I put all the sound events that require more or less the same treatment - be it noise reduction or what have you - together on the same track. (Same way I sort video events.)
Also, I use Graphic Dynamics with one of the compression presets instead of normalize. I never normalize. Only the mad needs that.
TIP: If you need to remove very much noise, do it in several small steps, not one big. You can apply several instances of Noise Reduction on the same effects chain.
Tor

Don't know what it looks like at your end. but the connection was bad, so I pushed the button too many times. All of a sudden I had submittet the same message three times. So I deleted.

It does seem strange that audio events can't have their own fX, per event, like video. I generally do a combination of what jetdv and TorS describe, but many times each event needs its own "tweaking" and it is a pain to have to switch back and forth between Sound Forge and Vegas. I can create lots of tracks like TorS does, but that gets unwieldly, having to scroll around so much. It sure seems like audio fX per audio event would be the way to go.

I just took a clip, split it into several pieces and applied individual audio FX to each one. I must not understand the question here.

I assume you right-clicked on the audio event and then selected "Apply Non-Real Time Event". You then select your even from the plug-in chain. When you are finished with your settings and click OK, Vegas then renders a WAV file with the new audio. This then shows up as a "take" for that event.

This is not what we want.

What we want instead is a real-time fX, like what you get on the video tracks, where you can preview (although sometimes at reduced resolution) without rendering. You can do this on the audio track fX, but not on a per-event basis.

If you've discovered a way to do this independently for each audio event on a track, without having Vegas render a WAV file, then by all means let us know how to do it. It wouldn't be the first time I was too stupid to figure out how to do something in Vegas (actually, for me, this is unfortunately a problem not entirely restricted to just Vegas ...)

John:

OK, I see what you're after now. No, I don't know of a way to avoid the render thing, but it happens so fast even on my pokey P3/933 that it's not an issue for me. But yes, it would be a good thing for Vegas to have.

It takes even less time on my P4 2.8 GHz system, but the real issue isn't the time, but the inability to "tinker" with the sound. If you open the sound in SoundForge, you can tinker all you want, and do it in real time (which is what I want), but if you later need to go back and do something else to the sound, then you have to go back to SF, make changes, re-render, etc. And, heaven help you if you accidentally use the wrong sound take.

DELETED

Zippy,

I usually consider myself the king of workarounds, but I don't have anything to recommend other than what you are already doing. What I usually do is this:

1. Right click on the audio event, and open a copy in Sound Forge.
2. Play with the audio in Sound Forge, because I can do interactive, real-time previews.
3. In Sound Forge, create a preset for each fX I'm going to use.
4. Close Sound Forge, but don't save the file. Thus, at this point, I've done nothing in Vegas (i.e., I haven't created a new take).
5. In Vegas, depending on how many different sound files I have that need to be fixed, I either create a new track and move audio events to that track, or I right click on each one and then "Apply non-real time Event Fx." In either case, I then assign the Fx that I used in Sound Forge, and once I get the dialog for the particular Fx, I just assign the preset that I created in Sound Forge.

This approach gives me the real-time preview that I find absolutely necessary to get good results. However it is a kludge, and I hope Sony adds event Fx for sound files in some future release. The fact that 4.0 doesn't have this seems more like an oversight than anything else, and hopefully won't be that difficult for the engineers to add (something that often helps decide what gets into a given release).

DELETED

The video event FX do not alter the video files on the hard disk, they apply the FX in realtime during playback. The non-real time audio event FX actually perform a destructive edit to the wave files. So apart from the reduction in flexibility in not being able to preview these FX in realtime, the other drawback is that the audio source files could be degraded unnecessarily if you just want to try out different FX. This is why you should make sure to save copies as takes and play around with the copies, rather than risk messing up the originals.

Richard Hunter

By default, Vegas saves the altered file under a new name and adds it to the timeline as a take.
Tor

Yea, when you do this, Vegas will prompt you for the name of the .wav file for the take so it should be obvious to you where it is going and what its name is. Vegas never overwrites the original file so you can make several of these and toggle between the takes to hear which one you like best. That’s the whole concept of takes.

I have to imagine that if Vegas could do this in real time it would but audio FX are rather CPU intensive along with all the other video streams that are running. I guess today’s computers could probably keep up but Vegas has been working this way since the early Pentiums and perhaps even 486’s. So some of this may be for legacy reasons. I could see where if you had 100's of events and applied an FX to each it would bog down the CPU pretty quickly.

It would be nice if a future version of Vegas let you make the decision of applying these FX in real time or not.

~jr

I have to imagine that if Vegas could do this in real time it would but audio FX are rather CPU intensive along with all the other video streams that are running.

The point is, Vegas already DOES do this, in real-time (so does Sound Forge). In Vegas, you can get exactly what we all want when using the TRACK fX. The problem is that, unlike the video events, Vegas 4.0 for some strange reason does not also provide fX control on a per-event basis. This forces all of us into the various workarounds already described in this thread -- the equivalent of the pre-rendering for video, but unlike video, this "pre-rendering" for audio is needed just to hear the results of the change. By contrast, if I apply an fX to an entire track, I can move the various sliders up/down, left/right and immediately hear what is going on. Believe me, once you've edited sound this way, you'll never be able to get back.

I'm sure that this is an oversight, and I bet the Sony engineers can easily add this one in the next release.

BTW, about CPU power, I still use Sound Forge on my 450 MHz PC, and it is able to easily handle all the sound fX that we're talking about here, in real time, without any problems.

I think the killer is trying to handle the workload of video and audio. With video you can dop the frame rate if the CPU cannot keep up and the result is still useble. Our ears are another matter, can you imagine the complaints if VV dropped frames from the audio on preview.

What might be a way around this is to be able to preview the clip with the fx applied in solo prior to rendering it. I imagine you wouldn't even need the video preview running while doing this.

> but unlike video, this "pre-rendering" for audio is needed just to hear the results of the change By contrast, if I apply an fX to an entire track, I can move the various sliders up/down, left/right and immediately hear what is going on. Believe me, once you've edited sound this way, you'll never be able to get back

This is the way I do edit sound in Vegas 4 today. I can add a non-realtime effect to an event and I can preview and tweak the audio all I want in realtime with several effects playing in the chain while I listen to how it sounds. When I’m done, it prompts me for a filename and does the render and replaces it as a take. So I don’t understand why you think you have to pre-render just to hear what it will sound like. Perhaps I’m just not following your point.

~jr

It's actually not easy, because of the nature of how it works. This has been a question since version 1.0 in Vegas. For instance, how does Vegas know to continue a delay past the end of the event? Because if it didn't, then the delay would just CHOP off at the end. That's just one small example. But it's been a 5 year question....not a new request. Were it easy, I guarantee it would be there. I've been part of discussions on this issue. Doesn't mean it can't be done, just that it's a fairly big challenge, and has lots of downsides associated with the upside too, in terms of engineering.

What are you guys using for noise reduction

I have a bunch of clips with computer noise under narration.

Sony's Noise Reduction is absolutely wonderful, awesome, terrific, great!

DELETED

The last question:
I assume you mean events (or clips which seems to be the fashionable word, with so many ex-premierers around). One tip is to make the audio overlap, Vegas will then automatically create a crossfade which helps in smoothing the change. If the change is still too dramatic, I put the two events on different tracks and try to tweak them into harmony.
If I must resort to using noise gate, (which silences everything below a certain dB point) I get ugly gaps of total silence. I cover them uo by adding a new layer of noise (atmospere, some relevant background noise or even music). That layer - stretched out over the entire scene will make the audio event appear to be more continous.

The first question:
Noise Reduction is pretty self-explaining - like its name. When you've read the manual (which is fairly short) spend one ot two afternoos experimenting. Remember it's often better to use it several times in smaller steps than to remove all noise in one go.

The second question:
Very tall order. I think I'll pass, expecting someone to come up with a good link or something. For starters, the Vegas manual has a graphic representation of the audio signal flow. I think it's on page 35.
Tor

DELETED

You add Noise Reduction more than once on the same bus or track. Each instance reduce the noise a little (the default is -12 dB I think) and leave something which the next instance can analyse more precisely.
Tor