Béla Bartók - Duos for Two Violins
Videographics - EQ - Recording - Video Editing (Camtasia)
Due to COVID, I figured that the best way to produce a recording of these duos was to multitrack myself playing both lines and then combine them all together. This leads into some technical challenges, both from a visual and audio perspective.
To begin with, I researched several practitioners who have performed these duos by using split screen. Here are three examples I found with an analysis of each:
1. Calligopoulos - The audio seems to have no panning or it's been recorded mono. This performance doesn't have a feeling of space and instead feels restrictive. I'm impressed by how he managed to keep in time without any headphones, but you can hear in between duos a clinging sound of a metronome that is mixed in with the low frequency noise in the background. The visual aspect is good, but there doesn't seem to be much interaction between the two players as a result of both of them looking in the right direction.
2. Bingham - The audio is stereo and has been panned here. It works really well, especially in the sections where the accents convey a conversation between two players. This is also reflected in the visual aspect where there are three cameras (one centre and 2 side) which can also show the conversation effectively. Even when the other player isn't playing, there's a real sense of performance in this. It's a good duo to perform because there isn't any tempo changes to worry about. The video being on a black screen means the split line is not visible, thus making it more likely the listener will be tricked visually.
3. Weilerstein - Similar to 1. where the interaction between the players is variable, mainly because of the way it is visually set up. Panning is fairly central but the transitions between the duos hasn't been considered and the change in space (due to various players home) can be heard and seen. (esp. 5:26). However, with the level of playing, there is excellent command to tempo changes and interaction through the audio.
My work goes beyond the three practitioners discussed above. On the audio side, I want to go beyond just simply a stereo mix of the performance but instead exploit the duo characterisitics and creatively edit the audio so that I can do tempo changes effectively and give the listener an unusual experience of space. On the visual side, I want to convey greater interaction between the two players, despite there only being one person recorded at a time. Thinking beyond the notes, I want the transitions between each duo to be considered, including the beginning and the ending of the video where I can show my consideration for this. Although in Bingham's example, I really like the idea of 3 views so I will replicate this.
I used a Rode NT2-A microphone - set using the cardioid function and angled so that the sound from the body of the violin was captured. The layout for the recording is shown in the picture below. I decided to pick a space that had a natural reverb so that the audio output didn't require any artificial reverb and less editing would have to be done to the sound. In terms of EQ, as there was not much change in the positioning of my playing, it was fairly similar on all the duos I recorded. The only adjustment required was to bring the dynamic of the loudest pieces down to a suitable level, as well as cutting the lowest frequencies to remove any background sounds. Cameras were positioned to take a central view and side views. These were rolling throughout the whole recording session so editing in Camtasia would be a breeze.
Some observations/considerations for certain duos:
Pillow Dance - I decided that a quicker tempo would contrast a slower Play Song rather than have two duos at the same tempo. Visually, I wanted to play around with the sleeping mask on both players. However, I got confused which sections required the mask meaning that I was on the wrong side when recording! To remedy this, I flipped and rotated the video in Camtasia - which was a good solution but for eagle-eyed viewers they can see an extra stool in the background and I also play the violin on the wrong shoulder!
Mosquito Dance - I decided to be creative on the panning for this duo. I imagined a mosquito flying around and then suddenly being swatted at the fortissimos. As a result, I panned it extreme (43L 42R) and then suddenly central at the climax. In terms of space, although rather unconventional, it worked well with the character. Notice on the screenshot, a further two tracks. These tracks were MIDI generated from Sibelius to help me keep in time, as the piece itself has some rhythmic challenges. An extra bar was added in the metronome clicks to accommodate for the pause in the middle of the piece and I decided in the end to cut two different takes together, using the fade in/out tool to help merge the two separate takes together in the lower part.
Ruthenian Kolomeika - The challenge in this duo was the tempo changes. In Calligopoulos' example, he ignored tempo changes but I wanted to keep them in and have the feeling of rubato in this particularly expressive duo. As a result, I decided to split the piece into three sections and record each section separately. This was a lot easier to do than setting a metronome to change at different points because then there is more flexibility in the amount of rubato you have between each section, controlled after by manipulating the recording. In terms of video, I was able to strategically switch to one of the side cameras to cover the two transitions so it seemed like one continuous take.
Reflection:
The clicking between each duo turned out to not only be a good transition between each duo but it was also really useful for lining up the video footage with the audio from the microphone - something to remember for the future! One thing that I wasn't aware of was how some video software deliberately compresses the sound. I was unsure where the compressing was coming from but after a quick comparison in Ableton (See comparison in amplitude from my edited audio Channel 1 and the compressed video audio Channel 2), I worked out that Camtasia was compressing the audio after I exported the movie file. In response to this, I then went back into Ableton and muted the compressed video and overlaid it with the non-compressed file. The result is a bright, clear recording that is enhanced by the visual element of the interaction between the players and the few humorous takes dotted around the video.
(1131 words)