space determines how a wave will be supported. If the wave is bigger than the space, then there will not be any additional support from the room. Weogo Reed: Thanks! That makes sense! Trooper Hales: Can you elaborate on the modal, diffusion, and pressure zones of a room and give us a bit of info on what’s possible as far as room treatment? Pat: Modes exist when the room dimensions reinforce certain frequencies (wavelengths). The result is coloration. Kind of like dead notes on a keyboard. Treatment of the modal zone requires frequency specific approaches. Visit www.rpginc.com. Peter D is the wizard on that, as well as diffusion. The model of sound behaving like light is only true for wavelengths much smaller than the space. This is called the absorption zone. This is the behavior that EASE assumes when generating reflectograms. It's a valid approach for big rooms, but falls apart for small rooms at low frequencies. Josh: Is superpostition the same as half and quarter space loading? Pat: Superposition is just adding waves. Space-loading is indeed superposition, but the waves are pretty much in-phase and simply add. Trip: how do you determine the distance at which you measure the rig? Pat: It depends on what frequency you are measuring, as well as the physical size of the radiator. You need to get in the far-free field of most loudspeakers to measure levels that can be extrapolated to further or closer distances using the inverse-square law. You can estimate it by ROUGHLY ten times the physical height of the array. Okay, five times to be practical and allow a little error. Of course, with a line array the whole point is to be in the near field. So it's okay to measure there. Chris: Can you offer any advice on how to deal with the common left stack - right stack arraying used in live performances? Pat: I'm not sure what you mean by "deal with." If it's a stereo system, then I try to make everyone hear both stacks. If it's mono, then I would like for each person to hear one of them. You're pretty much stuck with having to use two stacks in many applications. Just be aware that it causes some problems that you have to live with. I'm a big fan of LCR systems for this reason. I also do a lot of single source systems, mainly for speech. One loudspeaker is ideal for speech. Tucci: As long as you're talking array length... the line array 3db/distance doubling holds true for what relationship to the array height? Pat: Depends on who you talk to! My take would be about 10x the height. I use a 1.8 meter array in our seminars and it's good to about 20 meters. After that, you need a longer array. Trip: How much do you actually get to mix these days, band-wise? Pat: Almost never. I really miss it. I always liked mixing more than playing an instrument. Alan H: Please explain the difference(s) in your "Hands On Seminar" as opposed to the regular "Syn-Aud-Con Audio Seminar"? Pat: The "Hands On" seminar was created to allow people to actually do what we cover in the regular seminars. You can't do Hands On with 50 people. I've proved that to myself a couple of times. We limit the attendance to six. Also, the "System Setup and Optimization" seminar is a prerequisite to the "hands on" seminar. We don't cover the course material again. We just review it and do it. Annacoick: What are the key differences due to time constraints of live sound vs. installed sound system design? Pat: BIG difference. With live sound you've got to make it happen NOW. With installed sound you can stare at the computer screen and think. I don't do much live sound, but I respect the people who do. If I were doing it I would get the gain structure, etc. right on the system before it ever left the warehouse. Too many other things to think about on site. Moderator: Can you expand on setting up gain structure? Pat: Basically you've got a whole rack full of stuff that clips at different voltages. If you run the system at "unity" you will have more headroom in some components than others. No big deal, unless one of the components is short on headroom (usually the case). Gain structure is setting things up to clip simultaneously. Either everything clips or nothing does. Now whatever headroom is left above "meter zero" is there for everything in the chain. This has traditionally been done with a scope. We teach a method that uses a $5 piezo tweeter to get the same answer. You have to work pretty hard to get 96dB of dynamic range in a sound system (CD quality). Gain structure is a big part of the equation. Weogo Reed: You mentioned L/C/R systems. For small, L/R live systems how does a third sub in the middle help? What about five spread across the front of the hall? Pat: I think that there are too many variables involved for this forum. Usually spreading subs across the front is not a great idea because they all "superpose" (see I told you it was important!) down the middle of the room. Big hot spot. I like vertically stacked subs for this reason. The wavelengths coming from subs are so long that you are creating an array no matter how you stack them. Trip: Of course you don't want to clip anything, do you? More on the piezo thing, please. Trip: Why 400 Hz? (For the clip indicator) Pat: The 400 Hz fundamental will be inaudible in the piezo tweeter. If you pick too high a frequency, it will drive you crazy blaring out of the tweeter. When the mixer (or whatever) clips the harmonics fall within the bandwidth of the tweeter and are audible Josh: I’ve seen lots of articles on pad calculators, what do we use pads for? Pat: A pad (in today's scheme of things) is just a way to get rid of too much voltage. Pads allow a component to operate near clipping (for best S/N ratio) but still not overdrive the next component. In an ideal world (are you listening manufacturers?) all components would clip at the same voltage. You then wouldn't need pads. Dan: Is the pad in Don's book, SSE (Sound System Engineering)? Pat: Most books on pads assume that you wish to maintain an impedance match between the components. C.K.: Were you talking about pads with potentiometers, or the inline XLR type? Chris: Can you touch on the pin one issue? Pat: Pin one is a grounding issue and relates to how manufactures physically route pin one on their I/O connectors. Most experts (Neil Muncy, Bill Whitlock, others) agree that pin one should go to the chassis. I use three resistors in an XLR. Two series 1k (1%) and a parallel across pins two and three. The parallel resistor actually determines the pad value. You can replace it with a pot (10K) to make a variable voltage divider. Josh: Did the pin one problem occur when designers changed from tubes to transistors and op amps? Pat: I'm not really sure. Steve Macatee of Rane had done a lot of research on this. I would suggest that everyone get a reprint of the June 1995 AES Journal for a number of excellent papers on grounding. AES is Audio Engineering Society. It'll cost you a few bucks. Well worth it. Also visit the Jensen Transformers (and Rane) websites for more info on pin one and grounding. And don't EVER miss a chance to hear Neil Muncy speak on this. Josh: Is there any phase shift when using pads like in passive crossovers? Pat: Not really. Pads are pretty much a pure resistance. The 1-percent tolerance on the series resistors is to maintain the balance of the line. You can leave one of them out on an unbalanced interface. But then again, why would you want to use an unbalanced interface! :>) Charlie: What are your thoughts on Bob Katz proposed metering system to correlate better with the levels we are actually hearing? Pat: I'm not up on exactly what he is proposing, but I'm a big fan of monitoring levels. Dorrough has some excellent meters for this purpose. They give you both peak and RMS, so you can watch the crest factor in real time. Annacoick: Have you done any measurements on line arrays like the V-Dosc and VerTech, and do they perform and specified? Pat: No, I have not. I have measured the Intellivox arrays pretty extensively. They're more a speech only product, and they do exactly what they claim. The theory behind the V-Dosc and others is solid, so I have no reason to doubt their performance. PC: Regarding LCR installations, I've seen examples of systems with the subs positioned on either side of the center cluster between the L/R clusters. What potential problems/benefits could result? Pat: Sometimes subs are array in a certain fashion for purely aesthetic reasons, with no thought given to what the radiation pattern will be. If radio stations used that approach with transmitting antennas (basically the same thing) the FCC would be very busy putting people in jail. PC: What has been your experience with LCR systems using the cross-matrixed method (Vance Breshears)? Pat: Sub placements have to be evaluated for each application. Compromises abound. You may have to spread them to get them more equidistant from all parts of the audience, but then you have to live with the lobing. You can tight-pack them and create an array, but you make not be able to get even coverage from one position (except overhead). Pick your poison. I haven't used Vance's method, but he's a top-notch engineer. Teri Hogan: We run EV MT-2 cabs. From Monitor World, I sometimes have problems with certain frequencies resonating off the back of the cabs. Is there anything at all I can do about this? I notice these errant frequencies most when there is a solid surface within 10 feet or so behind the stacks. Pat: There are lots of things that can be going on there. Does the cabinet vibrate at the problem frequency? Lay a penny on it and fire up a sine generator. It could also be coupling between multiple cabs. Turn all of them off but one and see what happens. What can you do about it? Turn it down or get the vocalist closer to the mic. Also, try it outside some time and see if the problem persists. You have to "divide and conquer." Teri Hogan: This is only noticeable when a full band is performing. Pat: Sorry Teri. Too many variables. If it were mine I would try to schedule some time for experimentation. Feel free to contact me by e-mail for some thoughts. Trip: We touched on this during the Craig Janssen chat, Can you explain beam steering and its advantages? Pat: This takes us back to superposition. Beem steering is simply taking a bunch of loudspeakers that are arriving with a time offset (and resultant phase differential) at some point and using delay to "synchronize" them. This gives you in-phase summation and a higher level at the focal point. The loudspeakers are usually close-packed and in a vertical column. You can get rid of frequencies that don't cooperate by frequency shading with low-pass filters. Moderator: That concludes tonight’s chat. On behalf of ProSoundWeb.com I'd like to thank Pat for sharing his time and knowledge with us. Pat: The pressure was all mine. :>) Thanks for YOUR time. |