Tuesday, April 5, 2016

Use C-weighting to monitor levels in live sound

I continue to encounter live sound "professionals" who have not fully educated themselves on the science of sound and the frequency response of the human ear.  The topic evokes responses of borderline religious or political fervor, where discussion invariably devolves into personal attacks when defenders cannot produce technically accurate rebuttals. It's understandable that someone who has been doing sound for a long time would become defensive when learning they may have been doing something wrong the entire time.

I have seen all sorts of explanations about the correct application of A- and C-weighting.  Some professionals think that the difference has to do with the sampling rate, or think that C-weighting is a completely flat measurement.  Many people continue to use A-weighting to measure loud performances which is incorrect.

All this in the face of what is abundantly clear:  Many live sound "professionals" (and particularly the "DJ" subset) permanently raise the hearing thresholds of their audiences by running at high levels.  Each generation of concert-goers, once their hearing has been damaged, goes on to demand high levels from the next generation of live sound professionals, and the cycle never ends. Maybe we need one or two "reasonable sound level" shows during a run for audiences that don't want their hearing damaged.

I have personally resigned from productions when forced to exceed what I have set as my maximum sound level based on my own research. Not surprisingly, the individuals requesting higher levels had on prior occasions disclosed they had some form of hearing damage.  I urge you to do your own research and monitor your levels with an SPL meter. 

I don't want to hear excuses like "even though I'm at the mixing console I'm not the one who decides how loud the levels will be" - this sounds eerily similar to the excuse "I was just following orders" - and we know that whether your finger is on the trigger of a gun or on the master fader of a mixing console that you ARE responsible for the damage you cause.  As professionals we absolutely have the right to negotiate our contract however we want, and to exercise our right by walking off a production.

Here is a list of what facts you need to accept in order to reach my conclusions:
  1. The frequency response of the human ear is not flat
  2. The frequency response of the human ear changes at different overall sound levels.  At low sound levels, such as those of a quiet room, the frequency response curve is different than when in a loud concert.  (Fletcher and Munson)
  3. The A and C curves closely match human hearing at 40 and 100 phons, respectively.
  4. OSHA's inappropriate selection of the A curve for measuring loud performances happened prior to 1983. Once a standard is created, it is difficult to change to something better
  5. The appropriate curve to use on your SPL meter is based on the loudness of the performance (quiet lecture or loud concert)
My argument is that if you use the C scale and stay under Yamaha's limits you will certainly stay under OSHA's limits using the A scale, with the added benefit of causing LESS hearing damage.

References:

Given the sensitivity characteristic of the ear, the 'A weighted' curve is most suitable for low level sound measurement....In the presence of loud sounds, such as rock concerts, the ear has a "flatter" sensitivity characteristic...In order for the measured sound level to more closely correspond to the perceived sound level, one would want a flatter response from the SPL meter.  This is the function of the B and C weighting scales.  In apparent conflict with this common-sense approach, O.S.H.A. and most government agencies that get involved with sound level monitoring continue to use the A scale for measuring loud noises.  Since this causes them to obtain lower readings than they otherwise would, the inappropriate use of the A scale works in favor of those who don't want to be restricted...It is beyond the scope of this book to detail the 'danger' levels, but anyone responsible for sound reinforcement should be cautious about delivering levels above 95 db(C) SPL to an audience for any prolonged period.
Davis & Jones. pp. 30-31 Yamaha's Sound Reinforcement Handbook. 2nd ed. Milwaukee: Hal Leonard, 1989.
C-weighting is optimized for full-bandwidth sources at levels exceeding 85dB. A-weighting filters out the high and low frequencies and is optimized for lower volumes.
Gibson, Bill. p.46 The Ultimate Live Sound Operator's Handbook. 2nd ed. Milwaukee: Hal Leonard, 2011.
Pro audio equipment often lists an A-weighted noise spec − not because it correlates well with our hearing − but because it can "hide" nasty hum components that make for bad noise specs.
Tontechnik-Rechner
...the C-weighting correlates better with the human response to high noise levels.
Gracey & Associates
On a sound level meter (SLM)...and we are forced to choose between A and C, we should pick C.
Doctor ProAudio
When dealing with the uninformed, live sound folks can "cheat" the rules (or some poser's cursory knowledge) by using the A-weighted filter so that measurements appear to be lower in SPL than a standard or rule that is based on a Flat or a C-weighted limit.
Pro Sound Web

Wednesday, March 30, 2016

Big Blue Clear-Com Call Light

Rather than flashing every time a station presses their CALL button, this light only comes on when a CALL button has been pressed four times quickly. It starts out solid, then after a while starts blinking.

Perfect for getting a specific person's attention and ONLY when you need that specific person. I use this for the conductor in the orchestra pit. They just can't see the little call light on their intercom station.

This project uses a couple of modules and minimal soldering.

The first item is a step-down power converter. Rather than using batteries, let's power the unit off of the Clear-Com 30VDC line voltage. However, the components need 5VDC. An article about efficient power conversion led me to select the LM2596 which is available from Amazon.

The next item is an Arduino Nano. Be sure to order it from an authorized dealer such as Newark Electronics. The one I purchased from Amazon turned out to be a fake and would not work with the Arduino USB driver.

Deriving the CALL signal from the Clear-Com line was done by cheating. I simply purchased a Com-Bit from Illumineering and placed it in the enclosure, aimed at a photodiode. The photodiode is connected to one of the Arduino analog inputs which is held to ground with a resistor.

Then I selected and ordered a large blue illuminated pushbutton from Amazon. I changed the voltage drop resistor to work with the lower 5VDC used by the Arduino.

After a bit of Arduino programming I had a working call light. The light can be turned off by either pressing any CALL button once or by pressing the big button.


 

Thursday, March 24, 2016

Understanding Art-Net

While Artistic License publishes a detailed specification on the Art-Net protocol, I found very little additional documentation on the internet - including YouTube, to help configure an Art-Net fixture and light board to talk to one another.  So here is some info I'd like to share.

The light board we are using in this example is an ETC ION.  The fixture is an Elation SIXBAR 500.

First of all, the light board's network interface was already configured on a Class C network (192.168.0) with a wireless access point for use by ETC's iRFR iPad/iPhone app.  Users would use the app to do dimmer checks or advance through cues.

Since Art-Net uses the same network interface, it's best to reconfigure that Class C network to a Class A.  Art-Net defaults to the 2.0.0.0/8 network.  Anything on that network, including the wireless access point (and its associated DHCP pool) would need to be reconfigured accordingly.

ETC ION configuration:
IP Address: 2.0.0.2
Subnet Mask: 255.0.0.0
Art-Net Broadcast: Primary Art-Net
Art-Net Start: 1

Art-Net Start of 1 was necessary in our configuration because Art-Net universes are zero-based and ETC ION universes are one-based.  So in order for ETC ION universe 3 to equal Art-Net universe 3 the "Art-Net Start" should be set to offset by 1.

The Elation SIXBAR 500 uses Art-Net 2 not 3, so it only supports 256 Art-Net universes.  Basically if a fixture doesn't have configuration options for Art-Net 3 "Net" and "SubNet" then it is Art-Net 2.

The SIXBAR's "Universe" and "DMX Address" settings determine which Art-Net universe and channel it responds to.  Leave its IP and CTR addresses at 0.0.0.0 and it will automatically give itself a 2.x.y.z address in compliance with the Art-Net self-addressing specification.  In 43-channel mode, the SIXBAR set to Universe 3, DMX Address 101 is controllable from the ION (if configured as specified above) as 3/101-143.

Update:
I have found that either the Elation eNode or SIXBAR seems to have a firmware bug related to the Universe setting. With ION Art-Net Start of 1 the SIXBAR Universe=ION Universe but the eNode Universe=SIXBAR/ION Universe-1. With ION Art-Net Start of 0 the eNode Universe=ION Universe but the SIXBAR Universe=eNode/ION Universe+1. Without a third Art-Net device from another manufacturer for testing I am uncertain which product's universe selection is off by one.


Understand that Art-Net's Net, Subnet, Universe, Address and value data are all sent out over UDP port 6454 broadcasts.  They have nothing to do with the IP Address or Subnet mask of the network, fixtures or lighting desk.  As long as all the devices are on the same 2.0.0.0/8 network everything will just work.

In my environment the ETC ION was connected to a "control network" with other systems such as a Mac running QLab and a Soundcraft Expression mixing console.  All of these systems are controllable from a variety of iPad/iPhone apps using the wireless access point.  The problem is that Art-Net's chatty traffic may interfere with HiQnet - I had to separate HiQnet from Dante a while back otherwise the Soundcraft Visi Remote app would not work well with the mixing console.  If this is your situation as well, I suggest placing a router between the Art-Net network of 2.0.0.0/8 and the "control network" of 192.168.0/24 so that Art-Net broadcasts do not cross over to the "control network."

While implementing this solution, I discovered that iOS devices do not like to talk to private networks using public address space (2.x.y.z) so it was necessary to configure the router to NAT a private IP on the booth control network (192.168.0.20) to the new IP address of the ETC ION (2.0.0.2). I went with the Ubiquiti EdgeRouter ERLite-3 which is an inexpensive yet powerful three-port router with keyholes for easy backboard mounting.





Thursday, February 11, 2016

Don't be a nice consultant

From time to time people come to me with requests for custom software development projects on the side.  I continue to refuse due to bad experiences I've had in the past.

Back in 2013-2014 I attempted to work on a couple of projects and they were a disaster.  Not only did I lose my weekends and evenings I did not make any money.

At that time I had no experience doing software consulting and thought I would be generous in the terms of the relationship, such as not charging until the project was completed.

Here's what happened:  The clients always asked for more and more features than were initially expected.  The project could never be finished.

Other problems included poor communication; clients would assign handlers who didn't understand the project, had different goals, or did not respond to emails.

So if you are interested in doing software development consulting, make sure you are paid on the delivery of milestones, and make sure your project consists of several, material milestones that both parties can clearly identify the requirements for meeting.  The more milestones the better in my opinion.

By getting the client to pay by milestone they will be financially committed to each stage of development and will not want to leave the project unfinished.

I'm sure there are books on how to do software development consulting and these guidelines are just the tip of the iceberg, so be sure to do your research first.

For me, I'll stick to my salaried day job and enjoy my weekends and evenings.