I would like to start a discussion about reverb automation in Nuendo for scenes in an audio post context.
I still find it one of the most problematic areas of my work due to bugs and misbehavior of all candidates I have tested over the years.
It seems to me reverb plugs do not like parameter automation at all. Among the typical problems are: missing parameters (Virsyn reflect misses ER pattern automation), wild CPU spikes (Ircam Verb Session), missed linked automation (Phoenixverb), wrong parameter automation timing when bouncing (Altiverb) and general read/write parameter errors (Ircam again)
How do you guys solve these problems, what do you use and do you automate parameters or do you just open another instance for every room you need in a film? Which plugs do you use?
We usually have somewhere between 6 and 20 reverbs/rooms/places ready at hand.
Half of them for DX, the other half (copies or near the same spaces/and linked) for FX.
Any additional stuff we need, we create another FX channel, or we use an insert.
This procedure has covered about 99,9999999% of all the jobs we’ve done.
I don’t automate the presets. I sometimes do some automation on the plugin parameters though, but for switching scenes/rooms/presets … no.
It is just way too complicated and time consuming.
And I go even further. For animated series or dubs which are nicely recorded in the studio, I include the locked sends “punches” in the template.
So whenever the scene jumps takes part in -for example- the classroom, I recall and punch in the “DX_Classroom” sends automation for all the DX-tracks. That gives me about 90% of the needed reverbs (on DX) in that scene. All you need to do is some tweaking. There is more work in adding them one by one, than to modify a few.
Thanks for the insight on your workflow. It’s actually the same way I do it at the moment with a third reverb per room if I have the luxury of having foley for the show. Which plugs do you like?
Anyone else, who actually succeeds at automating reverb plugs?
In our lastest project (It’s Mega - Runs over a year) Exponential Audio Exclusively.
Not talking about the occasional specifics/exceptions.
For example, I’ve been making a lot of use of “Indoor” lately for placing Foley in a room.
The recently released plugin “INDOOR” by Audioease makes it possible to clickless automate the settings in your room-reverbs…although it can’t switch presets it comes in quite handy when switching perspectives between actors quickly. I think this new approach =although it’s not really there yet- will change the way we handle room reverbs in the future.
My wish would be for future releases of any post production software, to encapsulate more and more of repetitive tasks. Like in programming. You don’t code in assembler that much anymore, where you have to do everything manually. Also, C is slowly going away. More modern languages like C++ or even more modern like Swift or Rust have huge built-in libraries. You want to sort something? Call this method, somebody coded it already efficiently so you can just use it. Don’t do that stuff yourself.
And so you can concentrate more on what your software should do rather than figuring out how to sort lists efficiently. And this should also come more and more to software. Don’t do everything manually and set up tracks, sends, automate parameters for each track manually. If somebody could come up with a user interface where you could just define: This is a room. From here to here is a scene. Place these tracks in that room, switch with the scene. Camera moves away, turn all parameters in a way that is realistic for this situation. Less dry, more wet reverb, dampen the highs a little, oh the guy moves next door, make it muffled… Stuff like that, where you can concentrate on the sound and artistic implementation rather than having to turn every single knob, punch in automation, manage sends and dozens of tracks.
I hope this revolution should kick in at some point. It would help productivity. Currently, we’re stuck mostly with assembler, it seems.
-Assign all sends on your DX (and/or other tracks) and set them at the correct level.
-Punch the automation and save/lock/rename that punch.
-Do this for all your sends.
I have been wondering…
If there would be more sends I could have them all up all time. Then I’d have preconfigured reverbs waiting at zero level.
Then when I needed one of the preconfigured reverbs I’d just pull the volume up on that reverb…
I was thinking about that myself, and I think the problem is that if you do that you’ll be sending signal into all verbs at all time, which a) means they’re always processing and using cpu cycles, and b) when you then increase level of a verb you’ll be getting the tail from the content that went into it before the (video) cut. So if you have a scene in a larger space with a 2+ second tail, and someone was talking a second before the cut, then at the beginning of that cut you’ll have the tail of that person speaking despite having been in a different space (before the cut).
As for more sends I think an alternative could be to have different types of verb presets on different fx channels, but stack them and then bypass whichever you aren’t using. So perhaps you could send to Verb 1 and on that you have two-three instances that are all bypassed until you need them, and they can all have different presets. I suppose the only issue would be if you need to go from one preset to another and they’re both on the same channel.
I do it the same way, except I don’t pull up the volume of the reverb, I pull up the volume of the Sends.
Same thing, no difference.
Easy on Nuage (flip fader) and perfectly doable on other controllers too.
As you can see in my “Sends” pic, the first 5-6 sends are pre-configurated, the remaining two are for the occasional stuff. I don’t think I ever needed more than 8 different sends/rooms/spaces in a project. And if you need more, nothing stops you from recalling another preset from one of the reverb units. In fact, if you double-use each FX track (switching/automating between two presets), you have 16 sends. Triple them … and so on …
Exactly that’s what started my post.
I work the way you describe preconfigured sends.
Unfortunately I found out that most reverb plugins (the Steinberg and waves ones behave better than the rest) do not like automation at all. You will likely end up with really bad cpu spikes (IRCAM) and missed timing during rendering (Altiverb) or you will simply not find the parameters you want to automate (Virsyn Reflect ER pattern)
So another way of increasing the number of available rooms would be to write bypass automation for reverbs (which unfortunately some plugs like IRCAM/Flux do not support. I’ve filed a bugfix report almost a year ago at no avail)
It looks like some of you have hit the same wall as I have but just keep working around it (which is of course, what we all need to do if we want to satisfy our customers).
