A few years back while designing sounds for a horror thriller, I had the opportunity to do something interesting with sound. The reason was that the director had given me full freedom to do what I wanted to do. The film was not an extreme thriller, but since I had been given the freedom to do things the way I wanted, I let it all out. This post is primarily with Dialogue ideas I used then.
We are all used to hearing dialogues spoken to us in a normal way. Another thing that people had experimented was reversing. I thought that since these were already done with, what could be done next? It was by accident that I had reversed a dialogue and it sounded like a tune to me. I tried to sing it and then it struck me. I reversed dialogues whenever the psyching was needed, dubbed the reversed like as it is like a foreign language and then reversed my dub to sound like a recognisable one. The only thing is that while it was recognisable, it was also not natural. I had this in various delivery styles, stressing different words. The primary reason this psyches us is because of the way the breaths are. We are so used to hearing a language being pushed out of our oesophagus that when reversed, although the linguistic portion is understandable, the delivery makes us uneasy. Couple this with a few pitch variations and you have an amazing effect.
There was a scene where the spirit in the film appears as a haze and was also talking. I took a cue from the visual. After all thats what we let our art dictate isn’t it? As the haze and bluriness of the spirit varied, I put a notch EQ on the main dialogue and varied that. My movement was simple. There were 2 notches. One as the spirit approaches and departs from the camera. The second was for the blur on the visual effects. There were times when the notches overlapped each other and crossed each other. But the visual supported this movement and you now have a very dynamic movement on the dialogue to match the visual.
One of the scenes had the character go into water and become part of it and start talking. Now, my first instinct was to do an underwater effect. I decided against that later. It was too cliche. I analysed the scene. The character becomes part of the water. How would liquid talk? Well, we dont know. There was no way i could synthesise that too because it was dialogue. This is what I did. I created 3 tracks in Pro Tools. The first track had the dialogue. The second track had a copy of the dialogue with a mondomod plugin (from Waves) on it. I set the frequency to be varying and to give it a wavering tone. There was a third track with a stream of water that was not too bubbly but flowing. The second and the third track had a gate on them with sidechain. The sidechain signal was a bus from the main dialog track. The gates had different attack, release and thresholds. The idea was this. When the character talks, the gate opens and the process voice and the flowing water will playback. The threshold will determine how much of it will playback thereby giving a delivery based variation to the processing. This gave me an incredible technique of morphing sounds with dialogues that I have later used with many nature elements in films. So, you now have talking water.
There was a huge hall in the house (haunted obviously!). Now, once the activities start and towards the climax of the film, the dialogues take place in this space with the space becoming surreal. I did something really simple here. I had a huge long reverb (around 10 seconds or so of Reverb Time). To make this unnatural, I added a huge pre delay. There was also the regular reverb for the size of the room. The effect this created was that the long reverb would kick in after an amount of delay thereby throwing off a certain perspective. But because the natural reverb of the room was kept, it was easy to trick the mind. To add to this, I also had a ride on the long reverb based on the intensity of the dialog and the thinning out of the background score. I also inserted a pitch plugin on the reverb and tonally kept it moving in semitones or close notes. This created a dischord with the background, but was part of a dialogue and performance, yet very disturbing. So there you have, unnatural space.
The weak and the Strong
The spirit gets stronger as the film progresses and enters a person. There comes a scene where the person is possessed and the spirit talks through her. Now, the most common and quick thing to do here is to pitch shift the voice and mix it. I approached it differently. Before that, I will tell you my reason. The character gets rid of the spirit by will. So, she becomes stronger as it progresses. But the spirit is strong in the beginning of the act. There were two voices I dubbed throughout that scene. The spirit and the character. There was an additional thing I did. I made the spirit and the character say the same dialogues very monotonusly. I now had 2 tracks for each, the spirit and the character. This is what I did to the scene. In the beginning, the spirit was strong. So, I underplayed the character voice way below. I didnt vary that at all. I varied the monotonus spirit voice continuously. To add to this, i had an autotune on the Monotonus voice, so it had a constant note feel to it. Increasing and changing the volume levels on this gave a dynamism to the spirit. Considering the character, the monotonus voice was more prominent than the regular emotional one in the beginning. I forgot to mention that the spirit and the character talks in completely opposite emotions at the same time. When the spirit was aggressively saying a dialogue, the character said those same lines in a sad and emotional way. This helped me to bring in slight flickers of the person in between on close-up shots and play the viewer into the scene. As the scene progressed and the character became stronger, this balancing reversed finally ending in the character breaking the spirit. There as also a lot of play with reverbs in this but all the reverbs were mono and subtle to just glue the tracks. So, now, you have the weak and the strong.
Hi – I’m on the lookout for a sound design technique for “a talking river” and came upon this post. I like your idea! Do you have some sound examples of what you achieved? Would be great to even see the breakdown of the tracks and hear how everything reacts separately. My tool set is different – Logic and other modulation and gate plugins but I think I can try what you are doing. My initial thought was to somehow extract the volume envelope of the dialog and apply it to the water (instead of the gate) but I can’t seem to find a way to do it in Logic. Thank again for the inspiration
Hi Avi. Thanks! I dont have those sessions. But if you try it out using morphoder or a mondomod with a gate or even pro compressor with the compression monitor on you could get the effect.
Thanks for the ideas! I’m intrigued by Zynaptiq Morph 2 but the price is quite a bit higher. Will experiment this week. I only need it for a few lines. Thanks again
Sharon Kearney said:
Just discovered your blog tonite and really enjoying the wealth of information. Reading the Psyching Out piece above reminded me of something I read about what David Lynch did for the end of Twin Peaks. He had the actors learn the lines in reverse then shot the whole sequence in reverse, so the actors walked backwards and spoke backwards then the whole thing was reversed to look and sound forwards. The audio sounds similar to what you describe I think?
Thanks a lot Sharon! Yeah. It is similar to that I guess but was not something I knew before. 🙂 I had used this technique for an art installation before.