• Get involved.
    We want your input!
    Apply for Membership and join the conversations about everything related to broadcasting.

    After we receive your registration, a moderator will review it. After your registration is approved, you will be permitted to post.
    If you use a disposable or false email address, your registration will be rejected.

    After your membership is approved, please take a minute to tell us a little bit about yourself.
    https://www.radiodiscussions.com/forums/introduce-yourself.1088/

    Thanks in advance and have fun!
    RadioDiscussions Administrators

sweet spot Gentner prism II FM?

Personally, I'd like to keep the chain as short as possible and the following sounds a bit like overkill to me (IMHO). But I have a 8100/RCF2/Prism II FM combination and absolutely love it. But reading through this fantastic board I've run into several postings in which a Compellor is mentioned to keep the Prisms in their sweetspot.

Is this really 'recommended' and if so, what is the sweetspot?
 
richard.vanderveen said:
Personally, I'd like to keep the chain as short as possible and the following sounds a bit like overkill to me (IMHO). But I have a 8100/RCF2/Prism II FM combination and absolutely love it. But reading through this fantastic board I've run into several postings in which a Compellor is mentioned to keep the Prisms in their sweetspot.

Is this really 'recommended' and if so, what is the sweetspot?

Richard,

If you're happy, then carefully consider any changes you may try out. After all, once you're on the mountain top, every direction is down. :)

I've tried it both ways. It's my opinion that you only 'need' a Compellor if your levels are way out of control. Even then, the Texars have a built-in buffer that should cover most contingencies. The Compellor's sweet spot is not hugely wider than that of the Prism.

There are those among us who believe that using a wideband processor compromises many of the artifact-avoidance mechanisms a quality multiband processor provides.
But if you actually like the sound, then go for it. ;)

Kind Regards,
David
 
Back in the 80's and 90's I remember trying both scenarios. IICR, the compellor adds the dueling AGC problem that happens when you chain multiple boxes together. For years many stations in the northeast ran Monroville prism’s with WDVE mods, 8100, and CP803. It was hard to beat that combo until about 1997 when the original Omnia.FM was introduced. No matter how you modify the prism's they will have a hard time dealing with today's hyper compressed recordings. If you are happy leave it alone. If not, sell it on eBay and buy a modern DSP box. Good luck.
 
I think what's meant by the sweet spot is that the prism's built in AGC (called a buffer) only reduces gain when levels are too hot, otherwise it runs at full gain. The AGC of the prism, according to the manual, has a range of only 6dB. Some believe putting the compellor in line will prevent a grossly hot level from reaching the prism, and thus preventing distortion.

If you can borrow a Compellor, give it a try and see how you like it. Otherwise, as long as you can be reasonably sure levels don't exceed the 6dB range of the Prism's AGC, you probably won't notice much difference with one.
 
We ran the following and it kicked ass

DBX 168A (Orban designed unit while he was with JBL)
Prisms
CRL SEP-800 (-6 / Slow / Limit)
8100 (modified card 5 & 6)
CP-803

The 8100 is still on the air. The DBX, Prisms and CRL have been replaced with an Ariane.
 
@wgli

What exactly did you modify on the card #5 en #6?
I have a card #0 in my 8100A1, will an additional cp803 add extra loudness with acceptable distortion?

Regards,

Evert
 
fugazi said:
@wgli

What exactly did you modify on the card #5 en #6?
I have a card #0 in my 8100A1, will an additional cp803 add extra loudness with acceptable distortion?

Regards,

Evert

This 8100 did not have the card 0, which is why we had the cp803. If you have a Card 0, I wouldn't run a CP-803 unless it was at the other end of a composite shot and you needed overshoot protection.

I know we did some adjustments to the HF limiters on Card 6 to make them appear to sound brighter w/o sacrificing distortion (which it did very well), but to tell you what exactly we changed I would have to grab my friend's notes again since it was in 2000. I do remember using 2 cards as guinea pigs.
 
@wgli

Tnx for the response.

This 8100 is not used over a composite stl, so i don't need a cp803.
My end setup will be:

stereomaxx
gentners ( not in possesion yet )
8100A1 with card #0 and RCF 1
FM TX


Regards,

Evert
 
Well, if it was me, I'd consider using the StereoMaxx as a doorstop. To my ear, the Audio Prisms (and we used the original models with the marginal power supply and the -100 proc boards and 'Sharpuie' serial numbers) sounded best in front of the 8100 when the things were running full fast, and the COMOPRESS and EXPAND lights were in 'crossing' mode, that is, neither on longer than the other, and if you looked at them, they looked like the lights on a railroad crossing - same cadence. At the outset there weren't any replacements for card 5 in the optimod, so we ran ours set as Glen suggested. Put a clipper behind the setup, and set it topet maybe 1/2 - 3/4 dB extra loud. It's a sweet sounding setup.
This assumes your transmitter is capable of -45 dB or better synchronous AM... or about -60 asynchronous. If it isn't, you're wasting your time looking for nuances which will be lost in the crud of the RF system anyhow.
 
Run the Compellor as close to levelling and as far from compress as you can. The Compellor in this setup is to keep the jocks or automation from overdriving the Prisms. Yes, there is a buffer in the Prisms but it doesn't have a lot of capacity. you'll know you've got it right when the Compellor is not coloring the sound but just keeping the Prismas in the sweet spot.
 
radiosaur said:
Run the Compellor as close to levelling and as far from compress as you can. The Compellor in this setup is to keep the jocks or automation from overdriving the Prisms. Yes, there is a buffer in the Prisms but it doesn't have a lot of capacity. you'll know you've got it right when the Compellor is not coloring the sound but just keeping the Prismas in the sweet spot.
And with "in the sweet spot" you mean not having to use the buffer?
 
radiosaur said:
Run the Compellor as close to levelling and as far from compress as you can. The Compellor in this setup is to keep the jocks or automation from overdriving the Prisms. Yes, there is a buffer in the Prisms but it doesn't have a lot of capacity. you'll know you've got it right when the Compellor is not coloring the sound but just keeping the Prismas in the sweet spot.

We had some VERY sloppy board ops at Z100 yet I never felt 'compelled' to use another processor ahead of the Texars. I'm not sure why anyone would make such a tradeoff. I think it's an unneeded and generally sound-clouding crutch. IOW, the "cure" is simply worse than the sickness. ;)

And don't go saying "Yeah, well, Z100...distortion didn't matter" because reduced distortion was very much on our minds! Z100 would have sounded much worse had we not kept the audio as pristine as possible before applying a controlled, desired amount of clipping at the very end. And I can tell you unequivocally that that certainly did not occur at input to the Texars.

If all the pieces of equipment in the program line have their headrooms aligned to the same clipping point, the possibility for overload is greatly reduced. THAT is the cure that is no worse than the cause!

Kind Regards,
David
 
David Reaves said:
If all the pieces of equipment in the program line have their headrooms aligned to the same clipping point, the possibility for overload is greatly reduced. THAT is the cure that is no worse than the cause!

David,

You know what's interesting, is that most of the folks who carry on about analog, few ever considered this. Yet, the digital domain implores this, as there is no wandering headroom in digital. 0dBfs...is 0dBfs, with no exceptions, but few ever looked at analog this way.

Both analog and digital require a full understanding of "system headroom" or the audio will contain an achilles heel somewhere along the way.

-Frank Foti
 
Hear Hear! ;D

David, Frank,

Amen, Amen, Amen.

There is a LOT to be said for what you describe, though I have never found a paper to refer users to on this subject. Years ago I read an article about this in Broadcast Engineering.

Readers - here's an over view. Every piece of analog gear has an output "clip point" and an input "overload point" (maximum input level). They may or may not be the same!

Sometimes this information is available from the manufacturers spec sheet, but not always. Since the most common spec given is "nominal level" the issue of headroom comes into play.

The nominal output level of a device + headroom = overload point.

The nominal input level of a device + headroom = clip point

If the clip point of an output is matched to the overload overload point of the following input, maximum usable dynamic range is now possible. Any other scenario is non-optimal in terms of noise and/or distortion.

Complicating this scenario is that it became popular among board manufacturers to play specmanship with output headroom specs. Ostensibly this is/was due to the problem of sloppy board ops "clipping the board's outputs". However this "solution" to sloppy board ops was no solution at all, since the input to the next device would become over loaded (when board ops were sloppy) instead of the board clipping. Same result, but allows the board manufacturer to "point the finger elsewhere" as to the cause of the distortion.

Here is an example of the problem, and the solution:

Board has a nominal output of "+4 dBu" with a 21 dB headroom (e.g. clip point of +25 dBu).

Processor has a nominal input level of "+4 dBu" with 18 dB headroom (e.g. max input level of +22 dBu).

Problem - "Sloppy Jocks" can cause 3 dB of clipping in your air chain prior to the audio entering your processor. No AGC in the processor can fix this problem, because the audio is clipped at the INPUT stage to the processor.

Solution = A 3 dB pad between the board and the processor. The board ops may still be able to cause clipping, but the ONLY solution at that point is to teach them not to do it.

The reverse case also occurs, and means a few dB extra noise, but this case is not as easily remedied.

This stuff can really make a difference, particularly if the "operators" don't know how to operate the equipment. Sadly, as fundamental as this is, and as important as it is, it has become "lost knowledge" which is why I wanted to elaborate on David and Frank's comments.

And yes, engineers of lore used to painstakingly go through the entire airchain to eliminate these sources of lost headroom to PREVENT the GIGO effect going into the processing.

Final note: These concepts come into play with regards to "level discrepancies" when interoperability of digital equipment is desired.

For example, lets take a "budget codec" with inputs and outputs that have nominal +4 I/O with 15 dB of headroom (e.g. clip point and max input of +19), talking to a "hi-fi" codec with inputs and outputs with nominal +4 I/O and 18 dB of headroom (e.g. clip point and max input of +22 dBu )

For our analysis we will assume that the manufacturers of each codec have made the input overload point = to 0 dBfs.

We feed a tone of +15 dBu into both codecs.

In the budget to hi-fi direction here is the analysis: at the budget codec's input, +15 dBu = -4 dBfs. At the Hi-fi codec this digital information is received and it indicates a signal of -4 dBfs. The analog output of the hi-fi codec thus is at +18 dBu (4 dB below its max output level).

3 dB of "Gain" has occurred. ???

In the hi-fi to budget direction here is the analysis: at the hi-fi codec's input, +15 = -7 dBfs. At the budget end this is received and indicated as -7 dBfs. At the analog outputs of the budget codec the signal is output as +12 dBu (7 dB below its max output level)

3 dB of "loss" has occurred. ???

The solution is to make sure each codec properly interfaced at its *own* end.

So the hi-fi codec output must be padded (if necessary) to NEVER overload whatever it feeds.

And the budget codec may need a pad so that whatever is feeding it will never cause its inputs to be over loaded.

IMPORTANT stuff folks. Read and understand these concepts! 8)
 
richard.vanderveen said:
radiosaur said:
Run the Compellor as close to levelling and as far from compress as you can. The Compellor in this setup is to keep the jocks or automation from overdriving the Prisms. Yes, there is a buffer in the Prisms but it doesn't have a lot of capacity. you'll know you've got it right when the Compellor is not coloring the sound but just keeping the Prismas in the sweet spot.
And with "in the sweet spot" you mean not having to use the buffer?

"Buffer active" is OK.

"Buffer full" is bad.

You can hear it. Or at least I could. The Prisms sounded a little thin below the sweet spot, a bit fuzzy and mashed above it. I could really hear the audio become full and rich as the level got into the Prism's optimum range. Experiment, and I bet you'll see..er , hear what I mean.
 
radiosaur said:
"Buffer active" is OK.
"Buffer full" is bad.

You can hear it. Or at least I could. The Prisms sounded a little thin below the sweet spot, a bit fuzzy and mashed above it. I could really hear the audio become full and rich as the level got into the Prism's optimum range. Experiment, and I bet you'll see..er , hear what I mean.
That's definately worth a try!
 
We ran them sich that the BUFFER ACTIVE light blinked fairly often, but never stayed on, and NEVER let the BUFFER FULL light light at all.
 
Status
This thread has been closed due to inactivity. You can create a new thread to discuss this topic.


Back
Top Bottom