Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[hrogler]

hi everybody,

i'm working on my diploma - a radio play in 5.1. this is my first 5.1 production at all and by now i've experienced that when i try to pan a sound in 5.1 panorama, some diffusion occurs. (sorry for my english - my mother tongue is german.)

i'm not talking about panning "around" but "through" the room. which means: e.g. i pan a sound from Rs directly to L channel (NOT via the R and C channel or via the Ls channel). in this case it seems as the sound is not really locatable (which seems a bit confusing to me).

i thought it might be a phase problem. in addition the human ear is designed to localize sounds from the front better than sounds from behind. (also the surround speakers (Ls and Rs) have such a wide angle - might be better with 6.1 or even 7.1. but i don't have the facilities for those...)

has anyone of you experienced something similar?

thx,
herwig

[Loco]

hrogler wrote on Thu, 06 May 2004 06:50

i'm not talking about panning "around" but "through" the room. which means: e.g. i pan a sound from Rs directly to L channel (NOT via the R and C channel or via the Ls channel). in this case it seems as the sound is not really locatable (which seems a bit confusing to me).

You are trying to get the sound from your right wing to your left mirror through your head. That's a major thing. You need to use all of the speakers, just as if you were moving the sound inside an array of 5 microphones. that will help a lot your brain on the localization of the source. The other way would be using visual cues, but I guess you don't have that benefit.

There are surround panners based on reverberation algorytms that are specially good at this. Very DSP intensive, though. What DAW are you using?

[hrogler]

That's exactly what I was trying to do: moving the sound through my head. Thanks for the advice. I gonna try this.

I'm working on a rather poor DAW: P4 1,7 GHz 512MB with a MOTU 2408 on Cubase SX 2.01. I don't have any additional surround plug-ins than those that came with SX2.
Don't you know some free surround plug-ins that are of at least medium quality? (I don't have no budget...)

By the way: what are "visual cues"? I've never heard it.

p.s.: in addition I will use some EQs that open up when the sound gets closer and again decrease bass end high fqs when the sound's moving away. I think that will do a lot, too.

[Niels Elemans]

Well that's quite fortuitous. I'm a student Audio Design in The Netherlands and I'm studying 'placement in surround' for my graduation-project.

I've learned from an exprement from one of my teachers that when you try to pan something from front to rear with a standard (amplitude)pan you won't get the effect from movement. You just hear the front-speaker's level going down and the rear's speaker level going up. You should try using panning by time difference.

The ear can locate sources by level difference (ILD) and time difference (ITD). When you put a source on two speakers with the same amplitude and delay one speaker between 0 and 1.5 ms you'll notice that the source moves to the un-delayed speaker. When you do this with a front and a rear speaker this works much better than panning with level difference (standard pan-bottom).

I did some experiments with it on a front speaker and rear speaker and it works quite alright. This is probably because you hear both speakers all the time. When one is louder than the other, the louder one will mask the less louder.

I haven't tried it with crossing from LF to RS or RF to LS but I'll guess it should be possible this way. Also, try to use the other speakers to, it will can make the localization even better.

[hrogler]

very interesting. especially your project for graduation. are you focusing on music or post production?

i'm asking this, because i try to compare radio play production with post production and with music production (in 5.1, of course). what do you think about the usage of the LFE and the C-channel in music productions?

do you know the "rule of the first wavefront"? localization is not influenced by reflections that come to the listeners ear right after the direct sound - the "rule of the first wavefront" says, that these reflections even can be up to 10 dB louder than the direct sound as long as they are delayed for 5 to 50ms. localization will be uninfluenced.

this might be very useful for the application of 5.1 placement.

[Niels Elemans]

Quote:

very interesting. especially your project for graduation. are you focusing on music or post production?

I'm mainly focused on music production.

Quote:

i'm asking this, because i try to compare radio play production with post production and with music production (in 5.1, of course). what do you think about the usage of the LFE and the C-channel in music productions?

For music I tend not to use the LFE, only for SFX maybe. I let the bassmanagment do his work for feeding the subwoofer.

The center-channel always is a hot point of discussion. I feel when I put something (e.g.vocals) alone on the center-channel it doesn't really glue with the rest of the soundfield. There's just not enough coherence. But, the center can help to anchor something to the center. I do use the center in combination with the LF and RF. Then I put the center image usually about 3-10 dB lower than the LF an RF. I also the center to pan something between LF and RF. E.g. when you want to pan something halfway left, you can just use LF and RF, but you can also use C and LF; if you put an equal level on LF and C you get a phantom image halfway C and LF. Then, I can also use the RF for some divergence maybe.

Quote:

do you know the "rule of the first wavefront"? localization is not influenced by reflections that come to the listeners ear right after the direct sound - the "rule of the first wavefront" says, that these reflections even can be up to 10 dB louder than the direct sound as long as they are delayed for 5 to 50ms. localization will be uninfluenced.

That is the Haas-effect. What I use is 'phantomlocalization' which is working with delay shorter than 5 ms. When panning something to hard to one side 1.5 ms delay on the other speaker would be enough.

Quote:

this might be very useful for the application of 5.1 placement.

I don't know yet. How do you think that could be useful? Do you know the theories from Madsen?

[Loco]

hrogler wrote on Thu, 06 May 2004 16:54

By the way: what are "visual cues"? I've never heard it.

I you cna actually see the airplane going at you and go over your head, your brain will match the sound of the plane to the movement of the plane, even if it's just a mono source. Of course, moving the sound around makes a more dramatic effect.

[hrogler]

I don't know Madsens theories. Go ahead, I'm interested...

About the 1.5ms delay, etc. I experimented a lot know with placment based on time difference. The tricky thing is especially the animation of the soundsource, not only the placement. I tried to get the movement by using timestreching so I could swap the delay continuously between the different channels. A lot of effort if you don't have a special plugin for this purpose.

I have experienced that up to 5ms difference is much better than just 1.5. At 1.5ms you've got a major problem: in 1.5ms sound moves about 34cm. so if you're not exactly on the sweetspot the time difference may be destroyed. So, at least 3ms should be used. (In addition: even if you're sitting exactly on the sweet spot - IMHO - the localization is a little bit too low at 1.5ms difference.)

I also tried to combine loudness and time difference which worked out very fine. I used up to 10dB difference which increased the effect of movement but still produced a definite SURROUND sound and I had no problems with diffusion at all.

By the way: do you know any surround-pan-plugins that are based on time difference (or combine time and level difference)? This is very important for me to know for my diploma.

Best,
Herwig

[Loco]

hrogler wrote on Wed, 19 May 2004 06:29

By the way: do you know any surround-pan-plugins that are based on time difference (or combine time and level difference)? This is very important for me to know for my diploma.

Some panners from Kind of Loud had that thing. also, there's an excellent algorythm on TC system 6000. However, I use the built-in panner on DP4 more.

[hrogler]

Thanks for the hint. Just downloaded the manual for the TC System 6000 for further research. Do you know the URL of King of Loud? I can't find it. Or do you know the product name of a specific King-of-Loud-panner (for 5.1 of course)?

Herwig

[hrogler]

Right, Kind of Loud, not King of Loud... that's why I didn't find any homepage. But neverthelesse, the name of the company is Universal Audio.