What Average RMS setting do most Hollywood DVD's use?

[ChachiFace]

Hi. What is the standard average RMS audio setting for Hollywood movies on DVD? (I'm not talking about the Dialog Normalization)

I've noticed that >most< Hollywood DVD's have the same loudness. I would like for my homemade DVD's (NTSC) to have the same loudness as the Hollywood DVD's.

I am capturing analog to digital (AVI) and after encoding have a WAV file. Using an audio editor, what average RMS level do I want to Normalize my audio file to before encoding to AC3? I am using Sound Forge to tweek my WAV and Soft Encode to encode to AC3.

Currently, I just Maximize the Peak value to "0" as my Normilization but the overall RMS value is different from footage to footage. I would like to have some consistencey, like the Hollywood DVD's.


Any suggestions?

Thanks,
ChachiFace

[LeeBear]

Doom9 has an excellent guide about AC3 audio.

http://forum.doom9.org/showthread.php?s=&threadid=68300#post424067

You'll find a link there about how to encode to the proper levels.
The average is approximately -31 dBFS btw, but I suggest you read the guide to find out how it's derived.

-LeeBear

[ChachiFace]

LeeBear,

Thank you for your response but I have already read the guide and did not find it to help me in what I want to achieve.

The guide suggests to alter the dialog normalization in the AC3 encoder based on what average RMS level your source audio is (WAV). If this is the case, then why are almost all Hollywood DVD's found to use the same dialog normalization value? (-27db) Coincidentally, this is the default value in most AC3 encoders. (Soft Encode, anyway)

I have taken a sample AC3 audio from a DVD menu and checked the average RMS and it was indeed around -31db but I do not understand this b/c it sounds much louder than what a normal audio sample would be with an average RMS of -31db (WAV vs AC3).

All I know is when I pop in a Hollywood made DVD in my computer or DVD topset player, the audio levels are loud but not too loud. And my audio levels with my WAV converted to AC3 on my homemade DVDR's is never consistantly that loud from disc to disc (footage to footage). I just wanted to know what I should normalize my raw WAV file to prior to encoding to AC3. I have heard raising the dialog normalization value will increase the output volume but then why are Hollywood DVD's consistantly -27db? I believe the normalization to a set RMS value takes place with the audio file prior to encoding to AC3.

Since I am obviously not seasoned in the audio dept., hopefully someone who knows what I'm talking about can answer my questions.

Thanks for any suggestions.

And, again LeeBear, thanks for the link though I had already read it. That was a link that I had originally thought was going to answer my questions.


ChachiFace

[W_Eagle]

I have the very same question. Awaiting replies.

[Sillyname]

Most mastering that occurs uses an Exciter. You can get a directx plugin that will do it for Sound Forge but it'll cost you. I found an Aphex Aural Exciter at a pawn shop, still in the box, for $69. It's 2 channel but I like the fact that I can run audio through it in realtime and make adjustments as necessary. I think the plugins out there make you wait for the audio to render before you can hear it.

APHEX originally made the exciter for film audio production but it has been used in music audio for sometime as well. If you ever open a wave that looks like a straight block but sounds spacious and VERY loud, chances are that it has been ran through an exciter. BBE also makes one called the Sonic Maximizer.

[BJ_M]

most audio for film s not run through an exciter - some may be, but not a normal thing ... some dialog will be though during ADR ..

read this to get a good understanding how film levels were first arrived at and standardized ..

http://digido.com/portal/pmodule_id=11/pmdmode=fullscreen/pageadder_page_id=59

[Cornucopia]

Yeah, but I know for a fact that H'wood itself isn't really that consistent.

A mentor of mine (semi-retired C.A.S. member, so he ought to know) watches ~20 DVD's a week via Netflix. He checks them as they come in on his computer/sound rig. Gripes to me that they're wildly inconsistent (as if I could do anything about it). Dialnorm was supposed to help alleviate the problem, but there's lots of studios and post houses that don't necessarily follow the rules every time.

Scott

[Sillyname]

most audio for film s not run through an exciter - some may be, but not a normal thing ... some dialog will be though during ADR ..

read this to get a good understanding how film levels were first arrived at and standardized ..

http://digido.com/portal/pmodule_id=11/pmdmode=fullscreen/pageadder_page_id=59

An exciter is a way of cheating to get the same results a tweaker might be able to achieve. All of the processes in the literature above can be dialed-in to a certain extent using one. While the headroom of my levels are not as pretty to watch, the audio is "top notch"(pun). It's the difference between knowing and doing. I have to be a jack of all trades and if I can cut corners, I will as many out there do. I think sound engineers like to say they don't use an exciter because the shame of lumping so much of the mastering into one processor box denotes lack of knowledge in the finer details. It's still doing the same thing though, just faster.

[BJ_M]

well i am a sound engineer (as well as jack of all trades related to media and design) and i work on "Hollywood" movies and other projects including live sound, a lot of mastering projects and have worked in many major studios -- and i tell you , we dont use exciters that much .. may use BBE on some voices - yes .. sometimes aphex (not to much anymore) .... but sweeting and mastering is most often done by other methods ..

[Sillyname]

well i am a sound engineer (as well as jack of all trades related to media and design) and i work on "Hollywood" movies and other projects including live sound, a lot of mastering projects and have worked in many major studios -- and i tell you , we dont use exciters that much .. may use BBE on some voices - yes .. sometimes aphex (not to much anymore) .... but sweeting and mastering is most often done by other methods ..

I'm sorry. It's the baby head. It's making me say things as if I know what I am talking about.

When you use your BBE for voice enhancement, what are you trying to do to the voice if it is already seperate from other sounds in the mix? I guess I'm confused because I've been using it to clear up sound board audio from live concerts. Usually these are already thoroughly mixed and I use the exciter to bring forward certain bits that only sounded right when playing live at the venue but suck on the stereo.

So you "sculpt" voices alone with the BBEs? I should think it's the actors fault that their voices sound flat. Get a better actor. :P

Really, much of my current knowledge of the exciter, only comes from the history I have read on it. In the introduction in the manual, Aphex says that when it was originally created in 1975 for use in motion picture sound. Even won some awards around that time for it. I guess I was just assuming things hadn't changed. ASSUME=ASS of U & ME

What significant sonic difference does a BBE have over an Aphex?

[BJ_M]

for awhile - the aphex was used everywhere -- but as the industry moved to 96bit , it was apparent that those upper harmonics got nasty...

almost all dialog in film is added in later ...

I can understand for live sound mixed for cd/dvd (which ussually sounds pretty nasty at first anyway) ..
Also used for some older recording for re-mastering.

speaking of live audio - tommorrow i'm going to play with the new Digidesign Venue (which emulates a Midas in sound) http://www.digidesign.com/products/venue/

[BJ_M]

it also can run the aphex software plug-ins directly (as well as all other protools plug ins in real time)

[Kurt S]

An Apex Aural Exiter has absolutly nothing to do with the waveform being flattened out. That is a result of a compressor limiter.

[ChachiFace]

Most mastering that occurs uses an Exciter. You can get a directx plugin that will do it for Sound Forge but it'll cost you.

Just curious, what exactly is the Exciter directx plugin called for Sound Forge and where do you get it?


What does this do? (in simple terms for us simpletons)


Thanks,
ChachiFace

[Sillyname]

Most mastering that occurs uses an Exciter. You can get a directx plugin that will do it for Sound Forge but it'll cost you.

Just curious, what exactly is the Exciter directx plugin called for Sound Forge and where do you get it?


What does this do? (in simple terms for us simpletons)


Thanks,
ChachiFace

www.directxfiles.com

There's even a BBE Sonic Maximizer plug-in.

I was looking at these before I snagged the Aphex hardware:

http://www.dspfx.com/VirtualPack/html/effects.html

$149 for all of them! I almost got them. One of the people who endorse them was a former Black Flag producer, for what it's worth... Any friend of Henry Rollins is a friend of mine.

[Sillyname]

for awhile - the aphex was used everywhere -- but as the industry moved to 96bit , it was apparent that those upper harmonics got nasty...

almost all dialog in film is added in later ...

I can understand for live sound mixed for cd/dvd (which ussually sounds pretty nasty at first anyway) ..
Also used for some older recording for re-mastering.

speaking of live audio - tommorrow i'm going to play with the new Digidesign Venue (which emulates a Midas in sound)

Love those knobs! Quite different from the norm.

[Cornucopia]

Without being privvy to the inner workings of Aphex but having gained experience using it, the general idea is that the Aphex Aural Exciter is a Harmonics/Dynamics generator/synthesizer. It has its uses, but works best on old analog recordings & stuff which has had a good deal of band-limit processing. Not really as desireable as it use to be--good Tube Emulation probably works better for newer material.
AAE is supposed to be very proprietary, but I've come up with no less than 3 custom solutions as alternatives to it (will reveal them to those who ask nicely). All have pretty much the same strengths and weaknesses that AAE does. All should be used sparingly.

Scott

[BJ_M]

good guide to basic EQ

http://www.recordingwebsite.com/articles/eqprimer.php

[Sillyname]

Me likey:

http://www.directxfiles.com/manufacturers/cakewalk_ss.htm

I played with the demo for awhile but it only would process about 5 secs of your wave but it was cool! I wish it had a way you could load in your own environments that you build in a 3D program. If it does its calculations off of the physical measurements between sound source and reflective surface, then I wonder what would happen if you entered in a negative value.

[greymalkin]

reverse reverb like on poltergeist :P


mmmmmmmmommy.....

[Sillyname]

reverse reverb like on poltergeist :P


mmmmmmmmommy.....

I actually believe if you started with a wave file that had reverb in it and you knew the dimensions of the orig place it was recorded in, then built an exact duplicate of the environment you recorded it in in a 3d program, you should be able to change all x, y, z coordinates to opposite values and then place your microphones and waves at corresponding points in this "inside out" environment, it should cancel the reverb and leave you with a clean signal. I'd probably also have to swap positions between the speaker and the microphones. Basically everything would have to be setup "inside out" in order to recover the original sound from a wave that has too much reverb.

Make any sense?

[Cornucopia]

reverse reverb like on poltergeist :P


mmmmmmmmommy.....

I actually believe if you started with a wave file that had reverb in it and you knew the dimensions of the orig place it was recorded in, then built an exact duplicate of the environment you recorded it in in a 3d program, you should be able to change all x, y, z coordinates to opposite values and then place your microphones and waves at corresponding points in this "inside out" environment, it should cancel the reverb and leave you with a clean signal. I'd probably also have to swap positions between the speaker and the microphones. Basically everything would have to be setup "inside out" in order to recover the original sound from a wave that has too much reverb.

Make any sense?

Won't work the way you're suggesting:

1. The original reverb will now have a reverb of its own!
2. Your original signal has sound sources coming from many directions, now your only non-reverb source is mono or stereo (even 5.1 channels won't cut it).
3. Your original sounds didn't have the same frequency/dynamic response as a speaker cone. Neither do the frequency/dynamic responses of microphones.
4. Real reverb has a random quality to it. You can't duplicate and invert randomness.

Scott[/code]

[Sillyname]

1. The original reverb will now have a reverb of its own!

I'm talking about a totally inverse situation here. Mathematically it may work but I'm not sure it can be proven, if there is no program that will allow you to input a 3d object with inverse geometry. So, yes, I am talking out of my neck.

2. Your original signal has sound sources coming from many directions, now your only non-reverb source is mono or stereo (even 5.1 channels won't cut it).

Needed to clarify myself on this one. The output into the club is usually mono. There are very few bands that go out of their way to provide a live stereo mix.

3. Your original sounds didn't have the same frequency/dynamic response as a speaker cone. Neither do the frequency/dynamic responses of microphones.

The original sounds came from speaker stacks. In my post I mentioned swapping locations between mics and speakers (or performers as I believe this plug-in calls them.)

4. Real reverb has a random quality to it. You can't duplicate and invert randomness.

I wasn't speaking of a perfect situation, more theoretically. The more detail you pay attention to in the building of the 3d representation of your original venue where the sound was recorded, the more reverb you could subtract. I didn't say it was practical. When I start a post with "I actually believe...", I am having a little bit of trouble believing that I am believing it myself! :P


Another cool idea I thought of was placing multiple mics throughout a venue and recording a track from each mic, then subtract only the differences between tracks and what you should eventually end up with is something with less reverb. The more mics you use, the more samples you take from the environment, the cleaner your audio will be after subtraction of their differences.

[BJ_M]

would not work -- some reflections would be out of phase or somewhere in between and time domain differences and comb filter peaks between multiple (1000's ) of reflections can not be canceled out ..

[Sillyname]

would not work -- some reflections would be out of phase or somewhere in between and time domain differences and comb filter peaks between multiple (1000's ) of reflections can not be canceled out ..

Which one are you talking about, my first or second "theory" because I just did a quick search on "microphone array" and found that my second post was not a theory, at all. I was just reinventing the tubesock.

Here's something weird:
http://www.enablemart.com/productDetail.aspx?pid=973&dept=25&store=10