Qlab Sound Design for Zoom Performance

[gcooke]

Hi,

I'm working on rehearsals in Zoom for a virtual play performance. I am not 100% sure that the actual performance will be via Zoom, but I have some questions.  In Zoom, I have set Zoom to "Original Sound" and am sharing just the computer audio only.

1) "ZoomAudioDevice" in Qlab.  I notice this was created as a patch option in Qlab.  This captures the built-in output, correct?  So does the volume control for the built-in output control the output of Qlab?  My sense is yes.

2) Relatedly, there is a huge difference between what I experience to be the "correct" volume and what the performers experience.  My "correct" volume levels are too loud for the performers.

3) The Zoom setting to turn off the mic auto volume does not impact the computer sound, correct?

4)  Big issues with the automix/mic select algorithm in Zoom. It seems really difficult, for example, to setup "ambient" sound effects that run underneath dialog.

Thanks!!

-George Cooke

Technical Supervisor

Robsham Theater/Boston College

[Alexander (Mailing List) Taylor]

Hello,

I’ve noticed some weirdness with using the ZoomAudioDevice.  For me, the most reliable thing is to open Zoom, join the meeting, then launch QLab and make sure it’s patched to the ZoomAudioDevice.

The safe thing to do in terms of audio levels is to monitor from another computer.  So I join my “QLab" computer to the group then use another computer logged in as “Alexander” to monitor and control the zoom specific options.  I think zoom is trying to make everyone sound the same, which is messing things up, as you’ve discovered.

Alexander

--
Contact support anytime: sup...@figure53.com
Follow QLab on Twitter: https://twitter.com/QLabApp
User Group Code of Conduct: https://qlab.app/code-of-conduct/
---
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.
To view this discussion on the web visit https://groups.google.com/d/msgid/qlab/d6fae07b-a2f4-4043-aff7-1d8b85cd8e4c%40googlegroups.com.

The Right-To-Know Law provides that most e-mail communications to or from School District employees regarding the business of the School District are government records available to the public upon request. Therefore, this e-mail communication may be subject to public disclosure.

[Andy Dolph]

From my experience with Zoom, it seems to level everything rather hard and so doing something like an underscore is difficult.  It might be worth trying going out an analog out and the coming in through an analog in on an interface so zoom sees it as your "mic"  and then you could turn off the auto input gain and control it from the interface?

I'd be interested to hear what you find works - I'll second what Alexander said - I think it's critical to have a 2nd device logged in on the call as an "audience member" to monitor what they are getting - you can't tell from the machine you are broadcasting from - and my experience has also been that it ends up significantly louder for everyone else than for me.

I'll check with a colleague who did a play on zoom a couple of weeks ago and see if he learned anything he can pass on.

Andy

To view this discussion on the web visit https://groups.google.com/d/msgid/qlab/8580AB83-8865-446E-86D9-E1E1E7082080%40orcsd.org.

[Andy Dolph]

Here’s what my colleague had to say:

“ It is pretty raw.
Zoom is designed to only engage one speaker/sound source at a time. Someone said there was a way to override that- but I was never able to see any way to do that.
So- you are fine if a speaker and the sound cue are coming from the same computer.
If you are trying to have several people speaking and trying to include under music...no go. 
We ended up having to record the one section where under music and other sound elements were crucial. Beyond that- we had to have sound cues only go between people speaking. (But because of lag time- it sometimes sounded like crap- because if one mic came on too soon- it made the sound cue cut out.)”

Sent from my iPhone

On Apr 28, 2020, at 1:51 PM, Andy Dolph <acd...@gmail.com> wrote:



[Harry Butcher]

Interesting thread.

We’ve seen all services make changes in handling of audio improving use for core services. Their efforts are impressively effective. Rarely now do you get feedback loops and distractions from open mics but this has removed ease for outside the box applications like this.
I have always thought zoom was the worst in handling audio (and the best for achieving this).

It would be interesting to see a use case which works like standard broadcasting. Pulling all the streams in to a vision mixer before sending to YouTube / Twitch. I Guess having performers interact would be the challenge there, though not impossible. It would certainly ease the job of adding underscoring..

[Harry Butcher]

Just thinking about it, I’ve not used the new Facebook group calling?

I think you best bet is to find the app that has the worst audio setup for conference calling..

[Chet Miller]

I know Skype can do a per caller ndi out. I’d do that, pull it into xsplit or obs, then add the sound cues there. Bonus is that you can try to set up a dialog mix if you’re crafty. Wire your networks.

[Romo Hallahan]

For Zoom you can do some underscoring with people talking over, I currently have gotten this to work. Currently working on a project using it.

A couple things you have to do is  Enable original audio in settings and send something through the mic input. Also disable both suppress persistent background noise, and intermittent background noise so the algorithm and gating doesn't think think music frequencies are background noise. Zoom does allow overspeaking but you ha

On Tue, Apr 28, 2020 at 9:49 PM 'Harry Butcher' via QLab <ql...@googlegroups.com> wrote:

Just thinking about it, I’ve not used the new Facebook group calling?

I think you best bet is to find the app that has the worst audio setup for conference calling..

--
Contact support anytime: sup...@figure53.com
Follow QLab on Twitter: https://twitter.com/QLabApp
User Group Code of Conduct: https://qlab.app/code-of-conduct/
---
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.

To view this discussion on the web visit https://groups.google.com/d/msgid/qlab/1acd63ec-2db7-4713-824b-3bfb47f4b94c%40googlegroups.com.

[Gordon Nimmo-Smith]

I've been doing some basic messing around feeding a multi-person Zoom call into OBS and then Facebook Live, with me adding music and underscore as it passes through my computer. This seems to work well as I can time cues to what I hear on my computer, with that timing being replicated on the stream. The downside as you said is that this only really works for cues that the other participants don't need to interact with. Passing cued audio back into Zoom was problematic, due to timing glitches and Zooms internal audio correction, as already mentioned. I ended up downloading Audio Hijack to get all the audio paths through my computer working as I needed.

Thanks! This is a useful thread!

Gordon

--
Contact support anytime: sup...@figure53.com
Follow QLab on Twitter: https://twitter.com/QLabApp
User Group Code of Conduct: https://qlab.app/code-of-conduct/
---
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.

To view this discussion on the web visit https://groups.google.com/d/msgid/qlab/36a4672b-1956-4dfd-8549-814135ae13cc%40googlegroups.com.

--

Gordon Nimmo-Smith
US: +1 443 462 8505 / UK: +44 7764 756270
gordon.n...@gmail.com / gor...@gnsmith.co.uk

[Sean Doyle]

Try Audiomovers, it is done through the web. It requires an extra step for the both sides, and having a browser open. But the quality is good. Like legit good.

To view this discussion on the web visit https://groups.google.com/d/msgid/qlab/87242940-4EFB-4CAE-9BA8-C28FC9703BBC%40gmail.com.

[F Staab]

I just got done designing sound for a Zoom performance.

I’ll echo what others said - having a second computer logged into the Zoom call to monitor your audio is immensely helpful. Otherwise, setting levels becomes insanely frustrating.

I cued up everything in QLab, and then ran QLab through Loopback (from Rogue Amoeba) and then into Zoom. It gave me a little extra volume control without having to stop QLab in case something got too loud or weird. I didn’t change any of the Zoom mic settings, and just had my mic muted the entire time - I sort of assumed that “sharing sound” through Zoom bypassed all that mic processing anyways.

I found that I was able to do some very basic underscoring and still have actors talk over it - tones, some gentle waves, etc. But it couldn’t be too loud, otherwise it would start to take over and drop the actor talking. Zoom definitely behaves like a Thanksgiving Dinner - whoever yells the loudest will get heard. That obviously means that with any loud cues the actors will just have to wait for it. It also makes overlapping dialog virtually impossible.

The other annoying thing I found was the lag between hitting Go on QLab and actually hearing sound on Zoom. Some nights it was as much as 2 seconds of delay, sometimes it was 1 second or 0.5 seconds. So I recommend testing your “delay connection” and adjusting your GO timing accordingly. But it means you’ll never really get super tight cues - there’s always some variability.

Because of all the audio processing that Zoom does and the fact that most people are listening on small computer speakers, I also found that suddenly sounds that you wouldn’t think are similar become hard to differentiate. I felt it was an annoying limitation… 

Lastly, there can be weird things happening depending on everyone’s internet connection. Especially when the weather is bad. And satellite internet gives too much of a lag time to be useful. Generally, if you can plug into your router directly via ethernet, that’s the way to go.

If you’re interested in seeing some of the underscoring work I did, you can check it out here. Show itself starts about 15 minutes into the broadcast (-2:02:25). Anytime the woman with the tree background is speaking, there’s underscore.

https://youtu.be/2Y6bWEewcvQ

Feel free to reach out to me directly with any questions!

Cheers,

-Staab

—

www.florianstaab.com

—

To view this discussion on the web visit https://groups.google.com/d/msgid/qlab/CAMUGRHnXOydy7jJpvbKxN_Bu3jgRst7p_RAbsMjezya4a3v1Gw%40mail.gmail.com.

[gcooke]

Thanks, everyone, for your replies. The performance was this past Sat, and I will echo the comments of what the others have said here:

--Two computers: One for Qlab, the other to monitor

--Zoom settings: Original sound on, Disable both Background Suppression settings

--Settings levels for underscoring(rain, traffic sounds) is very sensitive and tricky, but can be done.  I noticed a range of 3-4db from inaudible to too loud(would suppress actors dialog slightly).

--I also noticed a slight degradation of quality in the actors audio when we ran the underscoring, so I ended up doing pretty quick fade outs for those cues.

-George

[talkingtobrian]

Hello, everyone. I've been quietly doing a lot of research on this, for my own future theatre's use (NVCC in CT) and to help my wife, who is TD/PM for Wesleyan University. 

They went on with their semester's show, "The Method Gun," and moved it online this past weekend. They really took the tech pretty far in a short amount of time.

As this is a Qlab forum, here's the nutshell of what I have to share:

Blog post on audio settings to use in Qlab and tweak Zoom (I used Soundflower, but in the process I moved to Blackhole, as that is the currently supported leader of free, virtual sound cards.): https://www.rocktzar.com/audio-playback-and-streaming-online-performances/

The resulting show, streaming on YouTube: https://www.themethodgunwes.com

The whole show was performed on Zoom, run through Open Broadcaster for some final polishing and control/lobby control, and then sent to Youtube.

I've been using Qlab in a similar case for my corporate day job, doing lobby music for and announcements for WebEx meetings and events.  (You think Zoom's audio is troublesome? At least they give you controls to improve your audio playback! Our install of WebEx is just abysmal for this type of thing.)

-Brian

On Tuesday, April 28, 2020 at 9:33:06 AM UTC-4, gcooke wrote:

[Kevin Sweetser]

Been doing a lot of digging into this. Just FYI the newest update to skype mobile app has a lot of audio issues. App volume doesn't mute for users. If you have folks trying to get better audio, IOS and android apps don't recognize external USB devices. 1/8" TRRS seems to work though. I did about a day worth of testing with about 4 devices. We wanted to get better video quality from phones for a live stream since webcams are sold out. 

[Rob Witmer]

Hey Brian, I enjoyed watching "The Method Gun", and reading about how you used Soundflower to get Qlab to output directly to Zoom.  Have you ever had actors try to sing along with a track you were playing? I imagine the latency would be difficult to overcome, even with wired LAN connections.  I know there are some software programs that are designed to minimize latency for musicians (like JamKazam) but I'm working with a Zoom prodction, and I can't really ask the actors to download another program.  It's starting to seem like if we want a song where everyone is singing that we should just pre-record and have everyone lip-sync. Thanks for any advice you might be able to offer!

[ivan birthistle]

Hi all,

There is some really interesting stuff going on here, great to see.

I thought in that context I'd let you know where I'm at.

I am Head of Sound and now Head of Digital Production and Lead Editor at the Lir Drama Academy (thelir.ie), for those of you who saw Normal People, that where a lot of the actors trained.

We are into week three of rehearsals for two Shakespeare shows, rehearsing on Zoom. One we will probably use the Zoom recording where the DSM is selecting the actors tiles for spotlighting etc and hopefully that will compile ok.

The other show we didn't want to have the raw Zoom look so we are using it as a tool for the actors to interact in real time while they simultaneously self-record in order to capture better quality video/audio, they then drop files at a central hub in Teams.

To preserve large chunks of our training we are aiming to treat this like a projected show and compile the pieces in QLab then Syphon out to (probably) OBS for recording, just treating OBS as our projection surface, then finally host on a private YouTube channel.

One of my main concerns at the beginning was lag and dropout during performance, I have seen many dramatic items recently and I just switch off if it looks and sounds like a regular Zoom call - I am spending 4-7 hours on Zoom each day already!

I am just working up template cues do make my life easier later, I know it's going to be a mental process but we will at least be in full control.

This is new stuff for all of us, we have hit several speed-bumps along the way and will probably hit a few more.

I'll keep you posted.

Ivan

p.s. it will be great to get NDI support in Zoom, I heard that Teams is also in that race and may get there first.

[Richard Williamson]

Hello

Sounds interesting. If it’s any use I’ve been working on a little project which makes it easier to get decent sound with zoom/YouTube etc - it’s a Mac app which creates its own virtual audio device (similar to loop back) and allows the user to easily create their own monitor and stream mixes. It doesn’t do anything revolutionary but is rather designed to allow non-technical users to simply and easily stream higher quality sound. Details are at www.stagesound.net - any feedback would be very gratefully received!

Sent from my iPhone

On 18 Jun 2020, at 16:32, ivan birthistle <ibi...@gmail.com> wrote:



--

Contact support anytime: sup...@figure53.com
Follow QLab on Twitter: https://twitter.com/QLabApp
User Group Code of Conduct: https://qlab.app/code-of-conduct/
---
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.

To view this discussion on the web visit https://groups.google.com/d/msgid/qlab/dca3cb1e-5542-4bcd-a90f-bcdf1bf8cd1eo%40googlegroups.com.