Inter-modulation

[mtglass]

Good day all,

TL;DR: For live sound, is inter-modulation a factor to address or ignore? The best home audio systems I've heard are planar speakers powered with a single amplifier, which says inter-modulation may not be a significant concern. My live band experience with sub-woofers says inter-modulation noticeably effects clarity.

I'm starting into the BFM world armed with a small bit of theory and a smaller bit of practical experience. I've been an amateur musician for 40 years, have a degree in EE, run sound occasionally for local bands, and have been using my collection of old PA gear for my current band. To me, the piece of sound equipment that has the greatest impact on quality is the speaker. I've been blessed to listen to very high-end home audio equipment, and really enjoy listening to a system that can create a wide and deep sound-stage that surprises me with details I haven't heard before. My goal is a live sound system that allows my audiences to have that same experience.

I know that's not possible. In my experience, good audio systems require exact positioning of equipment, listener, and reasonably acceptable acoustic rooms. In live sound I present to an audience that (a) won't necessarily recognize that they've heard good sound - but will intuitively react positively when they have, (b) are dispersed over an enormous area in noisy rooms with poor acoustics, (c) with sound sources that are noisy, cross-fed, and sometimes slightly less than artistic.

I was drawn to these designs because I was able to understand the engineering, they seem like I can reasonably build and maintain them, and the explanations in these forums match my experiences overall.

My experience is that running mono-sound live is superior to L/R because it gives more even coverage of all the sources to the audience. It's also better to separate subs from mains because I can separate some sound sources that have no business coming from the same speaker: vocalists aren't performing in the 20-100Hz region and shouldn't be in the subwoofer mix. I haven't bi-amped or tri-amped mains. I'm lead to believe that inter-modulation in speakers leads to systems that separate subs, lows, mids, and highs; and my subwoofer experiment seems to confirm this. To address inter-modulation, I might crossover between speakers:

• Sub: 20-100
• Low: 100-600
• Mid: 600-6k
• High: 6k-20k

So I can't make a live sound experience like a home-audio experience, but there are things I can do live that aren't practical at home. People want heart stopping bass and SPL is expected to be high. Good musicians can give more feel to the music than a recording. Integrating the senses of sight, sound, and feel can create a unique experience: and thank you to the beer-garden for taste and social encouragement. People at a concert are in party-mode, which is also unique. I think using these unique aspects of a live performance can leave the audience with an excellent experience.

Finally, I'm a huge fan of headroom and dynamic range because I think it adds clarity. Part of the reason I'm headed to horns is I expect their efficiency will increase headroom and I think that is very closely related to the inter-modulation question.

Thank you for your time and best wishes!

[mtglass]

Reserved for results when I get them.

[mtglass]

Test plan.

I've got a GRS 10PT-8, and I have some 4 Ohm 20W wirewound resistors.

I'll generate a 2.828V RMS 80, 160, 320, 640, 1280, and 2560, summed sine waves. I'll send that into an el-cheapo power amp I've been wanting to test out and through the resistors. I'll hook an oscilloscope up to the resistors (2 in series) and see what kind of waveform I get as a baseline. Not sure if I can export the DSO's capture, it's pretty old... Then I'll repeat by putting that signal into a single GRS 10PT-8 and observe the waveform. Will I be able to see any back-emf?

If I get any results at all, I'll move on and send an 80, 160, and 320 signal across the resistor network as another baseline. Then again apply it to a single GRS 10PT-8 to check the waveform. Then I'll do the same with the 640, 1280, and 2560. If I sum the two resulting waveforms, will I get the same results as putting all six sine waves across at the same time?

I think the difference will give some insight into inter-modulation.

Thoughts on improving the test? Is it a valid test?

[Bruce Weldy]

Sorry that no one has responded, but I honestly don't understand what you are trying to do. Are you building a system? Are you tuning a system? Are you going to be running live music shows with the system in various venues?

Or, is this just a scholarly endeavor?

[Radian]

+1 to what Bruce said.

Intermodulation distortion is an important parameter to control in the design phase of an audio system…but it’s not something to chase in an application.

Pick up a few dusty copies of some old AES publications and you’ll see there’s very little, if any, new ground to cover here.

[mtglass]

Thank you, great feedback. I'm sorry I wasn't clear, but not surprised.

I'm in the design phase of a PA to haul around for our rock and roll cover band Dusty Tones. In the end, of course, it's what everyone here says: it boils down to what you want to carry at the end of the night. Our current system is fine for most indoor gigs, but in Montana we end up outside in bad weather.

Currently building WH10s for monitors. I wish there was a drain in the bottom so beer poured into the monitor would drain out...

Have a couple Titan 48's, 24", in the works to add bottom end to my existing mains.

Then on to my question about mains: DR280's. Also 2' wide to match the cargo space. I selected DR280's so I don't have to haul Titan's to every venue, and when I do, I want to cross over around 100Hz so that Titan 48's will couple. We only want our PA to cover crowds of around 500. That's a big crowd for the venues we play. I've been shooting for around 100db @ 100', which is way more than necessary, but I'd much prefer a system that's idling that one that's working hard to keep up.

I have mixed experiences with using crossovers to split the frequency spectrum into smaller segments, and I was curious if inter-modulation was a primary factor, or something I can ignore.

To restate the question, can I expect noticeable / measurable improvement by cutting the frequency spectrum to each "stack" of mains in to segments of 2-3 octaves? So a stack of DR280s for low, a stack for mid, a stack for high? Or is it a waste? My expectation is "yes it's noticeable, but do you really want to carry all of that?"

I still think I'll measure some things per my test plan, just to see what pops up.

Best!

[Bill Fitzmaurice]

A total waste. In theory you can reduce IMD by using more drivers with more frequency passbands, but in practice IMD is so insignificant that it's safely ignored. If that wasn't the case three way systems would not be the benchmark, and audiophiles wouldn't swoon over single driver speakers. Plus when you do go multi-way you don't do it using the same speakers to cover different passbands. Each is optimized.

[mtglass]

Thank you.