Post-mortem for Sound Replacement Project
- purveshmehta98
- Apr 2, 2017
- 7 min read
Final Product
Project management & Teamwork
Reflecting on the weeks of our project, I can confidently say that there were very few issues that occurred during the project that we couldn’t overcome as a group. Generally, there were more positive aspects and new things that we learnt along the way while executing the project, guided by our preproduction plan that was well written with a detailed asset list.
Some of the issues we faced were mainly during the initial phase of the project and towards the end. At the start of the project, we struggled to find software that could convert our original video’s frame rate to one that Pro Tools was compatible with. Discussed in my previous blogs, the issue was quickly solved by me, Simon and Sujahn by each of us researching and finding a free piece of software that could convert the video’s frame rate. We each found different software and did some trial and error until video converted properly. However, these free programs had issues that we later discovered and couldn’t understand and realise until the end of the project. For some reason unknown, despite the frame rate being converted from 20 to 24fps, to allow us to drop samples onto the grid in Pro Tools, the final bounced out version didn’t sync correctly to our audio. Frustrated and confused we decided to go through endless forums trying to solve the issue, however, no one seemed to have a solution to our specific issue. What we believe could have happened, is that the conversion didn’t occur properly even though Pro Tools was reading the newly converted video at 24fps. What we did note was that when we put the converted video file and original audio into Pro Tools, the audio didn’t sync up with the video. We thought that we could manually put it back into place however there timing was way off in some sections and synced up in others. We decided that we were only going to use it as a reference as to what the original sounds sounded like so we didn’t care that it was out of time because that track wouldn’t be needed in the final bounce anyway. What we should’ve have done, instead of just leaving it, is actually find out why this was the case, we should’ve bounced out a version of the video with our samples that were in time with Pro Tools to see if the final bounce was also in time.
Despite this, we ended up bouncing the stereo audio mix separately, then syncing the video and audio up together using another free software like Movie Maker or iMovie however not really understanding how to use these programs and due to their limitations, being free software, the audio still couldn’t be synced up. I ended up deciding that since the video engine in Pro Tools played back the video in perfect sync with the audio, I decided to download a free screen capture software called OBS (https://obsproject.com), which is used by streamers and gamers that record their gameplay on their screens. I went through some basic tutorials on how to set up the software to capture only the Pro Tools video engine screen then imported that video into iMovie and placed the bounced stereo mix file of our samples into the software to get our audio synced up perfectly.
In the end, the final product was achieved and I was quite happy with my problem-solving abilities to find this solution to our problem. I did, however, brought it upon myself by promising the group members that I would deal with the issue and send them the final product before the due time, luckily I found a solution to the issue. Overall like I said in my previous blog, it was a pleasure working with my group and plan to continue to work with them on future projects in the near future.
Plans & Pitches
Our project plan, initially created by Simon, was something we believe was very clear and concise. We had planned out the following weeks to come and ended up completing tasks ahead of time and updating our asset list during the group recording sessions, the same time we would upload the session files onto the drive for us all to have access to. Although the asset list and the information regarding the exact time for the sample to be placed onto the timeline, this was not used at all because we ended up finding a better quality video which had different starting and ending times, this, however, didn’t affect our productivity as moving the samples into place didn’t take long as our workload was evenly spread between me, Simon and Sujahn and we had quick access to the Pro Tools session. We also communicated with each other almost every day, online, using messenger and keeping track of where we were all at, planning studio session times that suited us and also meeting in study rooms and outside of SAE. Overall the pitch and project plan was effective in guiding us in what approach we should take to build up each of the elements of the mix and polishing it before the due date.
Creative Work & Processes Used
Initially, like discussed in my session 1 blog, we recorded some Foley and clothing movement in the C24 post production studio. Most of this was fairly straight forward, just creating multiple audio tracks in Pro Tools and recording multiple takes using playlists. After this we recorded samples using a Zoom H4N which we placed in a separate folder within the Pro Tools Session folder, named H4N samples, these were also copied over into Pro Tools Audio files folder when imported into the session. As for processing, I was in charge of layering and processing kick movements, body impact sounds, blocking sounds, room reverb, hand wipe sounds, and hair ripping sounds. Simon teamed up with Sujahn to work on the ambience, cat sounds, Chuck and Lee’s ADR, Footsteps and then during the slow motions scene, we teamed up together to layers and process all those sounds, noting what it would sound like if it were to be slowed down.
Final Pro Tools Session displaying all tracks
Firstly, I colour coded the tracks and clips within the track to identify between light movements which would be created by Lee (purple clips) and heavier sounding movements for Chuck (light blue clip). After the clips were put in place, top, tailed and some consolidated. Corrective EQ was used to clean up the samples, cutting some low end, whereas creative EQ was used for the slow motion kick movement sounds, where the low end was emphasised. Varispeed was also selected for some of these samples to make each sample sound individual and different from each other, the TCE tool was used to stretch pitch down and bring out the low end while some were pushed together to pitch up the sound and make it sound thinner for some of the subtle movements.
For the body impact sounds, lots of EQ was done to create larger, heavier low-end impact sounds, using a low-pass filter to get rid of some of those unwanted high frequencies. All these channels were then bussed to a room reverb that was used as an overall, glueing reverb for all the samples.
As for the slow motion scene at the end, Varispeed and TCE was used to stretch out the samples and give it that slow motion feel. EQ was used to boost the low-end, also for the purpose of emphasising that slow motion feeling. ModDelay 3 was also used to give the samples some subtle echo effect, all these processes were completely experimental and purely instinctive, we didn’t do any in-depth research in sound design or anything, just using the knowledge we have and just experimenting and finding our own way, giving it a sort of authentic feel that you wouldn’t be able to achieve if it were completely reliant on pre-sets or following others methods.
Finally, after all the processing was completed, Simon and I played back the video at his home set up, monitoring it on his Logitech speakers for some final polishing and panning. For the panning, it can be seen that groups were created for layered samples that required identical panning to bring everything together and add to the stereo field. The panning was done by Simon, where he panned elements to where they would be prominent during that particular scene, for example, when Lee would kick from the right side of the screen, the samples would play back towards his side, say if he was on the left of screen, the samples would playback panned more towards the left channel and so on.
Taking on the feedback received, it was suggested that there could’ve better balance between the ambience sounds, which is something I purposely brought down in the mix because of some elements that stuck out to me which made it sound like it was looping and not fluid, to mask this however, another recording should’ve been used to mask it. The ADR was also apparent as being the loudest in the mix and still giving that “cheezy” Kung Fu film sound however, I believe this was due to Lee’s vocals sounding somewhat off because, no matter how hard you try to pull off that signature Bruce Lee fighting cry, there is no one that could do it better than Lee himself. A lot of respect to Simon for his performance and this was quite taxing on Simon’s throat, trying to scream like Lee does take a toll when you’re in a very cool and dry studio live room. Overall It could’ve been better improved if we sought feedback throughout the process and this will definitely be done for future projects.
Sound replacement for film was an enjoyable task, doing something out of your comfort zone and growing as an audio engineer, being able to do things that I initially thought was something I wouldn’t have the patience to do, ended up being a good learning experience. I Still plan to further my studies in music production and train myself to be better within that field as opposed to doing ADR and Foley which is something I'm not as driven to chase into the future. I’m more willing to chase something I'm more passionate about however I wouldn’t take this experience back and the processes I’ve learnt along the way, can be used to better improve my music production, organisation and workflow within Pro Tools as well as bettering my skills, working with a team.
Resources:
Obsprojectcom. (2017). Obsprojectcom. Retrieved, from https://obsproject.com
Comments