Using a 8100ST as leveler for Texar's

[stha]

I'd like to use some of my old stock processors to keep my Texar's input level in a good range. I know that Compellors are the most common choice, but I don't have any unit in hands. I want a very transparent leveler. What do you thing of using an Orban 8100ST chassis ? ( in wideband setting with clipper off) It's been a long time since I didn't used a 8100ST, but if I remember well, it's output bandwidth is up to 20 khz ? It's important because I use both FM & HD outputs of the Omnia6EXi that follows the Texar's, and a full 20Hz/20Khz bandwidth is prefered for HD.

 

[rorban]

I'd like to use some of my old stock processors to keep my Texar's input level in a good range. I know that Compellors are the most common choice, but I don't have any unit in hands. I want a very transparent leveler. What do you thing of using an Orban 8100ST chassis ? ( in wideband setting with clipper off) It's been a long time since I didn't used a 8100ST, but if I remember well, it's output bandwidth is up to 20 khz ? It's important because I use both FM & HD outputs of the Omnia6EXi that follows the Texar's, and a full 20Hz/20Khz bandwidth is prefered for HD.

If you do this, you should set the jumpers on card 5 the same as you would for XT2 operation. This slows the attack time and makes the 8100A's two-band compressor work like an AGC.

Bob Orban

 

[gtomas]

It's important because I use both FM & HD outputs of the Omnia6EXi that follows the Texar's, and a full 20Hz/20Khz bandwidth is prefered for HD.

Actually, unless you are using HD for simulcast only (no splitting for multiple channels) reducing audio bandwidth will most likely improve the quality of your broadcast. Even if you are only simulcasting and using full 96 kbps, you may find that slightly rolling-off high frequencies may reduce your high-end coding artifacts.


Regards,
Goran Tomas

 

[Sgeirk]

You're using Texar's in front of an Omnia6exi?

 

[wgliradio]

Why not? I'm using a Behringer MDX-4000 in front of an 8500. Back in the mid 90's when Behringer knew a thing or two about signal processing and multiband.

 

[stha]

Thank you very much Bob, I will think to set the jumpers on card 5, if you have another advice about card 5 settings, let me know. Goran, I'm not sure to understand very well, why should I introduce some rolling off with a 8100ST ? That's why I needed to be sure if a 8100ST output is not filtered (If I remember well, 15khz filtering is on the 8100A chassis). Sgeirk, Yes I'm using Texars in front of an Omnia 6Exi, I will replace with an Ariane as soon as I can, but today it's be best solution I've found. I will be happy to read your comments about this. I've read that many Texar/Prism users are defeating the limiters, what do you think about this and how to defeat the limiters easily on Texars ?

 

[rrsounds]

Thank you very much Bob, I will think to set the jumpers on card 5, if you have another advice about card 5 settings, let me know. Goran, I'm not sure to understand very well, why should I introduce some rolling off with a 8100ST ? That's why I needed to be sure if a 8100ST output is not filtered (If I remember well, 15khz filtering is on the 8100A chassis). Sgeirk, Yes I'm using Texars in front of an Omnia 6Exi, I will replace with an Ariane as soon as I can, but today it's be best solution I've found. I will be happy to read your comments about this. I've read that many Texar/Prism users are defeating the limiters, what do you think about this and how to defeat the limiters easily on Texars ?

If you slow the attack of the 8100 and remove the limiters from the Texar, I believe you will run the risk of clipping on transients.

Somewhere along the line, *something* needs to take care of the overshoots that get through slow attack AGCs. In my experience, while one processor alone usually doesn't create too much of a problem, two slow-attack processors in series may be asking for trouble.
Just a thought...

FWIW, one way you might defeat the limit function of the Texar is to remove (or lift one end of) R21, 22K on the four M101 boards. That's what I would try, looking at my 1984 vintage schematic. You may want to try selectively defeating limiting on particular bands, depending upon your preference.
By taking out R21, you still leave in the peak detection which will light the front panel LED. If you hear clipping coincident with the LED flash, you might want to put the resistor(s) back in.

Are you running your Texars in a matrix?

Kind Regards,
David

PS: thanks for considering an Ariane! ;D

 

[stha]

Hi David,
Thanks for your comments. I will try to defeat the Texar limiters. The Texar is in front of a french IDT sound design (used for 7 bands clipping only), and followed by Omnia6 (broadband AGC bypassed, very slow attack time on multiband AGC's). I'd like to remove the IDT, but I'm on a very competitive market, some of my competitors have 9 dB of MPX power using Arianne-Omnia6-STL and...Omnia6 at transmitter !

 

[The F Mister]

Interesting set-up you have going on. Any chance you can upload some processing clips? Considering the material you are using and the sound you are after a wild guess you're market is somewhere in Europe (france, italy)?
http://processingclips.gorantomas.com
userrocessing
pass:clips

 

[dspxscott]

Hi David,
Thanks for your comments. I will try to defeat the Texar limiters. The Texar is in front of a french IDT sound design (used for 7 bands clipping only), and followed by Omnia6 (broadband AGC bypassed, very slow attack time on multiband AGC's). I'd like to remove the IDT, but I'm on a very competitive market, some of my competitors have 9 dB of MPX power using Arianne-Omnia6-STL and...Omnia6 at transmitter !

It sounds like your competitors 1. Are trying to crack a nut with a sledgehammer or 2. Have money to burn. 3. Don't really know that much about processing.

You may be able to achieve the same effect with a pair of leds configured back to back across the input to the transmitter driven with 24dbu.

I am going to give you the benifet of the doubt but part of me thinks your pulling our leg. Can we get a photo of your setup?

 

[stha]

Dspxscott, you are probably right about the 3 reasons concerning my competitors. I never claimed that I'm a king of processing too, but my interest for processing since 25 years is certainely as respectable as all the rest of the group. That's why I will never "pulling your leg" here. Will you also suspect that I'm pulling your leg when I will shortly order a transmitter at BW ?

 

[gtomas]

Goran, I'm not sure to understand very well, why should I introduce some rolling off with a 8100ST ? That's why I needed to be sure if a 8100ST output is not filtered (If I remember well, 15khz filtering is on the 8100A chassis).

You do not introduce filtering with the 8100ST... The output of the 8100ST will go up to 20k (IIRC the first 15k filter in 8100 is on card 6, which you don't use in 8100ST chassis).

The filtering I'm talking about is done in your final processor in a section dedicated to final limiting and conditioning for digital broadcast. In-there you will find an adjustable low-pass filter. Without going into details of perceptual coding, you may find that setting this filter to a value lower than 20 kHz may reduce high-end artifacts of your digital broadcast. How much lower would depend on the bitrate of your HD channel as well as the processing going on before this filter.

The point I'm trying to make is that full 20k bandwidth is not always the best thing to throw at a codec, particularly low-bitrate one.


Regards,
Goran Tomas

 

[Sgeirk]

I just can't fathom placing a pair of Texar's in front of an Omnia 6. To each his own. I'm glad the stations in our area don't take such a "scorched earth" approach to processing.

I'm quite happy running Omnia6's barefoot.

 

[TomT]

I have ST's in front of my Omnia 3T--but only on a side chain for the commercials/liners coming off hard drive. Keeps levels consistent, especially for the liners over the main satellite programing, which flows straight through to the Omnia. I would concur that Texars in front of an Omnia 6 would seem to be a step backwards.

 

[RolfTaylor]

Goran,

With a straight perceptual coder you are correct - limiting bandwidth often will increase "fidelity" despite the loss of bandwidth.

However.... you are wrong in this case. Codecs using SBR (Spectral Band Replication) won't benefit from this approach, as the core codec is not processing that information (typically the core codec is running at 24 kHz sample rate so everything above ~10-12 kHz (Max) is handled by the SBR.

Rolf Taylor
Applications Support Engineer

APT North America

www.aptx.com

<SNIP>
The filtering I'm talking about is done in your final processor in a section dedicated to final limiting and conditioning for digital broadcast. In-there you will find an adjustable low-pass filter. Without going into details of perceptual coding, you may find that setting this filter to a value lower than 20 kHz may reduce high-end artifacts of your digital broadcast. How much lower would depend on the bitrate of your HD channel as well as the processing going on before this filter.

The point I'm trying to make is that full 20k bandwidth is not always the best thing to throw at a codec, particularly low-bitrate one.


Regards,
Goran Tomas

 

[gtomas]

However.... you are wrong in this case. Codecs using SBR (Spectral Band Replication) won't benefit from this approach, as the core codec is not processing that information (typically the core codec is running at 24 kHz sample rate so everything above ~10-12 kHz (Max) is handled by the SBR.

Rolf,

I'm very well aware that the user is using the SBR codec And I can understand where your thinking is coming from. In fact, reading about the SBR codecs I assumed the same as you do.

While I haven't had the access to HD Radio equipment to test the HDC codec itself, I've done lots of experiments and listening tests with HE-AAC (CT implementation). As these two codecs are quite similar, I assume the results might apply to HDC as well. And my experience has been that low-pass filtering can indeed help with SBR codecs. In fact, in some cases you can make a fairly drastic filtering with SBR codecs, while still not sacrificing your high end. This is because the highband of SBR codec is artificially synthesized and in practice (depending on the bitrate and the spectral content of the audio) the codec tends to put a lot of the noise and smearing around the high frequency components. I believe this is the one reason filtering helps slightly reduce the "metallic" artifacts of the high end as it simplifies the high frequency envelope for SBR part of the codec to describe. I haven't done enough experiments to actually claim this, but I believe it is what's happening. How much would filtering help depends on the user's sensitivity to SBR artifacts, the average spectral balance the preceding multiband processing produces, etc. That's why I suggested to OP to give it a try.

Btw, I would be VERY interested if someone has done real testing of HDC codec, to see how it actually performs and compares. For example the CT implementation of HE-AAC at 48kbps (which is a typical bitrate for HD2 channel) doesn't go above 16/17 kHz at all. If the HDC codec does the same, all this talk about the importance of full 20k bandwidth is a moot point (for HD splits, anyway).


Regards,
Goran Tomas

 

[The F Mister]

And how far can an average ear even hear? Is it really that important to go above 16Khz?

 

[rrsounds]

And how far can an average ear even hear? Is it really that important to go above 16Khz?

That's a highly subjective question!
Personally, I just go with the pressures of what the medium will bear. For example, with restricted data-rate signals, if lopping off above 15k can be proven to improve the sound below 15k for most people, I'm all for it. OTOH, if the data bandwidth is available to expand the range, I'm all for that, too. SACDs can sound simply monumental.

But really, it's a very good case of "YMMV".

Kind Regards,
David

 

[The F Mister]

And how far can an average ear even hear? Is it really that important to go above 16Khz?

That's a highly subjective question!
Personally, I just go with the pressures of what the medium will bear. For example, with restricted data-rate signals, if lopping off above 15k can be proven to improve the sound below 15k for most people, I'm all for it. OTOH, if the data bandwidth is available to expand the range, I'm all for that, too. SACDs can sound simply monumental.

But really, it's a very good case of "YMMV".

Kind Regards,
David

Help me out here, "YMMV" stands for....... ???

 

[TomT]

"your mileage may vary"..."all rebates to the dealer, annual apr 24% plus one arm and one leg; the representations of your vehicle salesman are not to be taken literally, and are subject to revocation/revival/ or recantation at any time"...