Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums > Product Reviews: Sound Reinforcement FUD Forum Archive

Amp Review

<< < (4/4)

Jeff Wheeler:
Langston Holland wrote on Sat, 20 November 2010 14:35
One advantage of this is that you incur less of a latency hit with this approach vs. using high slope FIR filters at the low to mid crossover region.

My understanding is that FIR delay is a function of sample rate and number of taps, and that you can basically throw CPU power at the filter to reduce delay while also making the filter more precise.  I could be remembering this incorrectly, though; my notes on this topic are trapped on a PC with a busted mainboard.

Langston Holland:
Hi Iain:

Thanks for the link to that 2007 post of yours - don't know how I missed it before - I'd probably been further down the road by now if I'd seen it. :) Might have just been due to how long that thread got at the time...

On that 3dB peak observation, that is true and what you'd expect from the measurements. The fact is that the long-term output increases by about this much as well with a hang of 6 boxes per amp channel (worst case in my rigs). The subjective difference is huge with larger arrays. Maybe I'll figure out a way to measure that someday that reduces the subjective (fun) component.

Since maximum output is tied to limiter settings, it would probably be helpful to describe my approach:

1. Start with the mfg. recommendations.

2. Adjust short-term limiting so that woofer slap is avoided using typical pink noise (12dB crest factor) and Keele's tone bursts through the passband. In the case of the KF730 low passband, I do this using the minimum number of boxes I'd ever use per amp channel (3), so I have the maximum voltage delivered to the boxes that they would ever see in actual use from that amp.

3. Adjust long-term limiting by ear using the now world famous "maximum musical output" test per a condensation of Tom Danley's life-long research into the deep realms of psychoacoustics (aka common sense).

The best objective long-term power test procedure IMO is Pat Brown's "Toaster Test". So far, I'm the only one in my neighborhood that's done it and it works. Yet if you set your long-term limiters by it you'll find the LF passband to be driven at least 3dB hotter than you'd like to listen to it with music, thus the TDMMO test always results in more conservative power limiting. It would be just as valid to use the toaster test for the HF passbands but I don't bother. The ear is obviously far more sensitive to distortions in this area and I fully subscribe to starting with the mfg's recommendation and them applying the "Ugh! Turn it down!" method for the final settings.

A final note - all of this is processor dependent. I'm into the Lake processors at present which have unusual limiter settings. Given the variation I've seen with something as simple as "Q" with parametric EQ between different mfg's processors, I'd expect standard limiter adjustments for "Attack", "Release", etc. to vary as well. Bottom line: measure everything that moves, if it stops moving, fix it and measure it again.

A final, final note - this all assumes you have two adjustable limiters on the processor output for the passband. The Lake processors including those internal to the Lab PLM amps, ITechs and optional Powersoft DSP have separate "RMS" and "Peak" limiters. You could also use the adjustable maximum voltage limiter on non-DSP amps such as the Lab FP+ series combined with a single compressor/limiter you find on a typical loudspeaker processor output. The coolest thing would be designing your own using a free wire processor per Charlie Hughes' article. Don't have time to experiment with that yet - wish I didn't have to sleep. :)

Bob Lee (QSC):
John Roberts  {JR} wrote on Tue, 16 November 2010 07:00
PS: You could do this null in real time, but many amps already do..... look for a flashing red LED.  


Thank you for that!

Jason Joseph:
Interesting read so is it safe to say that compared to the bigger boys the Camco Vortex 6 is better used for high/mid duty than for subs..?

Navigation

[0] Message Index

[*] Previous page