Dr. Future on 19 Nov 2000 20:55:39 -0000


[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

<nettime> Programming with a Paintbrush: The Last Interactive Workstation


Programming with a Paintbrush:
A Study in the Production Culture of the Moving Image

July, 1999

(An edited version of this essay appears in the next issue of Filmwaves
no. 12 This is the very long version...).


Since the beginning of the eighties the British company Quantel Ltd has
managed to maintain a reputation that has made it almost synonymous with
the limits of what is possible in digital post-production and broadcast
special effects work. More remarkable is the fact that this success is
based on an interface design that was introduced in 1981 and has remained
largely unchanged to the present day. The intention of the company to
provide a completely dedicated computer hardware that can provide instant
feedback and unequalled image processing speeds has resulted in a
completely different experience of the creative process for the user, and
one that is about to end in several important respects.

In the hierarchy of post-production equipment, Quantel machines are like
the Rolls Royce, their nearest rivals being Discrete Logic's Silicon
Graphics based Fire and Flame family. They are the machines that led to
the fashion in the eighties of designers carrying the distinctive Quantel
pens around in their top pockets when they went out to the local wine
bars, like a little calling card that could guarantee the respect due to
someone working at the dizzy high-end of mission critical advertising
schedules and high profile pop promos. These are the people who are
selected by facilities houses not only for their technical proficiency and
creative flair but also for their client manner, their ability to shoulder
the anxieties of pressured art directors and satisfy indecisive
advertising executives. A top Quantel operator justifies their telephone
number salary by selling confidence of a particularly rarefied sort. For
when you are working at a level of production which is premised on the
assumption that this is the best that money can buy and that technology
can dare then you are trading on a dream, a dream best characterised by
the Japanese media theorist Asada Akira as the meaning of technology
itself – "When we find something impossible, we do it".

Now we appear to be reaching the culmination of the desktop computer era
when large standalone systems like a Quantel look increasingly
anachronistic. But this is because of a lack of appreciation of the
production culture it was designed for rather than it being outpaced by
technological development. Desktop computer graphics systems have tended
to advance in terms of an increasing number of features which are
interactively controlled through interfaces such as layers and timelines
to assemble material like the fitting together of the pieces of a jigsaw
puzzle. Complicated image processing techniques are available as menu
options with familiar sounding names in order to provide the user with an
"intuitive" interface. This logic has become well known to most artists
and designers involved in digital production, but its premises of a tool
for every style are challenged by the less familiar history of dedicated
hardware based turnkey systems of which the Quantel range is the leading
example. For Quantel operation is both intuitive and technical, its range
of features is small yet its range of application is large, its interface
is highly interactive and gestural yet it requires a highly systematic
degree of working and logical forward planning to complete a job
successfully. The characterisation of these systems need different
criteria than that of being either interactive or script based, having an
intuitive or a logical work flow or in the sheer number of effects
available to the user. We need criteria that also allow a systems
technical design to be put into its cultural and commercial context more
clearly so that the force of its influence can be accounted for. When you
learn to use a computer graphics system you are learning a practice rather
than a toolbox of techniques as such. This practice is encoded into the
technology as input contexts, devices, menus, parameter values, controls
and so on. The practice in turn supports and expresses itself through a
production culture - its working methods, commercial imperatives and
aesthetic values. How these different levels relate to each other, whether
they are really separate at all and how a computer graphics system can
form its own aesthetic even when the technical potential of different
systems are basically the same are the questions we will explore here.

A single Quantel system can set you back anything between one hundred
thousand and half a million pounds, so it should come as no surprise that
they are not as common as "lower end" graphics systems. But this only adds
to their mystique of course. Normally housed in a large plush studio with
well dressed staff, comfy sofas and blue mineral water dispensers, every
effort is made to produce an environment in which you could believe that
the consummate achievements of human culture are being created. But
costing an average of 600 GBP an hour plus extras, a Quantel based
production still seems a distant possibility for the vast majority of
directors apart from minor work on the briefest effects shots. For many
years during the late eighties and early nineties the desire among many
independent film and video makers and animators to gain time on a Quantel
machine became a palpable envy leading to widespread embitteredness and
inferiority complexes.

This inevitable inaccessibility of the machine itself is paradoxically
combined with an extremely high visibility of its end product. The
majority of TV advertisements and nearly all of the ones relying on highly
stylised visuals have extensive work done on them by one or more of the
Quantel systems. Added to this we have graphic animation sequences on pop
promos, on channel idents and stings, also documentary and news graphics,
title sequences of all kinds as well as the corporate video sector. If we
remember that many stills images for advertising billboards, magazines and
posters are produced on the Paintbox system then we can start to see that
Quantel equipment is responsible for much of the landscape of visual media
we take for granted. As their publicity states, "Quantel is already
inescapably part of your life". Although the status of the company as a
cultural flagship in commercial image making, its eclectic origins and
historic battles with its rivals that at one point threatened the future
of half of the computer graphics industry itself would form an instructive
lesson in the cultural politics of media technology, my intention here is
just to focus on the technological basis of Quantel's success. For it is
possible even in the implementation of the basic processing functions and
the interface design of its equipment to find many of the characteristics
necessary to understand the practice it is intended for and its wider
impact on moving image culture. Its innovations in interactive interface
technology have had a particular influence on a very small but very
significant production sector, and its creative processes and aesthetic
values have in turn set cultural standards in commercial video production
for nearly two decades.

Although Quantel currently manufacture a range of systems from the newest
Editbox designed for online editing, the Henry for special effects and Hal
for videographic design, their basic functionality remains largely the
same. Their main differences are in certain features that change the
orientation of their production environments, but if pushed there are ways
of getting the same results whichever system you use, and such are the
similarities that after having trained up on one of them it is a
relatively painless matter to learn one of the others. In fact the most
noticeable thing about a hardware based Quantel system is that compared to
a software based system like Adobe's After Effects or Softimage's Eddie
the range of available functions and features is remarkably small.

Since the mid nineties developments in increased processor speed and disk
capacity have created the field of desktop video production and non linear
editing as a low end alternative to specialised workstations. Although
these systems cannot normally equal the fast rendering times, video
quality or integration and configurability of the high end equipment, the
software that they run can incorporate many professional features and
effects. A very common package used at this level of the industry is Adobe
After Effects which now comes with a vast array of image filters and
processes, keyframe animation facilities and compositing options for
special effects and post production. Given that a software based system
like After Effects can be up to one hundred times cheaper than a Quantel
system, it would seem reasonable to ask just what it is that people are
getting for their money. An advantage of software based systems is that it
is relatively easy to write new code to expand the package to include new
features, often sold as plug-in extras. On a hardware based system you
have to write the code and also design and build new logic circuits and
incorporate them into the existing architecture. Users of desktop based
systems are quick to point out that software based packages have an
extensive range of functions that far outstrip those available on a
Quantel machine. The most obvious difference between the two classes of
systems is speed – a hardware based system will always be faster than a
software based one, and building those specialised processors and circuit
boards dedicated to image processing is not cheap. But this difference can
be deceptive. We can argue that when we are rendering a special effects
sequence it can make little difference whether it takes two hours or
twenty hours – you will probably still have to leave it to render
overnight while you go home to bed. Even if this means tying up a machine
for a whole day while you wait for it to finish, PC and Mac computers are
so cheap nowadays that you could go out and buy another one to work on
with the petty cash. So if a half million pound Quantel Henry won't do as
much as a one grand copy of After Effects then what is all the fuss about?

If you tell a Quantel employee that your company does all its high quality
advertising campaigns on a Mac and a copy of After Effects then they will
ask you how many of those were "top end jobs". The implication is that at
the top level of video production there exist a somewhat different range
of priorities and standards. These take several different forms.
Reliability is cited as an important feature of production equipment, but
in this respect Quantel machines are really as vulnerable as any other.
Quality is also frequently mentioned, usually interpreted as being the
ability to work with uncompressed video and in the accuracy of various
image processing operations. But there are many software based systems now
offering high quality rendering and perfect image quality is not always
necessary in every production sector.

Although speed is a factor in finishing a project on time, in a top end
production environment it operates in a different context. Much of the
work at facilities houses takes place while the client is present, so the
speed at which the designer can respond to a director's instructions is a
primary issue. Sometimes the brief involves perhaps animating to a
storyboard and would involve designers working on their own on the
different elements ready to be presented and approved at a later date. But
the job is still broken down in separate parts as far as possible to allow
more flexibility to respond to changes later. This is a main advantage of
digital production, that a project is fragmented into different parts that
can be disassembled and reassembled over and over again without
sacrificing its structural integrity. And Quantel systems allow these
different parts to be swapped around, processed, re-coloured and re-timed
almost on the fly and in the presence of the client. A desktop based After
Effects production could not provide the rendering speed and instant
feedback necessary to keep up with a client who expects to be able to
direct the work as it happens. Every moment that the production is taking
place away from the direct gaze and supervision of the director is a
moment of anxious anticipation for him or her. Quantel equipment gives the
director more of a feeling that the programme is being put together under
their direct control, allowing constant dialogue with the operator and
receiving instant visual feedback. Therefore, at least while the client is
on site, an ideal of real-time production is always desired.

Once all the individual shots, effects footage, graphics elements and
animation layers have been prepared they are all brought together in the
presence of the client in a grandiose final creative act with the help of
a high end effects and editing system, typically a Quantel. The working
methods of facilities and production houses can be characterised then by a
de-centralised production process with a centralised final online event.
The machine's speed of rendering and of design and editing is therefore
crucial in maintaining this standard of client interface and the illusion
that the most pregnant creative moment takes place in front of the
assembled eyes of director, advertising agent and corporate client as they
lounge around on deep sofas and sip freshly squeezed orange juice. It is
this kind of environment and approach to creativity in a commercial
context that Quantel equipment has helped to create and maintain. The
simulation of creative control. The way that Quantel have developed its
technology to cope with these demands have lead it to a particular
approach to the implementation of the creative process, the aesthetics of
the interface and implications for the whole relationship between human
and computer.

The Quantel interface for video editing is not timeline based but presents
the footage as strips of video on a cutting reel or piles of frames on a
desktop. Details differ between each machine but editing is achieved by
cutting shots with a stroke of the pen or stylus, picking them up and
sticking them end to end. Layering is achieved by picking shots up and
placing them directly over other shots. The first impression a beginner
gets from using the machine is the degree to which the interface relies on
the physical gestures of the use of the pen. Depending on the context and
the particular way you move, the pen can pick up, drop, drag, swipe, cut,
stroke, paint, jog and shuttle. The tablet is also very sensitive which
means that you can use pen pressure to input numerical values into the
computer with a reasonable degree of exactitude. In fact, nearly every
parameter, numerical or character based, can be input into the system
using the stylus rather than typing them in using the keyboard. Many
software based systems offer this option as well by using slider displays
and dial icons, but in a Quantel system the ergonomics are quite
different. Instead of keyboard typing or adjusting graphical indicators
you are literally "painting in" the values by stroking the pen up and down
the tablet and watching the image change in more or less real time.
Combined with holding a somewhat less sexy handset of buttons or "rat" in
your other hand, this arrangement makes for an unusually high degree of
physical involvement with the machine which in turn reinforces the other
"hand made" qualities of its usage.

The second thing a beginner notices, especially one which already has
experience of other graphics systems, is that a Quantel system is
relatively small. It only takes about a week or two to learn nearly all
the menus and functions on a machine. A typical Quantel workstation like
the Editbox only has about four top level menus plus the main "desk". Each
of these menus will have about a dozen commands in them with very little
use made of further levels of hierarchically organised sub menus. You have
a Viewer for adjusting edit points and transitions, a Mix/Effects for
compositing, transformation based animation and applying filters, a Track
menu for doing motion tracking. There's not much else. Many of the
functions are repeated in several different menus for convenience. It is a
small "toolbox" of basic editing and image processing functions and you
are left to work out what to do with them.

The reason for this sparseness is mainly to do with the technological
issues of hardware based systems. It is expensive to build processors to
perform specific graphics tasks, so there are limits as to the number of
operations that can be hardware accelerated at a particular stage in the
development of the technology. Quantel systems use completely uncompressed
digital video so the amounts of data to be processed are huge (and vastly
more so for a film resolution system like Domino). Therefore the strategy
is to provide a small number of fast and very high quality functions which
can hopefully be combined together to achieve more complex effects. It is
a bit like being handed a stick of charcoal and a sheet of cartridge
paper. You can sharpen the charcoal and draw with a point, shade with it
on its side, rub the dust with your finger, sprinkle water over it, rub it
out and so on – it is up to you. Within these limitations an enormous
amount can be achieved in terms of effects design. Although it only takes
a week to learn the system it can take about six months to become a
skilled user. In keeping with the ideal of working from first principles
Quantel systems have displayed an unusually high degree of constancy over
the years. Seemingly impervious to the pressures that oblige other
software companies to continually update and redesign their interfaces,
Quantel menu design has remained largely unchanged since their beginning
in the early 1980s. It is as though the Quantel system is based on
functions so fundamental that a change in their design is inconceivable,
as unnecessary as redesigning the shape of a pencil. Once Quantel has
solved a problem it stays solved.

Let's look at some of the implications of this interface for the user and
how it feeds through into the final product. Because so much emphasis is
placed on the Quantel designer developing their own working methods (there
are training tapes provided to give you a start with this by the way),
each user quickly starts building up their own way of doing things based
on the simple basic operations. As an example we might consider the common
process of "keying", whereby a top layer of video is composited onto a
background by making part of it transparent. Often this is performed with
a live action shot where the subject is against a blue background. The
blue background is removed by "pulling a key" to produce a "matte" so that
the background shows through. In a software based system like After
Effects this is done by simply applying a Keying effect to the video layer
and setting four numerical parameters - defining the blue colour and its
tolerance, the edge softness and edge width. In a typical Quantel system
however this task takes on the character of conducting an orchestra. First
of all the method of defining the blue keying colour is quite different.
It is performing by tapping down on the blue of the video image directly
until most is gone, combined at the same time with adjustments of other
parameters like Softness and Value. The experience is that you are burning
a hole in the video image as you race around the blue areas with your pen.
Very satisfying. After you have pulled this "main key" you can apply a
"spill key" to help remove reflected light but usually you need to adjust
the area of the matte a bit further. A very frequent problem is "matte
lines" around the edge of the matte where it has not quite excluded all
the blue background from the edge of the subject. In After Effects this is
usually dealt with simply by adjusting the edge width parameter to
"contract" the matte. In Quantel it is the beginning of a whole art form.

In the training videos Quantel discuss no less than three ways to contract
a matte and suggest a fourth. It was not long before I discovered a fifth
way which became my personal preferred method. They all involve many
different separate operations of rendering out, blurring, tonal graduation
and re-combining footage. Most of the time it is difficult to discern an
objective difference in quality between these methods but they quickly
become identified with the users "style" of working. It must be noted that
After Effects also provides the same functions which could be used to
mimic the Quantel processes of keying, but in practice an After Effects
user would never consider working in this way. In After Effects you use
the Key effects to do keying and that's that. For many keying jobs the
difference in quality between Quantel functions and After Effects effects
filters are negligible although for more complicated and difficult keys
the Quantel practice of applying simple functions bit by bit will afford
you more control. But as just mentioned, you could use After Effects in
this way as well, which leads us to an important point. A big difference
between hardware systems like Quantel and software based After Effects is
not in their technical quality nor rendering speed as such but the working
methods that the system as a whole gives rise to. These working methods
contribute to completely different artistic practice as well as defining
how well they integrate into each level of commercial production
practices.

The increased number of separate operations that need to be applied to
pull a good key in Quantel would seem to imply that the system is slower
to operate than the all-in-one approach of something like After Effects.
There are at least two reasons why this is usually not so. Firstly the
decreased amount of control available in adjusting the parameters of an
After Effects Key effect often means that much more fiddling around is
needed when dealing with all but the simplest of keys. But more
importantly we must remember the impact of the Quantel interface on the
users operating abilities. Apart from the fact that a pen interface is
faster to use than having to roll a mouse all around a screen, the higher
degree of integration of the users manual dexterity into the Quantel
system means that the speed of editing is limited more by their physical
operating skills than by the response times and rendering power of the
machine. This means that the more dextrous the user becomes the faster the
whole process goes. There are other factors related to this as well.
Because the process has been broken down into small separate operations it
is frequently possible to find new combinations of these which can
accomplish the task a little quicker. And as experience increases it is
easier in a Quantel system to fine tune the necessary number of operations
to the particular demands of each individual editing task, skipping some
steps when not significant and spending longer on others which will have a
greater impact on the result. The more you use a Quantel the better and
faster it gets. The industry encourages a professional pride in being the
fastest Quantel operator, as though they are racing drivers fired up on
the adrenaline of post production. When a computer system offers this
degree of physical involvement in its interface and is designed to be used
in a bottom-up approach to building complex effects out of simple
operations, then much of the uniformity, reproducibility and expediency
commonly associated with computerised design systems disappears. However,
it would not be true to think that this implies that Quantel systems are
perfectly transparent and neutral technologies that exert no influence of
their own on the user as we shall see a little later.

As mentioned before, practically all of the parameter input into a Quantel
machine takes place through the pen and tablet interface. This practice of
stroking the pen across the screen until the right value has been reached
keeps the creative process at the visual and intuitive level rather than
that of comparing and checking the numerical values of different
functions, although that too is certainly still an option. Sometimes this
means that you cannot remember how you achieved a certain effect, and
usually ensures that you never get quite the same results every time. The
interactive interface is not entirely informal either – it is a digital
system and the entered values and pen moves can be recorded and displayed.
But the extent to which these systems rely on manually editing and
repainting the image is quite great. Instead of applying extra filters to
adjust the colour of a sky for instance, a Quantel operator will
frequently and more speedily just paint over it in a few brushstrokes. If
you discover that your matte is not quite accurate enough over a couple of
dozen frames then it is often easier just to paint over the mistakes frame
by frame rather than to redo it. It often comes as some surprise to
students to learn that the glamorous special effects that they have seen
on TV or in films have been achieved only partly through complicated
mathematical processes and sophisticated animation systems. Much of the
detailed work and blending of elements together necessary to attain a
completely convincing effect is through the painstaking retouching and
adjustments performed frame by frame by hand. It is only on the fastest
and most sensitive graphics systems that this becomes a practical
possibility. There are often repeated cycles of manual painting and
manipulation of footage until the particular desired effect is achieved.
It is because the video footage in Quantel is completely uncompressed that
it is possible to continually apply more and more adjustments in
individual stages without the quality of the image deteriorating through
repeated "cascading" recompression. This prominence of hand work should
remind us that computerisation is not just about processing more things
automatically under software control, it is also about integrating and
optimising manual skills like painting as part of the production process.
This in turn can have the effect of reintroducing qualities like
authorship and originality above operator skills, returning us to
traditional values that we might have thought were becoming less and less
relevant.

The tactile qualities of the interface create the experience for the user
that they are personally kneading the footage into shape like a piece of
dough, prodding at some bits with the pen and then stroking and rubbing at
others. The operation becomes very physically engaging and finally
addictive. There is a strong similarity with becoming absorbed into the
space of a computer simulation game. The experience is further reinforced
by the pleasure involved in solving a technical problem like pulling the
perfect key. The degree of control offered by the sensitivity of the
interface makes the pursuit of technical perfection all the more exacting
and intense. This in turn combines with other factors to produce a
particular kind of aesthetic in the final product which testifies to the
particularity and the idiomatics possible through the subtlety of
interface. It perpetuates a notion of quality based on precision of
control and evidenced through a smooth seamless finish.

There is actually a Quantel "look" that is instantly recognisable to an
insider, despite the scope for individual stylistic signature that the
interface makes possible (or perhaps because of it). This is partly due to
the early design of the first Paintbox. The research team at Quantel
wanted to work with a "traditional" artist to help design and test the
interface and chose to work with an illustrator called Martin Holbrook who
commonly worked on posters and record sleeves, of aeroplanes and fantasy
landscapes. In response to his input they designed a system that used a
very responsive stylus to put down very smooth tonally graduated
brushstrokes and airbrush marks. This has remained in subsequent systems
both in the inclusion of the original Paintbox functions and developed
into the widespread availability of blurring and softening tools. Most of
the main menus and functions now have their own softening or smoothing
filters which allow disparate pictorial elements to be seamlessly
combined. It is quite surprising how many "mistakes" can be magicked away
by the application of a few strokes of the airbrush, how many incongruous
scenic additions can be made to blend in with a soft edge matte. These
abilities also make it very good at producing various glow effects which
are so frequently used to make adverts and music promos more glamorous,
romantic or just softer and less harsh.  It is an aesthetic of continuity
and naturalness built into the very heart of the system which concurs
perfectly with a corporate design style that has remained dominant until
the present day. This is not to claim that Quantel is responsible for a
whole cultural milieu, but it is an aesthetic which gives form and
direction to the tendency of the interactive interface to emphasise the
particularities of individualistic treatments [1]. For these reasons the
freedom of individual treatment that the technology allows can lead to a
preoccupation with more precise control rather than to artistic diversity.

Quantel systems typically render down each adjustment into to single clip
making it very difficult to "unpick" and separate the various processes
that have been applied. You cannot analyse how the image has been
constructed unless you exhaustively save and keep track of every
intermediate stage, which is not normally practical. There is no "undo"
function on a Quantel system. Each operation you perform requires a
similar kind of commitment as does physical media. Saving previous
versions will work up to a point but does not retain the dynamic of
reversibility that is almost standard on other software based systems.
Quantel cannot put Humpty Dumpty back together again. This also means that
as you build up your image step by step you have to have a fairly clear
idea of whether you are heading in the right direction. Quantel warn their
operators of getting stuck down "dead ends" – of applying a whole series
of processes to an image and then discovering that you have forgotten an
important factor which forces you to start all over again. There are very
few "live" events in Quantel, which means that you must normally "commit"
one edit or effect until you can proceed to the next. In a software based
system it is almost taken for granted that you can go back and re-edit a
particular parameter of a particular filter applied to a particular layer
of video again and again. In software based packages, even after rendering
any changes that you have made, the material is still organised into its
data structure and can be recorded in some kind of separate project file.
This accepted paradigm of arranging all your visual elements and effects
into a total framework or data structure in which each part is in an
interchangeable relationship with the others is effectively unknown in
Quantel. The hardware has simply not been able to support such
complexities until recently.

But the fact that Quantel systems have never offered the user a overall
layout in which to organise all their visual elements and processes does
not mean that a systematic method of working is not possible. It means
instead that it is the user themselves who must supply the organisation –
the data structure is not in the computer but in the user's head. This
internalisation of the logic necessary to work effectively with a digital
system means that such interactive workstations that seek to reproduce the
dynamics of traditional physical media are the very opposite of the
intuitive and spontaneous pattern of working that we would expect. The
Quantel user must know what they are doing without the help of extended
layer displays, timelines and associated parameter lists. The logic of
digital processing must itself become intuitive to a far greater extent
than with other systems that rely on reference to a visible and constantly
updating data structure.

The most common model for software design is object orientated, based on
the idea/expression dichotomy which aims to externalise an artist's
working methods. The software system seeks to mirror the internal creative
process by organising it into an external data process or structure. The
software is a system of menu commands and options which seeks to match an
internal model of creativity as a process of decision making that seek to
approximate an ideal artistic goal. The more functions and the more
parameters that are made available the more successful this will be, the
more exactly the software system will be able to match the artist's
intentions and give them what they want. But of course most artists and
designers do not really know what they want before they start – the
creative process is actually a process of playing and "visual thinking"
that leads to a variety of "solutions" and modifications of the original
"problem" or brief. If software architects assume a decision based
idea/expression model then the number of options they strive to make
available in a software system will proliferate with the effect that the
decision making process will become overwhelmed and arbitrary. It is as
though the exactitude of the software parameters threaten to exceed the
exactitude of the artists creativity resulting in the whole process
defaulting back to the exploration of a subject orientated smooth space of
the unknown. One example of what this leads to is the growth of plug-in
modules to provide common effects like lens flares, explosions and
rippling. In order for these effects not to become too standardised and
familiar looking they usually come with a huge number of editable
parameters which the artist can use to customise the result. But it is
very common just to give up on the endless experimentation needed to learn
all these options and to accept the default parameters, or else to find
oneself continually "fiddling" with the values until you simply run out of
time. In the arena of digital media operating under the rubric of artistic
vision, individual expression becomes customisation.

The Quantel system and interface by contrast has been more subject
centred. This has been an effect of the many characteristics discussed
above that tend to reject the second guessing of artists working methods
with pre-packaged effects and instead reduce the system to a bare minimum
of functions. This restriction of means is partly dictated by the current
technological development of the hardware and partly by priorities in the
design of the interface. But despite this we can still point to other ways
that a design aesthetic asserted itself through the design of the
technology and the "tools" themselves as well as by the commercial
production environments in which it has been used. Such a simply modelled
system still produced images that were both remarkably complex and
remarkably similar. The sensitivity of the Quantel interactive interface
and bottom-up functionality totally absorbs the user physically and
mentally, and the aesthetic emphasis on seamlessness gives direction to
this level of user engagement. In this context the artistic model of
individual expression tends to become fetishised in some ways but
privileges the agency of the subject in others.

Graphics systems including Quantel are now becoming increasingly object
centred. Despite their close integration with the top of the commercial
production sector, the economics of a changing production culture has been
turning against them and favouring software or hybrid systems like Flame.
This year (1999) the new Quantel machines like Editbox version 7 and Henry
Infinity include considerable enhancements to their hardware to allow many
more layers and edit decisions to remain "live" and un-rendered before
finally committing. These much more complex data structures are now
technologically feasible in the real-time environment in which Quantel
operates and begin an inevitable shift in emphasis from the subject to the
object in their human computer interface. This is accompanied by new
trends such as Quantel's aggressive marketing of its Editbox range, aimed
at the relatively lower end online editing market, where the demand for
sophisticated high-end effects is far less and Quantel speed can be used
to turn around longer form television programmes. Quantel systems now also
boast an "open architecture" for the first time, meaning that third party
developers can use programming languages like Java to develop new effects
plug-ins. This means that many common effects like glows and complicated
transitions will no longer have to be constructed from the ground up in
Quantel but simply loaded in and "customised". Quantel have themselves
also been developing much more sophisticated functions to perform
previously piecemeal processes in one go, such as their new one-step
keying function which is intended to create perfect keys in one press of
the button. Over the past decade other ways of maintaining the users
agency have emerged, like the growth of scripting languages as part of a
standard software package and the inclusion of Software Developers
Toolkits. These will allow some aspects of Quantel's subject centred
approach to continue albeit in quite different contexts. Each of these
will bring their own particular pressures to bear on visual aesthetics and
on wider cultural issues and their commercial implications.

By a careful analysis of a graphics system it is quite to possible to see
how a manufacturers particular technological development can have an
impact on moving image culture, both through particular aesthetic biases
and through its relation to the values of the commercial environment it
has been designed for – its production culture. Quantel’s intention to
design for a real time production environment lead to a fast hardware
based system but was limited by the current state of the technology to
implementing a range of basic functions. The combination of these simpler
function along with a highly sensitive user interface derived from
illustration resulted in the highly idiomatic creative practices and
aesthetics described here. This analysis is apart from a cultural one that
takes place perhaps at the level of corporate policy and identity,
marketing, labour relations and working practices and the interpretation
or theorisation of the end products as texts. In fact, in order to get
some leverage on the rationale behind the direction in which the
production culture is heading it is essential to penetrate to the level of
technical practice to get a full picture of the forces which are
determining its future. It is not excusable to relegate these forces to
the status of neutral technologies that are exploited as tools for some
higher purpose, nor to present them as ideological technologies that
embody a premeditated agenda within their very structure. Technologies are
now composed of both machines and people which work together in quite
specific and complex ways.

The fact that it is possible for a Quantel operator to use a software
system like After Effects in ways that an After Effects operator would not
have thought of indicates that the technology of media is located in a
practice and that this practice is formed within the context of a
particular production culture. But trying to use a Quantel system like
After Effects would not be so successful. The object centred practice
relies on being able to define design problems in terms of the available
technical functions and in expressing the creative process in an external
data structure. This creative process is pre-empted by the traversal of
menu options and navigation of data bases. We might conclude, therefore,
that all technologically based art practices are not equal. As long as
there is access to a range of basic processing functions, the subject
centred approach is always workable, even outside of its native production
environment. Why is it that After Effects users do not generally know how
to best fine tune their effects work to get higher quality results and
Quantel users do? After Effects users are not stupid, they just develop a
different practice in a trajectory set by the design of their software.

As we leave a generation of interface design and computer graphics systems
behind and fully enter the age of the object centred system, it is
necessary to remember that alternatives once existed. With hindsight it
may appear that the highly interactive graphics workstations and the
highly systematic text based computer programming languages that once
seemed to be diametrically opposed to each other actually had more in
common than was ever imagined.



Notes

[1] There are some other factors that contribute to the Quantel “look”
that are not based directly on the dynamics of the interface. One very
noticeable feature is the preponderance of saturated shadows in a Quantel
treated video. This comes about because the function that darkens and
lightens an image uses a colour space that tries to maintain the same
degree of colour saturation as in the original image. This is presumably
because it is felt that it is more intuitive to alter tone independently
of colour. The effect is to give dark areas a slightly psychedelic look.







#  distributed via <nettime>: no commercial use without permission
#  <nettime> is a moderated mailing list for net criticism,
#  collaborative text filtering and cultural politics of the nets
#  more info: majordomo@bbs.thing.net and "info nettime-l" in the msg body
#  archive: http://www.nettime.org contact: nettime@bbs.thing.net