How come no realtime audio fx?

[Douglas Spotted Eagle]

Potentially, but it also causes many more headaches. It's got nothing to do with loss of productivity; this has been discussed since the early days of Vegas. Vegas is intended to:
a-operate like an analog suite operates in the real world
b-provide logical interface for workflows.
Audio can have temporal effects applied to the media. What happens say...when you have a 5 second event to which you want to apply a 10 second reverb? (Very common application)
The 5 second file has ended. A new file must be written.
This is just one example.
Yes, the engineers at Sony *could* (and probably easily) write code to make this work. But it's at a cost that likely isn't cost-effective in terms of various aspects of the way the app works.
In the "real" world, an FX bus is an FX bus. Vegas works this way, just like an analog system works. Interfaced with a HUI, Vegas functions identically to an analog console setup in a live or studio environment.

[Jarrod Whaley]

Originally Posted by Douglas Spotted Eagle

Potentially, but it also causes many more headaches. It's got nothing to do with loss of productivity; this has been discussed since the early days of Vegas. Vegas is intended to:
a-operate like an analog suite operates in the real world
b-provide logical interface for workflows.
Audio can have temporal effects applied to the media. What happens say...when you have a 5 second event to which you want to apply a 10 second reverb? (Very common application)
The 5 second file has ended. A new file must be written.
This is just one example.
Yes, the engineers at Sony *could* (and probably easily) write code to make this work. But it's at a cost that likely isn't cost-effective in terms of various aspects of the way the app works.
In the "real" world, an FX bus is an FX bus. Vegas works this way, just like an analog system works. Interfaced with a HUI, Vegas functions identically to an analog console setup in a live or studio environment.

What headaches? And what about the headaches I (and surely many others) currently have by virtue of the fact that this functionality is not available? If you needed this functionality, I'd bet you'd be arguing in favor of it.

Video can have temporal effects too, by the way.

In the "real world" you don't have to do the equivalent of rendering out to a new file. If you were mixing on a console and had to record your sound to a tape every time you made minor tweaks to the effects configuration on it, I bet you wouldn't be happy.

I suspect you'll rebut by saying that if I were to use automation or envelopes, then I'll be operating the software like I would operate a console in the real world. But we're talking about software, not the real world. When I use a word processor, I don't expect to have to listen for a "ding" at the end of each line of text. When I am listening to an MP3, I don't expect to have to rewind at the end.

Software doesn't have to emulate the "real" world when certain aspects of the real world are a huge pain is the hindquarters; the entire point of using computers is that they are supposed to make things easier, cheaper, and less time-consuming (*everyone laughs uneasily here*). This particular aspect of Vegas makes things harder, more expensive (due to lost productivity), and much more time-consuming.

Originally Posted by Josh Bass

So is there any chance a future Vegas would change this?

My guess is that the Sun will turn purple and start whistling showtunes before that ever happens.

[Douglas Spotted Eagle]

What DAW packages out there don't emulate the real world?
Could you identify temporal fX for video that go beyond the end of a file without writing a new file.
I'm not defending Vegas, I'm curious as to how you'd rather see this realistically work.
'nother example...
you a door slam/Foley. File size/length is 1.5 seconds. You want it to reverberate for say....15 seconds. Since it's a file-level door slam (which cannot be accomplished in the real world), where do the additional 13.5 seconds of reverberation come from without writing a new file on the fly?

[Jarrod Whaley]

Originally Posted by Douglas Spotted Eagle

What DAW packages out there don't emulate the real world?

I don't know which ones, because Vegas is what I use. My point, though, is that there's no reason why a piece of software should have to be exactly like the real world. I realize that by emulating a console workflow, Vegas retains a familiar workflow that has worked for a long time. Believe me, I do see the value in that. However, if something is a pain to do on an analog console, why should that pain be translated directly into software, where anything goes?

Originally Posted by Douglas Spotted Eagle

Could you identify temporal fX for video that go beyond the end of a file without writing a new file.

I realize it's not quite the same thing, but it's close in many ways: slow-motion requires a change in the length of the event. When you use an event velocity envelope, or change the playback rate in the event preferences, Vegas truncates the effect by retaining the original length of the event--you have to stretch it out manually if you want to get the entire clip as you originally cut it. Worst case, things like reverb could function like this. It would be a little bit of a bother to have to stretch out the event, but it would be better than juggling massive numbers of tracks containing single (or two or three) events and/or messing with complicated envelopes. Again, this would be the inelegant, worst-case solution; a better way of doing this is described below...

Originally Posted by Douglas Spotted Eagle

I'm not defending Vegas, I'm curious as to how you'd rather see this realistically work.
'nother example...
you a door slam/Foley. File size/length is 1.5 seconds. You want it to reverberate for say....15 seconds. Since it's a file-level door slam (which cannot be accomplished in the real world), where do the additional 13.5 seconds of reverberation come from without writing a new file on the fly?

Since all we're talking about is real-time preview here, all Vegas would have to do is to continue playing the reverberating sound after the event ends, saving the actual rendering for the point at which the entire project is rendered. Again, if Vegas can play back reverb on the fly at the track level, why not also at the event level? How hard could that be to implement? And what issues (technical or otherwise) could that possibly cause? It would require more processing power, but the difference would be pretty marginal when you think about all of the processing that already goes into working with video in a real-time environment.

[Douglas Spotted Eagle]

Originally Posted by Jarrod Whaley

I
Since all we're talking about is real-time preview here, all Vegas would have to do is to continue playing the reverberating sound after the event ends--just as Vegas doesn't write a new video file when you add effects to a video event. How hard could that be to implement? And what issues (technical or otherwise) could that possibly cause?

I surely wish I could have a dime for every time someone says "It can't be that hard." If it's not that hard, why isn't everyone doing it?

Event is 1.5 seconds. 15 second reverb. The next event on the timeline is only 6 seconds downstream from the event that has a 15 second reverb on it, on the same track. How would you see Vegas managing this at at Event level?

[Jarrod Whaley]

Originally Posted by Douglas Spotted Eagle

I surely wish I could have a dime for every time someone says "It can't be that hard." If it's not that hard, why isn't everyone doing it?

We seem to be asking the same question, but with different rhetorical intentions. :) Why indeed.

If I can apply color correction, masks, keyframed motion, and keys all to a single event without having to render it before I can see those effects depicted on the timeline, I should be able to do the same amount of tweaking to an audio event and be able to see the changes reflected on the timeline without having to render out first. It's a no-brainer.

Originally Posted by Douglas Spotted Eagle

Event is 1.5 seconds. 15 second reverb. The next event on the timeline is only 6 seconds downstream from the event that has a 15 second reverb on it, on the same track. How would you see Vegas managing this at at Event level?

It seems that I can't properly express what I'm suggesting. We're just talking about real-time preview--Vegas could simply play both events back with reverb applied to them, with no need to render them yet. Again, if you put a "dry" event on a track that has reverb applied to it, Vegas will apply reverb to the event on the fly with no need to render it out, right? Why in the world can't the same thing happen with an event?

Theoretically speaking: when the playback reaches the event, it plays the event as if it has reverb on it, and the reverberation continues after the event ends. Simple. Or, alternatively, Vegas could truncate the file like I was saying before, making a manual adjustment to the length of the event necessary.

Could you address each of these scenarios directly and tell me why, in your opinion, they would not work?

[Douglas Spotted Eagle]

In your color correction example, the color corrected clip isn't playing over top of the next clip in line. It requires a new track should you wish to play both events at the same time, and have them both be visible.Temporal-based video events do not spill beyond the temporal boundaries of the file/event.
Again...two events, 3 seconds apart. First event has 15 seconds of reverb on it. What happens to next event? What happens to the reverb during the next event, because the reverb should be carrying over it. On a track level, this is possible. But on an event level, it's not. I guess what seems fairly simple to grasp from my perspective may not be so easy to understand, or rather, I'm not explaining it clearly enough.

[Jarrod Whaley]

Originally Posted by Douglas Spotted Eagle

In your color correction example, the color corrected clip isn't playing over top of the next clip in line. It requires a new track should you wish to play both events at the same time, and have them both be visible.Temporal-based video events do not spill beyond the temporal boundaries of the file/event.

I realize this, and I see the difference. The point here was just to illustrate how frustrating it is that audio effects are harder to deal with than video effects, and that this is completely baffling to me.

Originally Posted by Douglas Spotted Eagle

Again...two events, 3 seconds apart. First event has 15 seconds of reverb on it. What happens to next event? What happens to the reverb during the next event, because the reverb should be carrying over it.

Again... Vegas plays (because playback in real-time is all that's necessary) the first event as if it has reverb on it, and plays the second the way you have instructed Vegas to play it. The reverb from the first event goes on after the event ends and is "mixed" on the fly with the sound in the second event. This is all just playback for preview purposes, mind you.

Originally Posted by Douglas Spotted Eagle

On a track level, this is possible. But on an event level, it's not. I guess what seems fairly simple to grasp from my perspective may not be so easy to understand, or rather, I'm not explaining it clearly enough.

I feel the same way. What I'm suggesting here seems to me like a very simple thing, but I've ended up saying it four or five times. I'm clearly not explaining this very well.

[Jarrod Whaley]

Originally Posted by Douglas Spotted Eagle

Temporal-based video events do not spill beyond the temporal boundaries of the file/event.

Sorry for the double post, but I thought I'd point out that this is, in fact, one of the two models for real-time audio effects that I have suggested.

I'd also like to add that even if, in your example, I had to put the event with reverb on it onto a second track so that it could overlap an immediately following event, I'd still prefer this method to the current situation, in which you have to create an entirely new track for the clip with the reverb on it. Instead, I could put the event with reverb applied to it on an already existing track.

Part of my problem with the way Vegas handles audio effects is that you either:
A) Have to render out tons of tiny wav files for each event to which you want to apply effects;
B) Have to wade your way through a project that contains an absurd number of tracks in many situations; or
C) Have to contend with extremely complicated and easy-to-screw-up envelope and/or automation schemes.

All I'm saying is that none of these scenarios lends itself to an efficient use of one's time and/or resources, and that a more efficient model than any of the above should by no means be too much to ask.

[Glenn Chan]

It seems that I can't properly express what I'm suggesting. We're just talking about real-time preview--Vegas could simply play both events back with reverb applied to them, with no need to render them yet. Again, if you put a "dry" event on a track that has reverb applied to it, Vegas will apply reverb to the event on the fly with no need to render it out, right? Why in the world can't the same thing happen with an event?

The post on the Sony forum somewhere explains why this is difficult:

If it were only as easy as you believe...it isn't.

- Not all FX tell us when they are done.
- FX can't be asked consistently for information like 'tail length' produced. Not being able to know exactly when an "event" is not longer producing data is a problem.
- Output can be variable if an FX has parameter automation. (If we let you put FX on Events, I cann't imagine that you wouldn't want to automate them over time...)
- Some plugins - UAD - can be very problematic when used in any way out of how they define them. (They define their own variance and rules on the VST SDK/Spec.)

For example.

You cannot just feed data into a UAD plugin when you want. You must process ALL UAD plugins at the exact same time relative to the current sample frame all the time other wise the hardware yells - loudly. So let say you have 10 events, each with a different UAD plugin. You must feed silence into all of the UADs so that the hardware limitation of the UADs is met. Sure, you could disable and enable the plugins on the fly, but you can't do this arbitrarily. It would still have to happen on sample frame boundaries. Disabling and enabling the UAD plugins has an overhead associated to it as well.

While it seems trivial on the surface, it is not. Sure we could guess and try to determine the output of a plugin in realtime, but now we are spinning CPU cycles on listening for "silence" or detecting tails.

WRT limting which plugins can be used:

I could just hear those users that have Brand X or "Type X" plugins that we choose not to support as Event level plugins. Ugh! If we can't make it work for every plugin, then we are just digging a different hole for ourselves.

I am not saying this can't be done or that we wont do it, but understand it is never as trivial as it looks on the surface. To do it right and make as many users happy is no small task.

FWIW: CDArch is not really giving you Event level FX. They GUI just makes it appear this way.

Peter

http://www.sonycreativesoftware.com/...ssageID=482604

2- Feature film mixes can easily have 40+ tracks.

Vegas has shortcuts for collapsing tracks so they don't take as much real estate... check the shortcuts sticky at the top of this forum, I think it has them in there.

3- To me, a somewhat more sensible solution is if the non-real-time FX engine is tweaked so that it remembers your audio plug-ins... that way it sort of behaves like there is eventFX. Though reverbs won't work properly, and that might be confusing if you don't know what's going on.

[Jarrod Whaley]

Thanks for digging up that post, Glenn. At least now I have a better idea of *why* audio effects are incapable of being handled in real-time.

The poster from the Vegas forum mentioned that CD Architect's GUI makes it seem as if effects are being applied in real-time... I've never used CDA, but it sounds like a similar tweak to the Vegas GUI would satisfy me, because it sounds like at least in some ways, I'd never know the difference on any practical level.

You say that feature films routinely use 40+ tracks... this is in line with my experience with Vegas on long-form projects. Unfortunately, in an NLE, you can't see all of those tracks at once like you can on a console, for example. The problem with minimizing tracks is that it doesn't really go too far toward solving the problem of keeping everything in view if you do have 40+ tracks (which I often do), and if (make that when) you ever need to move a video clip and all of its corresponding audio clips, you're still likely to miss some little event down on, say, track 43. Yes, grouping events can help avoid this problem, but it doesn't apply to every situation, and you can still miss an event when you go to group them.

Also, even though the bit about not knowing when an audio effect will end almost kinda-sorta makes sense (especially since Sony can't know what 3rd-party plug-ins are going to do), I still don't understand why Vegas couldn't, as a slightly inelegant alternative, truncate the end of a reverb application at the end of an event, so that you have to think ahead and allow for silence before applying the reverb--or at the very least, allow us to use a real-time preview on some effects. I have no doubt that quite a few hours of my life that I have wasted wading through tons of tracks would be returned to me if Sony had adapted either of these models from the beginning.

I guess I just have to learn to live with this, as painful and time-wasting and counterintuitive and absurd and ridiculous as it is. Surely there has got to be a better way, but it sounds like we're never going to actually get a better way to handle audio, because Sony has their stock excuse and their minds are closed on the matter.

Anyway, thanks again for digging up the info, Glenn.

[Brian Standing]

This may not help with your central concern, since you will still need to render out a .wav for an event-based Audio FX. However, you can help manage all the resulting "tiny .wav" files, at least on the timeline, by using Vegas' "take" function.

If you add the rendered .wav file over the original as a take (try right-mouse-clicking and dragging the file to overlap the original), you can easily return to the original, or cycle through multiple takes, by pressing the "T" key. You can also set Vegas to automatically display the take name, if you need to remember what you did to each one.

If they're truly tiny events, the rendering time isn't much, and once rendered they don't need to be again.

Just a thought.

[Jarrod Whaley]

That is a helpful suggestion Brian, and it does address at least one part of the issue I personally have with non-real-time effects.

Thanks.

[Josh Bass]

Oh yeah, I wanted to ask you guys about that. . .can you make Vegas 6 tell you the name of the clip/event in the timeline? 4 Did it, and I can't find an option anywhere in 6. What I'm talking is about is actual text written on the event in the timeline ("John CU take 6").


Also, doesn't FCP have the realtime audio FX?

[Douglas Spotted Eagle]

Vegas does have realtime audio FX, at two different levels, whereas FCP only offers it at one level. Problem is, Vegas offers event-based FX as well, but they're not real time. Therefore, it's a problem for some folks.
Coming from the audio world, it's not an issue for many of us.
To view take names/event/audio file names, select VIEW/ACTIVE TAKE INFORMATION or CTRL+SHIFT+I.