I am wondering, once I get the dialogue nailed throughout the film and do all the foley and sound scape, what is the procedure to mix this down and master? Do I need to work with smaller scene files (5 to 10mins) or do I only have one long feature length file for my dialogue foley, sound scape and score ?

  • 1
    Notice that if you plan to have chemical film prints of the feature (instead of DCP or file only), you might be required to deliver as several 'reels', with some tricky short overlapping between reels for the optical sound printout. (due to the mechanical distance between projector's lens and optical soun d reader)
    – audionuma
    Commented May 24, 2016 at 17:10

4 Answers 4


You would create stems Dialogue, Ambience, Foley, SFX for the entire length of the film


Like Pierre de Fermat comment in the margin in regards to his "last theorem", it may be difficult to do justice to a question such as this in the size of this text box... However, we'll try.

To summarise,

  1. Work on a reel-by-reel basis. All your editing and premixing should be done on reels no more than 20 minutes long. This one trick will make editing and conform much simpler. Re-conform of a Locked Edit (there is no such thing any more) is much easier if you have shorter reels to manage.

  2. Edit - Premix - Final Mix. Your process should be Edit all your sound, Premix to Stems - final mix. You should probably premix reels, but perform your final mix with a fully conformed film as there is likely to not be gaps where you would like there to be reel changes. Premix as much as you can at a reel level and then final mix once you have conformed all the reels together.

  3. Maintain consistent plugin changes across the premix and final mix using templates. Make sure you maintain consistency with all the processing across all the reels that you have.

  4. Keep all elements separate. Don't mix up SFX with Foley, ATMOS/BG and Production effects. These are separate elements and should be kept separate for final mix. Definitely keep all dialogue separate as this will become critical if a "Music and Effects" mix is required for international dubbing.

  5. Room tone is not ATMOS. Room tone is simply used to smooth dialogue edits. ATMOS is very different to this. Try not to get them mixed up or confused.

  6. Mixing for Cinema is different to mixing for TVC or TV or even Short films. There is a significant difference as to what constitutes a good mix across these platforms.

  7. Mix for emotional impact. Don't mix too loud as you may well find it difficult to recover enough dynamic range or headroom when you need it for effect - for instance gunshots. Gunshots should have a significant emotional impact that cannot be obtained if you are constantly mixing at a high level.

  8. If you're mixing for 5.1/surround, make sure you premix for 5.1 as well. There are situations where you can upmix certain elements from stereo to 5.1, but don't rely on this as being your first choice. Ensure you have enough panning in your SFX elements across your 5.1 mix. Don't be afraid to put elements in rear channels or play with phasing - these are all valid choices.

  • I agree with everything apart from premixing stems. It's seems a pointless step. I buss all my tracks to sub groups/vca's which is essentially the same thing, but I don't bounce them to audio tracks. If your doing the final mix yourself and not passing it along then I don't see the point. And even then in today's Pro Tools world if the mixer is using Pro Tools which they will be, you may as well just send the session across. The only time I would bounce stems is for an M+E. Correct me if I'm wrong but...
    – Melloj
    Commented Dec 21, 2016 at 18:11

Depends on what software your using and how good you are at Micro Fades when Placing Audio into the timeline of the Video Track . I do smaller Bites of audio to keep things clean in the timeline but Lock each piece of Audio to the Section of Video before bouncing everything down to a final out put .

Conversely if you are working in a DAW , with just a Time track then render whole files to keep everything synced up properly .


You simply bounce it down to either a stereo file or If it's surround you need to run it through a surround sound encoder so the resulting wav will know its multiple panned channels. You only need to do stems if it's required for being redubbed in a foriegn language.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.