I’ve been spending the past few years trying to learn audio engineering through YouTube videos and message boards and wanted to ask others what’s their preferred order of operation for applying changes to audio?
Are there certain changes I need to be making first before other changes (ie compression before/after EQ)? And if I use a noise gate, am I just wasting my time with noise reduction? Do I DeReverb and DeEsser before EQ’ing and processing? Should I cut down an episode (misspeaks, uhms, etc.) before or after changing the sound of the audio?
In short, is there a preferred load order that allows each change to work as it needs to while keeping the highest level of audio quality?
.
What I apply to my audio in order: noise reduction, loudness maximizer, dynamic noise gate, dereverb, debreath, normalize, cut down episode, loudness maximizer, dynamic processing, deesser, normalize, hard limiter.
In short you should always strive for as close to perfect audio at recording as possible so the first step shouldn’t be needed.
The first step is to repair the audio, such as dereverb and a noise gate. As a note I’d avoid dereverb if possible as in order to remove reverb it degrades the audio to what I’d call unacceptable quality. If you have to use dereverb at more than 5-7% don’t use it and re-record eliminating the reverb at source.
The second step is to process the audio. By this I mean add your EQ. It’s a personal thing, everyone’s voice sounds different and even one person’s voice can sound different when recorded by different mics meaning different EQ is needed. Once you’ve got your EQ sorted you can save it as a preset and then if you’re recording in the same environment with the same mic you can use this preset.
The third step is to compress the audio. For a podcast you don’t want the audio sounding too compressed so avoid slamming the threshold and pushing the ratio or you’ll end up with an over compressed sound.
In terms of chopping and editing I would put the effects into the effects rack of a multi-track track and then you can edit, cutting out the bits you don’t want non destructively and mix down to an output file.
Finally don’t forget to match your output file loudness to the podcast standard using the match clip loudness window.
Thank you so much for the insightful help Mark! I appreciate you taking the time out of your day to share your knowledge. I will take your advice into my next recording session. Thank you once again!
You would certainly start out by repairing the audio before you go ahead and enhance.
Start with DeNoise, Noise Reduction, DeReverb and then head into EQ followed by compression. If you do the latter first you’re essentially enhancing any noise that you don’t want in the recording so repair first. Even better, record as clean as you can at source.
The noise gate you mention is a little different to noise reduction/restoration. It doesn’t touch your dialouge it simply makes everything quiet when there is no speech. A great thing to place on after you have done as much noise reduction as you can.
Thanks for your helpful information Mike, I will add your notes to my notes!
As for a noise gate; I understand the basic principles of setting one up to help eliminate background noise or mic bleed, but how do you use the additional settings (Attack, Release, Hold) to make your gate sound more natural? You briefly mentioned upon how those worked in one of your livestreams a few years back, but I still have trouble wrapping my head around the idea of how they work.
Great qeustion @AdamWilliam about noise gates. Here are some answers!
Noise Gates: Attack, Release, Hold
Noise Gate Attack
This is how quickly the gate will open once your audio is above the set threshold. You’ll usually want this to be as low as possible.
Noise Gate Release
This is how long it’ll take for the gate to close after your audio goes below your set threshold. Leave a bit of space on this to allow natural word endings while working with dialogue.
Noise Gate Hold
How long the gate stays open after it exceeds the threshold.