Aural Stage Studios is a team of audio designers, producers and engineers experienced in all aspects of audio, including theatre sound design, audio post-production, sound effect recording and design, voice recording, Foley and studio recording. In addition to the crew at Aural Stage Studios, we have access to a wide array of talented writers, voice actors and musicians who produce exceptional results for our award-winning audio dramas or just about any project we wrap our warped, little heads around.

What does Aural Stage Studios do?

Aural Stage Studios showcases writers and performing artists in an inexpensive and easily distributable medium with potential for world wide exposure. The crew and I stand by it, and we continuously strive to do so in unique and interesting ways.

Aural Stage Studios has seen far too many artists sell their talents short. The “dime-a-dozen” model for acting has convinced actors that acting for free is necessary. Worse, it has convinced producers that actors, designers and talent can be paid last, if there is enough money to do so. We’d like to help remedy that.

How does Aural Stage Studios accomplish what it does?

New Media allows us to distribute world wide with a few quick uploads. Mp3 players, Itunes, Last.fm, Amazon mp3 and other easily accesible media services make it possible to reach audiences in unexpected places. Online collaboration services like Google Docs, Skype make it possible to work and record with actors worldwide with outstanding clarity and convenient functionality.

How is Aural Stage Studios different?

Artists associated with Aural Stage Studios are paid for their work. While we may not be able to yet afford scale salaries for talent, our attempt is to pay actors what we can afford to, as well as provide an opportunity for professional experience and worldwide exposure. We also attempt to provide our casts and crew with the tools and services they need to work and promote themselves professionally, whether it is access to studio quality equipment or professional sound design and audio editing for reels and demos.

Why should people fund Aural Stage Studios?

Aural Stage Studios pays for it’s talent, but we currently do not charge our audiences to listen to our work, ensuring a wider dissemination in a culture of consumers who like their free. That means that the producers at Aural Stage Studios are picking up the tab.  And that’s where we can use your help.

While audio productions are inexpensive to produce, they’re not free.  A professional audio drama can run anywhere from $2000 to $10000 to produce.  While that’s a fraction of the production costs of other mediums, it takes it’s toll on our bank accounts.  Many of our crew have been gracious enough to volunteer their time and talents in exchange for services and support, affording us some of the more expensive aspects of production for a very low cost.

Funding through donations, advertising pand in-kind contributions helps Aural Stage Studios continue to make award-winning programs with some very talented artists, as well as pay our entire crew, including those who so have so graciously donated their time in the past.

• Aural Stage Shows

  • 1918
  • Aural Traditions

• Subscribe

  • Add to Google
  • My Yahoo
  • Add to iTunes
  • Add to Zune
  • Twitter
  • Google Profile
  • Facebook

First, were going to need a few tools to simplify our journey.  You’re going to want to grab is called flashget .  Flashget has a lot of nice features, but the main feature I enjoy is the ability to quickly navigate, filter and mass download based on file type.  I also use this to mass download free stock images and anything else I need to download mass amounts of files from.  For you mac users, you’ll have to find a program equivalent, but the idea is to be able to mass download.  There are also Firefox extensions like Download Them All and a host of others.

When you install Flashget, an icon floats on your screen that looks like or similar to this:

Right-click on the icon and select Site Explorer:

Then the fun begins.  In the address bar, type http://www.soundjay.com:

Sound Jay has a ton of royalty-free sounds and music.  The library is always expanding so check back often to update your library.  Check out his terms of use which are extremely reasonable.

I highly recommend creating a folder structure similar to the Sound Jay setup.  You may need more or less folders, but as your library grows, it’s extremely important to stay organized.

So we’ve pulled up the Sound Jay page and now we’re faced with a bunch of html page we need to sort through.  The nice thing about Flashget, is that it does that for you in the directory tree on the left.  Yay Flashget!  As you click on the directories more directories come available as the site shows more pages. It all stay fairly organized and keeps everything pretty accessible once you get used to it.  It may be helpful to run a browser with the site running in it to help you understand what’s happening.

Click ambient-sounds.html.  First, you’ll notice a new page (ambient-sounds-2.html) and a jumble of files on the right.  Two possibilities here:

1) You can click the Type column to sort by type, then slect the files you want to download (I highly recommend WAV over MP3, but that’s your choice).

2) Click the edit menu and select filter.  In the box that follows, type the file extensions you’d like to filter separated by semicolons (.wav;.mp3).

From there, you can select all the files to download, or control-click to select specific files to download.  Click the green play icon, select file > download, or right-click download to start the download dialog.  Select the folder you want the sounds to download too and click OK to download:

Now continue for the remaining folders to build yourself one heck of a base sound library.

But wait!  There’s more.

Navigate Flashget over to http://www.pacdv.com/sounds/ and grab a ton more.

Still not enough?

How about 100 free professional sound effect from Blast Wave FX, owned by legendary sound effects man Ric Viers?  What What What?  Oh yes.. Check it out here: http://www.blastwavefx.com/Free-SFX-Download-Pack-%231/t3/articles.html

Additionally, if you register for contests at the Blast Wave FX site, they give you not only give you a bank of royalty-free sounds to work with, but also a code to get one of their download packs free (or credit toward a more expensive pack).

Sound Ideas, a standard for broadcasting and theatre companies doesn’t have free sounds.  What they do have is quite possibly the most useful piece of software in my sound arsenal, MetaDigger.  MetaDigger allows you to categorize and organize your sounds based on whatever keywords you enter.  Additionally you can search by keyword in order to find all the audio in you collections that meet a specific criteria (“splash” “blurp” “sploosh” “sword” “clank”) once you have entered those terms into the files metadata using the program.  It also allows you to audion those sounds and send them to a workfolder for use in your editing software. Pick it up at http://www.sound-ideas.com/metadigger.html.

Got more suggestions?  Need more ideas?  Send me your comments below and we’ll compare notes.

Someone Who Cares was inspired by a film project by 1918 actor Patrick Pizzolorusso (Henry) and Director/Producer Samantha Mason.

Sure, if you’ve got a bundle of cash to spend, you’ll want to invest in professional grade materials from Auralex, ATS or Soft Sound (among others), but for those of us in the home studio who don’t have a bundle of cash, there are alternatives.

1) Fill your room with fabric – the heavier the better.  I cover my windows with some one inch fiberboard (masked with fabric for more absorption and better looks) and then supplement that with some heavy curtains.

2) Pillows and cushions – A big old comfy couch give you a place to relax when your spending ours in the studio, is a place for the talent to sit and it absorbs a lot of low to high frequency sound.  The more pillows the merrier.  I have a futon in my studio, which is also a good place to crash when you just can’t edit anymore.

3) Quick, cheap mid to high frequency acoustic panels:

You’re certainly going to eventually put in some bass traps to handle low frequency absorption.  But getting a room under control is as simple as hanging some acoustic tile in your problem areas.  For me, I have a big chimney running through my studio which causes some awful rings between it and the opposing wall.  For the most part, these are problems that occur in frequencies above 1kHz, which acoustic tile handles quite nicely.

The one problem here is that acoustic tile looks absolutely awful.  Mine came with the kitchen, which I quickly replaced with luan paneling, which looks much better. However, there was plenty of extra panels in my rafters, so I saved a few bucks there.

Simple solution?  Cover them with fabric, the thicker the better.  Lay out your panel on a corner of the fabric to allow about an inch and a half of fabric overlay. Cut the remaining two sides with about an inch and a half of overlay.

Fold the overlay of fabric over the back of the tile and staple or glue in place, working frist on one side, the the opposite side to pull the fabric taut to get rid of any wrinkles.

Looks better.  Doesn’t it?

That’s really it… Except, of course, hanging the boards.  Have an assistant help you fin your problem spots, by clapping to hear the ringing areas, and placing the board until the ringing decreases or is eliminated.  It’s a goo idea to also place a few from any blank wall that faces another wall.  You don’t have to cover the entire wall, but get a good surface covering going to reduce reverb and ring in the room, which will ultimately effect your recording and editing by adulterating the sound.

Here’s my finalized wall with the curtains and futon.

Have Fun!

Check out this option:

Start with one Accu Grip Handheld Tripod -  $14.94 (plus shipping)

Add one 1/4 inch female to 5/8 inch male adapter - $4.99 (plus shipping)

And finally, one universal shock mount - $14.99 (plus shipping)

Put it all together to get:

Total Cost (w/o shipping):  $34.92

Plus! The parts can be used for other configurations.

Half the cost… twice the application.

Let’s look at out reverb module:

With some variation, these are the basic settings for most reverb modules.  Some are much more expansive. Others only include a delay time.  Still others only include presets (which aren’t going to do you much good if you’re trying to build for yourself.  Avoid these like plague since they are rarely accurate or intuitive.) I’m using Adobe Audition’s basic reverb.

Pre-delay time:  These are your early reflections, as calculated in the second part of this series.  If you use the delay like me, leave this at zero.  Otherwise, you can enter your calculated pre-delay here.

Diffusion:  While I use diffusion, I rarely calculate it.  The idea is that diffusion is based on the shape, size and contents of your room.  If a wall is flat, there is going to be relatively little diffusion.  If there are lots of materials of different sizes and shapes (like a library wall lined with books), diffusion is going to be higher as it changes how the sound is reflected around the room.

Perception: This covers how absorbent or reflective your space is.  In general, EQ covers most of this, but you can use the perception slider to adjust for more metallic rooms(reflective), rooms with more curtains (absorbent) and anywhere in between.

Decay:  I saved this one for last, as it requires the most care.  This is your bread and butter in the reverb module.

A room’s reverb is measure in terms of RT60.  That is, the time it takes for a sound to fall 60 dB.  You can do the math if you want… but there’s no reason to, since we have the interwebz to calculate for us.  I typically use this calculator as it gives multiple values for different frequencies.  That said, for most purposes the calculated values for 1000 Hz should serve most purposes.

Enter you dimensions and go!  If you want to add elements such as windows and doors, the calculator will take those into consideration as well.

For our 10 x 10 x 8 room, I got an RT90 time of 1.46 seconds or 1460 ms.

Set your wet to 100 and your dry to zero.

Trust trust trust trust trust your math.  Resist the urge to over reverb a space as it will make your room sound empty and your voices sound hollow and far away.  You can bump up your perception to give things a little more dramatic reverb, but to keep things natural, don’t over-saturate your sound.

There it is… your room is set:

Dry Track

With Delay

With Room Eq

Full Room (With Verb)

]]> http://www.auralstage.com/2011/08/creating-realistic-environments-%e2%80%93-part-five-%e2%80%93-reverb/feed/ 0 http://www.auralstage.com/2011/07/creating-realistic-environments-%e2%80%93-part-four-%e2%80%93-absorption-and-eq-pt-2/ http://www.auralstage.com/2011/07/creating-realistic-environments-%e2%80%93-part-four-%e2%80%93-absorption-and-eq-pt-2/#comments Tue, 26 Jul 2011 16:24:32 +0000 Aural Stage Studios http://www.auralstage.com/?p=584 Once we have our base EQ laid in, it’s time to use our coefficients to figure out what else to remove.  You will get much much much much much much better results if you resist the temptation to add to the EQ.  I nearly always subtract from.  Adding EQ get’s muddy very quickly and should be reserved for layered sounds, and giving sound effects or instruments weight and presence.  For natural voice purposes.. and especially for purposes of creating a natural room, we substract.

Let’s go back to our 10 ft x 10 ft room (hardly an ideal sound listening environment, mind you, but the math is easy for illustration).

For purposes of simplicity, we’re going to concern ourselves only with the two parallel walls.  For more accuracy you would want to consider the reflection and absorption of as many surfaces as possible, but you can get effective results by synthesizing just a basic environment using the direct reflective surfaces.

Let’s go with the basics:  Sheet rock walls.  This really helps to narrow our numbers down as three of the surfaces are going to have the same coefficient and effect the sound relatively equally (in an ideal world), leaving only the carpeted floor to deal with.  Keep in mind whether your room is insulated (and what type of insulation is used.), popcorn ceilings are going to diffuse sound just as they do with light, and underlayment under the carpet is going to also affect the sound of the room.  Let’s get some numbers:

Walls and Ceiling (insulated no popcorn):

Next we add the elements of the wall surfaces together:

Now we need one more number to determine our EQ settings, the dB Range of our EQ.  This is typically 24.  I set mine manually to 48 for more dramatic results and greater control in the margins.  There are two ways to determine your range.  Some allow you to select a range, which means the range is typically displayed on the interface somewhere.  Some are hardset, but you have the numbers along the sliders as a guide: a range between -24 and 24 dB indicates a 48dB range.  Now we half that (24) for use in calculating our EQ settings.  To do so, all we need to do is multiply half the range (24) by our total coefficients.  This is how much we will reduce our EQ for each frequency:

Next, we plug our numbers into the EQ we have set up on our FX bus so the EQ only effects our reflected sound and not the direct sound.  Round as needed.  It may require some tweaking to get the sound to “sound right” since we’re really using a very empty room basis for our EQ settings:

You can just set these levels as-is in a 10-band EQ and roll with that.  It will still get fairly effective results.  Personally I like to smooth out the edges with a 20 or thirty band by rolling off the top and bottom ends and smoothly rolling the in-between frequencies to form a nice curve for blending:

Here is the result:

Dry Track

With Room Eq

One final trip into reverb to go.  And we’ll have the basics of creating an environment from scratch.

The level of frequencies a material absorbs is referred to as the absorption coefficient.  The coefficient of a material is determined by calculating the difference between frequencies through a microphone with an extremely flat response and the same frequencies filters through the measured material into the same microphone.

Fortunately for us, this has pretty much been done for us.  There are several sites with charts that show these.  I usually use this one.

Now comes the hard part… the math (Oh no!)

First we need to know a few variables.

Since the typical instrument we’re dealing with is the human voice, we can start by rolling off our eq to eliminate everything out side this spectrum.  The human voice ranges roughly between 80 hz and 1100 khz.  Vocal harmonics range up to about 2500 hz and the sybillant frequencies (think t’s and s’s) range from about 3-7 kHz.  Therefore we can fairly safely roll off everything below 75 Hz and above 11kHz on our vocal tracks.  This eliminates a great deal of ambient hiss from the high end, and rumble, hum and plosive popping at the low end.

These numbers will vary with other instruments and sounds, so I highly recommend having a chart on hand if you’re dealing with other sounds that illustrate the frequency spectrum.  I also recommend mixing EQ through a good set of source monitors, as headphones tend to make everything sound good. A headphone mix can often sound hollow and empty through a different set of speakers, where a good set of flat response monitors will produce decent results regardless of the speakers you listen to the final mix on.

With our basic rolloff applied to the direct track (as opposed to the FX bus we set up), you should hear very little difference in the recorded voice and the EQ’d track.  Any changes heard, should vastly improve the aural qualities by blocking out frequencies that are above and below the natural voice range.

Dry Track

Rolloff EQ Track

Depending on your setup, EQ is either the first or the second effect in your FX chain.  It’s important not to confuse this EQ with the EQ for you track.  Track EQ affects the sound and color of the instrument, voice or sound effect coming from the track and is either typically placed inline just after the inserts.   This EQ, however is inserted as part of the bus effect chain we created in Part Two.

Where you place the EQ is based on which way you choose to set up your effects chain.  If your like me and want to control your delay seprately, then you should place the EQ directly after the delay and before the reverb. If you prefer doing everything in the reverb module, then place the EQ in the chain first, followed by the reverb.

Why?  Our EQ in this case has less to do with coloration and more with absorption (though it obviously does color the room sound).  What we are simulating with the EQ is how much of each frequency is getting bounced from the surface and how much is getting absorbed by the surface material. So we are essentially carving away those frequencies that get absorbed.

Whether you use parametric EQ or graphic EQ is strictly up to user preference.  Many engineers prefer parametric EQ for the precision control it gives over sound.  I use Parametric when I need to really dig out problem frequencies in a sound, or when I need to highlight or roll of frequencies for a desired effect (bringing instruments forward, controlling sibilance, eliminating plosive pops, etc). For general sound shaping (such as room environments), I use a 20 or 30 band graphic eq, which allows me to visually shape the sound quickly, without having to dial in each frequency first.

Reverb, choice is also completely up to you.  For the most part, I have an affinity for ProTools DVerb for vocal tracks.  In Audition, I tend toward the reverb that has the least options, since I tend to control most of the options using other effects, so I lean toward the standard Reverb rather than Studio Reverb or Full Reverb.

That about does it for the effects chain on this one.  Short and sweet.

Next time, we’ll be getting into more math with some absorption and EQ settings.

To calculate the delay time, we need the same two bits of information: the distance the sound travels and the time per foot (.9 ms).  Like before, we’re simply going to multiply the distance and time together.  But we have to do some math to get the distance.

What we need to figure is a reflection point on the wall about halfway between the source and the receiver.  Since we don’t have this information, we need to calculate it using the information we do have: the distance from the source and the receiver (6ft), and the width of the room (10ft).  Since our source and receiver are centered, we can use 1/2 the width (5ft).  And since we’re calculating a point on the wall halfway between the source and receiver, we can half that distance as well (3ft).  Now we use the Pythagorean theorem to calculate the distance from the sirce to the halfway point on the wall.. and double it to get our length:

Pythagorean theorem: a² + b² = c²

5² + 3² = c²

25 + 9 = c²

34 = c²

c = about 5.83 ft

So our distance to the wall is about 5.83 feet.  We double that to get the distance to reflect back to our receiver, about 11.66 feet.

Now we do our simple math:

distance: 11.66 ft x time per foot: .9 ms

about 10.5 ms

Now we have our delay.  Right?  Not quite!  But we’re almost there.  What we have is the time it takes for the sound to reflect.  Our delay is the difference between this time and the direct sound (which we calculated earlier).

So our delay time  is:

10.5ms – 5.4 ms = 5.1 ms

an even 5 ms will work just fine

So these are the numbers you would plug into your pre-delay or your early reflections.  If these are not an option, you can add a delay to the chain before the reverb and eq. I prefer to use a separate stereo delay than to use the reverb’s built in early reflections anyway, as it gives me more control over the type, position and decay of the delay, but for most, the pre-delay or early reflections will work fine.

The volume will of course also change over the distance, again decreasing by 3 dB for every doubling in feet.

So for a distance of 11.66 feet (round to 12) we’re doubling about 3.5 times (2,4,8.. then 1.5 to 12) for a volume shift of about -10.5 dB.

Plugging the numbers in:

Alright.. now the technical stuff…

Create a separate effects bus, this allows us to put all of our sounds into one space, which creates a more natural sound and reduces processor burden.

Create a send in the source track that goes to the new bus. Set the send type to post so that changes in volume effect the volume of the send as well.  Also set the Send volume to match your change in dB calculated for the delay (-10.5dB):

Add a delay to the FX chain of your bus.  Then apply your calculated delay time settings and set the mix to 100% (your mix is controlled by your calculated volume that we applied to the send):

That’s it!  The difference is (and should be) subtle.  It’s about getting a natural sense of the space rather than sounding affected.  And a room this small will have an especially small difference in sound.

Here are samples of the difference based on out calculations:

Dry Track

With Delay

Still to Come:  EQ, Reverb and Tweaking

In reverb modules, there is often a control for early reflections.  Early reflections are the  first sounds the receiver hears reflected after the direct sound.  These are typically reflected from surfaces only once or twice before they reach the listener and arrive after the direct sound.

Longer reflection times indicate a larger space.  Conversely, smaller reflection times indicate a smaller space.  If we know the distance from the speaker to the listener (or the source to the receiver), it is fairly simple to calculate our early reflections.  Don’t get put off by my lengthy explanations, it’s really simple. I’m just verbose and want you to understand what we’re doing with the numbers.

For illustration we will take a small room, 10ft by 10 ft.  We’ll place the source at the center of the room about a foot from the front wall.  The receiver will stand about 6 ft from the source.  This can vary depending on your needs.

First we need to calculate the time and volume change for the direct sound.  Volume change can be omitted for most purposes, but since I like to move my characters throughout the space, I typically account for the drops in volume from the start and then adjust as I need to for clarity in the final mix.

To calculate our direct sound, we need the distance from the source to the receiver (6ft) and the speed of sound (about 1116 ft/sec).  Actually what we need is the length of time it takes sound to travel one foot… so you can just write down or memorize .000896 seconds, or  about .9 milliseconds.

Now we multiply our distance 6 ft x .9 ms (time per foot) to get an initial time of 5.4 ms.  That was easy.  Right?  This is the time it takes sound to travel from the source to the receiver six feet away.  We’re not going to use this number since it won’t affect how the audience hears anything without a visual reference. but we’ll need it later.  So we’ll just set it aside.

Now the change in volume…  For every doubling in distance, our volume drops 3 dB.  I use a standard reference of 1 ft and calculate from there.  In this case, we’re doubling 2.5 times.  (1 x 2 = 2, 2 x 2 = 4, 4 x 1.5 = 6) for a total change of -7.5 dB.  If we set our reference to 0 dB at 1 ft away, then a source 6ft away will sound 7.5dB quieter.

You can either dial these in (reduce the volume for the track by -7.5 dB) or ignore this entire step, depending on your needs.  I typically ignore it, UNLESS, I have characters moving through space (Frau Hera’s footsteps in the beginning of episode 1 of 1918 or if I have multiple characters at different distances in the same space.

On to the next section–>

“Are there any reverb options with programs like Audacity or other various plug-ins available that can create smaller sounding room environments such as simulating the interior of a car or a low-ceiling basement, etc? It would be cool to have the capability of creating a realistic acoustical environment for the various scenarios that are recorded without actually putting my actors in a car.”

Here are some options with various levels of math and complexity to use as possible solutions:

Option One – Location Recording:
My friend Fred Greenhalgh over at Finalrune Productions would (and did) say to go ahead and put those actors in the car?  Location recording certainly takes a lot of the guess-work out of recording and Fred has certainly had a lot of success using the technique.  However there are some engineers (including me) who like to have as much control over the mix as possible.  It’s nothing against the technique, so much as my own personal lack of willingness to give up control.  There are some engineers (like me) who hit record only to have a B-1 ineveitably pass overhead during the take.  There are still other engineers (um.. yeah.. me too) who just can’t access some of the environments (believe me.. I looked for a ww1 plane in the area) that our shows take place in.

Option Two – Worldizing:
Worldizing is a pretty decent balance between location recording and control over a mix.  I frequency worldize to get just the right nuance of a specialized environment.

Worldizing involves placing a set of monitors and microphones in the intended environment to produce the desired sound.  Set the monitors at the source position (say the front seat of the car) and use a balanced stereo pair of mics in and X/Y position (or a Zoom H2/H4) at the receiving position (maybe the back seat or the passenger’s seat).

Play the tracks you wish to worldize  through the monitors and record using the stereo pair and voila! You’ve created a worldized track.  Then all you need to do is worry about the bombers that keep flying overhead when you hit record.

Option Three – Convolution:
I find convolution reverb especially useful when I’m working in post-production for film, but I’ve used  them in a pinch to generate alleyways and lounges.  Convolution uses what’s is called and impulse response to create reverb.  An impulse response is a sample of a wave form taken in a particular environment, usually a gunshot, hand clap or broadband white noise.  What’s nice about convolution is you not only get the reverb, but typically the absorption, eq and delays associated with an environment.

If you’re looking for a good convolution verb, you have a few options.  If you’re running Audition, you’re in luck, as it already has a convolution reverb built in.  You can also check out industry standard Altiverb , SIR (Sir1 is a free simplified version), or I personally use Liquidsonics Reverberate (Again, I use the full version but there is an LE version that works great, albeit with some screen nagging).  They take some getting used to, but once you get the trick of using the settings very subtly and conservatively, you get some very good results.

Option Four: Math
I get very good results by doing the math.  The math is pretty easy, but there’s a lot of it.  And you have to take  all the variables into consideration.  For instance in small environments, the decay rate is so fast, you almost never have to use a reverb.   For outdoor environments, most of the reflecting sound gets absorbed, as do most of the eq frequencies between the source and the receiver.  Once you get used to it, though, the process is not all that difficult and I create tables with the settings I like so I don’t have to recalculate every time I attempt the same environment.

The downside to this  format (other than doing the math) is that math can make things a little too perfect to the numbers, but not to the really real world where chaos gets involved and creates subtle changes.  Another factor is that not all effects modules work or sound the same, so they don’t produce the desired result right from the start.  Becuase of these factors, some measure of tweaking is involved.  I’ll break this method down in the next article.

Option Five: Tweak it until it sounds good
This is my audio engineer Mike DeAlmeida’s preferred method.  You start with a general setting in DVerb or some other reverb module and then just tweak it until it sounds right.  This is by far the most common mnethod, bt sometimes you just can’t dial in the settings like you want to hear things.  That’s when I turn to the other methods.

More soon…

Mr. Khan,

Are you ready to meet our monster?

Question:
I’m laying out the voice tracks (there are three) and weaving them together for the Enchantress. Once I have them all pieced together is it better to equalize just the voice track and then slowly add all the sound effects and the music an different tracks and then equalize the whole thing? I ask because there is some yelling and my lead has a strong bass and when he lets go… boing.
And then, of course, how the hell do I do that?

Sable

Now in Production:
An Elemental Horror – A Rebecca Bowen Mystery
In 1890′s London the highest paid private detective is a woman.

Watch for:
A Jewel of the Gods Adventure: The Amber Room Hunters
full-cast audio drama: www.sablejak.com <http://www.sablejak.com/>

also:
Writing The Fantasy Film, Heroes and Journeys in Alternate Realities www.mwp.com <http://www.mwp.com/>

Answer:

Proper EQ and mastering should be the goal no matter what the medium, especially if you plan to sell CDs. Treat each track as a separate instrument. I tend to use grouping to keep sound effects with their respective characters:
In Auditon 3:

1. Hold down the Ctrl key, and click each clip you want in the group.
2. Choose Clip > Group Clips. Alternatively, right‑click any clip in
the group, and choose Group Clips.

For instance, I would put the lead voice, with his footsteps, any Foley effects and spot effects associated with the character on different tracks grouped together for quick reference.

EQ: There should be separate eq for the characters and a master eq for the entire production. So Davis would be eq’d all as one track. treat each track like it’s own instrument with it’s own sense of space. When you start mixing down effects into the tracks, you lose the ability to edit individual details.

Alright… methods for controlling the boomy voice:
1) When you record, make sure your actor backs off of his microphone a bit. Close proximity creates a resonance in the low frequencies (proximity effect) that artificially adds bass to the voice.
2) Drop the 250hz eq slider by about 3 to 6 dB
2) Roll down frequencies below 500hz. Roll down means stepping down by about for every frequency below that range: There’s a great resource on eq ranges here for reference: http://www.digitalprosound.com/2002/03_mar/tutorials/mixing_excerpt1.htm
3) You can also try a multiband compressor, but these can sound unnatural if not carefully used. If you want to try this, let me know, and I’ll set you up, But EQ should handle most of your tweaking needs.

Rolling down:
I tend to use a 20 band graphic eq for tracks and 30 for the master. You can easily get away with using a 10 band for voice tracks. The method is the same either way. I’ll give it to you for 20 band:
1) Remove all frequencies that aren’t part of the instrument’s natural range. For voice, anything below 125 can be cut:

2) Rolling off: This is easier to do if you start with at 125. Cut the 125 slider by half (-12dB). Set the 180 slider to half the difference (-6dB) and again for all the frequencies below 500 (250 to -3dB, 355 to -1.5dB). These down have to be exact.

3) Tweak: Now tweak the voice to your desired levels.
Fuller voice: Bump up the 125 slowly until the richness you want is desired?
Boomy: Cut the 250 down a little at a time
Clearer voice: Bump the 5.6 and the 11.3 up very slightly
Dull voice: Bump up the 8k
Sibilance (t’s and s’s are too present): bumps the 8k down slightly.

There is a lot of information on the net regarding materials and how they can or can’t be used for acoustic treatment.  Many of the opinions against seem to be sucking down the rhetorical kool-aid of the companies that sold them their overpriced acoustic foam materials in the first place, claiming that it’s the only right way to do things.  That is simply not true.  As Ethan Winer stated on his site, there are a number of contractor-grade rigid fiberboard options that out perform acoustic foam at a fraction of the cost (Owens-Corning 703 and 705FRK, specifically.

Being the skeptic that I am, and having time to kill since I was planning on moving into the house this weekend, I decided to run some tests on a few materials available at Home Depot and Joanne’s, just to see what numbers we got.  We tested only mid frequencies between 125Hz and 4kHz, just to get a feel for the materials and weed some out for future testing.

As a baseline, we compare everything to Ethan Winer’s numbers for Acoustic Foam:

The speaker I used was kind of mediocre, but was more than capable of producing the frequencies for the test.  Albeit it didn’t achieve a flat response, but as long as we maintained consistent use of the same settings, the math should still work out.  Here’s the data for the open mic recording:

Wow.. those numbers are really ugly for an open mic recording with only about 2 ft of space. But as I said, as long as conditions remain the same, the data is viable. We tested several materials: 2-inch foam rubber, standard acoustic tile, Styrofoam, bamboo fill, 1-inch fiberglass board, floral foam, cotton batting, utility fabric (foam backed), standard pink insulation r-19, 1-inch rigid foam board (bead board), and standard carpet padding.  Here are our coefficients:

A number of the materials had difficulty in the 500Hz range and the 2kHz range. Several, actually made values in the 2kHz range increase, which is only useful if you room id dull in these ranges. Most rooms are not.

The fiberglass board and the rigid foam board out-performed the  acoustic foam in the low-mid frequencies, but didn’t really accomplish much over 500Hz.  This may be in part due to the thickness of the materials used, which will be our next test.  The carpet pad (only about 1/4″ thick) had decent enough results, especially in the 2kHz problematic rang to warrant further testing, especially since the material can be very densely packed.

We’ll be taking the best results and doing some experimentation on combining these to create some boards on the cheap.  In the end, if you can afford it, the Owens-Corning 703 may be your best bet, as it rings in at about $75 for six boards.  So depending on performance, if this ends up costing more than that, it is hardly worth the expense.

Hopefully I’ll be able to get some room data soon, so the design isn’t completely in my head.

Ultimately, I won’t know until I have access to the place. For now, I’m just going to run some numbers based on some average ceiling heights in order to come up with ideas for space manipulation. There are a number of different ratios for optimal design. I’ll be using the set developed by L.W. Sepmeyer as they are based around the golden ratio, a long-used architectural and artistic mathematical precision. I’m particularly attracted to the idea and would like to use it if possible.

Right… so..  the average height for ceilings in my current house is 9 feet.  Base on this the dimensions would be:

As much as I’d like to use the third set of numbers, the likelihood that I have a room close to 21 feet is pretty low. While the other sets of numbers look doable, they stray away from the golden ratio.  One possibility here is to drop the ceilings.  This all depends on the current size of the ceiling, of course.  I’m not looking to  contort myself to fit into the studio.  And at 6’5″ I need some room to move.  My son, who will also use the studio, is bound to be taller.  So we need to be practical.  Let’s try 8 feet.

That sets us up with better numbers. Ultimately, I’ll need solid numbers to play with, but for now, this is gives me an idea where to start.

There are a few different way to set this up.  I’ll try not to confuse you or myself in the process.  As I said in the previous article I my degree is in English and Theatre, not Math.  So, we’ll deal less with the why things work and more with the how.  That being side, I’m bound to mess something up here, so if you have information to contribute, and can explain it, I’ll try to clarify it here for those of us who are trying to connect the dots.

Let’s look at the Numbers:

0,1,4,2,2,4,1

Were going to be arranging these numbers in a grid based on our chosen prime, in this case 7.  Therefore, we’ll be arranging the numbers in a 7 x 7 grid.

We could start by plugging in the numbers in a linear format by putting 0,1,4,2,2,4,1 across the top row and down the left column.  However dong so is likely to give us a shape that is less symmetrical.  It would work, but for my project, I’d like something that does a little framing.  Also, I’m pretty sure it will be easier in calculations and putting the thing together at the end.  What you do with the info is up to you.

Anyway, while I can’t claim the idea as my own, I like the idea of having a zero space in the middle, both as a a design choice, and so maybe I can put something in the middle squares.  I’m all about the multifunctional thing.  So with that, there are two ways to arrange the numbers, diagonally and across and down, using 0 as the pivot point.  Either way 0 is the center, so the sequence becomes:

2,4,1,0,1,4,2

We’ll start with the cross grid.

From here, you’d think filling in the numbers is pretty straight forward, starting the sequence from the next number.  You’d think wrong.  While the idea is that the sequence does continue, it doesn’t use the same numbers previously calculated.  Instead we recalculate the numbers the number at the ceter of the row or column were and offset.

A method that works is to take the number from the row and the number from the column and add them together.

This is probably the easiest method that involves the least amount of the simplest calculations.  You can also recalculate using the original formula by adding the differences as offsets, but it is more complicated.  The easier method, however does require some adjustments.  Let’s fill in the rest of the grid:

You’ll notice I highlighted some of the numbers in red.  This goes back to Part one where we discussed the number system being based on the value of the prime and always being one less.  In a 7 prime, values 7 and above need to be converted in order to work in the system.  Why?  Ask a mathematician.  Here’s the how:

Step 1: Subtract the prime from the number.

Okay.. that’s it.  Really. In this case we take the out of range number 8 and subtract the prime 7 from it.

8-7=1

If you want, you can calculate it the hard way, but the numbers will come out the same.

That’s pretty much it. All that’s left it to convert your numbers to measurements.  If you want your diffusers to run 6 inches deep, then the numbers are good to go.  Those of us trying to conserve some space will have to run some more numbers.

Most commercial-based diffusers are only 3 inches deep.  That makes the math easy as we only need to half everything.  For those with other dimensions in mind, just take the desired depth (we’ll go with 4 inches for illustration) and divide it by your largest numerical value in the system (6).

4/6 = 2/3″

So each step in value will be 2/3″

1 :   2/3
2 :   1-1/3
3 :   2
4:   2-2/3
5:   3-1/3
6:  4

That’s all there is to it.  I’ll post more when I get a house to build them in.

In the interim, I’m trying to plan out a studio that I don’t have the numbers for.  I worked out the math on some diffusers.  One of the keys to this studio that I haven’t mentioned yet is wrapped up in the experiment of this house: if we can do it ourselves, we will.  Anyway, the math on quadratic residue diffusers is slightly tricky, but I think, with the help of a book entitled Acoustic Absorbers and Diffusers: Theory, Design and Application and a few other websites, I understand how the math is applied.  I also have a lot of information from Acoustic Treatment and Design for Recording Studios and Listening Rooms on room sizes, placement of elements, including diffusers and absorbers to optimize the room for music listening and production.  Unfortunately, without the numbers from the room, it doesn’t do me a hell of a lot of good.

Here is the math on the quadratic diffusers:

s_n = n² (mod N)

Those of you who are looking at this equation cross-eyed are about where I was when I first looked at it.  Let’s see if I can break it down to as much sense as it makes to me:

n represents a well in the diffuser.  Specifically it is calling a sequential well in a linear diffuser.. that is a diffuser that has only one row.  Our diffuser is going to have more, but I will get to that in a moment.

s_{n represents the depth of the well, which we are attempting to calculate with this formula}

n^{2 tells us we are going to take the number assigned to the well and square it}

N represents the number of wells we have.  In this calculation, N must be prime.  Why? Because the book said so.  My degree is in English and Theatre, so while I have no fear of mathematics, there’s a point where I try not to delve into the theory too much and just act like a technician, leaving the “why” to the mathematicians.  This is one such case.

mod (short for modulus) is best expressed as the remainder.  For instance 17 mod 3 is the remainder of dividing 17 by 3.  Remember elementary mathematics when everything you divided had a remainder?  Well that’s what the mod is.  It’s pretty useful in mathematical calculations.  The difference here is that we throw the answer away and just keep the remainder.  17 / 3 = 5 with a remainder of 2 (or 5 R2) .  In this case we’re going to ditch the 5 and keep the remainder.  Why?  Ask a mathematician.

A few things to understand here:

1) Mathematics typically starts calculating from 0 as this is the base. This means our calculation will start with 0, which in terms of our diffuser means the wall with nothing added to it.

2) The MOD numbers will never be higher than one number lower than your prime number. So if your number is 7, your numbers should never climb higher than 6?  Why?  Because of the last rule and because that’s how the remainders work out when you divide by 7.  Basic math: If you divide by 7 and you have 7 or more left over, you divided wrong.  While 14 / 7 is  1 R7, it is more properly expressed as 2… in this case though we go for precision and say  2 R0, so we can use that 0 for our answer.

Another way to understand it is by using our base 10 counting.  Based ten counting uses ten numbers, 0 through 9 to define the system.  In the mod 7, it is as if we are counting in base 7, so our number system is 0-6.  If that is too confusing, just ask a mathematician.

3) There will be a number of answers equal to your prime number before the system repeats.  I your number is 7, there will be seven answers.  17 will have seventeen answers.  And so on.

4) You’ll end up with a symmetrical sequence.  That is, halfway through, the sequence will reverse itself before repeating.

Confused yet?

Lets start playing with the numbers.  We’re going to do a base seven sequence because it is easier to calculate and easier to build when the whole thing is completed.  I’ll give you the numbers for base 17 and base 23 at the end, so you can use them or just see how the numbers work.

Let’s say we’re going to create a linear diffuser similar to the one above which has 14 boxes.  We’ll include the wall as our zero box (represented by the whitespace) for a total of 15.  If we label these boxes as 0 (the wall) through 14 we have our numbers to plug in to our formula.

Now we take our formula:

s_n = n² (mod N)

and start plugging in the numbers:

well₀ (or the wall)  = 0² (mod 7)

Hopefully if you have read the above, this is starting to make sense.  If not, let me know so I can clarify.

Time to run the numbers:

0 squared = 0

so: wall = 0 (mod 7)

then we figure out the mod:

0 / 7 = 0

or more precisely 0 R0

so we drop the answer and keep the remainder (0) as our completed equation

wall = 0

Make sense?

Let’s try the next one.

s_n = n² (mod N)

plug in the numbers:

well 1 = 1² (mod 7)

run the numbers:

1 squared = 1

so: well 1 = 1 (mod 7)

then we figure out the mod:

1 / 7 = 0 R 1

so we drop the answer and keep the remainder (1) as our completed equation

well 1 = 1

Next one:

s_n = n² (mod N)

well 2 = 2² (mod 7)

2 squared = 4

well 2 = 4 (mod 7)

4 / 7 = 0 R 4

keep the remainder (4)

well 2 = 4

And so on:

well 3 = 3² (mod 7)

well 3 = 9 (mod 7)

9 / 7 = 1 R 2

well 3= 2

————————————-

well 4 = 4² (mod 7)
well 4 = 16 (mod 7)
16 / 7 = 2 R 2
well 4 = 2

————————————-

well 5 = 5² (mod 7)
well 5= 25 (mod 7)
25/ 7 = 3 R 4
well 5 = 4
(notice that the numbers have started to mirror here)

————————————-

well 6 = 6² (mod 7)
well 6 = 36 (mod 7)
36 / 7 = 5 R 1
well 6 = 1

————————————-

well 7 = 7² (mod 7)
well 7 = 49 (mod 7)
49 / 7 = 7 R 0
well 7 = 0

Once we have reached 0 again the numbers repeat with zero as the hinge.  You can run the numbers, if you like.  Or you can take my word for it.  I always advise running the numbers yourself, just so you can see that it works.  Our sequence in a 14 block sequence becomes:

0,1,4,2,2,4,1

for a 14 block sequence at the 7 prime we would see:

As promised:

N = 17:  0, 1, 4, 9, 16, 8, 2, 15, 13, 13, 15, 2, 8, 16, 9, 4, 1, 0

N = 23: 0,1,4,9,16,2,13,3,18,12,8,6,6,8,12,18,3,13,2,16,9,4,1,0

Next time, we’ll look at how to plug these numbers into a 2D grid and how to get some aesthetic and variety out of it.