Frequently Asked Questions
Below you’ll find answers to some common post sound questions. There are many approaches; these are my opinions based on my own experience. I have tried to keep my answers brief, although the issues are sometimes complex.
What materials do you need from the filmmaker in order to do the mix?
- AAF – A file exported from your editing software which contains all of the audio tracks that the picture editor was using, including any volume automation and fades that may have been done. Please include at least 6 seconds (144 frames) of audio handles. Embedded or referenced AAFs both work fine. For the audio format, 24-bit/48 kHz is recommended. Some workflows (such as “nesting” in Adobe Premiere) lead to issues in the AAF exporting process. OMF is another file type that works, but these days AAF is more common and reliable.
- Reference video – 1080p Apple ProRes Proxy codec works well with Pro Tools and is what I prefer. Avid DNxHD also works. If you need to have a small file size to send via the internet, H.264 is fine, but I will need to convert it. Ideally, your video will have timecode burned in and a 2-pop before the start of the program. It is good practice to have your first frame of picture start at 01:00:00:00, with any leader and 2-pop starting before the 1 hour mark.
- Production audio and sound reports – For the purposes of accessing all available mics and looking for better sounds in alternate takes. Sometimes a copy of the script can be helpful, especially for feature films.
How should I organize my audio tracks?
It is best to group all of the dialogue and production audio on one set of tracks, sound effects on another set of tracks, and music on another set of tracks. Please include any audio associated with the picture, even if you don’t think it’s usable. Scratch camera audio is not needed unless it’s the best audio you have. Lower the volume of unused mics if needed (try not to mute — it may not come through in the OMF/AAF). Sometimes I can make it work or find another usable section of the clip in the handles, and even if I cannot, I want to be able to make that decision.
For some projects, the picture editor may be working mainly with the mix track from the production sound mixer. This is a good workflow so long as all timecode metadata is carefully maintained and comes through in the AAF. In this case I will access the different mic isos through the Pro Tools Field Recorder workflow. If this workflow is unfamiliar to you, it’s best to include all mic channels in your edit timeline/AAF.
What is dialogue editing?
Dialogue editing is cleaning up all of the production sound recorded on set. A dialogue editor smoothes out the dialogue and production tracks by selecting the correct microphone or combination of microphones, filling holes with clean ambience, and fading clips in and out to create a seamless dialogue track. Clicks, bumps, buzzes, and other unwanted noises are removed or reduced. Often a dialogue editor will search through alternate takes from production to find cleaner words or syllables to drop in and help solve audio problems. Volume automation, EQ, and noise reduction are often done by the dialogue editor, although sometimes this is left up to the re-recording mixer.
What is the difference between foley and sound effects?
Foley is a category of sound effects used to enhance or recreate sounds relating to the movement of a person or animal onscreen. Footsteps, prop movements, and clothing rustle all qualify as foley. When a project’s budget permits, foley is recorded specifically for the program by a foley artist. Other times, foley is pulled from a sound effects library and cut into place. Sound effects is a wider category that includes ambient backgrounds, sounds of cars and machinery, doors opening and closing, abstract sound design, etc.
What is mixing? What is re-recording?
Mixing is the process of bringing together the various sound elements together so that the program sounds the way the filmmaker wants it to. This is accomplished using any number of techniques, including automating volume, equalization, compression, noise reduction, reverb, delay, and panning. Re-recording is recording the final mix into a smaller set of tracks that can be played back in a theater or on any video-playback device (e.g. 6 tracks for a 5.1 mix, 2 tracks for a stereo mix, etc.). Re-recording can be done in real time, so that the mixer can stop and make changes as needed.
Do I need a 5.1 mix? Do I need a stereo mix? What about Atmos?
Will your project be shown in a theater? If yes, you need a 5.1 mix. In a theater, you want most of the dialogue to come from the center of the screen, not the sides, and this is accomplished with 5.1 mix. Plus, you get added sonic depth from the sides, surrounds, and LFE. Streaming services, Blu-Ray and DVD can also playback 5.1 audio, so you want it for those applications as well. Any 5.1 project will also get a stereo mix. The stereo mix is suitable for putting your project on the web or playback on any 2-speaker setup. Dolby Atmos is an immersive audio format that incorporates speakers mounted in the ceiling, as well as object-based panning. For larger projects, we can mix in Atmos at Soundcrafter.
What levels do you mix to?
This is a complicated issue, but the basic idea is that movie theaters are calibrated to playback at a certain decibel level, and our mixing rooms are calibrated to that same level. Sound mixers use their ears and loudness meters to determine the correct levels to mix to. Broadcast levels are roughly 6dB louder than theatrical levels, but must meet CALM Act, EBU R128, or other loudness regulations. For web videos, anything goes, but I tend to deliver 10-12dB louder than theatrical levels.
How much can you clean up noisy dialogue?
I can almost always improve audio, but I may or may not be able to make it sound perfect. A great recording is the first step to a great mix, so take great care to record clean audio on set. You’ll save money and have a better sounding project. There are many tools and techniques to reduce unwanted noises, including EQ, spectral processors, AI-based dialogue isolation plug-ins, and multiband compressors/expanders. Taking out bumps, wind noise, birds, bugs, clicks, hums, buzzes, and excessive reverb can all be done with varying degrees of success. In general, the broader the frequency content of the unwanted noise, the more difficult it is to remove, and the more isolated the frequency of the noise, the more easily it can be removed.