Editing Sleep Talking Podcast (Workflow)
- Make a copy of the source audio (unedited) to keep them in case you want to restore original recording.
- Clean audio recordings in Audition.
Effects to apply: Denoise, Single-band Compressor, EQ, Dynamics (AutoGate) Dereverb (be very attentive with this one).
You can apply a preset, the problem is even if you have the same person with the same microphone the audio will be different, because of the distance because of the environment because of the air. You need to tweak all settings in order for the audio to sound good (Ex. Check Denoiser can you lower the value or maybe you need to add another Denoiser if there is a lot of noise) AutoGate - can you lower it or maybe you need to make the threshold bigger. Note: The order of effects matter!
Pay attention at AutoGate - it can clip some of the lower sound of voice. Example when laughing or whispering. Some of these words might get cut by AutoGate, double check at the beginning and move the Threshold if needed. Also you need to compress audio in order to have enough room to Match Loudness at -17LUFS.
IF needed you can apply a second compressor with lower ratio to get enough room.
- Use Amplitute Statistics to check where exactly are Peaks and what are loudness meter at the moment.
See image below. Peak amplitude -10.74 dB with ITU-R BS 1770-3 Loudness at -27.18 dB. That means in order to get to -17 dB loudness we need 10 dB headroom for peaks in order for them to not be clipped. And also its good to have 1.5 dB extra. So we need -11.5 dB and we currently have -10.74 We can use another compressor to quickly achieve this and then next step Match Loudness.
- Match Loudness -17 Lufs. Typically if the audio was recorded in a noisier environment or if a guest was very quiet it is acceptable to get -19Lufs or even -21 Lufs. They will be quieter then others but sometimes its better then having audio with artifacts and noise.
- Clean noises which weren’t picked by AutoGate and Denoiser. Often times the guest or host will make some noise by moving in chair or putting something on the table or coughing. Some of these noises will be stopped by AutoGate but not all because we have a weak Threshold (Remember we don’t want any words to cut "that’s why its weak"). Now you have to go over the file and quickly spot any unnecessary sounds. It doesn’t have to be perfect clean but you have to remove some of these nuisance sounds. You can go x4 speed since you don’t need to pay attention to content here.
- Import files into Premiere. After you cleaned up all audio of the host and guest and you normalized it to a specific loudness import all files into Premiere. Inside of the project You should have a template Sequence and a bin with every single episode which contains all raw files (audio corrected but not yet cut) 
- Export the Raw mixdown. Export a file with all speakers (host+guests) to give it to someone who will review it and make notes about edits (cuts) needed. See image below for an example of the notes.
- Start cutting the tracks according to the timecode inside the Raw sequence. Don’t delete anything just mark with red parts which will not be included and yellow parts which need to be shifted into a different place. (Add yellow marker to the place where they need to be included) See image below. This is a better workflow because later if you need to bring something or check parts which you deleted it is easier too see what exactly you removed.
- Start compiling the final product. Create a bin with current episode and create a sequence from template with current episode and start copying good files from the RAW sequence. (which aren’t red) and shift the yellow in the position they need to be. Before copying any parts mark them with green so you know they were imported into final sequence. (you might want to leave yellow as they are to know you they were shifted) Now you need to check all cuts how they sound. Is the space too short or too long. Make Fine Cuts.
- Add podcast advertisement on the locked tracks (same track which has intro outro and which is locked so you don’t delete or move it accidently) Then move the outro all the way to the end and lock all tracks again. Don’t keep them unlocked if you don’t need them!
- Mastering. Before exporting add a Mastering effect (or it should be there if you used the template but definitely check to see if its there) Mastering effect on master track will make sure to have a limiter at -1dB and will add a slightly EQ with a boost in higher frequencies to make voice a little crispier. It’s a very subtle effect and you might not always spot the difference but it is definitely there. Listen to the entire track to check if everything is alright.
- Export. Final step is exporting this mixdown. Export mp3 at 320 kbps. (Check with podbean when they will accept wav file)
Final Notes: If you are not sure about some sounds or interruption which you want to remove in Audition, you can add a marker and then you can listen to the mixdown in Premiere (all speakers) and see if you really need to make that part silent. The markers from Audition will import into Premiere. See image below. 
Was this article helpful?
That’s Great!
Thank you for your feedback
Sorry! We couldn't be helpful
Thank you for your feedback
Feedback sent
We appreciate your effort and will try to fix the article