A few weeks back I took on a new project, a 10-minute short action/thriller, which Sam is helping me out with. It’s proven to be a very time-consuming project and as it is my first proper post-production ‘job’ I do, my inexperience has definitely come in the way more than once.
As I wasn’t present from the start of the project, nor was any other audio student, I’ve walked in to very poorly recorded audio which has not been in a focus of getting good sound but rather, to have some audio available. What this has meant for me is that I have somewhat of a struggle to get the audio to have a more-than-decent quality.
The first issue, which is slightly embarrassing that I didn’t remember as I’ve researched it before, is that I didn’t import AAF files into Pro Tools from Premiere but rather was handed all the files and started to work out their placement myself. After about four days into this process the light bulb went on and I was thinking to myself that there must be a way to import the audio with timecode printed…
Yes Sebasthian, there is.
After bouncing out AAF files and importing into Pro Tools It’s been a week of steady progress. I started out with chopping up and editing the audio to fit the scenes, resynced a lot of it, and found more suitable takes where necessary. Doing this and getting to a state where I feel somewhat satisfied has taken me up until last Wednesday. It’s been a quite stressful experience so far as the deadline for the project is next Wednesday and my lack of experience with post-production workflow and techniques have not been helpful at all.
While editing the clips I came to realize that there were more than a few sections where none of the takes worked out as desired. During one of the scenes, there’s a zombie growling behind a fence, the original audio was nowhere to be found and we couldn’t find suitable replacements. This led me to take a few random screams from the same actor and cut, paste, time-stretch, pitch, and EQ until I got it to work. This was not a completely new process to me as I had to do similar work with my Mac ‘n’ Cheese sound replacement.
In a similar sense, I’ve had to go over one of the other zombies growls and, mainly, pitch them to fit the visual content and the aesthetic feeling that the scene is trying to convey.
After the foundational grunt editing was done, I moved on to making a rough mix and sorting out the levels of the clips. One big issue here has been that as everything was only recorded with one microphone, all the audio is summed. This has given me severe issues with balancing certain parts of the project, however, after a quick feedback session with my lecturer, I’ve come to understand that RX spectral repair can actually remove parts of the audio. This essentially allows me to duplicate clips, remove the unwanted audio as far as possible before layering them back together and processing the one that needs amplification. Sadly, I’m not entirely sure that I will have time to go very deep with this as the whole project could benefit from this process. That is, however, a lot of work for someone whos never done it before, and I’d believe that even for a professional it would be quite time-consuming. There are although, two clips that especially needs it done and I will hopefully have that accomplished after my S6 session on Sunday.
Naturally, after pushing all my levels up, there were a new set of issues appearing. Noise, and lots of it as well as clicks, pops, distortion, and awkward reverberation. This is the first time I’ve had to work with noise reduction plugins and they have proven to be extremely useful. With iZotope’s RX I’ve been able to a 98% degree remove all unwanted of the above-named issues. Yet again, a very time-consuming process as I had to learn how the different tools worked. I do have a last, very tricky, clip to edit which will most likely have to be done in the spectral repair in combination with some corrective EQ.
I’m really happy with the result and overall the audio is just so clean, however, this created another issue for me. I was always aware the this was a process that I would have to go through but the result of it is that I’ve lost quite a bit of the actual room tone. Now, this wouldn’t have created an issue with an audio student on site from day 1 as room tone would have been recorded, in this case, it wasn’t. Now I got a few scenes where there’s audibly room tone shifting due and I’ll have to figure out the best way to cover this up. I’m thinking of either, preferably, recording new room tones in similar areas or covering it up with white noise.
The next step on the road to completion is to add the missing sounds, yes there are heaps of them too! I’m missing zombie growls, footsteps, body impacts, punch and gore sounds, however, I’ve recorded everything that I need as well as being provided with samples from Sam. Yesterday I went into the C24 with three members of the film crew and we recorded a freaky sounding toy, zombie growls, and a wrench. There’s a good amount of work left to do but I feel confident that it will be pulled off in time.
Preferably I’ve been wanting to use iZotope Insight for my mastering process of the film, in terms of mixing to the OP59 broadcast standard, however, at home I’ve only got access to the HOFA 4U meter which I’m not as comfortable with. I find the UI of Insight to be much more descriptive and easy to read seeing as they provide a visual representation as well as numbers and metering. Thankfully, uni has Insight available so I will do my final mixing on site together with the director next Wednesday evening so that he can provide some final feedback before delivery.
To me, this means that I have to ensure the mix is lying as close to -24 integrated LUFS, as well as not peaking past -2dB true peak. To ensure this is not occurring I’ve put a brick wall limiter on my master at -2.8dB as I don’t have access to a true peak meter as well as checking my mix, making sure I stick to -24 LUFS. I have yet to know if I’m only delivering one summed ST print of the audio or if I’m delivering separate M/D/E stems as well. This could, however, prove to be a bit complicated due to the summed state of the recordings.
I’ve had four meetings with the director and film editor where feedback has been given and acted upon immediately during the session or shortly after which has progressed the work immensely as I’ve been ensuring that I’m going in the right direction.
Seb's Audio Journey 