The Signal Chain

While VAC and other programs can handle differing sampling rates, it's generally advisable to keep them similar throughout the chain.  If they're different, you have to convert; conversion takes up CPU cycles and potentially can degrade a signal.  It may not make much difference with some programming genres (such as the voice based ones), and powerful hardware, but not everyone has that luxury.

On Artisan's Statement

"...differing sampling rates, it’s generally advisable to keep them similar throughout the chain."

Very important, and KDX keeps the whole string of software running at 44.1 kHz, except that audiofiles from different producers have all kinds of sample rates and channel numbers so they go through a conversion to 44.1, which at least happens without skips or drop-outs, and I cannot be sure how the quality changes due to the process.

We could re-sample all the programs but that is also a way of losing quality, and of course would be time-consuming.

Signals in Chains

This morning during a cool rain it was time to re-visit the question of compression for the radio programs produced by KDX Worldround Radio. This gets complicated.

Besides wanting our productions to "jump out" and sound "professional", there is the fact that when they get played by radio stations our shows go through compressor/limiters used at the transmitter end for boosting loudness on the dial. This adds a second compression layer.

The reason this issue came up right now is that the other day one of my programs followed another program and sounded weak by comparison. Some programs use very nicely planned compression, others do a poor job. Therefore the well compressed shows that come along bear investigation to observe what to look for in the end result.

The ratio I've been applying is 2.5/1 which is more delicate than the more common 3/1, and we know about a few 10/1 instances, which is very severe compression.

We are still puzzled by the interaction between the "threshhold" and "noise-floor" settings, mainly because we haven't figured out what to listen for.

Send comments by way of Aijet Pai.

Well, I'll just say his about all that:

I've been working full time, on the air in commercial broadcast radio since 1973.  During this time I have worked in many production studios producing commercials, programs, interviews, etc.  At no time in 45 years have I EVER worked in production studio that had ANY processing for audio being recorded, be it a commercial, or a public affairs interview.  Never. Every production studio I've worked in was basically a decent mixer with various audio inputs, be it turntables, CD players, computers, etc..  and a decent mic, and direct output to the recording device -- in the old days a reel to reel recorder or cart machine, and today a computer.

Why? Because everything produced in these studios was going through the stations on-air processing, whatever that may be. I've worked at stations with nothing more than limiters to prevent over modulation to insure legal operation, to the stations I work for now that have the latest machines from Orban.  Double processing is not necessary.

I have, for the past 10 years or so been busy offering my service as a voice over artist and commercial producer. I apply zero processing to this work, unless it's requested.  I think in thousands and thousands of cuts, maybe 3 have requested mic processing, a hint of reverb, etc.

As for the air chain of my Part 15 -- the audio comes straight out of an iMac and into the audio module of my Procaster, and any processing applied is done with the onboard processor that comes with the Procaster transmitter system. That's it. And it sounds as good as any commercial station on the AM band.

Same thing when I produce programs (such as "The Oompah Hour") although I'm using a different mixer, recorder, mic etc on a different side of the studio, it's just a mic, mixer, turntables, etc. straight into a digital recorder.

Also worth noting that most of this is done with a dynamic mic, not a condenser.  Both at work and my own show productions the trusty Electro-Voice RE20 is the mic of choice, although I do use a condenser mic for voice over work.

TIB

Different Radio Ages

In 1959 when I entered broadcasting AM was still the dominant medium and was being slaughtered by television which had been established for 10-years.

AM stations then used very blunt compressor limiters to protect their modulation and boost the average level.

Commercials arrived on transcription disks, and there was a deep science involved with producing commercials that cut through the bad frequency response and static.

In my collection I have a Coca-Cola jingle transcription with a very crisp highly produced sound that features emphasis in the main hearing range 1 kHz to 3 kHz with lows rolled off and nothing above perhaps 6 kHz. It really sounds sharp.

FM didn't enter maturity until the early 70s, and refined processing gear for either AM or FM was still getting established.

Tim is 100% right about today's professional radio landscape, but the game had different rules when I learned the trade.

In these recent years, knowing that some of my shows were destined to be shared on part 15 stations, I have doctored the sound quality to account for station operators not having the skill or means of processing their air sound.

Of course a well recorded un-processed recording contains all the potential needed to add flavoring later in the signal chain, which matches Tim's message.

Chefs have many kitchens.

I Almost Forgot

There is a significant reason for shaping the compression within an audiofile and somehow it slipped my mind the other day.

In many cases listeners hear my radio shows as audiofiles not being broadcast over a radio station, and a well-balanced dash of compression sparks the program into sounding better.

Since talking about this days ago many inroads have been made and I now know as much about setting compression as I've let people think I knew over the years.

I think I was able to slide by because of owning several hardware racks.