Roles based mixer

[Charlie Austin]

[Oliver Peters] “A lot of us would like to see Apple develop a roles-based mixer as a feature within FCPX. “

They recently had a patent approved for exactly that. Who knows when/if it’ll be added, but it appears to be on their radar…

EDIT: Braden Storrs has a good writeup for those who haven’t seen it… https://thefcpxeditor.tumblr.com/post/137629234928/fcpxrolesmixer

————————————————————-

~ My FCPX Babbling blog ~
~”It is a poor craftsman who blames his tools.”~
~”The function you just attempted is not yet implemented”~

https://www.tumblr.com/thefcpxeditor/137629234928/fcpxrolesmixer

[Walter Soyka]

I should probably read the patent, but I’m curious how a roles mixer works with presto change-o rearrange-o magnetism, because I think the implicit relationships in an audio mix are much more complicated than the cut-here video clip relationships.

When you start slinging clips around the timeline, how does recorded mixer automation know what to follow? For example, if I duck music to allow dialog to come through, then move the dialog in the timeline, how does FCPX keep sync sacred?

Walter Soyka
Designer & Mad Scientist at Keen Live [link]
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
@keenlive   |   RenderBreak [blog]   |   Profile [LinkedIn]

[Mathieu Ghekiere]

[Walter Soyka] “When you start slinging clips around the timeline, how does recorded mixer automation know what to follow? For example, if I duck music to allow dialog to come through, then move the dialog in the timeline, how does FCPX keep sync sacred?”

Wouldn’t this be the same with tracks? (I don’t use a DAW, so excuse me for if it’s a stupid question)

Anyhow, I’m very curious to see how Apple will tackle this. I think it’s a hard problem to tackle, user-interface wise, but at the same time, I think the concept of Roles is pretty genius. The idea of within this trackless timeline, to come up with the idea of having stuff labelled meta-datawise based on their CONTENT and being able to easily export it in seperate tracks without doing (what Charlie calls) Track Tetris… I still admire that and I think Roles are one of these things that a lot of people who haven’t tried out X yet, don’t know the existence of, while I think it’s one of it’s most genius features.
I would like that it comes with visual coloring and (optional) automatic grouping like described here:

https://disproportionatepictures.blogspot.be/2014/05/roles.html

And automatic presets you could slap on dialogue and music to get dialogue come trough when the mixer sees that there is music underneath, and so you just have to adjust instead of doing it from the start. Well, this is maybe a big sin to people who do audio design, but as a tool for the editor that’s optional, it would be a great way to start off an audio mix.

To answer Oliver’s original question: it seems that if you want it interfere with the user interface of X itself, that I don’t think it can be a plugin.
I do think better audio is on Apple’s roadmap, and I’m still thinking it will be one of the big focus points of the next big feature update.

https://mathieughekiere.wordpress.com

[Jeremy Garchow]

[Walter Soyka] “When you start slinging clips around the timeline, how does recorded mixer automation know what to follow? For example, if I duck music to allow dialog to come through, then move the dialog in the timeline, how does FCPX keep sync sacred?”

Ask Adobe. The only had track level automation on a mixer for YEARS before they came up with the “clip mixer” and the automation stayed put when the timeline clip slinging commenced. So now, there are two mixers in Pr, at the clip level and the track level.

I could imagine a similar (but hopefully better) system in FCPX where the Roles act more like a bus and you can apply a “group level” automation to the Role itself. If the timeline change-o rearrange-o’s then you have to rearrange the automation. Need a new bus? Make a new Role and separate the material.

Then clip level actions would happen on the top clip level on the timeline and the mixer. Component level stuff would have to be done with the mouse.

[Walter Soyka]

[Jeremy Garchow] “Ask Adobe. The only had track level automation on a mixer for YEARS before they came up with the “clip mixer” and the automation stayed put when the timeline clip slinging commenced. So now, there are two mixers in Pr, at the clip level and the track level.”

I get Premiere’s dual (dueling?) mixers, but I think FCPX needs more than that to preserve magnetism.

In the example I gave above, ducking music for dialog, the mix automation is applied to the music role, but should be anchored by the dialog clip.

[Jeremy Garchow] “I could imagine a similar (but hopefully better) system in FCPX where the Roles act more like a bus and you can apply a “group level” automation to the Role itself. If the timeline change-o rearrange-o’s then you have to rearrange the automation.”

That’s the thing that wouldn’t feel very FCPX to me. A role does not exist on the timeline, so where does the automation data live? Role/bus/track-level automation is a function of absolute time, which does not really exist in FCPX. How do you add a role mixer without giving up some element of magnetism?

Walter Soyka
Designer & Mad Scientist at Keen Live [link]
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
@keenlive   |   RenderBreak [blog]   |   Profile [LinkedIn]

[Peter Gruden]

Ducking is not normaly applied using automation. Most of the time it is performed by sidechaining a dynamic processor. This way there would be no issues moving dialog or music on the timeline. In FCPX example, a dialog role would feed the sidechain input of the processor inserted in music role.

Unfortunately, Noise Gate and Expander in FCPX do not have sidechain input like they have in Logic Pro X. The reason is that sidechain source is normaly an audio track, or role bus..

How to display automation, if not on a linear track? Good question.

[Jeff Kirkland]

at a basic level simply being able to duck one or more selected roles by x% whenever the volume of the dialogue role exceeds a certain threshold would work for me.

Jeff Kirkland | Video Producer | Southern Creative Media | Melbourne Australia
http://www.southerncreative.com.au | G+: https://gplus.to/jeffkirkland | Twitter: @jeffkirkland

[Bret Williams]

iMovie solves ducking very simply. With a checkbox and a slider. Just apply it to any clip (like VO or SOT) and it will “lower other clips volume” by that amount. Pretty slick. And sad that such a thing isn’t in X. And then a checkbox to “turn ducking into keyframes” or something.

I think they used to call it ducking but probably nobody knew what that ways. So no it literally says “lower other clips volume.” But if you look at the undo cache, it’ll say undo ducking.

[Walter Soyka]

[Peter Gruden] “Ducking is not normaly applied using automation. Most of the time it is performed by sidechaining a dynamic processor.”

Clearly I have chosen the worst possible example!

My point was that in a mix, the automation is relative to the confluence of mixing elements, not to any one member in isolation. As soon as you start moving those elements around, you’ve very likely invalidated the mix — right?

My questions are, how do you store a role mixer’s data in the FCPX paradigm, and how do you maintain the devil-may-care malleability of the timeline once you’ve mixed roles?

But I see another interesting question coming out of the other responses. What if mixer automation were actually, you know, automated, instead being a recorded manual mix? Would you be happy with roles mixing parameters that let you prioritize audio elements, with the software automagically mixing to meet your priorities?

Walter Soyka
Designer & Mad Scientist at Keen Live [link]
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
@keenlive   |   RenderBreak [blog]   |   Profile [LinkedIn]

[Jeff Kirkland]

I’m not expecting FCPX to suddenly get all the features of a DAW. I see roles becoming pseudo tracks feeding into master busses.

I’d still be working like I do now, finessing volume, adding effects, etc on a clip by clip basis if needed, which will move around the magnetic timeline as always. The addition would be a mixing panel that had a global master fader and a channel for each audio role I’d created.

So basically you have the option of working with individual clips, the roles as if they were tracks, or a timeline master bus. If you need more than that, there’s still the option to export to Pro Tools or whatever. Then as I mentioned earlier, give me the option somewhere of ducking one or more roles based on the volume level of other roles and I’d probably never have to send audio outside of FCPX.

Jeff Kirkland | Video Producer | Southern Creative Media | Melbourne Australia
http://www.southerncreative.com.au | G+: https://gplus.to/jeffkirkland | Twitter: @jeffkirkland