Stop audio but effect "rings out"?

[mikaelka...@gmail.com]

Certainly there is a clever way of doing this that eludes me...

I´m running a loop of a telephone ring. Just the one ring looping. When the actor puts his hand on top of the phone I want the ringing to stop. Simple enough with a Devamp.

Here´s the tricky part. I´ve added AUMatrixReverb for ambiance. And of course that ambiance should not die with the audio.

Solution?

If a use the fade curve for the phone sound, the reverb fade as well. Same thing if I use a fade cue.

The parameter I actually would like to fade would be the audio input level feeding the AUMatrixReverb.  

To me the scope of this issue would include most use of ambience effects.

Best regards, Micke

[Christopher Ashworth]

Hi Micke,

The stop cue will stop the cue but not the tail of the effect.

-C

[Jeremy Lee]

Use a reverb plugin on the device matrix and use it as a send.

Jeremy Lee

- A thumb is a terrible speller. Please forgive my trespasses.

On Aug 5, 2013, at 4:31 PM, mikaelka...@gmail.com wrote:

--
--
Change your preferences or unsubscribe here:
http://groups.google.com/group/qlab
 
Follow Figure 53 on Twitter: http://twitter.com/Figure53
 
---
You received this message because you are subscribed to the Google Groups "QLab" group.
To unsubscribe from this group and stop receiving emails from it, send an email to qlab+uns...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[John Leonard]

No, the right way to do this is to use an effect of the tail of the phone ring and do a near-instant cross-fade when the phone is picked up. It takes a bit of experimentation to get it right, but it's much better than using reverb.

Been doing this for years and years and years...

[John Leonard]

And apply the effect to the 'ting' as well.

[mikaelka...@gmail.com]

Wow, it´s that easy?! Almost to obvious! =)

Thanks! /Micke

[mikaelka...@gmail.com]

That would solve things. Or to just mute the sound in the audio file to get the cut-effect. But for me, the whole point of adding effects in Qlab, as supposed to in the audio file, or even the DigiCo console, is simplicity.

[Rich Walsh]

Simplicity is not the same as accuracy though: and it will pull your audience out of the moment. Bell phones ring out when you pick them up, even mid ring. The echo of a bell is not the same thing as the decay of a bell (unless you are convolving through an IR of the bell…).

Rich

[Jeremy Lee]

I agree if naturalism is key. I thought he was going for something different...

Jeremy Lee
- A thumb is a terrible speller. Please forgive my trespasses.

[John Leonard]

Oh, very probably. Still a neat trick, though.

John

[micpool]

Of course if naturalism is the key one would actually make the phone bell ring.

The complication with recorded phone rings  with tail samples is if the phone is picked up between rings. it's very difficult to come up with a technique that work with all scenarios particularly with a dry sounding bell. The crossfade might be required during the full ring, at any point during the decay or in the silent part of the gap between rings. There is probably a complex method that can be devised in Qlab to track the timing of the main phone ring cue and arm and disarm alternative tail cues depending when the phone is picked up. 

I'll outline a complex method of ringing a phone bell from Qlab and making the actual  bell much louder from a speaker placed at the same position as the phone in a separate post.

Mic

[micpool]

Here is my phone ringing technique. This is a description for a UK phone so the actual frequencies, voltages and timings will vary depending where you are in the world

WARNING: The voltages used to ring telephone bells are quite large and need to be treated with the same respect as mains electricity!

Record an audio cue consisting of a  25Hz tone at around -10 in the required cadence of your choice. (UK 400mson/200ms/off/400ms on/2000ms off)

Play  this from Qlab to an amplifier with a 4 ohm output

Connect a 70v/100v line transformer between your amp and the telephone

In the UK Canford do a lovely toroidal which is a CLOUD CXL-100T 100V line driving transformer 4ohm 100W (UK canford part 23-491).

Here's another advantage of ringing a phone from Qlab using this method.

Inevitably your phone ring will be too quiet. You will then mic it up, the mic will be too close for a natural sound and you won't be able to place the speaker right next to the phone because of  feedback.

However if you make an audiophile recording of the actual phone you are going to use, rung off your 25Hz wav file, then you can play the 25Hz tone and your recording of the actual phone it is ringing, in sync. i.e the 25hz tone wav and the audio recording of the phone in a fire all group cue or as a 2 ch wav.

The audio can go as loud as you like and the speaker can be literally under the phone.

You will have to decide how you are going to stop the audio recording/ play a tail sample/ etc when the phone is lifted as the recording will of course continue when the bells cut on the phone microswitch. This can be as simple (operator visual) or as complicated (MIDI generated from microswitch) as you require.

Mic

[Jeremy Lee]

It's not really that hard to do in QLab. 

Jeremy Lee

- A thumb is a terrible speller. Please forgive my trespasses.

[micpool]

Er... what isn't. 

And whatever it is, that isn't, there are always refinements that can be developed using more complicated techniques.

Mic

[Andy Leviss Lang]

On Mon, Aug 5, 2013 at 6:05 PM, Rich Walsh <rich...@mac.com> wrote:

Simplicity is not the same as accuracy though: and it will pull your audience out of the moment. Bell phones ring out when you pick them up, even mid ring. The echo of a bell is not the same thing as the decay of a bell (unless you are convolving through an IR of the bell…).

As John Leonard, Rich, and Mic all touched on, for a phone bell to sound natural, the *ring* needs to decay, not an echo of a ring. What you need to do is slice a single ring into two parts, one with the "body" of the ring, one with the decay. Then, you need to conditionally vamp/devamp it, so that it loops Ring/Tail/Ring/Tail, and then when you take the cue on the actor picking it up, if it's in the middle of the body of the ring, it immediately jumps to the decay tail, but if it's already decaying, it just finishes the decay and then stops.

It's one of those easy in concept/tricky in execution things, but with some clever use of arm/disarm cues and devamping, it's very doable. Rich actually helped me solve this problem a couple years ago, here's a link to his illustrated post with an example, you just need to play with the timings to suit your particular ring sounds:

http://bit.ly/ringvamp

or

https://groups.google.com/forum/#!searchin/qlab/rich$20walsh$20andy$20leviss$20ring/qlab/U0jL97dmdcU/Ds90-vrlahcJ

Hope this helps,

Andy

[micpool]

That's a good method. For a UK bell perhaps it needs further refinement, which potentially adds considerable complexity, to deal with the fact that there are 2 decay portions in a single ring cycle (phone not wagner). 

If a US ring requires16 lines of programming, whats the minimum number of lines for a UK ring? Should we have a competition? Is there a Qlab3 method which is more compact?

And although "it's not really hard",  perhaps we should acknowledge that it's sort of intermediate level difficulty!

Mic

[Andy Leviss Lang]

On Tue, Aug 6, 2013 at 9:42 AM, micpool <m...@micpool.com> wrote:
>
> That's a good method. For a UK bell perhaps it needs further refinement, which potentially adds considerable complexity, to deal with the fact that there are 2 decay portions in a single ring cycle (phone not wagner).

--snip--

> And although "it's not really hard", perhaps we should acknowledge that it's sort of intermediate level difficulty!

I suppose that's a more elegant way to rephrase my "easy in
concept/tricky in execution" statement :-)

I'm at the shop right now, so can't quite sit down and wrap my head
around it at the moment, but theoretically it should be possible to
create a little standalone script applet that would let you enter the
cue numbers for the ring and the decay, or multiples thereof, and
would do all the heavy lifting to generate the sequence for you. If I
get a chance over the next few days, and nobody else is more motivated
to do it, I'll see if I can sort it out, but no promises. Between the
day job at the shop and working on a bunch of pending Duck's Echo
projects that are soooo close to finished, I can't guarantee to get to
it for a few days.

-Andy

[micpool]

I think  a UK ring  also might require 2 main samples, as you would need a sample consisting of the first of the 2 rings in the cycle that rang out fully, as if the phone had been picked up before the second ring in the cycle,

Mic

[Sean Dougall]

If you're in a hurry, and okay with each ring being slightly unnaturally consistent, you could do it with slices:

The audio is a single ring, edited down to remove all but a couple cycles of the hammer. If the sample has lots of high-end hammer noise, the repetition can get robotic, but I'll offer it as my entry into the competition. :-) Plus, a LPF might help mitigate that effect a bit.

And for a two-ring cycle, this seems to work:

The "second ring" cue is a duplicate of "first ring" -- same audio as the US ring version, but with fewer repetitions on the slice. It's not perfect, but a fun exercise at least.

Sean

[mikaelka...@gmail.com]

OK, everybody. I don´t mean to sound ungrateful, but we seem to have gone way of topic. Forget about the phone. Forget about "how to pick up a phone".

Let´s talk about ANY audio cue. I play a wave-file. To this wave-file I´ve added the AUMatrixReverb. This is how I want to do it. I could use an external reverb plug. I could use the reverb in the console. I could edit the wave-file. But I don´t want to.

For me, the whole idea is to do non-destructive playback, the simplest way possible.

If I mute the audio cue with a volume change, the Reverb mutes as well. There will be no "tail".

If I stop the cue with a STOP cue, the reverb stops as well. There will be no "tail".

However, if the audio cue plays till the end, there will be a "tail".

But this way I´me bound by time. The end will occur whenever the cue has reached it´s end. If I want to stop the cue as a response to any given action on stage  - there will be no "tail".

To summarize: Fade doesn´t work, stop doesn´t work. Any ideas to what will work, other then play the cue till its end?

[Dave "luckydave" Memory]

To summarize: Fade doesn´t work, stop doesn´t work. Any ideas to what will work, other then play the cue till its end?

You can fade any parameter of an AU effect, so it sounds like the thing to do would be to fade the input of the effect.

http://figure53.com/qlab/documentation/fading-audio-effects/

-- 

luck...@figure53.com

[Chris Ashworth]

(mobile)

On Aug 7, 2013, at 12:57 PM, mikaelka...@gmail.com wrote:
>
> If I mute the audio cue with a volume change, the Reverb mutes as well. There will be no "tail".
> If I stop the cue with a STOP cue, the reverb stops as well. There will be no "tail".

Ah, that sounds like a bug. Stop is supposed to work for this. I'll check it when I'm back at a computer.

C

[Sean Dougall]

A stop cue works for me in 3.0.8. That's different from the Stop command, though (hitting S)—you need to actually program in a stop cue and trigger it with a GO.

Sean

[mikaelka...@gmail.com]

Hello again. Tried Stop cue, but for me, that also stops the tail.

Den måndagen den 5:e augusti 2013 kl. 23:18:07 UTC+2 skrev Christopher Ashworth:

[micpool]

Works fine for me using stop CUE

The purpose of the stop command [ESC] is to silence  all audio gracefully in the time set in the preference.

There is a possible bug in that if you use the stop command all audio fades but the cues with the reverbs do not reset until after the reverb time. If you Panic with [ESC ESC] the cue with the reverb resets immediately

Mic

[micpool]

OK I've replicated this now

It's all seems  to do with how the reverb plug in tells Qlab it's played it's tail out which is indicated by thegreen play symbol changing to a little grey triangle

With Apple Matrix reverb when you stop the cue you get the grey triangle. On my rig same with Vienna Miracle

With Ozone the grey triangle never appears and the reverb curts off hard.

What plug in are you using?

2 other things that can confuse your testing

Apple Matrix reverb doesn't seem to work when inserted on mono audio cues. 

As previously reported Audio Ease Altiverb never seems to tell Qlab it's finished its tail and is therefore unusable for all practical purposes.Double S on the cue or double escape will reset the cue.

Mic

[Christopher Ashworth]

On Aug 7, 2013, at 2:24 PM, micpool <m...@micpool.com> wrote:
>
> Apple Matrix reverb doesn't seem to work when inserted on mono audio cues.

The apple reverb doesn't work with single channels, annoyingly.

-C

[mikaelka...@gmail.com]

Runing Qlab 3.0.8.

Using AUMatrixReverb.

Of course we are talking about Stop cues. Not stopping playback with Esc. You wouldn´t use Esc unless something´s gone wrong...

So far only stereo files. So the Reverbs does work. It is the cutting of the tail that is the problem at hand.

Cheers! /Micke

Den måndagen den 5:e augusti 2013 kl. 22:31:50 UTC+2 skrev mikaelka...@gmail.com:

Certainly there is a clever way of doing this that eludes me...

I´m running a loop of a telephone ring. Just the one ring looping. When the actor puts his hand on top of the phone I want the ringing to stop. Simple enough with a Devamp.

Here´s the tricky part. I´ve added AUMatrixReverb for ambiance. And of course that ambiance should not die with the audio.

Solution?

If a use the fade curve for the phone sound, the reverb fade as well. Same thing if I use a fade cue.

The parameter I actually would like to fade would be the audio input level feeding the AUMatrixReverb.  

To me the scope of this issue would include most use of ambience effects.

Best regards, Micke

[micpool]

On Wednesday, August 7, 2013 7:31:23 PM UTC+1, Christopher Ashworth wrote:

The apple reverb doesn't work with single channels, annoyingly. 

This therefore also precludes its use on cue outputs in edit device.

Has anyone found a good serviceable reverb that works well with Qlab.

Altiverb is a no no, Vienna Miracle is ridiculously expensive if you have no need for the actual MIR package. Apple matrix won't work in mono.

Guitar rig is the same as Ozone with regards to cutting the tail  and  it takes about 5 secs to load any cue its inserted on. Has anyone got any other reverbs that form part of rack type plugins. The fact that Ozone and Guitar rig both cut instantly tends to suggest they work differently in Qlab to dedicated reverb plugs

Mic

[Christopher Ashworth]

On Aug 7, 2013, at 2:51 PM, micpool <m...@micpool.com> wrote:

On Wednesday, August 7, 2013 7:31:23 PM UTC+1, Christopher Ashworth wrote:

The apple reverb doesn't work with single channels, annoyingly. 

This therefore also precludes its use on cue outputs in edit device.

Correct, until cue outputs support multi-channel at some point in the future.

-C

[mikaelka...@gmail.com]

Hi Dave!

Which parameter in the AUMatrixReverb would need to be adjusted to access the input?

Cheers! / Micke

[Christopher Ashworth]

Hi Micke,

On Aug 7, 2013, at 2:48 PM, mikaelka...@gmail.com wrote:

> Runing Qlab 3.0.8.
> Using AUMatrixReverb.
> Of course we are talking about Stop cues. Not stopping playback with Esc. You wouldn´t use Esc unless something´s gone wrong...
> So far only stereo files. So the Reverbs does work. It is the cutting of the tail that is the problem at hand.

Hm, not sure what to say. Just tested it and it works perfectly for me.

-C

[mikaelka...@gmail.com]

GOD DAMMIT!!! =)

Since everyone were saying that the stop cue version works fine, I decided to try it again. Of course it worked this time!!! 

I even tried different presets within the AUMatrixReverb, different order for adding the stop cue, effect, etc. Worked fine every time. Very annoying, indeed.

Well, well. For me that solves the problem! Let´s hope that it doesn´t return! =)

Cheers! / Micke