Web and Video

Flash CS3 and Soundbooth CS3 : designing content with great audio — mostly brought to us courtesy of Jason Levine, with some assistance from Greg Rewis near the end to cover the Flash bit. Session was mostly on features in Soundbooth (which are quite awesome). So on to the notes… please excuse the mess - haven’t had time to clean as the session ended about 15 minutes ago and I am just hitting post

Soundbooth — Audio for flash people. How easy it is to do great sounding audio, export as FLV/MP3. Ways you can manipulate and modify audio are simpler than ever before. Don’t need to know a lot about audio to make things sound great.

Like to showcase “non audio creative professionals” are the target audience for Soundbooth. Just means your main focus is not audio - but you occassionally need to do certain audio tasks.

Identify things visually and fix them visually. Use your eyes

Five most common tasks are available in the Tasks panel (such as change pitch and timing, add sound to a file — “autocomposer”, clean up audio, create a loop, remove a sound). Purpose of loop - make a seamless loopable file without a click.

You’re going to use Photoshop knowledge to do all of these tasks visually - using “spectral frequency display”. You can see frequency over time in Soundbooth - frequency is on vertical axis and time along the horizontal. Illustrates what sounds look like in the spectral view in SB. Represents kinds of noise. Color is amplitude — closer to yellow/white, the louder the amplitude, opposite for black. Orange is at a softer amplitude, so it is harder to hear in the sample file being used. Going to repair this file so you can hear the orange stuff - which is a voice.

Goes to clean up audio in tasks panel. SB can scan the file, and try to remove noise automatically. Showing Noise dialog. How aggressively is SB going to try and remove those sounds, and reduce by what. So then you have a very quiet audio after they clean it up. So then you can try to bring up the audio that is underneath  — now you can hear what is underneath all that noise.

More advanced - you can understand where the noise is by itself “capturing a noise profile”. You can right click when you have a selection and capture the noise priont (take a profile)  with right-click. Applies to everything - hum, rumble, hiss. Then you click “remove a sound” - opens up a new view. Then go and highlight the entire file and you can increase the wave form just by drawing on the screen.

Removing individual sounds:

Opens a tape recording of Beatles, and it has hiss in the background. We’re able to take a profile, but instead you can use adaptive noise reduction and adjust it in the Noise dialog until you have the sound you want. You can preview your settings as it plays, and make it sound like you want. If you have noise by itself, you can use it for a profile and get much better reduction.

Photoshop tools for sound removal:

Use remove a sound to take out something like a squeaky door. Opens spectral display, so you can actually *see* where the squeak is. No other way to do this - can’t use filters/equalizers to do something like this. Pitches begin to look like images so you can remove things specifically.

So how to remove? If you use the marquee - can’t get tight enough, would remove too much noise. So you use the lasso instead — probably want to use a wacom tablet or something to be more exact. Checkbox “play selected frequencies only” so you know exactly what you’re removing. Test it - and we have only that squeak. You can take it and paste it elsewhere to clone the sound. Reuse the sound as  a sound effect. To get rid of it, take the volume and drop it down and the squeak is gone.

Another way to do this. Use the healing brush to remove the squeak. Have to be precise with the selection - limitation on the brush size. You can use auto-heal - analzyie the audio and then heals it. Doesn’t effect anything on bottom channel either. And it’s totally gone. And didn’t require any audio knowledge either.

Another couple restoration examples:

Piece with clicks and crackles (vinyl) — wireless, you’ll get lots of little clicks. Use the click and pop filter, use the slider to remove them, then click OK to process the file.   Same thing for removing rumble (mic rumble and pops) - another filter. Easier to get in there, fix, and get out.

Live piano recording, and there’s a cell phone goes off. You can pretty much highlight the ringtones - identify them digitally and remove them easily.

Dialog:

Lots of noise in the background, cell phone in the background. Same thing, select, drop the volume down, auto-heal.  Use adaptive noise reduction to remove enough of the background noise.

Sounds close and boxy after you remove to much noise. Using some effects to fix that. They were using a boom mic, but the volume of his voice is all over the place - soft and loud. One of the first things you can do to make volumes the same is highlight a section, and use the onclip volume tool to visually match the audio levels by redrawing the size - will just have the same consistent level by just matching it visually. If you can get it visually approximately the same, after you compress it for the web, people won’t notice.

Then you can go to the history panel to look at what the changes were. Non destructive so you can step back in time.

Fade in / Fade out.

If you do a fade in and fade out you are respecting the 0 cross - it ramps up from negative infinity and ramps down to negative infinity so you don’t have a click.
You can trim off the beginning and end, grab the fade handle to draw the fade curve for how the audio fades in and fades out.

Presets/filters:

Apply them non-descructivity. Only apply when you export them to FLV or whatnot.

Have a ton of presets for reverb, mastering, voice, EQ and so on. Stuff to give you clarity, fix muddy voices, clarity, prsets for male and female voices. Preset for sivvelence (”de-esser”) that takes the “s’s” out of a voice.

You can stack the presets to make layered effects, so you can apply multiples. So, if a voice sounds like it doesn’t have an environment (if you pull out too much noise, going to make file sound artifacted). So you can apply simple effects - reverb. You get a slider to set how much reverb you want to apply. Or you get a bunch of presets instead and use them instead — you can choose “Dark room, bright room, clean room” and so on. Or you can use an advanced reverb where you have a bunch of sliders that you can use if you know what you;re doing (or want to play around).

Then you can apply the effect (which is descrutive).

Now make the whole thing louder — at the bottom right of the interface, there is a louder button that you can just click to make things louder.

Now adding some cuepoints to prepare for Flash.

Greg Rewis.

Opening Flash and AS3. Selects an MP3, but don’t want to import the MP3. Do not do that, because if you put the sound into the Flash file you bloat the size of the SWF.

Hard way - ActionScript.
- Writes some code to import flash.media and then URLRequest the sound file. Then create the sound object. Say that it will hold the file that you’re requesting. Then you play the sound.
<didn’t quite get to cuepoints>

(2 votes, average: 5 out of 5)

 Loading ...