Subscribe on iTunes, SoundCloud, Google Play or Stitcher.
Support Sound Design Live on Patreon.
In this episode of Sound Design Live I talk with the Product Manager at L-Acoustics, Scott Sugden. We discuss the automation built into the new M1 measurement software, why you can’t trust measurement data above 10kHz beyond 80m, and the specific number of measurement positions you need to represent an audience for proper output EQ choice.
I ask:
- How did you get your first job in audio?
- What are some of the biggest mistakes you see people making who are new to L-Acoustics systems?
- What’s interesting for me about the P1 processor is that you have managed to automate some procedures that we would normally do manually. Would you talk about one or two of those procedures and maybe why L-Acoustics decided to pursue this kind of automation?
- Tell us about one of the biggest or maybe most painful mistakes you’ve made on the job and how you recovered.
- From FB
- Kevan Atkins: During a demo I attended for L-ISA, he made a claim that it addresses the problem of comb filtering in system design but didn’t really expand on it. I’d be curious to hear him talk about this in more detail.
- Haniel Trisna: Explain the idea of boosting high mids in the middle boxes to air compensate for long throws instead of the top box (generated by auto FIR in Sound Vision), and where can we learn more about working with the auto FIR and auto splay.
- Primož Vozelj: Are they working on vertical processing of their line arrays (like ArrayProcessing, MLA etc.)?
- Calum Young: I’d love to know what sort of measurement equipment / facilities / testing procedures they go through while developing new units. Is most of the work and decision making done in simulation software pre building prototypes, or is there more extended testing / voicing of units?
- Steve Knots: If Greek amphitheaters were designed to put the audience in the best place for sound, why are we not creating clubs, theaters and control rooms in similar architectural style?
- Roy Sputtz: Is the idea of true stereo in live sound a myth?? And also why isn’t anyone making an all weather line array?
- Ockert Marais: Are they planning on supporting, mic correction curves and z-weighted weighting curves on the P1?
- What’s in your work bag?
About 8 microphone positions distributed evenly in the center mass of the coverage of a loudspeaker is pretty representative of the overall. The likelihood of a poor EQ choice because of that is pretty low.
Scott Sugden
Notes
- All music in this episode by Bodo Felusch.
- Justin Vernon and Bon Iver
- Cadac J type
- L-Isa, Soundvision
- Workbag: iSEMcon 7150, DPA 4007, Digigram Cancun 442
- Book: The Signal and the Noise: Why So Many Predictions Fail–but Some Don’t
- Podcast: 20,000Hz, 538 Politics, Science Vs, Planet Money, Freakonomics, The Flophouse
- Quotes
- A common mistake made with a lot of systems is the expectation that you can solve your problems of a bad design after you install it.
- The lack of knowledge; it’s hard to be aware that you don’t know something. That’s one that only comes with time, experience, and making mistakes.
- If you have just one mic at FOH or one at FOH and one 20ft off stage of that, the likelihood of an EQ choice not being representative of the audience is really high.
- We have taken measurements outside in atmospheric conditions that are good and found that at 80m from measurement to measurement on average is ±5dB at 10kHz. This means that if you see a measurement with a bump of +1dB at 10kHz, you can’t know if that’s the reality or it isn’t. Even a long average doesn’t help.
- We tend to look at measurements and think they are some hard fact, but if you’ve used any measurement software outside at distance, you watch the curve move around and when you think it looks good you hit store.
- The best thing we can do is use the modeling environment to find the best result, especially at distance, and then get outside and verify behaviors.
- It’s hard for anyone to be an expert at all things. It’s an important part of career growth to identify what you can or want to be an expert at and then support yourself by surrounding yourself with other people that reinforce those skills, not reproduce them.
- It’s better to figure out what you like and hone down on that expertise than to try to cover every little thing.
- Once you get used to the workflow [of the P1/M1] the savings in workflow time and organization of your data is a 10x increase.
Nathan and Scott –
I agree wholeheartedly with the discussion regarding mic calibration files. I did a demonstration last month at an audio meetup that I thought you might find relevant: two measurements taken from the same spot, with and without a calibration file (pink / red) and then another measurement taken by moving the mic a foot away (blue). The deviations that a file corrects for are dwarfed by the spatial variation.
I think the statement I like best is “if your mic is good enough to come with a calibration file, it’s good enough not to need it.” To me, the file is important for a different reason: it means someone actually measured the response of the mic, so we know there’s nothing crazy going on.