The author is VP of global sales for Orban Labs. This is from the ebook “Optimize Your Air Chain.”
An air chain consists of anything you want to transmit. It starts with audio sources like automation systems, mics, satellite feeds and so on. This audio then runs through your “plant,” which may consist of consoles, routers, EAS gear, PPM encoding and maybe processing (if it’s not at the transmitter site).
It then takes a ride on a microwave or some other method to go from the plant to the transmitter site. Once there, the audio might encounter an EAS box and/or processing.
This is all part of your air chain, and there are lots of opportunities for audio to become damaged along that circuitous route.
Over the years, I have diagnosed many issues with air chains, both at the plant and the transmitter site. How do you diagnose these issues?
Step 1 is to realize you have a problem. Audio corruption is subtle and may have many sources. The best way to start looking at an air chain is to make a block diagram. Start at the output of the plant and, working backward, put your hands on every inch of the system.
In our travels, we have found gear patched into air chains hiding in the bottom of racks under piles of cables, hiding in closets and stuffed in all kinds of weird locations. We hunted down pairs of cascaded processors under a pile of wire 12 inches deep at the bottom of a rack. Those were running a combined 40 dB of compression before the main processor, and it sounded like Satan’s ShopVac.
Key message of this step: Air chains tend to accumulate gear over the years, and if you are relatively new to the station it probably makes sense to see what’s been added by your predecessors.
Step 2: While you are digging around, look for unnecessary digital-to-analog and analog-to-digital conversions. Even though A /D and D/A have come a long way in terms of performance, having six sets of them in an air chain isn’t a good plan for great audio.
Step 3: LISTEN to your audio. A good set of closed-back headphones with an appropriate headphone amp is a great addition to your “bag of tricks.”
I was hunting a pernicious distortion problem at an AM facility and found that the IP-based STL was having its input clipped at the studios. I found it by carefully listening to the STL output … and that clipped audio was ugly.
Step 4: Don’t forget about the STL. Microwave STLs can have all kinds of issues, especially if they are old enough to have a driver’s license!
When was the last time you looked at the RF link margin and/or receive level? Do those numbers look like they did when it was installed, or have they degraded? Have you looked at the error rate on the digital STL lately? Did the link come up within a couple of dB of the predicted path when it was installed, or was it off by 6 dB or more?
Dish misalignment from high winds, water in the RF connectors, N connectors tightened with a pliers can all contribute to an STL problem. This becomes critical if you are right at the edge of the link budget.
Step 5: Check to see if there is AGC before the STL. These devices are a throwback to the days of analog STL. If you have a digital plant and a digital STL, and there aren’t a raft of D-to-A-to-D conversions or gain being added, it’s probably unnecessary. FS should be FS everywhere in the system. The air chain is going to sound better with the fewest number of devices in the path between the console and the transmitter.
If you can’t get around using one, make sure that the AGC in the processor is turned off or the two of them will be at war with each other. A typical symptom of this is major pumping and breathing.
If you’re using an Orban processor and you don’t have a preceding AGC in your chain, you should be running somewhere in the range of 9 dB to 12 dB AGC action with “nominal” program level. This is the number one problem I see in the field.
Step 6: Don’t forget about Nielsen PPM encoding. PPM encoding requires the cleanest, highest density audio you can provide. Bad audio or poorly processed audio will negatively impact PPM encodability.
Field story: We ran into a PPM issue with a client who had a remote host. Every time the host was on the air talking, the PPM numbers plummeted. We did some sleuthing and found that the remote host was using mono MP3 encoding at 64 kpbs. We changed the encoding to mono AAC at 256 kbps and the PPM encoding problem went away.
At a minimum, PPM encoding should be done after the AGC. It’s significantly more effective to encode PPM after all of the multiband processing is done and before the safety limiter. That requires dual PPM encoders if you are running HD.
Audio processors with onboard PPM encoding are more effective than external outboard hardware encoders even with “PPM Enhancers.” Do not use lossy codecs on PPM-encoded audio, and be wary of lossy compressed composite schemes too. Testing by one of our largest customers found that these can seriously degrade PPM encoding and damage Nielsen numbers by upwards of 20 percent!
Step 7: Check those microphone preamps. It’s really hard to find well-designed mic preamps!. Many sound OK at low gain but fall apart at the 60 dB to 70 dB of gain that dynamic mics typically need in a broadcast environment.
Listen for lack of high frequencies; elongated, unnatural sibilance; and excessive noise. You aren’t going to be able to fix those problems with processing. Take a listen to the output of the morning drive console without any processing. You might have the wrong mics, or the preamp isn’t up to par, or both.
Your assignment: Make a block diagram and then look at what’s in the air chain … and get rid of everything that absolutely doesn’t need to be there. With air chains, less is always more.
Read more in the ebook “Optimize Your Air Chain.”
