Stage 5: Beat movement edit

 Sound matching beats

With any fast pacing thriller, sound is one major way to contribute to the general atmosphere. This could be donw by matching the footage movement to the sound, to create a coherent action beat

Sound draft (29/3)

For not having enough time to record the sound orignally as we wanted, I created a mock up of the color graded version with stock sound online, to see how the mood of the opening sequence would go with sound in it to add more contextual layer.

    I edited this version in the span of 2 (unconcentrated) days, as I was hoping around Adobe Premiere Pro (having the school account) and Adobe Rush on my personal device. The most time consuming part is navigating all the stock sound unzipped to place it in the correct segment of the editing timeline. As Rush is not as supportive with importing and displaying imported material, it took me a while to scroll through different files in the folder, with similar title as backup for similar sound. 

    Moreover, as ambient sound is prevalent in suspense, I was becoming susceptible to loud, disorientating noises after hours putting together the audio. The differences in windy sound, from city bustling atmosphere to pure, cinematic controlled environment wind or a stock suspense rush, were blurred and melted together after a while, while contrastingly enhanced to me. I was able to notice to their small differences. 

    Audio output quality was one of the concern as well, as I was switching back and forth of my own headphone to earphone to the school ones, I lose some of the intricacy of mulitple audio layers in lower quality earphones. Moreover, as higher quality headphones can differentiate better audio in both left and right ears, I accidently select one sound file that can only plays on the left channel, which ultimately did not come out sounding as immersive. 

    Overally, there are some flaws that I identified with this draft, as well as learning experiences for next version:

• Sound stock for running with boots is no applicable - this is the first priortity of us to record the sound ourselves. The sound stock online was mostly running on sand or wood, else it was not very depicable through recording, as well as the running step was very consistent and patterned, while our running scene is quite disorientating with different pace. I compromised by changing the volume of each running segment, and/or clipping running soundwave together to increase pace, or randomly cutting them up. 
• Layering of atmospheric sound, which is crucial to depict the urban atmosphere without having organic noises. In this version, I layered 2 wind stock, and 1 base layer of eerie presence ambience sound. In later version, I'd consider to layer more layer of vehicles and rumbling background music
• Characters talking over phone - we thought it would be easy to find distorted voice over the phone, or mumbling at least, but there was no direct match to not make it being too comical. I've decided in this version to cut out the detective (Nat) mumbling at all, and thus only hearing the wind whistled until we can record it ourselves. 
• Building suspense ambience is complicated as well, as ambience sound stock may not have the progression we wanted. Some ambience sound was really contextual to our opening sequence, but not having enough substances to use over the span of the running scene, or too long but not having the enough build up if trimmed down. Moreover, I had to source 4 difference ambience to connote change in physical environment and emotional state of the character, to have enough distinguish but not too "left field" to be noticeable. 
• Lyrical song in audio, was intricate as matching the song context to the action at hand, thus I was cutting down the song, having the peak emotional state with the first line of the chorus, then transitioning to the last line of the chorus then merging the immediate bridge and outro together to be more convincing when the last note of the piano lingers. I am most proud of this decision and precision I made in this segment, as it sounds so seemless listening back. 

 

➡️➡️   I was pretty pleased with this version, as we have made do with what we have at the moment, and learn quite a few bit for later edition. 

________________________________________________________________________________

Sound matching beats 2.

(10/4/23)

The newer version is indefinitely better and smoother, as I sat down straight for 4 hours to complete this task on Premiere Pro, which is significantly more optimizing than Rush. 

    With the new rough cut transferred by Rene, I went on and did a semi-final version of soun editing, to match sound effect with the footstep in step printing and more intricate detail in general. It was definitely easier than the previous attempt, as I have laid out some rough estimation for each sequence in the sound mock on 2/4 (link to blog post). I finetune some moment, add and compare to online sound stock, find more ambience sound to enrich the atmosphere and insert effects on top of the sound level. 

        However, the sound quality issue with high quality audio output is exaggerated as we streamed the opening sequence with default laptop speaker. It was almost inaudible, some sound layers are undistinguishable, while other being distorted due to loud volume. It was vastly different. Therefore, I had to be mindful while adjusting sound level and complicated layers and fading effects to not get distorted.

        I showed the semi-final draft to my teammates and here are some of the feedback I received:

        Here are some of things I learn out this time around:

• Speed/Duration alteration: In Premiere Pro, there is adjustment for speed for audio, thus making it easier for me to match footstep and dialogue with mouth movement of the character. I used it a lot to stretch out any faster pacing walking sound for step printing sequence, instead of having to duplicate it or trim individual footstep. However, increasing the duration would decrease the pitch significantly (as noted as "different gendered breathing voice 😭)
• Pitch correction: So that I asked my teacher to assist me with correcting the pitch accordingly. It entails in using Adobe Audition, and adding more effects on pitch shifter. A trade off with this is that the increase pitch (assumably higher voices) would be stretched out and becoming "grating, automatic. However, this issue is hardly noticable if placed amongst other audio layer.

 

(Adobe Audition)                                                                      (Reverb panel in Premiere Pro_

• Reverb:  Admittedly, I did pretty overdone this, as ambience sound and distant sound would be more realistic and overwhelming with added reverb. I used this for bulding on ambient "whoosh" sound when the antagonist appears. the gunshot and the body falling down to create a more distanced and heavy hearing, like the viewer is in an observant role. However, adding on more reverb could also lower pitch and making it inaudible even with volume turned higher on.
• Audio transition: Is the most useful tool for an amateur editor like me. Blending the sound together, especially environmental sound (wind, city bustling, chattering) to ambience building (ambience background music, suspenseful sound effects) is significantly easier with the Constant Fade feature in Effects. Moreover, it helps me to create sound bridge of step printing to normal close up angle, as I subdue the music down by fading transition to low volume, and increase in Constant Gain of ambience whoosing sound. However, it would be better if I know how to control the degree of the fade, instead of all fading into 0% volume.

However, I was proud of this rendition. I would treat this as a milstone reference point for my sound editing, and making minor changes towards to final cut (with title and idents and all). 

🌱