The Orban Optimod (Austereo)

[SurveillanceP]

Hi guys,

Issue 66 of Audio Technology is finally out for all you Aussie's out there- Hooray!

I'm intrigued by the article about the new Austereo setup, in particular the use of the Orban Optimods.

I'm a bedroom kid musician, though I do sometimes get my tracks played on the australian radio statoin jjj.

The engineer interviewed explains how insanely configurable these things are, so I understand that there is no cure all "prep" approach to getting a radio mix ready that can emerge unscathed from them. He also mentions how loooong it takes to tune them, tiny changes over months.

General specs for Optimod

Pauly mentions in this thread how he wouldn't have given the details of his settings out willy nilly and I can see why that makes sense.

He also says this:

Always keep in mind that radio guys are compressing and limiting a lot..the less you do, the better it sounds on the air...generally that is

Which is congruent with what I thought was true about use of compression, it being multiplicative and all.

However in the Audio Technology article (I'll try to quote as little of it as possible) the station engineer has this to say:

"The Optimods are so brutal on-air that they'll just snap down any transients they see, so we have to be very careful to control any extraneous transients before they reach the on-air limiter."

and,

"All those Harvey Norman ads stick out like they do because they're slammed to the hilt before the signal gets to the Optimods"

I'm confused and there is alot here I obviously don't understand. From what the austereo guy is saying though, it seems that sending an arty layered song with detailed and varied dynamics would result in a train wreck.

Or- Slam it before radio does, and it will only get slammed twice over and still be a train wreck.

I've read words by big wigs I respect, some saying clever limiting, even individual stem limiting, can make a mix emerge fairly unscathed through these monsters, others saying that this is a myth and you should limit as little as you possibly can because the optimods are going to have their way with it anyway. I don't think I've ever read of anyone suggesting a compromise middle road type solution...

Any wisdom? I'm interested in thought about radio prep, and also any more information on these optimobeast thingies lol.

And now I'll slink back over to the low end forums where I belong


P.S.stikestikestike
(Plug: Buy Audio Technology- its sometimes late but always great)

[Coyoteous]

Austereo = austere + stereo?

- not the friendliest clever hybrid word out there, is it?

- interesting info, though...

[Jules]

I can only throw this into the pot..

voice overs - are different from music

What might make a voiceover ad or a 12 second 'jingle' REALLY loud - might sound disgusting on a 3 min piece of music.

I have had tracks of mine on the radio 'dip down' on the choruses.. not good......

Interesting topic..

Carry on

P.S. I know of an industrial music drummer who can get an fantastic drum sound using just one stereo PZM mic crushed to death with an old Orban Optimod. (But he is a very skilled drummer to start with)

[SurveillanceP]

haha yes I believe I read about this drummer in a related thread. Sounds like an expert at beating "anything" until it sounds good be it a skin or a machine!

[SurveillanceP]

edit: I see this thread is getting lots of views but not so many replies, understandably I guess seeing as radio broadcasting is kind of a niche of the wide audio world.

But even if you haven't slept in a bed next to the Optomegatron for 40 years I'm still keen to read your words lol.

[Tom H]

It's like you said; Optimods can do so many things, it's just the way you set them up. But most stations I know do heavy processing... wich means the loudest stuff wins.

I know some jocks rely a lot on that processing while doing their shows. Overruling records with voices and jingles; that's the way programs are glued together. My own experience spinning records as a DJ on air is that it can actually sound pretty cool to hear the processing work while mixing.

In a relation to mixing and mastering; well mixed and good mastered records should have less side effects due to radio processing. So I guess yes, this means less dynamics in mixes for the big radiostations. I know this sounds like blasphemy but I think that's the way a lot of A&R guys think.

[SurveillanceP]

Update: For curious thread viewers- I've found a Ton of further information on radio broadcasting kindly posted by Thomas. W. Bethe on a nearby thread:

Radio mastering

Looks like I have many happy reading hours ahead

[Bob Olhsson]

Some of the loudest audio you'll hear on radio and television is the totally unprocessed mike feed of an announcer who has great mike technique. On many stations the DJ is louder than the music for this reason.

The idea that you can make broadcast audio louder by limiting and compressing it is simply not true, at least according to Bob Orban and Frank Foti.

[Ben F]

I can't listen to JJJ any more, their compression is just too full on for my ears. In my experience the most slammed masters nearly always sound worse on JJJ.

[Coyoteous]

Originally Posted by Ben F

I can't listen to JJJ any more, their compression is just too full on for my ears. In my experience the most slammed masters nearly always sound worse on JJJ.

- sounds like they're living up to the "austere" part of the name then, at least in the "severe" sense.

[SurveillanceP]

Ben F, I had always thought that compared to its competitors it was fairly conservative- relatively speaking of course... Do you think thats no longer the case, or indeed never was?

[Ben F]

Originally Posted by SurveillanceP

Ben F, I had always thought that compared to its competitors it was fairly conservative- relatively speaking of course... Do you think thats no longer the case, or indeed never was?

No it's slammed just like the rest. I've watched the Optimod in action at the ABC, it works better with music that isn't slammed (some dynamic range), at least there is space to breath. I've always found the release on the compression is set up for dialogue (slow) rather than fast for music, hence the pumping affect often heard with faster indie rock.

There is also a lot of multiband 'enhancing' give the output width and more top end.

[Led]

Originally Posted by Coyoteous

- sounds like they're living up to the "austere" part of the name then, at least in the "severe" sense.

Not that it really matters, but just for info's sake, it's a merging of the 2 words - "Australian" and "Stereo". It's a national network.

[Coyoteous]

Originally Posted by Led

Not that it really matters, but just for info's sake, it's a merging of the 2 words - "Australian" and "Stereo". It's a national network.

Oh, I got that too... guess the austere reference just wasn't enough override someone's cutesy wordplay - not to belabor my lack of a real contribution to this thread.

[SurveillanceP]

Originally Posted by Bob Olhsson

Some of the loudest audio you'll hear on radio and television is the totally unprocessed mike feed of an announcer who has great mike technique. On many stations the DJ is louder than the music for this reason.

The idea that you can make broadcast audio louder by limiting and compressing it is simply not true, at least according to Bob Orban and Frank Foti.

Aha yes Bob, I'm doing some of that reading now- "THE VOICE OF GOD" lol.

If no one minds I'm going to keep reading this mind bending stuff (so dense...) then post back here when I have got some tentative conclusions to see what people make of them!

[SurveillanceP]

Here is my current understanding based on those articles------

http://www.orban.com/support/orban/t..._Truth_1.3.pdf


1. Radio systems cannot be tuned for hypercompressed modern music because a) they also have to player older music and b) There is now known way to tune for degraded material anyhow.

2. Radio systems react to clipped material by exaggerating the clipping, adding no noise- just "grunge". (Is that the new "vinyl" sound of the new generation or just plain bad, make of it what you will....)

Additional thought- The "Look Ahead" Limiter function has basically got to be turned on to avoid this, or you have to be watching the meters closely.

3. Having lots of energy above 5Khz is in the final master is asking for trouble, because the pre-emphasis/HF limit process boosts it alot then squashes it all back, leaving the high end "dull, distorted or both". Note: is this then an area of critical difference between a CD for home and a radio edit?

4. Slow Compression is fine- Fast is very bad. Leave this to the Station.

5. There will be a high pass filter. The figure of 15k was mentioned. I imagine that stations tweak this themselves, and that the more aggressively loud stations pull it down the most. Preset slammed top end "sparkle" in Tracks isn't going to do the song any favours on air. both for this reason and for that made in point 3.

Thats all I can get right now. Any thoughts? I realise the knowledgeable really do have better things to do with their time, but really, if Bob Olhsson is willing to reply to a schmo self educating on the net, this sites got to be pretty special hey?

P.S

Hey Ben F, in regard to Triple J being slammed just as hard- I think I might have been fooled by the lack of offensive Harvey Norman ads and bleepy chirpy station tags?

[Adam Dempsey]

Originally Posted by Ben F

I can't listen to JJJ any more, their compression is just too full on for my ears. In my experience the most slammed masters nearly always sound worse on JJJ.

Their needless commercialism and repetition too much for mine. And is where I compared Ramstein in full flight vs Men at Work on another station; the latter sounded so much louder, fatter, punchier. At least the J's 'smiley' EQ curve and over-hyped highs seem less than it once was.
Two other non commercial stations here, apparently with same pre-transmitter signal chain, sound vastly different to each other.. one like a lesser bandwidth cassette tape in comparison.

Or- Slam it before radio does, and it will only get slammed twice over and still be a train wreck.

Exactly. And the gain reduction is the product of - not the sum of - the two.
And from the pic in the article, the low bands are getting 6dB compression (way more than any half decent mastering would apply) and 'comp' signal is metering at 100%...
(nobody ever said ME's can't also have an eye for detail..).

[DominicWyeth]

I shudder when I think about the austereo broadcast signal path and the guys who use to set it and run it. Set for numbers not music. I use to work in OB's for Brisbane austereo - the production guys were cool but i worked for the techy guys...and they were all that tech's not music techs.

cool guys and fun company don't get me wrong - but its about repeatability, conformity and sales...not music.

[Adam Dempsey]

Originally Posted by DominicWyeth

its about repeatability, conformity and sales...not music.

Yes, these entities are businesses.. they exist to earn income from advertising. The content is the filler.
How that applies to Triple J - a taxpayer-funded station that's been known to take out full page ads in street press to flog particular albums and tours - I'll never understand.

[turtlerock]

the biggest drama we all have with radio hasnt changed much in history of radio limiters ie
if you have a quiet intro and somewhere in the first 30-secs to one minute of a song you have a loud section that comes in as a dynamic hit ( like a big chorus or a guitar stab.. drum hit etc ) - well radio has always had trouble as the compresser grabs the quiet intro and turns it up to stop "dead air" and when the hit comes in things get "turned down" and keep down for anywhere from 2 secs to 30 secs .

if you in the oz radio part of the world listen to "run to paradise" by the choirboys next time its on your radio , its a classic example of what not to do !
-err... i mastered that track and in its day i made a specially mastered radio version which sounded great an made it a radio hit , nowadays they play the album version ( which i also did but was only for the album as i knew radio would slay it and it just kills me every time i hear it:(

if you not here in oz near a radio , carry on folks nothing that will make sense to you particular with that reference track
i only mentioned it because its still on the radio 20 years later and sounds worse then it used to but
maybe somebody can post there own that will be more unlocalised

[SurveillanceP]

[SurveillanceP]

ech too newb to post photo's

Here is an attatchment of a snippet of a Josh Pyke song, recorded off Triple J.

Two things that interest me about the graphic are;

-The filter at around 15khz looks like it means serious business.

-the (inaudible?) line of energy at around 19khz. I'm sure I've read an explaination for this but can't remember it just now.

I also recorded samples from a local Austereo station, and both the observations above were present there too.

--------------

Rick,
Would you mind elaborating on how your approach differed between mastering "Run to Paradise" for the album and for the radio? I think I can guess the general idea but I'd love to here more

[turtlerock]

probably not the right place to talk about the specifics of a local 20 year old radio hit
i just mentioned it so you would understand the real problem with radio limiters
and to save you getting caught up reinventing the wheel

[SurveillanceP]

Ah yes, fair point. Thanks for the info

[dcollins]

Originally Posted by SurveillanceP

Here is an attatchment of a snippet of a Josh Pyke song, recorded off Triple J.

Two things that interest me about the graphic are;

-The filter at around 15khz looks like it means serious business.

-the (inaudible?) line of energy at around 19khz. I'm sure I've read an explaination for this but can't remember it just now.

FM broadcast has a sharp filter at 15kHz and the 19kHz signal is the "pilot tone" that indicates a stereo signal.


DC

[SurveillanceP]

OK I am obviously a newb that doesn't understand what you mean by a "pilot tone" at 19k. Care to elaborate?

In all fairness I did read something about this recently but didn't really come to grips with it

[Jesse Graffam]

Here's a hot tip, since I have NEVER seen this mentioned "out loud" anywhere online yet. And this only really works well for Optimods. It does relate somewhat to other broadcast processors depending on how they are tuned... but is especially relevant to the Optimods (which are not the #1 processor in the world, Omnias are)

So uhm yeah... check out Orban's free Loudness Meter, here:
ORBAN Loudness Meter

Match the integration time of the CBS and the ITU BS.1770 meters, and the "test" loudness, here's my settings:
http://img165.imageshack.us/img165/7...9orbanloud.png
And Bob Orban agrees on the CBS Loudness Gain in order to match the ITU levels, on the average program material. I asked him myself.

That being said...

Optimods use the CBS Loudness algorithm to drive their AGCs, which almost completely effects how they set the average levels. The Orban Loudness Meter is an implimentation of the latest CBS Loudness algorithm, as hear in the Optimod 8500.

ITU BS.1770 (aka LKFS) is widely considered to be one of the most accurate subjective loudness algorithms, and despite it's relative simplicity it's scores in testing are only beaten by one much more complicated psychoacoustic algorithm by TC Electronics.

The CBS one you can judge for yourself, but in my experience it isn't anywhere near as accurate as LKFS, except in some cases with some speech-old content.

That being said...


You can use this software to tell how an Optimod is going to judge it's loudness (relative to the settings of course) and set the AGC accordingly, in comparison to LKFS which is very close to it's actual loudness.

It's by all means not fool-proof, no noooo. But it should give you a better idea of how the algorithm behind the Optimod's averaging works, and more importantly what kinds of sounds, dynamics, and frequencies it consideres louder or not.

There's more information online on how CBS Loudness algorithm works, as well as patents, if you really want to geek yourself out on the details.

Cheers thumbsup

[dcollins]

Originally Posted by SurveillanceP

OK I am obviously a newb that doesn't understand what you mean by a "pilot tone" at 19k. Care to elaborate?

In all fairness I did read something about this recently but didn't really come to grips with it

The 19k pilot tone inducates it's a stereo broadcast so your tuner will know to decode it as such.


DC

[SurveillanceP]

Originally Posted by uncajesse

Here's a hot tip, since I have NEVER seen this mentioned "out loud" anywhere online yet. And this only really works well for Optimods. It does relate somewhat to other broadcast processors depending on how they are tuned... but is especially relevant to the Optimods (which are not the #1 processor in the world, Omnias are)

So uhm yeah... check out Orban's free Loudness Meter, here:
ORBAN Loudness Meter

Match the integration time of the CBS and the ITU BS.1770 meters, and the "test" loudness, here's my settings:
http://img165.imageshack.us/img165/7...9orbanloud.png
And Bob Orban agrees on the CBS Loudness Gain in order to match the ITU levels, on the average program material. I asked him myself.

That being said...

Optimods use the CBS Loudness algorithm to drive their AGCs, which almost completely effects how they set the average levels. The Orban Loudness Meter is an implimentation of the latest CBS Loudness algorithm, as hear in the Optimod 8500.

ITU BS.1770 (aka LKFS) is widely considered to be one of the most accurate subjective loudness algorithms, and despite it's relative simplicity it's scores in testing are only beaten by one much more complicated psychoacoustic algorithm by TC Electronics.

The CBS one you can judge for yourself, but in my experience it isn't anywhere near as accurate as LKFS, except in some cases with some speech-old content.

That being said...


You can use this software to tell how an Optimod is going to judge it's loudness (relative to the settings of course) and set the AGC accordingly, in comparison to LKFS which is very close to it's actual loudness.

It's by all means not fool-proof, no noooo. But it should give you a better idea of how the algorithm behind the Optimod's averaging works, and more importantly what kinds of sounds, dynamics, and frequencies it consideres louder or not.

There's more information online on how CBS Loudness algorithm works, as well as patents, if you really want to geek yourself out on the details.

Cheers thumbsup

Totally awesome. Thanks uncajesse!

Dcollins- Thanks for the info on the pilot tone. I had seen that from taping radio and looking at the frequency display- going "wtf is that thing doing there?"

[Jesse Graffam]

Thought I could add more to this topic.

Probably the main thing in original recordings that I'm hearing that causes a broadcast audio processor to do this "quiet on the chorus" business, is when you're already hitting into the limiters hard enough so that when a section of music comes in with a louder midrange... the compression/limiting that you used on your track has to work harder to keep the levels, and you end up losing the average levels of your lows & highs.

Compression of buses in mixing or mastering compression makes this more apparent, because you're creating louder transients which do hit the multiband compression of the broadcast audio processor differently at higher volumes, even if the preset (or user preset) it's running is using Inf:1 ratio there. Plus you are very likely limiting those compressed verses a bit which is increasing the average level without increasing the peak.

Then along comes your chorus which requires your compression and/or limiting to reduce the average levels of your bass & treble to allow the midrange to fit. The peak level reduction of the same frequency areas often does not get reduced linear fashion to the average. And there-in lies the problem.

If it the average levels of bass and treble are getting reduced, and the crest stayed the same (in those frequency areas), almost all broadcast audio processor setups can easily deal with that, and will correct for that pretty darn fast. It's when the crest (the difference between average and peak) at certain frequency areas change, when the sound of your track will change on-air will change regardless (at this point in time) of what processing is being used, to some extent, and your track will leave the goal the station's processing consultant (if they are smart enough to use one, nudge wink) had in mind.

Another important thing to remember is that actual signal level & crest in a frequency range, and it's perceived loudness & crest... are not totally the same thing. As much as you can "trick" human hearing, there are ways to "trick" broadcast audio processing to an extent as well.

But like pauly, I'm not going to start giving out too many trade secrets that I've developed over the last decade of coming to understand broadcast audio processing code, people in the scene, etc... and how it related to my mastering.

Hopefully I pointed someone in the right direction.