Monday, May 09, 2011
Misbehaving Clocks: A Primary Pathology of Timecode Troubles
by Gautam Pemmaraju
What then is time? If no one asks me, I know what it is. If I wish to explain it to him who asks, I do not know.
- St Augustine
Of the many professional vexations that I have encountered, there are a few that remain implacable. They appear unannounced, lurk in the shadows, thief-like, and seek out opportune moments to manifest, bringing a unique set of anxieties, afflictions and injury. There is a quality of mystery to some: their appearance is seldom anticipated, the torments they unleash may or may not be prior detected or prevented, and their severity may not be accurately assessed until after the damage is done.
So when, a few months ago, I took the tapes back from the two-day Mahindra Blues Festival to the edit studio for post-production (the multi-camera TV production of which I had directed), I was to soon realise to my utter dismay, the anguish that was in store for me. A multiplicity of timecode issues – drift, break, sync, control track – appeared on the master tapes and I was confronted with the horror of the loss of automated synchronization amongst other devilry. This perfidy cannot be overstated – the prospect of trying to achieve/repair sync, the flaws of which are in turn mischievously asynchronous, begins with the acceptance of many, many edit hours of painful remedial work. Someone or something fucked up and I had to pay for it. I need also to mention here that generally, post-production suites are vile, dank, freezing holes-in-the-wall inhabited in many instances, by overworked, underpaid editors with frightening dietary habits and appalling personal hygiene. Editors and directors, as in other symbiotic partnerships, have no alternative but to rely on one other and any breach of protocol or even some unknown impedance in their delicately calibrated fellowship, can lead to disastrous consequences.
The technical/historical aspects of this revolutionary innovation1 are reasonably well chronicled but there are literary and artistic ideas in the invocation of timecode – from its utility, its flaws and tempers, the consequential effects thereof, to its intriguing presence in mediated reality.
Timecode is essentially a labeling system for video, film and audio material wherein each frame has a unique identification address in Hour:Minute:Second:Frame format. As binary coded media metadata, timecode formats (standardized SMPTE, EBU) are practical ways to identify, locate, access and then manipulate recorded audio/visual data. My own introduction to the various formats, the history of innovation and development, standardization principles and processes, and the strange, complex ideas associated with this technology began way back in graduate school when the ever eager TA of our music production class, thrust upon us his painstakingly compiled monograph, cutely titled, A Brief History of Timecode. Needless to say, he was mighty chuffed about it and was understandably annoyed with the mixed responses. The class was not emphatic in roundly praising his work, mostly, as I recall, due to his fulsome manner.
It appeared to me, as I watched the legendary Buddy Guy on tape go progressively out of sync, that I was confronted with a case of timecode drift. As is the process in live (or ‘as live’) TV production, the inputs of multiple cameras are received simultaneously (and synchronously by Genlock) and are edited in real time to one recording source (on-line master). There are other recording sources known as Isos (short for isolation), which are fed the input of an assigned camera so that errors or corrections that have occurred live may be set right through these isolated recordings. All these sources, the live edit tape and the isolation tapes, must have the same timecode. In the days of linear editing (before computer based non-linear systems), these multiple sources would be locked through a controller by the reading of their identical timecode, and the corrections would thus be made onto a fresh source (a new tape). Now of course, all material is digitized to a non-linear editing platform (AVID, FCP, etc) and with the help of internal protocols these multiple sources are again synchronized with their identical timecodes.
In my case here, there was an offset between the timecode on the edit master and the Isos (or the ‘slaves’ since they are ‘slaved’ to the ‘master’ through a controlling signal). So when I tried to cut away from a long duration wide-shot into a close-up of the keyboard player, I found that the cut did not match. I had extrapolated a piece of real time recording and placed it onto a timeline of the very same recording, but at a different time in the timeline. In order to calculate the offset, the editor and I manually synced a crest in the audio waveform in the vicinity (a snare shot if I recall well) on the two sources, ‘scrubbing’2 back and forth over it with fixity, adding/subtracting frames to locate precisely, the 3 second 12 frame differential. At that time.
Drift is here a loss of sync. The clocks drifted apart, or more precisely, the ‘master clock’ was not able to consistently enslave its subordinates. And the drift itself was a variable, not a constant. Its value changed over time, bringing up the rather curious idea of a drifting apart of time over time. Multiple clocks, meant to be synchronously tethered to one another instead achieve a sort of frisson, a momentary excitement or perturbation, unannounced and governed perhaps by an incalculable whimsy as they break away from their moorings, leading then to schismatic clocks and parallel, fractured times.
Pondering on the phantasmagoric nature of these interlinked drifts, of time and perception3, leads me to consider paths crossing, or crossed paths, missed opportunities, and ultimately, an eschatological stoppage – a grinding halt of time. Auden’s famous melancholic plea to stop all the clocks is widely quoted with appropriate solemnity but I recall here his words from As I Walked Out One Evening:
But all the clocks in the city
Began to whirr and chime:
O let not Time deceive you,
You cannot conquer Time.
The sensational 2005 case of the disappearance of American teenager Natalee Holloway in Aruba, involving the Dutch youngster Joran Van der Sloot and his two step-brothers was widely reported, with predictable fanfare and salaciousness. The confessions, sting operations, trial, interviews that followed was accompanied and amplified by blogosphere chatter. Following his arrest and indictment for the murder of Stephany Flores Ramirez in 2010, the interest in Van der Sloot and the earlier case was reignited with much fervour. One peculiar ‘ambiguity’ was pointed out by several sources and thoroughly discussed. There was a discrepancy in the time stamps of the security cameras of the Peruvian hotel where Van der Sloot was staying and had brought the girl to. The hotel receptionist found the murdered girl in Van der Sloot’s room on May 30 2010, which was ironically, the fifth anniversary of the disappearance of Natalee Holloway.
The time stamps show Van der Sloot leaving his room at 08:56 AM and passing by the lobby on his way out at 08:45 AM. Several bloggers queried how such a reversal of time was possible, posting both the video that was made public and the still images of his entry into the hotel, his room, and then his exit.
zed wrote: Can anyone make sense of this? The camera time stamps at 5:20 go forward ... lobby time 5:20, entering room 5:33, but time stamps 3 hours later go backwards ... leaving room 8:56, lobby time 8:45.
Barfdoll Posted: Sun Jun 06, 2010 9:05 pm
hallway camera timecode reads: 08:58:40 when his door opens on his "departure"
Front desk camera timecode has an additional field for the "frame" (30 frames a second for "normal video), and so it reads: 08:45:06:18 …Ive wondered aloud here a few times whether or not the camera systems timecodes are properly synced or if there is a 10+ minute difference (the hallway camera is possibly set to a later time).
In this case there appear to be conflicting clocks. Ambiguity is a technical term used in relation to timecode issues in GPS systems (and telecommunications). Timecode generators and receivers can be in conflict, there may be issues with variable travel times of the signal, and any computation is done with a degree of ambiguity. Here I use the term more generally, with literary meaning, alluding to a conflict between multiple clocks. In Van der Sloot’s instance it seems to be the case of two different timecode generators stamping data according to their internal clocks. The BITC (or Burnt In Time Code) that one sees on screen is linked directly to the generator, but there is also code written into the visual signal (video watermarks, Vertical Interval Time Code or VITC) aside from timecode sent across a wire. This ambiguity is then related to the absence of a ‘master clock’ calibrated precisely according to a set of prerequisite parameters (Time Zone, temparature, etc) that generates data for multiple receivers. Now there are digital generators but previously analog devices had to be ‘warmed up’ an hour or so for them to achieve a steady, optimum functioning. The oscillators needed a head start else they were ‘poorly calibrated clocks’ - which could well be the title of a La Monte Young composition.
The American standard definition broadcasting system NTSC is mockingly expanded as ‘Never The Same Colour’, since it is seen by many to be inherently flawed and inferior to the PAL system used by most other nations. With lesser lines of resolution and a different frame rate, there is some truth to the system’s deficiencies (HD has brought many changes). There is a peculiar difference between American timecode clocks and clocks of other nations that contributes to continental contempt (amongst other factors) for what is perceived as a misguided sense of competitive nationalism (in their adherence to an inferior standard). The utility frequency of electrical supply in America is 60 HZ as opposed to 50 HZ in most other nations. This determines frame rates for images – whereas in the PAL system each second of video is composed of 25 frames (or 50 fields) and makes for an easy correspondence in technical translation, the NTSC frame rate, meant to be 30 frames (or 60 fields), is actually only 29.97. This had presented a whole range of curious problems and is linked to the development of a uniquely idiosyncratic system, a way to work around the history of compromised convention – Drop Frame Timecode. Since the 29.97 frame rate is not a precise integer, the translation into a timecode clock introduced a lag and clocks would, over time, at the end of a working day for instance, lose ‘time’. This glitch was addressed “by regularly bumping forward the value of the frame count to make the slow-running clock catch up with an actual clock”.4 Dropping frames meant dropping the counting of certain frames (frames 1 and 2 are ‘dropped’ every 66 2/3 seconds) so that conflicting clocks achieve synchronous states.
The idea of clocks in conflict speaking varying times, lagging behind and then tricked/urged into thinking they are in step, suggests to me inherent deformities. In the evolution of technological standards (the Edison-Tesla wars are well chronicled), the choices of those who seek to impose a world standard, one whose universal patent assures powerful dominion, are based on wresting control. Expansionism and neo-imperialist processes are predicated on such broad ideologies of control - from pharmaceutical products, food products to fast moving consumer goods, we see how global control and monopolistic practices are conducted. This brings to mind the idea of conflicting ideological clocks – some seeking to gain control whilst ticking away furiously, producing globally mobile goods and capital that follow grand rhythms of transport and distribution, while others resist being overwhelmed by ‘alien time’ which challenges autonomy and threatens subservience.
From the movement of capital as each financial market opens and closes, to migratory flights, solar/lunar cycles, the blooming of flowers, the monsoon rains, peak time traffic, airports, power looms to nuclear plants, how many clocks are at conflict? And our own circadian rhythms, psychological cadences – what is to be done when these internal clocks, or kala-chakras, are in collision with one another and with external clocks?
I turn here to P.D. Ouspensky in Tertium Organum:
The idea of motion of any kind, also the idea of absence of motion, is indissolubly bound up with the idea of time. Any motion or absence of motion proceeds in time and cannot proceed out of time. Consequently, before speaking of what motion is, we must answer the question, what is time?
Time is the most formidable and difficult problem which confronts humanity.
III. Timecode Break/Loss of Control Track/Fluency disorder
I was faced with breaks in continuous timecode and Control Track while editing Shemekia Copeland’s performance. Such discontinuities result in momentary loss of signal, poor signal, bad transportation of information, etc. Control Track breaks cause gaps in the audio and video information due to a loss of tracking. Poor magnetic material on the tape can also cause ‘drop-outs’ and loss of Control Track (LOCT), which cannot be corrected on the source. The material can be duplicated onto another source with continuous timecode and Control Track, but that results in a ‘generation loss’.
These ‘glitches’ are not just of discontinuity or disruption but instead suggest disfigurement (and scarification when intended) and fluency disorders (discussed in speech pathology). The ‘clutters’ and ‘stutters’ adversely affect the loquacious tranport and retrieval of code and information. In this case, as the editor and I tried to work around the loss of information, the interruptions (and as I began to formulate this essay in my head), I wondered inwardly if we were not faced with stuttering clocks.
Interestingly, these visual/aural scars, cicatrices and defects have become part of popular idiom (especially in visual and music culture). In the realm of music television, with experimental shorts, idents and music videos, a whole range of aesthetics developed in the early to mid 90’s that embraced and innovatively incorporated the many flaws generated by machines used in video/film production. The flawed/reversed/incorrect transport of a Telecine machine (which transferred optical film to videotape) generated a whole set of curious images. The random (or perverted) manipulation of chrominance/luminance and RGB factors of video signals also produced interesting results. The opening sequence of Mike Figgis’ 1999 experimental film Timecode incorporates out-of-phase RGB visuals alongside images of timecode counters, VU meters and other electronic (both analog and digital) paraphernalia. Erratic visual rhythms, interrupted/fractured visuals, leaking-shifting-phasing colours and forms, iterative sounds (such as stuck CD’s etc) began to be used very intriguingly. The work of the British duo Hexstatic (with Coldcut) has to be mentioned here, particularly Deadly Media, which plays around brilliantly with synchronization, syncopation and perverted volubility. I also recall here the highly sophisticated avant-garde work of Hoon: ESU, Yoshihiro Hanno and Ryuichi Sakamoto. Art of Noise, Kraftwerk are relevant here, as is the late Stockhausen. These are but a few of the many artists working conceptually.
Sam Zaman, frontman of the iconic UK band State of Bengal, while discussing this with me points to occasional problems with MIDI timecode, and systems such as Qbase freezing up (for several reasons). He recalls using ‘phantom’ sounds that were not meant to be there but played back due to signal issues, routing, etc. Sounds played out of time, wrong sounds played out and several such ‘fuck-ups’ made their way back into the mix.
It is this idea that I will conclude with. How then do we internalize our torments? What do we do with our grievances, anguishes and vicissitudes? What are the conceptual dimensions to pain? I’d like to think there are many creative ways to reign in demons, tame our torments and in doing so we perhaps seek/find pathways to newer intellectual and artistic realms. Regardless of whether we are (mis)guided by drifting, conflicted or stuttering clocks or by well-behaved ones.
1 For more detailed reading look at Ratcliff J, Timecode:A User’s Guide, Focal Press (1999)
2 Scrubbing is the movement of the cursor/joystick/controller over an audio waveform. The sounds produced through such an action sort of mimics vinyl scratching. Virtual turntables now use timecoded vinyl plates to replicate scratchback and other actions that are then encoded as processes onto digitized music on the computer and the effected (‘wet’) sound is sent out to the monitors.
3 French theorist Paul Virilio has written extensively on technology, perception and visual culture and he argues that telecommunications are ushering in the "invention of a perspective of real time" which results in "some kind of choking of the senses, a loss of control over reason".
Posted by Gautam Pemmaraju at 12:35 AM | Permalink