Don't delay codecs

I have to comment on the Audio Codecs article in the November 2000 issue. I agree that broadcasters today never had it so good with on-demand, bi-directional 22kHz stereo digital service. The article would have been greatly enriched if the author had taken a look at the apt-X codec algorithm; a coding solution that has been universally accepted as probably the best there is. In terms of signal delay, it made me grimace when I looked at some of the delay figures in the table. I pity the interviewer with some of those quoted delays. Equally, what about the poor presenter expected to monitor the backhaul leg or off-air feed from a transmitter with an STL codec in addition to a codec with some of those quoted delays. I believe that 20ms delay is about just how much delay a person can reasonably handle.

In live scenarios, where program continuity is desirable, delay is crucial. In a 15kHz stereo circuit, the standard apt-X algorithm will introduce 7.5ms delay, while the new enhanced version of apt-X reduces the delay to 6ms. It gets even better for 22kHz stereo where delay is a mere 2.5ms, now improved to 2ms.

What happens to the codec when the data network nasties rear their ugly heads? Some codecs just lie down and mute. Apt-X can soldier on, securely delivering acceptable audio even with 1 bit error in 1,000.

How does the audio stand up to additional passes of coding, especially with algorithms with compression ratios greater than 4:1? The effects of concatenation or tandem coding must be considered when the same audio file is expected to endure further stages of compression. It is widely accepted that 4:1 is a fairly safe bet if further coding is anticipated. One apt-X user is happy for me to quote that on occasions he has successfully subjected his audio to some forty passes of coding. A mighty boast, but it makes the point that, when it comes to repeatability, apt-X is the best.

Fred Wylie
technical consultant
Audio Processing Technology
Belfast, Ireland

Popular parasitics

John Battison:

I was just looking through my new issue of BE Radio [RF Engineering, January 2001]. Great article on parasitic radiators. Very interesting stuff. I always look forward to reading your articles.

Arnie Clawson
Chief Engineer
Clear Channel Communications
Ashland, OH

John Battison:

With great interest I just read your article about Parasitic Radiators. I am a very active ham radio operator and very much interested in the low bands, especially 160 meter.

We just finished a contest on the 160m band last weekend. We used a portable setup with military surplus mast that consisted of two verticals spaced 130 feet apart, 70 feet high with 3 top loading wires 45 feet long and a slope angle of 45 degrees. For the switching we used a home-built hybrid coupler to make the artificial delay to switch it in three directions: infire, endfire and broadside. These are shunt-fed, grounded verticals. In this case, we used a slant-feed system 11 feet high and 11 feet out of the vertical. An omega capacitor system with two vacuum variables were used for the tuning network.

After some fine tuning, I was able to get both towers to 1:1 SWR. The front-to-back ratio on this array was amazing; about 35dB.

I have shared the electronic version of your article with my fellow ham radio amateurs on my website at www.topband.net. The website is devoted soley to 160 meter applications. Please let me know what you think. I welcome any advice you can offer.

Willem A. Angenent
K6NDV
Lancaster, CA