How Do You Finish Your Audio? at DVinfo.net

Steven Schuldt · August 8th, 2005, 12:12 PM

All,

Does anyone have a recommended or preferred set of audio filters they'd suggest running on audio in post or another audio finishing strategy? In my current project the basic audio is fairly good, but these are primarily vanilla dialogue tracks and the levels are often somewhat low and between different scenes can be uneven. Also, much of my audio is monophonic (we were using a single Schoeps CMC641 on a boom). I'd like to give the audio a professional, finished gloss. This is on a Mac and I have Final Cut Pro/Soundtrack Pro/Logic etc.

I was thinking something along the lines of exporting to Soundtrack Pro from FCP and running some series of filters like:

Spread Stereo Spectrum/Multipressor/Normalize

- but I have no idea what would be optimal as far as compression, normalization and enhancement. I could make some guesses, but hopefully some of you have been down this road before!

All ideas welcome. Many thanks!

Steven

Troy Tiscareno · August 8th, 2005, 01:06 PM

No one can really give you too much advice with the details given. What is your project? What do you have recorded? What current problems do you have with the current audio that you are seeking to fix? Etc.

-Troy

Steven Schuldt · August 8th, 2005, 01:12 PM

Okay then, specific problems:

- The audio, while very clean, has levels that tend to be low, so I assume they need normalization at the very least.
- The audio has no compression, so the dynamic range is extremely wide resulting in quieter moments potentially being too quiet.
- The audio is (often) monophonic.
- The audio has no "sparkle" - as would audio put through a multipressor, enhancer or at least a compressor.

Sorry if I'm asking an unanswerable question. Maybe there is simply no standard finishing technique and I'll simply need to use my best judgement. That in itself is useful to know if it's the case.

Thanks!

Steven

Steven Schuldt · August 8th, 2005, 01:21 PM

In case this matters, here are project and tech details: Independent feature film, charcter driven, some incidental music but no lush orchestral soundtrack or anything. Mostly dialog with some musical performances (dubbed). Few sound effects - primarily ringing phones.

Project was recorded to Sound Devices 744t field recorder in 24bit, 48Khz mono. Out of 744t went to in of Sony Z1u HDV camera. Rough cut of the film has primarily relied on the Z1u capture from the out of the 744t - so it appears as 16bit PCM in FCP timeline. I can selectively pull pristine 24bit audio from the original 744t recordings - although the audible difference is marginal as the Z1u's audio (yes, destructively compressed, I know) seems generally acceptable.

Hope that helps!

Jose di Cani · August 12th, 2005, 01:15 PM

Here is what I had done if I was a beginner/semi profesional today.

check out:
www.kvr-vst.com

then look for free vst plugins windows and you will have a great amount of amazing fx plugins (synths, special effects synths, you name it). With that you can add profesional things to you audio. NOt to much please....a little is good. Compare the sound you have with movie sounds. KEep comparing. BUy better monitors if you gonna compare...it will be all good.

Then you need a sequencer software (cubase, sonar, cakewal, logic, magic studio, etc etc) so that you can use lots of audio tracks at the same time and edit everything real-time, adding your free plugs in them and move audio samples and reverse them and add background noise, and add extra vocals...change EQ.....well, sequencers are like lighing for video. You can't work without them.

Glenn Chan · August 12th, 2005, 03:56 PM

Here's some things you can do:

A- Mix all your sounds so you can hear everything and there's no large differences in volume if there isn't supposed to be. The audio usually shouldn't call attention to itself (i.e. you switch between lav and boom, which sound very different without EQ and reverb)... that would kind of be like a continuity error.

If you are targetting for TV, you can't have subtle noises because they will be lost on your audience (most people usually have high background noise when watching TV). Film festival is difference, as are other target formats/audiences.

You need to do this by ear on a good set of monitors- meters don't tell the whole story.

A compressor can help you even things a little more.

In the end, you want your levels are a specific dB level if you are making a master.

Jay Rose's book Audio Postproduction for Digital Video has a lot of information on this, I recommend it. see dplay.com on how to buy it for $30.

B- Sound Design
Add in ambience and key SFX so the mix sounds realistic. Movie realism is different from real life. Car crashes and gunshots sound a lot better in movies than they do in real life.

Key SFX would be things that appear on screen that make noise.

You can also use sound design to create moods, add style (I'm thinking like requiem for a dream), and or enhance the meaning/point/goal of your movie.

Quote:

- The audio is (often) monophonic.

I don't think that's a problem?

THere's threads here discussing stereo dialogue (panning dialogue depending on the talent's screen position). From what I've read, it's just confusing to the audience.

Zach Mull · August 12th, 2005, 04:28 PM

I agree with Glenn about stereo panning. I would keep the dialog pan at the center unless you have a line that requires a panning effect (e.g. I sometimes use this for offscreen lines that cause an onscreen character to look in a particular direction). I also don't think you can gain anything in dialog by using a filter to spread a mono track to stereo.

August 8th, 2005, 12:12 PM	#1
Steven Schuldt Regular Crew Join Date: Jan 2005 Location: USA Posts: 90	How Do You Finish Your Audio? All, Does anyone have a recommended or preferred set of audio filters they'd suggest running on audio in post or another audio finishing strategy? In my current project the basic audio is fairly good, but these are primarily vanilla dialogue tracks and the levels are often somewhat low and between different scenes can be uneven. Also, much of my audio is monophonic (we were using a single Schoeps CMC641 on a boom). I'd like to give the audio a professional, finished gloss. This is on a Mac and I have Final Cut Pro/Soundtrack Pro/Logic etc. I was thinking something along the lines of exporting to Soundtrack Pro from FCP and running some series of filters like: Spread Stereo Spectrum/Multipressor/Normalize - but I have no idea what would be optimal as far as compression, normalization and enhancement. I could make some guesses, but hopefully some of you have been down this road before! All ideas welcome. Many thanks! Steven

August 8th, 2005, 01:12 PM	#3
Steven Schuldt Regular Crew Join Date: Jan 2005 Location: USA Posts: 90	Details Okay then, specific problems: - The audio, while very clean, has levels that tend to be low, so I assume they need normalization at the very least. - The audio has no compression, so the dynamic range is extremely wide resulting in quieter moments potentially being too quiet. - The audio is (often) monophonic. - The audio has no "sparkle" - as would audio put through a multipressor, enhancer or at least a compressor. Sorry if I'm asking an unanswerable question. Maybe there is simply no standard finishing technique and I'll simply need to use my best judgement. That in itself is useful to know if it's the case. Thanks! Steven

August 8th, 2005, 01:21 PM	#4
Steven Schuldt Regular Crew Join Date: Jan 2005 Location: USA Posts: 90	Tech Details In case this matters, here are project and tech details: Independent feature film, charcter driven, some incidental music but no lush orchestral soundtrack or anything. Mostly dialog with some musical performances (dubbed). Few sound effects - primarily ringing phones. Project was recorded to Sound Devices 744t field recorder in 24bit, 48Khz mono. Out of 744t went to in of Sony Z1u HDV camera. Rough cut of the film has primarily relied on the Z1u capture from the out of the 744t - so it appears as 16bit PCM in FCP timeline. I can selectively pull pristine 24bit audio from the original 744t recordings - although the audible difference is marginal as the Z1u's audio (yes, destructively compressed, I know) seems generally acceptable. Hope that helps!

August 8th, 2005, 01:06 PM	#2
Troy Tiscareno Regular Crew Join Date: Sep 2004 Location: Bay Area, California Posts: 53	No one can really give you too much advice with the details given. What is your project? What do you have recorded? What current problems do you have with the current audio that you are seeking to fix? Etc. -Troy

August 12th, 2005, 01:15 PM	#5
Jose di Cani Major Player Join Date: Oct 2004 Location: Holland, Europe Posts: 214	Here is what I had done if I was a beginner/semi profesional today. check out: www.kvr-vst.com then look for free vst plugins windows and you will have a great amount of amazing fx plugins (synths, special effects synths, you name it). With that you can add profesional things to you audio. NOt to much please....a little is good. Compare the sound you have with movie sounds. KEep comparing. BUy better monitors if you gonna compare...it will be all good. Then you need a sequencer software (cubase, sonar, cakewal, logic, magic studio, etc etc) so that you can use lots of audio tracks at the same time and edit everything real-time, adding your free plugs in them and move audio samples and reverse them and add background noise, and add extra vocals...change EQ.....well, sequencers are like lighing for video. You can't work without them.

August 12th, 2005, 04:28 PM	#7
Zach Mull Major Player Join Date: Feb 2005 Location: Portland, OR Posts: 273	I agree with Glenn about stereo panning. I would keep the dialog pan at the center unless you have a line that requires a panning effect (e.g. I sometimes use this for offscreen lines that cause an onscreen character to look in a particular direction). I also don't think you can gain anything in dialog by using a filter to spread a mono track to stereo.