|  IP: Logged

"Then there's the question of ganging the channels (most pro units are designed to be ganged and have a loopthru option -- that wouldn't be a problem). BUT, ganging presents its own set of very audible artifacts -- namely, you don;t necessarily want ALL the channels to duck just because one has reached the threshold. That is a bad thing even in a two channel stereo system; I contend it could REALLY sound disturbing in a 6 channel system -- get a loud explosion in the left channel and the rest of the fronts channel and the surrounds drop 10 dB. for a second. That kind of pumping will be quite noticeable. To avoid this, using multiple stereo units would have the advantage that you could gang whichever channels you choose -- for example, the system could be configured so that this kind of "everybody duck" artifact is minimized -- if one stereo unit is set up to limit L&R independent of the other channels and another for the Surrounds and a third for the center channel and subs (this one can be decoupled so that the stereo channels are running independently), you would get better sonic results."

Ganging of compressor channels in a stereo system is standard practice throughout all facets of the audio industry for one very, very good reason. Almost all stereo sound tracks (BTW "stereo" means multi-channel, not just two-channel) make use of what is termed a "sound field". A sound field means that each channel does not only contain it's own signal, but also varying proportions of signals which are also present in adjacent channels. This creates the "stereo effect" where you are convinced that sounds are not only coming from each individual speaker but also from the area in between.

Now if a multi-channel limiter/compressor is placed in the signal path without ganging the channels together, the sound field floats back and forth sideways between speakers, whenever the compression threshold is exceeded more in any one channel than any other. A typical example is a two-channel stereo system in which a vocalist is panned mid-way between left and right (as in probably 90% of all popular music). Whenever one channel compresses more than the other, the vocal shifts sideways towards the channel with the least compression. The result is that the vocal floats back and forth sideways and depending on the severity of the compression, the effect can be serously annoying. Ganging the compresors together ensures that the sound field remains positionally stable, regardless of how unbalanced the dynamics of the soundtrack may be.

A properly designed and setup limiter/compressor will not produce "ducking" as Frank calls it. This is a characteristic of cheap AGC circuits that are not used in modern professional audio systems. In a 6-channel system, a loud explosion in one channel which causes the limiter/compressor to compress will indeed cause compression of all the other channels if they are ganged, however a signal which is loud enough to cause significant compression in only one channel will also be loud enough to mask the compression effects in the other channels. (Psycho-acoustic masking is the basic principle upon which the various Dolby processes are based, from Dolby A through to Dolby Digital).

Furthermore, in a properly designed and setup multi-channel limiter/compressor, the compression caused by that explosion will follow the envelope of the explosion closely, in fact as closely as attack and decay times will allow whilst keeping distorion inaudible. This results in an inverse envelop or "hole" in the other channels that closely matches the envelop of the explosion and this actually aids the masking process. Anyone who doubts this effect only need to listen to any Dolby soundtrack where acoustic masking is constantly busy at work hiding all manner of horrors from your tender ears.

If ganging was not employed, the soundfield would shift away from the channel with the explosion and then suck back again as the explosion subsides. (Recall that classic footage of the nuclear blast as it first blows the house in one direction and then in the other).

Of course if you go out and buy a whole bunch of studio compressor/limiters as Frank suggests, then you are bound to end up with horrendous results, unless you are an audio professional and know how to set up and optimise the many variable parameters these devices offer. So the Smart device has already been setup and optimised for you and provides a useful degree of dynamic range compression with minimal audible artefacts.

BTW I am not advocating (or otherwise) the use of compression in cinemas.

------------------
Ray Derrick
President/Chief Engineer
Panalogic Corporation Sydney, Australia
Phone: 61 (0)2 9894 6655 Fax: 61 (0)2 9894 6935

You are partially right about ganging compression channels -- depending on the program material, the parameter settings, especially the way the compressor attacks and releases the signal, you certainly can create very audible, undesirable artifacts, specifically the distortion of the stereo soundstage. But these can be just as bad with ganged compression. With regard to film sound, I was talking about threshold limiting rather than linear compression. Soft threshold limiting, correctly and minimally set can be applied to non-ganged channels without causing noticeable artifacts, again, depending on the program material.

In recording sessions I have used this type of limiting for the sole purpose of preventing tape saturation. Done judiciously, it is very effective and practically transparent with no audible shift in the stereo soundfield. In fact, I used it on an album that I engineered, RUBBER NICKELS (which won the Phoenix Award in 1975 for Best Quadraphonic Recording of an Original Musical). ANY kind of compression, ganged or non-ganged, would be most apparent on matrixed (4-2-4) material as was this album. The trick is to use your ears. Compression of any kind is always a dubious, although sometimes necessary, distortion of the original signal; a good engineer uses his ears as the ultimate tool to determine what parameters and what mode of compression works best and least intrusive for any particular sound source.

Personally, for film sound, I would not recommend ANY tampering with the soundtrack -- whatever signal processing needs to be done on the tracts should have already been done in the mix. But if there HAS to be additionally processing because of shortcomings in the playback system, as you point out and as I agreed, the Smart single button would be the best bet.

"With regard to film sound, I was talking about threshold limiting rather than linear compression"

So was I.

"Soft threshold limiting, correctly and minimally set can be applied to non-ganged channels without causing noticeable artifacts, again, depending on the program material".

It would have to be VERY minimally set. Maybe the album you mixed was very mono heavy, or had exceptionally even dynamics, or maybe you were just lucky, who knows, but you cannot raise a single specific instance and apply it as a generalisation. The fact is you cannot change the laws of physics, and reducing gain in one channel MUST shift the position of the soundstage. If the reduction is subtle then the shift is subtle, but there is a shift none the less.

"The trick is to use your ears".

Using your ears is not a trick, just common sense.

"Compression of any kind is always a dubious, although sometimes necessary, distortion of the original signal;"

If compression is dubious distortion then so is equalisation, so do we remove all the equalisers and dynamics from mixing consoles?

"a good engineer uses his ears as the ultimate tool to determine what parameters and what mode of compression works best and least intrusive for any particular sound source".

Couldn't agree more. But good sound engineers are not normally employed in cinemas.

"Personally, for film sound, I would not recommend ANY tampering with the soundtrack --"

Frank I would agree with you, except when your patrons start walking out of the theatre en-masse and ask for a refund because the loud parts of the movie are deafening, you do have to start to question the director's judgement. There are no doubt times and circumstances when the "judicial" use of compression in a cinema may be warranted, particularly when your continued viability depends on it. Under those circumstances it is important to chose a system that is discrete and as free from artefacts as possible. To simply say "don't do anything because that is what the director intended" is like saying the director also intended to put my cinema out of business because no normal human being can stand the loud parts of this film.

Somewhere along the line you may have to decide if playing a film the way the director intended is more important than going broke. And if you do have to modify the performance, then do it correctly and discretely with a product properly designed for the purpose. That is all I am saying.

------------------
Ray Derrick
President/Chief Engineer
Panalogic Corporation Sydney, Australia
Phone: 61 (0)2 9894 6655 Fax: 61 (0)2 9894 6935

Rolling stock ads (on video) -- loud.

MovieTunes to accompany slides -- loud.

Trailers -- loud. Some distortion in the trailer for Road to Perdition.

"Blade Runner" snipe (original version) -- loud with the dynamic range all but taken away.

Feature -- sounded perfect. Turns out the movie is heavy on dialogue (Sunshine State).

 |  IP: Logged

Cheers!

------------------
Greg Mueller
Amateur Astronomer, Machinist, Filmnut http://www.muellersatomics.com/

 |  IP: Logged

With the afterburners on, I never had a single problem. Not one. I never noticed a sound rolloff of any kind, and my ears never bled.

I give kudos to smart on this particular product.

I am very pleased that you are having good results with the Afterburner. You said, "There is no sign of volume riding or other objectionable byproducts, in fact, it is quite transparent in operation." This has been the general experience with this product from many customers.

As you have noted, yours is in line between the DTS and the CP. This is the original intention for the Afterburner, to be used on digital soundtracks. It is now more commonly placed between the CP and amps so that ALL material, whether digital or optical (especially loud trailers) can pass through it.

Thanks also to Dave and Gordon for the very nice comments.

------------------
Oscar Neundorfer
Chief Engineer
SMART Devices, Inc.

 |  IP: Logged

Since this is a topic on limiting and compressing, i will get back on topic.

www.behringer.com
When feeding audio to a radio transmitter, there is a volume level "ceiling", over which you either can't (in the case of AM) or shouldn't (for FM) exceed. The problem with an improved drive-in sound system is that now you have a much greater dynamic range to your sound. Things like heavy bass content and hot passages (like explosions) can either cause your transmitters to distort OR force a single-channel compressor/limiter punch huge holes in your sound. The dynamics (the difference between the loudest and quietest sounds your system will pass) may also be too wide for car radio listening... causing your customers to turn their radios up when things are quiet, then back down when "the dinosaur roars"! FM transmission poses an additional challenge, in that high frequencies are designed to pass through the system at higher levels than mid & low frequencies. Many theatres address these issues by using single band compressor/limiters. These are certainly much better than nothing. However, a single-band unit may add its own audible artifacts to the sound you're sending to your customers. Consider that the operation of a single band processor is very much like manually adjusting your volume. Imagine that you have your hand on the master volume control knob in your booth. You've turned the volume up because the actors are whispering at each other. All of a sudden, here come the hot bass beats of a rap song. Since you absolutely can not allow your volume to go above a certain level you would have to (quickly) turn your volume control down far enough to let the bass pass at or below the 100% level. Unfortunately, you just turned everything else down as well. This is basically what a single-band compressor is... a single automatic volume control. They can be effective, but can also punch big enough holes in your sound that your customers CAN hear!

Most modern radio stations have moved away from single band processors for much the same reasons described above. In their place has emerged the Multiband Processor. These devices are something like taking a multiband equalizer and putting a compressor on each band. If a hot bass beat goes through one of these systems, the channel that handles low frequencies reduces the bass while allowing the mid and upper bands to continue at their own preset levels. In effect, instead of one automatic volume control over the whole audio spectrum (bass to treble), these multiband processors chop the spectrum into a number of channels (6). Each channel is "in charge" of a portion of the audio band. It's something like automatic bass and treble controls on your stereo, except in this case we have 6 controls instead of two. It is a bit technical to set up, but is extremely stable & can be left alone, once you have it sounding the way you like.

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

http://www.autoramadrivein.com