Month: July 2004

Now’s the chance for the people of Los Angeles to see a bit about what those of us in the sound industry do:

Los Angeles moviegoers will have an unusual treat this summer!  The Motion Picture Sound Editors and American Cinematheque will co-present “Big Movie Sound Effects: Behind the Scenes and Out of the Speakers.”

Here’s your chance to see – and hear – how those cool sounds for big science fiction movies are made.  Dane A. Davis, MPSE, and Gary Rydstrom, MPSE, will present excerpts from their Oscar and Golden Reel Award-winning work as supervising sound editors/designers/re-recording mixers on THE MATRIX and JURASSIC PARK.  In addition to discussing the processes they used to create the unique aural effects, our guests will also play sequences from both films with the sound effects only in order to give a clear and rare demonstration of the craft of motion picture sound.

So come and give a listen to what remains in these classic movie soundtracks after the dialogue and music are removed.  Experience the rippling waves from slowed-down bullets, the roar of a Tyrannosaurus Rex, and every little detail that gets lost in the final mix; as well as the stories behind them.

“Big Movie Sound Effects: Behind the Scenes and Out of the Speakers” will be presented in the Lloyd E. Rigler Theatre in the Egyptian Theatre at 7:30 p.m. on Wednesday, July 14.  The theater is located at 6712 Hollywood Boulevard, Hollywood, California.

General Admission:  $9
Seniors 65+ & Students w/valid ID: $8
MPEG and other guild & craft organization members w/valid ID: $6

You can find out more about the Motion Picture Sound Editors at our web site: www.MPSE.org and the American Cinematheque at www.americancinematheque.com.

Made possible with support from Universal Pictures, Warner Bros. Pictures, Worldlink Digital, DTS, and Dolby Laboratories, Inc.

Program subject to change.

You call yourself an Assistant Sound Editor and you don’t use BBEdit for EDLs and change notes? If you say that you’re still using Vantage, I’m going to smack you. Microsoft Word is what you’re using? Well, I guess I understand. We’ve all kind of grown up with that whole “Word is the program you use for typing things”. But seriously, yuck. Maybe you just don’t realize what you’re missing.

How about this spiffy little header it puts at the top of all printed pages? Pretty nice, huh? Ever dropped a stack of Vantage change notes and then tried to figure out the correct order again?

I barely know where to start with all the great text-fixing features it has built-in. Just take a look at all the options on this menu. Pretty great stuff.

But it’s real strength is in it’s support for Regular Expressions and AppleScript. Regular Expressions are complex find and replace searches that are much more powerful than a simple “Find ‘cat’ and Replace with ‘dog'”. In my last post I gave an example of a Regular Expression that could be used in Soundminer to split apart text with the format of "Title Description" into two individual “Title” and “Description” fields.

Just today I was trying to assemble some reels of dialogue using Titan. The source DATs were loaded with the filenames like “001 004/05.L.wav”—meaning Sound Roll 1, Scene 4, Take 5. Unfortunately the EDLs listed the sound rolls without leading zeros. So it would say “1” not “001”. Titan needs the EDL and filename to match so that things can be linked properly.

No problem with BBEdit and Regular Expressions.

Bring up the Find window. Make sure that “Use Grep” is checked. That’s what will turn on Regular Expressions.

Type this in the Find field without quotes: “^(d{3}s{2})(d{1}s+)”

And type this in the Replace field without quotes: “102”

Click the Replace All button.

In English you just told BBEdit: Starting at the beginning of a line (^) look for a group of characters, 3 digits followed by 2 spaces (d{3}s{2}), then look for a second group of characters, 1 digit followed by 1 or more spaces (d{1}s+). Return the first group (1), add a “0” and then return the second group (2).

This will find any single digit sound rolls in an EDL and add a zero to the head. Now you need to turn 2 digit sound rolls into 3 digit ones.

Type this in the Find field without quotes: “^(d{3}s{2})(d{2}s+)”

And type this in the Replace field without quotes: “102”

Click the Replace All button.

Now all of your sound rolls have 3 digits, leading zeros as necessary. All in about a minute. Try typing all those zeros in by hand and see how long it takes you. Plus the Find window allows you to save those Regular Expressions as Patterns that you can call up any time you need them.

I said that the second strength of BBEdit is AppleScript. Nearly all of its functions can be called from simple AppleScripts, including Find and Replace with Regular Expressions. That means you can make a simple “Add Leading Zeros To Sound Rolls In EDLs” Droplet. Drag your EDLs on to it and they’ll all be fixed in seconds.

Other ideas for BBEdit, Regular Expressions and AppleScript are removing those ugly boxes that show up at the end of every line of change notes made by the new Avid Meridien systems. Or splitting Picture and Track changes notes into two files. Or even interacting with other programs like Word, Excel or Filemaker. How about searching a change note for the new LFOA and entering it into a chart in Excel? All kinds of things can be done if you just take a little time to learn.

This is based on a little tip I posted on the Soundminer beta tester message boards:

How To Clean Up A Sound Effects Library Ripped In MTools And Imported Into Soundminer

Anyone who has ripped the sound effects from a commercial CD library with Soundminer’s Ripper program knows how great their data is. Anyone who has moved to Soundminer from MTools knows that the MTools data isn’t anywhere near as cool looking. Don’t worry, you can get things looking nice without too much trouble.

Make sure that you’re displaying the following fields: Filename, Description, Source, Category, Notes, Designer, Library, and possibly RecMedium.

Make a new database and scan a folder to work with. If you screw it up, you want to be able to go back to your original stuff.

Bring up the Admin window from the Misc menu. Set the Designer field to the creator of the CD library—most likely either Sound Ideas or Hollywood Edge. Set the Library field to the name of the CD library—Citi Trax or Impact Effects or whatever it is.

Copy Filename to Source. If the CD library is from Sound Ideas the filenames probably all start with “SI-” and end with “.L”. If they are from Hollywood Edge they probably start with “HE-” and end with “.L”. Remove these from the Source field. You can either Remove 3 Characters From The Start and Remove 2 Characters From The End. Or do a Find and Replace with nothing. The remaining data though not identical to what Ripper writes will be very similar. For example, a file from Impact Effects that was ripped with MTools will probably be named something like “SI-IE01_04_01.L”. After removing the extra characters the Source field would be “IE01_04_01”.

I also like to note which files were done by MTools for future reference. I set the RecMedium field to “MTools” for that reason. (A good-sized chunk of Cameron’s library was converted from Waveframe over to Pro Tools. I put “waveframe” in the RecMedium field of those files for the same reason. You never know when that information might be useful.)

If the description is all uppercase you can use the change case to Title Case function to make it much more readable.

With these few simple steps. You’ll be very close to what a Ripper ripped sound file looks like. It’s definitely a good idea to backup your data to the files themselves.

There’s one other thing you can do to make things look even better. Most of the CD libraries from Sound Ideas use the format "Title [Lots of spaces] Description" in the Description field of their database. All of this information including all those spaces—which make the printed catalog look nice but which are messy in a digital database—wind up in the Description field when the CD is ripped by MTools. You can put the “Title” into the Category field and the “Description” into the Description field (and dump all those extra spaces) by following these simple steps.

This process will give you about a 95% success rate. The biggest problem you will encounter is data that is not consistantly formatted. I found that the Audio Pro library is a big offender in this area. This process looks for 2 or more spaces in a row. If there aren’t 2 or more spaces between the Title and Description, it won’t work correctly.

Again from the Admin window, copy the Description into the Notes field. (Or another long field that you’re not using for anything else.)

In the Find / Replace box, check the RegEx box.

Type this into the Find field “(.+?)ss+?(.*)” without quotes.

Type this into the Replace field “1” without quotes, and click Ok.

This will return just the “Title” part of the of the Description. Take a second and check to make sure it worked. Like I said, if there aren’t more than 2 spaces or tabs in a row, this won’t work. You might have to fix some by hand. When you’re satisfied, continue.

Copy Notes into Category. Copy Description into Notes.

Type this into the Find field “(.+?)ss+?(.*)” without quotes, again.

Type this into the Replace field “2” without quotes, and click Ok.

This will return the Description part. Again, check it to make sure that it returned everything correct. (You might wind up with extra spaces at the head of this new description. You can use the Remove Characters From Start function to get rid of them quickly.) When you’re satisfied, copy Notes to Description and erase the Notes field.

That’s it. Definitely back up all this data to the sound files.

Those funky looking things you typed into the Find and Replace fields are called Regular Expressions. They can be a bit tricky to learn but they are amazingly powerful. Definitely worth a bit of study. You have a great command-line Unix version installed with OS X called “grep”. If you fire up your terminal and type “man grep” at the prompt you can read all about it. Google would be another great place to look for information. There’s also a book by O’Reilly “Mastering Regular Expressions” that might be useful.

Yesterday, Boing Boing posted a link to an article about surround sound research at UC Davis.

This new technology is a very interesting extension of binaural recording. Binaural is a two channel format that mimics the pick-up pattern of the human ears.

Stereo is two channel as well but it recreates a “wall of sound”. To record in stereo, you take to two mono microphones and position them in a “v” shape. The point is the head of the microphones. They should be at a 90 degree angle to each other. The heads should (obviously) be pointed at the sound source.

When played back from stereo speakers, a stereo recording creates, as I said, a “wall of sound” that is projected out from those speakers. Closing your eyes and facing the speakers, you could imagine the sounds happening just in front of you. Listening to a stereo recording on headphones places that plane of sound in your head. Put on a record that features a lot of stereo panning like Pink Floyd’s “Dark Side Of The Moon”. On headphones, you’d swear the sound was moving right through the center of your head. Here’s a little bit of “On The Run” from that album so you can see what I mean.

Binaural recording, as the article points out, typically uses a dummy head with microphones placed in the ears and positioned at the same angle as human ears. There are also “stealth” binaural recording rigs that put tiny microphones in what appears to be walkman-style headphones. Wearing the “headphones” puts them in the proper location for binaural recording.

The biggest difference between stereo and binaural recording is that binaural can only be played back on headphones. Listening to binaural recordings is like really being there with sounds going on all around you. It records in 360 degrees, so recording a sound of someone walking behind the dummy head would sound exactly like someone was walking behind you when listening on headphones. If you play a binaural recording back on stereo speakers it sounds really strange. It’s hard to exactly put your finger on what’s wrong but you can tell that something isn’t right with it.

This new technology from UC Davis records 8 or 16 microphones positioned in a circle, and during playback mixes the relative strength of each signal in real time based on the positional data from a worn sensor. With normal binaural, the previously mentioned sound of walking behind you would always be behind you no matter where you turned your head. This new technology would allow you to turn your head around to “see” who was behind you and you would then hear the sound of walking as if you were looking at it.

Fascinating stuff.

I worked on an IMAX movie called “T-Rex: Back To The Cretaceous” that had a limited amount of binaural sound in it. Being a 3-D IMAX movie, you were given goggles to wear which would make the 3-D images look correct when you sat in the theater. These goggles were much larger than the standard red and green paper ones that you often get for these kinds of things. There were actually tiny speakers in them that were positioned directly over your ears when properly worn.

When we were on the dub stage mixing the movie, we would actually have to wear these goggles (minus the 3-D lenses) so that we could properly mix the sound for that channel. For those who are curious, it was actually wirelessly beamed to the headset via an infrared signal. The idea was that 3-D visuals plus binaural sound should really place you in the movie. However I found that the soundtrack was so loud with the music and the dinosaurs stomping around that even with the speakers right next to your ears, they were easily drowned out by main speakers in the theater. It was an interesting concept that wasn’t quite realized.

In March 1999 I recorded a rock show at a little club in Los Angeles called Dragonfly. A band I knew said it was ok and I was very excited about the prospect so I showed up with 2 different DAT rigs and a video camera. Two of my friends came along to help me manage it all. One of the DATs included a stealth binuaral headphone setup. During one of the opening bands, I had my friend who was going to use the stealth rig, go practice using it. He didn’t quite understand the concept and he moved around alot. Put on some headphones and you can hear the vocals and instruments move from ear to ear in this recording as he looked around from side to side. (One of the best reasons to use a dummy head instead of a live body.) You’ll notice two things from the recording of Candy Ass covering The Runaways’ “Queens Of Noise”. First, unlike the Pink Floyd song above, when the sound moves it doesn’t pass through your head—it “rotates” around you. Second, if you listen on speakers, it doesn’t sound right.

It’s not a very good example of binaural recording. I know. But it’s the best I could come up with easily.