Cascading Audio Processors

[GeneSavage]

Especially back in the 80s, but I'll bet before and after too, some of the biggest stations in the U.S. had "secret blends" of audio processing to create their signature sound.

In some cases, it might be hot-rodding an Aphex Dominator (& eventually starting your own company); in others, it was 2, 3 or more boxes in a row.

I know an "all-in-one" solution is always best because the various processors were designed to work with each other... but I'd love to hear stories of the best combination of boxes that shouldn't have worked, the biggest number of processors in a row you ever saw, and maybe your favorite blend of yesterday's processors.

In particular, I'd love to know what some of the flame throwin', hot rockin' top 40s of yester-year were using to sound the way they did: Z100 in New York (1983 - 1988), KIIS-FM in LA (1981 - early 90s), WLS in Chicago (AM('74 - '86 & FM '80 - '91), and Houston's 93Q (especially 1987 - 1991).

I know one of the most common combos requiring no special hot-rodding was feeding an Aphex Compellor into a couple of Texar Audio Prizims, and feeding that into an Optimod 8100/A with the XT/2 limiter chassis. The best example of that chain that I ever heard was, in all places, Joplin Missouri, on a rimshot called CD98.3. I also heard it locally being used to slur and crunch everything fed it. :

At the time, these were heavily-guarded secrets, with some engineers allegedly setting up "mock" processing to show anyone touring so they wouldn't know the "real" secret to their sound... but it's been 30 years and the big boys have (I assume) all gone digital...

...What Frankenstein processing chain did you see (or create!) that "sang?" What was your "trick" to getting that "sound?"

(In case you wonder why I'm asking now about then: I've begun sequencing a couple of DSP audio processors on my internet radio station, and am trying to learn the best and worst practices in setting them up. Plus, I've always been fascinated by radio stations that had a "signature sound" as you tuned across the dial... going back to the 60s, at least!)

 

[FFoti1]

In particular, I'd love to know what some of the flame throwin', hot rockin' top 40s of yester-year were using to sound the way they did: Z100 in New York (1983 - 1988), KIIS-FM in LA (1981 - early 90s), WLS in Chicago (AM('74 - '86 & FM '80 - '91), and Houston's 93Q (especially 1987 - 1991).

I can answer the question about Z-100, NYC. Interesting you asked about the time period of 1983-1988, as the person in charge of the Flame Thrower's processing was a left-handed kid from Cleveland, who now has a penchant for live steam locomotives in 1/8th scale!

The air chain consisted of:

Howe Phase Chaser (used to maintain consistent mono, because the format was all cart)
Texar Audio Prisms (early version, with single LED bargraphs) Hand delivered by Mr. Glen Clark himself!
EXR Exciter (like an Aphex Aural exciter. Used for sparkle) This was permanently "borrowed" from WMMS, Cleveland.
Barefoot Optimod 8100 (left over from WVNJ)
CP-803 composite clipper (Jim Somich came up with the idea of allowing the red LED to illuminate!)
Continental 802A exciter

That was the original airchain, and remained that way until 1986, when the same left-handed kid from the land of WMMS, began modifying the Aphex Dominator, model 700. Also, the EXR exciter was replaced with an early version Barcus-Berry, we added a Rev-5 reverb to the chain, and experimented with stereo enhancement.

When the modified Dominator was added, later known commercially as The Vigilante, we eliminated the guts of the 8100. At first, we used the Optimod clipper cards #8 & #9 and its stereo generator, but later on, the same kid from Cleveland built his own clipper.

-Frank Foti

 

[TomT]

Gotta watch those kids from Cleveland--they're subversive!.

Must be that Lake Erie water!


(TPT from Lakewood)

 

[GeneSavage]

Frank, thank you for the response! I remember seeing the ads for the Vigilante in the BSW catalogs... I'm guessing I hadn't yet been exposed to the magic of Z100 (living in Oklahoma and not having the internet yet), so when I later found out about the legend, I didn't connect the two.

How COOL! The composite clipper was the one piece I had forgotten about. There was a while those were REALLY popular... and seems like another manufacturer came out against them pretty heavily... I want to say Orban, but it might have been CRL or somebody else.

I'm not familiar with the EXR Exciter, but I of course with the Aphex Aural Exciter... what an amazing box. I wish I had one to play with today; I tried a plug-in that was supposed to emulate it once, but it seemed to just generate high-frequency "grunge."

I saw a guy running a sound booth at a performance venue take a hissy accompaniment cassette provided by the upcoming singer... he rolled off the highs on the board, getting rid of the hiss but making it dull. Then he dialed in the AA... and the highs came back... without the hiss! I was SOLD.

Interesting that you were running the Prizims into the Optimod WITHOUT the XT/2 limiter chassis... was it not available, or did you feel like it was just too much?

On a related note: I've always heard it said that multiband before single band creates pumping everytime... but it effect, that was the Prizims into the Optimod, wasn't it? Or were you basically using the 8100's limiter?

Interesting that you were able to do so much with "so little." I suppose that's the sign of a superior engineer.

I'm trying to make do with free software, and am feeding Stereo Tool into Sound Solution 1.31. I tried the 2.0 version of Sound Solution but found the dual-band AGC in the front too pumpy... I've been pretty happy with the preset I developed for Sound Solution, but it isn't very loud and didn't always pull up everything like I wanted.

Stereo Tool never gave me the sound I wanted, either, and now playing with it in front of Sound Solution there are things I like and things I dislike about the sounds. I'm using it a bit like a slow dynamic equalizer, and if set right it also tends to even out the RMS of songs before they hit Sound Solution... a GOOD thing.

Thanks for taking time to reply with these details; the sound you and others created on the radio dial were truly "larger than life!" It feels like we've moved away from that, now... for better, and for worse...

 

[LA_Guy]

We did similar in Boston...

I had an Aphex Compellor (old style). We removed the output stage and ran right off the buffer (balanced no less). Rechipped, realigned and recapped it (actually only had input caps-there were DC servos in it). That fed the Marti Dual mono STL-10 (set for 25 us), again rechipped and recapped and both input and output transformers removed. That went into a pair of Prisms (again totally went through) and an 8100 Optimod (again, rechipped and recapped-and IC310 fixed) a CP803 (rechipped and recapped, regulated power supplies added and grounds fixed) and all that went into a BE FX-30 (again, rechipped and recapped). My favorite opamps at the time were AD711s and AD 712s.

Wasn't it funny that the yellow light on the 803 was fed from a 555 timer? It really had nothing to do with clip level. We used to set the clip level with our QEI mod monitor. I'd null the pilot and then bring up the clipping until the crud around the pilot was at around -40 db. Yes, the red light would occasionally come on that way-but I knew that the clipping was around 1 db this way.

The board we had was an LPB (again, rechipped and recapped-I used an OP-37 in the program amp and came out unbalanced. CD players were Phillips/Magnavox (also reworked).

I used to buy military spec. film caps from a company in NJ called Elcon Sales. I bought around 1000 .1 uf 50 volt polypropolene mil. caps for 50 dollars. Used 'em for years.

I had a BBE box (the one with the big knob on the front) and used it for a while but later took it out. It made too much distortion.

The format was alternative (WFNX) and no one could touch our little 1650 watt signal-we sounded like a 50 kW station! 3 db louder then 'BCN and twice as clean!

When I was in Los Angeles I ran a similar setup there on an AAA station there (just ran a reciter for an STL-rechipped of course). Replaced an Omnia 5 (sorry Frank ). The sales person from Aphex wanted me to try their processor and he was planning to drop in on his way back from San Diego. He never showed-and told me later that KACD/KBCD sounded great and his box wouldn't be able to make then sound any better then they did.

Also ran similar on KRQQ in Tucson in 2000-2002.

Did I mention that I'm (also) left handed?

 

[FFoti1]

Re: We did similar in Boston...

Wasn't it funny that the yellow light on the 803 was fed from a 555 timer? It really had nothing to do with clip level. We used to set the clip level with our QEI mod monitor. I'd null the pilot and then bring up the clipping until the crud around the pilot was at around -40 db. Yes, the red light would occasionally come on that way-but I knew that the clipping was around 1 db this way.

Actually the 555 did have something to do with the clip level. It was part of the duty cycle monitoring function. If the clipping duty cycle became too long, it triggered the red LED.

-Frank Foti

 

[radiovigilante]

Frank
What did you run when you set up the Airchain at Pirate Radio?

 

[WNTIRadio]

It wasn't going to win any loudness wars, but the Ariane and recapped/chipped 8100a combo is real clean.

If you want loud, put a Cobalt Blue card in it and run about 2 or 3dB of gain reduction. Still sounds open, but more competitive with the digital boxes. With the card 0 installed, it comes close in loudness but can still breathe.

Also ran the Compellor+Texars+Optimod combo at some other places too. Took a lot of fiddling, but if set up properly could sound decent. The biggest problem was keeping the L-R null set in the Prisms which tended to drift all over the place. Remember running pink noise late at night for a minute or two every few weeks to get it back to a -50dB null.

I also put an Omnia.11 on a station this year and as much fun as it was tinkering with the old stuff, this thing sits up and barks... without all the bad side effects of the cascaded boxes.

 

[Kmagrill]

1986: Texar Audio Prisms driving a slightly hot-rodded 8100a feeding a Moseley PCL606 composite STL with a stock CP803 feeding the exciter. Moduation was monitored on a QEI 691 because it could ignore peaks shorter than 5ms. When setup properly, that configuration is jump-off-the-dial loud, yet amazingly clean sounding. I've heard stations still using it today and it's still competative, though not jump-off-the-dial loud by comparison to today's stuff. The key to success in analog processing is limiting and distortion canceling clipping. Compression produces a busy sound that is fatiguing and IMHO, does not contribute much to loudness. Stations using Aphex Compellors driving Prisms always had a signature busy sound to my ears that never beat any of my stations in any of our markets (we were in 7 medium & major markets at the time). I know for a fact that several engineers at competing stations got blasted by their GMs for not being able to make their $12,000+ air chains sound better than our $9,000 ones. I also could never get an XT chassis to not sound too busy. A barefoot set of Prisms, operating as a dynamic content equalizer (think low to medium density) more than as a compressor gives a very even sound that can be fed to your 8100 which does most of the loudness work. The composite clipper, in addition to being essential to handle bounce on any composite STL, also played a key role in loudness at the expense of adding some L-R noise, but the magic was that the clipping action was mostly inaudible. I used one of Frank's composite filters (I forget what they were called) to clean up HF spikes after the clipper and that bought another percent or so of loudness. We used to laugh at the baseband meter on the exciters which usually quivered just below the 100% line, looking almost like it was reading DC instead of audio. What fun that was!


For Internet radio, the requirements are slightly different because clipping then coding to a lossy streaming format can have audible effects. For that reason, good digital processing is preferred, though well-tuned analog processing can still be used.

 

[WNTIRadio]

I always set the Compellor to 8:30 on the "level/compress" control to keep it slow and feeding the Prisms a controlled steady level. I found that if you didn't keep them in the sweet spot, especially with some of the board monkeys we had, they would either go into blinking too much audio mode or the gate would activate and then the spectral balance would be off. Ran the density control on the Prisms at 9 o'clock, and used the band mixes to tailor the output to the Optimod.

Anyone remember stacking modified (rechipped, recapped etc.) CP-803's? Run the first one light and the second one a little heavier?? Did this on a mono FM station and it could blow the doors off your car. Of course, no L-R to trash up. Well, it was trashed up on the spectrum analyzer, but nobody heard it since it was mono. Ahhhh, the bad old days!

The composite filter was the Dividend.

It's bad to use any type of clipping in front of a codec. Lookahead limiting is much better.

 

[Sgeirk]

Interesting to see the BBE mentioned. I bypassed one in 1995 and picked up about 15 miles of extra fringe coverage on a 100kw FM.

 

[FFoti1]

Frank, thank you for the response! I remember seeing the ads for the Vigilante in the BSW catalogs... I'm guessing I hadn't yet been exposed to the magic of Z100 (living in Oklahoma and not having the internet yet), so when I later found out about the legend, I didn't connect the two.

How COOL! The composite clipper was the one piece I had forgotten about. There was a while those were REALLY popular... and seems like another manufacturer came out against them pretty heavily... I want to say Orban, but it might have been CRL or somebody else.

I'm not familiar with the EXR Exciter, but I of course with the Aphex Aural Exciter... what an amazing box. I wish I had one to play with today; I tried a plug-in that was supposed to emulate it once, but it seemed to just generate high-frequency "grunge."

I saw a guy running a sound booth at a performance venue take a hissy accompaniment cassette provided by the upcoming singer... he rolled off the highs on the board, getting rid of the hiss but making it dull. Then he dialed in the AA... and the highs came back... without the hiss! I was SOLD.

Interesting that you were running the Prizims into the Optimod WITHOUT the XT/2 limiter chassis... was it not available, or did you feel like it was just too much?

On a related note: I've always heard it said that multiband before single band creates pumping everytime... but it effect, that was the Prizims into the Optimod, wasn't it? Or were you basically using the 8100's limiter?

Interesting that you were able to do so much with "so little." I suppose that's the sign of a superior engineer.

I'm trying to make do with free software, and am feeding Stereo Tool into Sound Solution 1.31. I tried the 2.0 version of Sound Solution but found the dual-band AGC in the front too pumpy... I've been pretty happy with the preset I developed for Sound Solution, but it isn't very loud and didn't always pull up everything like I wanted.

Stereo Tool never gave me the sound I wanted, either, and now playing with it in front of Sound Solution there are things I like and things I dislike about the sounds. I'm using it a bit like a slow dynamic equalizer, and if set right it also tends to even out the RMS of songs before they hit Sound Solution... a GOOD thing.

Thanks for taking time to reply with these details; the sound you and others created on the radio dial were truly "larger than life!" It feels like we've moved away from that, now... for better, and for worse...

Thanks for your nice comments. There was something 'magical' about Z-100, both the format and sound. I've tried to keep that 'sound' in my head and use it as a guide for what is now Omnia.

To your question why were we not using the XT/2...There's a story to that.

In 1983, while still working at KNEW/KSAN in SFO, I 'knew' Bob was working on a multi band adapter for the 8100. It wasn't out just yet. So, even though it wasn't released in August of 1983 when we launched the "Hot rocking' Flame Throwin' Z-100," and we were running with the barefoot 8100 we got from WVNJ, I told Scott Shannon I'd bring back a new toy from SFO. (Had to head back there to sell my house, which I'd lived in for *only* three weeks.) While back in the bay area, I called my buddy at Orban, Jesse Maxench (miss you still Jess), and asked if it was possible to test the 8100 'adapter' on-the-air at this new CHR in NYC? Jesse was a good friend, and very candid. He told me that Bob didn't want to have something that was still in Beta, out of the area. I respected that position.

So, I had to go back to New Jersey, without a new processing toy. Knowing Scott would be asking me first thing upon my return, I needed Plan B. That turned out to be the Texar Prisms. I'd seen them earlier that year at NAB, and called Allied Broadcast to get a pair to test. Later that day, a fellow by the name of Glen Clark calls to ask if I wanted to demo these boxes? I asked if he could send them overnight? His reply....Well, if you could pick me up at the Newark Airport, I'll be there in a few hours!! That night the Prisms hit the air in NYC, and I was able to deliver on my promise to Scott about bringing back a new processing toy!! Hence, we never felt the need to use the XT/2.

As far as how radio sounds today, it's more than just the on-air processing. Much has happened - for better or worse - in content generation. We've been doing work to some effect in hopes of undoing some of the nasties that happen to source audio.

-Frank Foti
Z-100's original Chief Engineer (1983 - 1988)

 

[FFoti1]

Frank
What did you run when you set up the Airchain at Pirate Radio?

Think of it as Z-100 West. The only additions were we used the "Tailor" from HIT Design, along with the air chain described prior for Z-100. If there's a place for pix, I do have some of Z-100's chain, along with Pirate Radio. BTW: I always felt the LA market was set to process a bit less intense than NYC. One night we put the Z-100 'settings' into Pirate Radio, sand it was way over the top!! Yes...we backed it down.

-Frank Foti

 

[rrsounds]

Re: We did similar in Boston...

Did I mention that I'm (also) left handed?

I think that quite a number of us in the industry are left-handed/right-brained. Especially evident among those who, paraphrasing Steve Jobs (another left-hander), want to connect the technical with the artistic.

Kind Regards,
David

 

[LA_Guy]

Re: We did similar in Boston...

Wasn't it funny that the yellow light on the 803 was fed from a 555 timer? It really had nothing to do with clip level. We used to set the clip level with our QEI mod monitor. I'd null the pilot and then bring up the clipping until the crud around the pilot was at around -40 db. Yes, the red light would occasionally come on that way-but I knew that the clipping was around 1 db this way.

Actually the 555 did have something to do with the clip level. It was part of the duty cycle monitoring function. If the clipping duty cycle became too long, it triggered the red LED.



-Frank Foti

You are correct. I remember now. If you put a tone into that unit the red LED came on way before the yellow one did.

 

[gtomas]

BTW: I always felt the LA market was set to process a bit less intense than NYC. One night we put the Z-100 'settings' into Pirate Radio, sand it was way over the top!! Yes...we backed it down.

Which just proves what is also true today - that setting up processing is very market dependent. What works in one market is just too much, or not enough, in the other...


Regards,
Goran Tomas

 

[GeneSavage]

It's always tough to tell when you're listening to an mp3 of a 25-year old cassette of a radio broadcast, but I noticed what I heard of airchecks of Z100 vs. say KIIS-FM, Z100 sounded pretty "pumpy," actually more jumping up and down than I remember CHRs of that era doing (which very well may be the result of how it was recorded), vs. the aircheck of KIIS-FM which sounded... controlled, but a bit flat... gain riding, yes, but not a fire-breathing dragon on the tape I heard at ALL.

If anybody knows of any (especially unscoped!) airchecks from that era that pretty faithfully reproduce the sound of these stations over the air, I'd love to hear them. I have cassettes of a local CHR from 83-85 that I LOVED, but they used the CRL chain. They sounded great, but nothing like the mega-markets I'm sure.

Later I worked for one of the local "flame-throwers" on the weekends, and really got familiar with how you could take these boxes and stomp the life out of a song. Gritty, grungy, slurred highs, crunchy, just fatiguing... oh, it was a "signature sound," all right!

When the station got a new engineer, he tried to "clean it up" while keeping us as loud or louder, spent a whole night tweaking the whole chain from studio to transmitter... the PD came in the next morning grumbling about how the engineer had "ruined" our sound, and a week later we sounded just like we always had. :

Just a year or two later I'd hear KJMZ-FM 100.3 Dallas (100.3 Jamz) for the first time, and while it wasn't "on fire," it was a very smooth and lively processing setup; it sounded like I thought big-market radio should (although by this time the 80s were on their way out, and "nailing it to the wall" wasn't the prevailing theory anymore).

It seems like we started the decade with the theory that the processor should constantly adjust your focus to whatever was the dominant sound at the time, riding the gain to keep everything "up" without clipping it to death. As we went through the 80s into the 90s, that kind of "re-focusing" of the listener's audio perspective got a bad rap because of pumping and breathing... and we stopped as much constant gain riding and instead went to limiting and clipping the snot out of the source to make it as loud as possible without moving the image at all.

Now we have a local AC whose processing completely "controls" the sound. It's not distorted. It doesn't pump, it doesn't breathe. It does, however, flatten the life out of every song played through it. The peaks are gone, and the snappy rhythms of the 80s pop tracks they play are driven into the background. The songs just... sit there... lifeless. If I was hearing an 80s tune for the first time on this station, I'd be confused as to how it could have ever moved anyone.

I slowed down the attack times on my audio processing and got back that analog "jump" that really makes drums and the like pop... and the songs sound as exciting as I remember them sounding on the radio!

That's a tangent, I suppose... but I was always impressed when you could feed audio through multiple boxes without losing those transients.

There was also a "sound" to the AM blowtorches of that era; I was never sure if it was eq, or processing, or both... but I KNEW when I had 89 WLS for example. There was a texture to their audio that was unmistakable. I suspect they were running an Optimod 9100, but I don't know that for sure; whatever it was, it was set up RIGHT... even for someone listening on skip to Chicago at night from Tulsa.

 

[richard.vanderveen]

I really love to read the stories about the old chains and the thoughts behind them ;D
The loudness war reached Europe considerably later, one of the preferred processors at the start was the Inovonics 250 because it had such a clean sound. But once the wheel started spinning the French really loved it and went overboard to create a special sound. Orban 222's followed by ModSci MYB-2's were 'common', 222 for the depth, MYB for the width!. Also cascading Orbans and Omnia's were on air. Crazy, but it did put the 'French sound' on the processingmap.

 

[stacker]

Air chain then in use at WVWA. (9!)

An Audimax (slow release); fed into a Volumax; fed into a Gates Level Devil.

 

[rorban]

How COOL! The composite clipper was the one piece I had forgotten about. There was a while those were REALLY popular... and seems like another manufacturer came out against them pretty heavily... I want to say Orban, but it might have been CRL or somebody else.

That was Orban.

The reason was that once you band-limit a composite clipper (to protect RBDS and/or SCAs), the process of lowpass filtering the output at 53 kHz seriously compromises stereo imaging. Composite clipping produces quite a bit of energy around 76 kHz (second harmonic of 38 kHz) when the program material has a lot of L-R energy. When this energy is removed by linear lowpass filtering, the composite stereo waveform "lifts off the baseline," which indicates that stereo separation is compromised. This filtering also introduces substantial overshoot.

I developed a composite limiting technique I call "half-cosine interpolation" that avoids both of these problems. It allows a linear 53 kHz baseband lowpass filter to be used, which, unlike nonlinear filtering techniques, does not add further distortion in an attempt to control overshoot. This has been offered in all recent Optimod-FM processors starting with the 8400.

For more info, look up U.S. Patent 6,434,241 at patents.google.com.