Audio Processing Hardware

You could use a Nanocompressor. They're on Ebay for about $60.

Frank
www.easthillradio.com

Before you spend any money, check the settings on your TV. Mine is several years old, but in one of the menus, it has a setting to keep the volume at a constant level (normalized, compressed, whatever). I don't remember what that setting is called, as I don't use it. Perhaps your TV has such a setting.

Neil,

You can buy these units on Ebay for $20.00 sometimes and you can buy these at RadioShack too... They work great and is a very good DSP for the money...

http://www.audiovox.com/webapp/wcs/stores/servlet/ProductDisplay?catalogId=10001&storeId=10001&productId=14166&langId=-1

Thank you all for your advice. Your help has pointed me in the right direction.

I'll let you know how it works out.

Neil

I got a Nanocompressor and just installed it. All the suggestions were good and appreciated, but this seemed to be the most usable in my particular configuration.

I am just getting started with this but it sounds good so far and it solved my problem.

Thanks again to all,

Neil

Hello again,

As I stated, this hardware solved my problem with movie audio but I did not tell the whole story.

The connection with Part15 FM is that I use a Ramsey FM-25 with a 4" aerial (reduced from my previous posts) as the link to the Sony Walkman to which my headphones are attached.

I was able to calibrate the transmitter and the Nanocompressor to yield 75 kHz. deviation when the front panel LEDs are in the top yellow range. If nothing else, this serves as a poor man's VU meter. This system is not set and forget, but I now have a means to roughly set my modulation rather than flying blind as I previously had with my calibrated eardrums as the meter.

If anyone is using a FM transmitter without a means to monitor and limit the audio, this is not a bad way to go.

Neil

Hi all,

Several weeks ago I posted a question which you will see at the top of this thread. I obtained a Nanocompressor on a web auction for $31 and am completely happy with this installed in my audio chain just before the FM 25A. Thanks for the advice.

The point of this post concerns audio processing. Tonight I changed my operating frequency because a new station has been granted a CP in my area, so I thought I would move now rather than later. I listened around the "clear" frequency using a Sony FM walkman and Sony CD30 phones. What I heard from the licensed stations was nothing but over compressed junk. It really did not sound good at all! Then I found my signal modulated at 75 kHz. deviation (confirmed with a Motorola service monitor) and what a difference. It sounded almost CD quality.

This is probably not due to the Nanocompressor which is very nice, but rather it is due to me not overcompressing my audio and over modulating my transmitter as I assume the other stations are.

A friend was the CE at a very high power FM station and I visited the station many times. He told me about their compression scheme and how "proprietary" it was. We also took a look around the band with the station's spectrum analyser hooked to a roof antenna. Many of the signals from other stations were overmodulated, exceeding the allowable bandwidth and the audio sounded as if it was clipped. No wonder they sound like (technical term) crap. It seems they equate loud with range. Why spend $$$ for a home or car system when the signal at the antenna is junk?

My experience on part 15 FM is that if you don't push too hard with the audio, it is excellent as compared to many of the "big boys" on the band.

Neil

I remember the days of Grant Fiekert, the inventor of the AM radio "Audio Pilot". It was the first AM audio processor that provided phase correction between music and voice as each went through the processor. The entire machine used electron tubes; no semiconductors at all. The box had multi-band compression, hard-knee limiting, phase detection/correction and a final hard clipper to provide assymetrical modulation. I took up over half of a 6 foot tall control rack.

But today, I have a analog/digital audio processing chain that will do all that and more...and the audio is nearly transparent. Of course, the notable exception is the NRSC standard audio equalization and shaping. It takes up 3 inches of rack space and its all solid state. It's a combination of 2 pieces of equipment.

My FM processor is a digital "Ultramizer" from Behringer. It has nearly a hundred pre-sets. Audio processing today can look forward and selectively compare audio levels already through the processor and audio about to be processed for loudness, shape and phase correction.

The primary reason for audio processing in either AM or FM is to improve the listenability of broadcast audio inside the cab of an automobile and the high level of ambient noise. The age of the table top radio passed a couple of decades ago.

As to the overmodulation measured on a spectrum analyzer from a fixed outdoor antenna. I have worked with FCC inspectors over the years to track down the most likely reason for over modulation, especially on FM. The reasons include base band clipping of the FM composite signal, mulitple signals at the receive antenna as a result of multipath and incidental AM riding on the FM signal. These reasons, and a few more were found to give incorrect indications on monitoring equipment used by the FCC. When the signal was measured directly from the transmitter, the over modulation disappeared. Actually it never really existed. My stations were given warning calls and inspections for over modulation. None ever received a citation because high powered stations in large cities and in mountainous terrain produce an incredible amount of multipath. In TV, its called "ghosting". The bottom line is, there are bad actors in very aspect of society. But, for the most part many stations do their very best to produce the highest quality signal possible.

With the eventual advent of HD radio and the digital audio processing used with these systems, broadcast audio will be absolutely transparent. However, no one operating a Part 15 AM or FM will likely have the vigorish to buy the exciters and audio processing necessary to sound as good or better. The days of single channel compression, limiting, phase correction and equalization are but a fleeting memory. Today's "smart" processors will supply AM and FM broadcast audio at the receiver that will exceed the quality of compact disks. Atmospheric noise of mulitpath will no longer be a factor. Golly Gee Batman, what ever happened to Flash Gordon anyway?

Marshall Johnson, Sr.
Senior Pastor, President
Rhema Christian Fellowship, Inc.

Rhema Radio - The Word In Worship
AM 1660 - FM 93.5
http://www.rhemaradio.org

Today's "smart" processors will supply AM and FM broadcast audio at the receiver that will exceed the quality of compact disks.

Much good information in your post, Marshall, but the above statement from it applying to digital broadcasting may be a bit too enthusiastic.

AM and FM IBOC/"HD Radio" broadcast transmitters use data compression, which adds artifacts to the original program source material detected by the receiver. These artifacts produce quite significant audio harmonic and intermodulation distortion when measured by standard audio test equipment. But the artifacts (in theory) are not very objectionable to the average listener.

In contrast, a compact disk is not data-reduced, but a linear, 16-bit digital medium. It is capable of a 96 dB noise floor and harmonic/IM distortion levels so low they are difficult to measure accurately.

So unfortunately, a smart audio processor will not be able to compensate for the basic limitations of the HD Radio digital standards so that digital AM/FM broadcasting will "exceed the quality of compact disks" at the output of the receiver.
//

Since I've mostly been dealing with getting our stream going, I thought I would comment on this. It is absolutely critical to have processing on your audio before it hits the encoder. If you are using something like the lo-fi windows media codec at low bit rates, the difference can be night and day! If people are interested there are a couple of good articles on the Orban web site. Some of the most important things are compression (or at least limiting), and frequency dependant limiting. Spikes in certain frequencies really take up a lot of bandwidth.

http://www.orban.com/support/orban/techtopics/

I haven't read all the articles yet, most of it is from trial and error with our Aphex chain (compressor, aural exciter, Dominator). And if the supplier ever ships it, I'll probably be moving things to the Orban 1100 Optimod PC card that works in conjunction with the encoder that we are using. I'd suggest that you take a listen, but I'm still waiting for IT to connect to the 'net. You can listen to other streams that are encoded with the same software at http://www.opticodec.net/ . Check out the 32K and 24K streams.

Rich,

There has never been an accepted broadcast transmission signal process that has had a minus 96 db noise floor. The stereo FM broadcast protocol used for the past several decades has been accepted as the "best so far" audio standard. With the advent of composite clippers, baseband processing and exciter and antenna design, the weak link will forever be the exciter.

I remember when the new silver bullet was the detection of AM modulation on the FM carrier, and the phase shifting control supplied by adjusting the output tuning and antenna tuning of the station. It reduced the weak signal noise at the receiver, thereby extending the listenable range of the signal.

And finally, even though HD-R FM out of the exciter will not reach Cd quality audio delivered to the receiver, the audio processing chain of most stations certainly will. That, of course, is ignoring the quality of the D/A and A/D converters required to allow the audio to be processed in the first place. I guess we haven't reached perfection, yet. I still want to know where Buck Rogers went to.

Marshall Johnson, Sr.
Senior Pastor, President
Rhema Christian Fellowship, Inc.

Rhema Radio - The Word In Worship
AM 1660 - FM 93.5
http://www.rhemaradio.org

There has never been an accepted broadcast transmission signal process that has had a minus 96 db noise floor.

We were writing about the noise floor of a compact disk. However a digitally modulated FM exciter fed by AES-3 digital audio has performance very close to this: 94 dB S/N, and harmonic/IM distortion less than 0.007% in monaural mode. In stereo the S/N drops to 83 dB due to the wider r-f bandwidth used, but the harmonic/IM distortion is still 0.008% or less. All of these levels are inaudible during programming.

The complete specs for this exciter are available here http://www.freefilehosting.org/public/18765/DIGIT%20CD%20Specs.pdf (disclosure - I was the original author of that spec sheet).

With the advent of composite clippers, baseband processing and exciter and antenna design, the weak link will forever be the exciter.

By many orders of magnitude, composite clippers and baseband processing add significant distortion to the incoming program material. The reasons they are used are to give the station a characteristic "sound" desired by the management, and to raise the average modulation in the exciter while not exceeding FCC peak limits, so that the station will sound louder and more competitive. Clean, low-distortion audio is not as important to broadcasters as is being "loud."

As you will see in the numbers in my text above and in the spec sheet I linked to, the signal degradation of the FM exciter itself is virtually nothing, in comparison to what the audio processors and clippers generate.
//