Two years ago Lene Henningsen approached me with a text that she imagined might be turned into some kind of ‘sound theatre’. It didn’t quite fit the mould of physical theatre, and she was reluctant to change the text radically enough to suit the stage. Were there other avenues that might be explored?
The substantial manuscript revolved around the relationship between Oskar Kokoschka and Alma Mahler, and while Kokoschka’s expressionism and Gustav Mahler’s late romanticism aren’t immediately my cup of tea, I was struck by the stubborn life-energy of the two central characters who managed to pick themselves up again and again in spite of some incredibly hard life-experiences.
We first imagined the piece as a kind of extended concert; with two actors, a handful of musicians, sound and video, and began with recording the actors1 reading the text as a kind of skeleton around which the music and sound design could be built.
Once a first recording was completed and edited together2, it became clear to us that we needed to take a closer look at the dramatic form of the text and how it might be communicated through sound. The opera director Tine Topsøe was invited in as a dramatic consultant, and together we set about dissecting the text and investigating how it might be reassembled in different ways. After many attempts, including various re-assemblages of the voice recordings, we eventually arrived back at something very close to that we had started with, but now with fresh insights, and a clearer idea of how to implement Lene’s vision – one that embraced the poetic aspects that might, under normal circumstances, be shied away from with staged productions. It was at this point that we abandoned our original idea of presenting it all in some kind of concert format and decided to focus on doing it as a podcast.
Out in the World
It also became increasingly clear that our original studio recording of the dialogue was too stiff – one could somehow hear that the actors were standing in a confined studio space. (There are some interesting similar observations in an Aeon article on Orson Welles’ War of the Worlds.) The next step was re-record some of the scenes out in the world, and the wonderful Hirschsprung Collection kindly allowed us to use some of their spaces. We set about making new recordings ‘on location’ and with a better idea of the dramatic flow of the scenes.
A practical aspect, after having made a new set of dialogue recordings, and completing the long process of sorting though them, was cleaning them up – mostly removing things like unwanted noise from the lavalier mics, and in some cases also adjusting or removing background reverberations. This brought with it a deep appreciation of those that edit film dialogue, with all the intricacies in vocal nuances and background ambiences that need to be adjusted and matched.3
With a basic, cleaned-up dialogue skeleton in place, the next task was to once again start placing the voices in acoustic spaces. Podcasts, at least many of the more professionally produced ones, typically opt for a dry sound without any reverberation or sense of the room they are being recorded in. This helps keep the dialogue clear, and when listened to with headphones, as podcasts typically are, gives the impression of the voices being ‘in one’s head’. In this case I kept Lene Henningsen’s ‘meta’ commentaries in this dry, ‘neutral’ space, but chose to keep or (re)create the spaces in which the characters unfold their dramas: An art gallery, Oskar’s apartment, Alma’s summer residence, etc.
One of my first ideas for the sound-design of the piece was to acoustically recreate something along the lines of the swirling brush-strokes of Kokoschka’s famous Bride of the Wind in terms of sound: to find a way in which the characters words could swirl around the listener in a way analogous to the paintings.4 This could either take on an agitated form, as in the argument between Oskar and Alma in Scene 2, or a more gentle aspect, as in Oskar’s description of Bride of the Wind in Scene 9.
Another aspect was the addition of straight forward sound effects such as explosions in the scenes in which there are descriptions of war and the like. This involved a deal of research and much time spent on the Freesound site and Soundly service, and as with the dialogue, brought a newfound appreciation for this kind of work. On the one hand Lene imagined a role for the poetic aspects of sounds such as those of war, wind, and water, but I quickly found myself steering way from a something that was too illustrative, too much of a soundtrack.
After initially adding various layers of sound effects in Ableton Live, I got stuck – probably largely due to the aspect of it all becoming a little too illustrative (as mentioned above), and decided to try another approach: mainly using AUM and the SpaceCraft granular synth on my iPad. FabFilter also released a suite of their wonderful plugins for iOS which put some high quality audio tools within my grasp, and with the advantage of a new context in which I couldn’t fall back on old DAW habits. Somewhat inspired by Hainbach, I took to thinking of the soundscapes for each scene as based on a set of (virtual) tape loops, with both large chunks of audio and minute grains within SpaceCraft (and with the help of Bram Bos’ MIDI LFOs to sweep across the larger audio expanses) or simple audio loops using AUMs file player – all shaped and placed in spaces using the FabFilter plugins, and the relative volumes of each loop ‘played’ by hand on AUM’s faders.
Part of my thinking with the ‘performance’ set-up on the iPad was to find a way of working somewhat analogous to the way Kokoschka created many of his paintings and drawings: themselves small (or large) performances, rather than being built up from preparatory sketches, that embraced the imperfections that arose as part of the story being told. With my collections of loops faded in and out by hand it was difficult to place sounds and effects with the precision that one might when working with a timeline in a DAW. Instead broader textures could be created with close attention to the relative volume of the words and the sounds – I ended up mostly recording it all to a single stereo track, giving up the possibility of endlessly adjusting it all at a later point, as one might with a multitrack setup.
I also found that working with AUM helped focus my attention more closely on the sounds, without the visual distraction of a timeline laid out from left to right, as practical as that might be. I ended up doing many takes of each (sub)scene, simply choosing the takes I felt worked best overall: In some cases this meant giving up on details that I might have liked to place sightly differently, in favour of the overall result.
Another breakthrough in finding a path for the sound-design was setting up a noise-floor: I settled on a basic (white) noise continuum ranging from the roughness of radio static and crowd cheers to ocean waves and the gentle filtered white noise of wind in the trees. This provided a kind of ground canvas out of which the dialogue and other elements could emerge and return to – something that could remain as a presence once the various dramas had played themselves out.
An aspect that also took many painstaking runs to get right was the balance between foreground and background, between the dialogue and the sounds and voices surrounding it. Often it felt right to keep that background quite low in volume with the ‘musical sounds’ (orchestral samples and the like) kept quite subtle in relation to the voices. In retrospect that’s an aspect that makes the piece a little more tricky to listen to in noisy environments – there is simply too much information that gets lost, and this cuts out being able to properly listen to the podcast during a commute in a car, bus, or train. There’s unfortunately not much that can be done about it: Oskar & Alma is best listened to with a good set of headphones in a quiet environment.
Caught in a Rut
One aspect that I had a little fun with was EQ matching some of the dialogue recordings against historical recordings such as wax cylinder recordings of Gertrude Stein and T.S. Elliot. In the end it was too much to apply those profiles directly to the main dialogue, but those experiments lived on in some of the low circular cylinder noises and background EQ profiles.
Another aspect that I considered was in some way reflecting first on the wax cylinder and gramophone as the means of recording available during between the world wars, and tape as something that emerged after the second world war.
The gramophone and cylinder loops could also be taken as a counterpart to the recurring emotional and thought patterns Oskar and Alma found themselves stuck in. In the case of tape I was thinking of layers slowly being erased over time, or the technique (or technical deficiency) of additional layers added over previous recordings not entirely erased.5
Finding Its Form
The piece ended up taking on quite a different character after changing paths and deciding to do it as a podcast rather than some kind of extended concert piece. The podcast form made it possible to approach the sound-design in a manner quite different to that I had initially imagined, ending with a result I now find more suited to it. I’m also enthusiastic about how certain aspects of the podcast medium can add to the piece: The podcast artwork, the ability to jump to a particular chapter, and especially the episode or ‘show notes’ with links that can be followed up on as expansions on points of interest referred to in the podcast.
There are however some aspects that don’t quite fit what one normally expects of a podcast. John Gruber sums up some of those assumptions and expectations with a nice description on Daring Fireball:
A podcast, to me, is a series of audio episodes available over the web. At a technical level, it’s an RSS feed, and the RSS feed has entries for each episode, and each episode has links to the actual audio file (in MP3 or AAC format, but usually MP3) and other metadata. RSS is an open format that can be used to serialize anything.
Oskar and Alma isn’t conversational in the sense that many podcasts are, and we decided to present it as a single episode – as a ‘one-off’ rather than a series of scenes unfolding over a number of weeks. In that way it’s perhaps closer to an old-fashioned radio drama, but presented in a form that can easily be downloaded to one’s phone, computer or tablet, and with the added advantages of the chapter art, notes and links, as referred to above.
Oskar & Alma is a podcast simply in the sense of an audio piece available on the web that can also be accessed via RSS and listened to in a browser or some kind of podcast player.6 At a point in time at which ‘audio is the new black’ and expectations become increasingly set with podcasts carefully designed and optimised with commercial interests in mind (see also Gruber’s article above), perhaps it’s also a good junction at which to remember the freer, less structured, more artistic possibilities of the medium: time for some poetic podcasts.
* * *
Oskar & Alma can be found on the Poetisk Podcast site as well as via Apple Podcasts.
Mette Frank and Ulver Skuli Abildgaard ↩
It also involved a dive into the software used to accomplish these tasks. I started with demo versions of the widely used iZotope RX7 and perhaps lesser known Acoustica, eventually settling on the latter due to its more attractive pricing and the fact that I found it easier to use, despite the interface not being quite as beautiful as iZotope’s. ↩
If you have any thoughts or comments you can reply via Twitter @RudigerMeyer, send me an email, or send a response via the webmention form below.