pouët.net

The art of visual editing, flow, rhythm

category: general [glöplog]
Quote:
Editing in film is not really the same, as it is a longer form and also because in film they have (compared to us) unlimited assets, so tend to work on a higher conceptual level.


Editing in film and television usually has to respect narrative continuity, i.e. its aim is to keep the viewer engaged with the story being told. So yeah, insofar as we aren't in the business of telling stories, there are differences, but I still do think there's a lot of stuff in the tradition of continuity editing that you could apply in the process of making a demo. Matching formal characteristics of the images you're cutting is one obvious thing that comes to mind; for example, in the film world, dialogue sequences shot in close-up are usually 'eyeline matched', that is, the shots are aligned symmetrically so that the actors' eyelines are roughly at the same distance and height so that the characters talking to each other also look like they are looking directly at each other. The underlying principle can be extended much further though - I recall reading that Michael Mann lined up the bars of a prison cell when shooting a dialogue scene between a cop and an inmate for Manhunter.

But yeah, that's just off the top my head. Establishing formal symmetry between shots seems so elementary that it's probably something already developed by filmmakers like Eisenstein and Griffith way back in the silent era. "In The Blink Of An Eye" by Walter Murch is a good practical read regarding film editing in general, and the recent debate in film studies about the state of continuity editing in contemporary audiovisual media is interesting as well. The gist of it is that a lot of films these days do not honor the principles of classical continuity editing, and yet they seemingly succeed in immersing the viewer in their stories, which begs the question of whether or not narrative coherence needs continuity editing in the first place. So I guess if filmmakers can break the rules of audiovisual storytelling, it's not immediately obvious why demosceners should follow them to begin with. But maybe there's a few tricks that can be learnt from a hundred-year-old tradition of sequencing moving pictures together.
added on the 2022-04-19 09:16:55 by aegis aegis
As with everything, there's no right or wrong. Demos aren't movies, but some demos attempt to build a narrative; in which case it pays to adhere to both the tricks of the trade and the traditions of movies. Leading the viewer with expectations/undercutting expectations and flow of motion is important.

But there are also reasons to do none of that; for example when a demo is just about a hypnotic experience. X-Mix sure as fuck would be horrible if it was cut like a Robert Eggers movie (as in; not at all). Cutting on the beat is necessary fairly often in demos that rely on a tight music/visual sync, but do it because it's the right thing to do, not because you don't know of any other way to do it. That's just lazy.
added on the 2022-04-19 10:30:55 by gloom gloom
@aegis and @gloom, I am guessing I am somewhere in between your perspectives. I am aware of some of the debates about film editing, but my point was more influenced by the needs of demos working on a platform with 128K of memory in total, so that one can have at best 10-20 fullscreen images. Hence, the issues like continuity of shots tend to be replaced by matters of the compositional symmetry of what is on the screen and the anyway visuals are relatively static compared to what we now expect from a quality production on a modern platform.

This does not mean that experience of film editing is not relevant, it just exists in a somewhat parallel universe. However, music videos often work more abstractly and with fewer assets, so they solve problems more similar to ours (even if their reasons for this may be different). This is where I disagree with you @gloom, when you say that demos defy the traditional expectations of continuity editing and flow. I don't think they do, they just operate differently. It is just that they focus on a different kind of continuity and a different kind of a flow from what people do in films. Instead of focusing on "getting people to think" to links the scenes on the screen together, one can go and focus on matching/reflecting/supporting tempo, rhythm, intensity of the underlying soundtrack, so the production becomes less of an musically supported picture and more of a synesthetic experience where visuals and music are intentionally equal in importance. Rules of editing in film are closely related with the needs to preserve narratives, but if narrative is abstract the rules for traditional narrative become irrelevant.

This is why I find a lot of "abstract animation" or "absolute film" or whatever else they call it, from mostly the middle of the 20th century, almost the closest genre to what we sometimes call hypnotic experience or abstract story in demo productions. See e.g. An Optical Poem by Oskar Fischinger (1938) or Permutations by John Whitney (1966) or Lines Horizontal by Evelyn Lambart & Norman McLaren (1962).

People played with these ideas a lot, and many years earlier than we did. They just did not have very efficient means of production for this kind of stuff.
added on the 2022-04-19 11:42:08 by introspec introspec
Quote:
This is where I disagree with you @gloom, when you say that demos defy the traditional expectations of continuity editing and flow.

I.. didn't say that? :)
added on the 2022-04-19 13:03:33 by gloom gloom
It was partly a rhetorical exaggeration, but yes, I did get carried away and exaggerated a bit too much there.
added on the 2022-04-19 13:20:34 by introspec introspec
I was thinking of this conversation and remembered that in one of our previous prods, we did something a bit unusual and which I haven't seen much elsewhere. In one of the final scenes of G – Level One, there is a tunnel scene in which the geometry is "synced" with the music beats. As in, the mesh was modelled after the music.

See at 3'54": https://youtu.be/Ex1k2r8C_V4?t=234

The tunnel first has as many pillars as there are beats, then a rapid succession of arches matches the guitar, and finally there are as many pillars on the walls as there are drum beats before leaving the tunnel.

The scene was fun to make and I was happy with the result, but I haven't really had the opportunity to do something similar since.
added on the 2022-04-19 14:07:29 by Zavie Zavie
Great topic, thanks! :)

In my optinion, when we are looking at editing and flow in demos and intros, one of the most important aspects is pacing. Whether it is a flash+boom style heavily synced bassfest
or a narative driven story-demo, the pace of the whole prod and the pacing of single scenes/effects has a really big impact on how you feel watchig a prod. It is also one of the
elements which can be applied to any form of media which shows/tells/shouts something over time and a lot of the aspects which make for great pacing in film/television (and also comics and novels) can also be applied to demos and intros.

A demo can feel boring even if there are fantastic effects to be seen. It can feel sluggish even if there is a banging soundtrack playing and it can feel rushed, even if it spends a lot of time on details.
It's pacing which makes the difference in those cases. It's the decision to show the rotation of this cube another time, change the camera angle, or switch to an entirely new scene.
Pacing defines how fast you scroll through your text, if a scene transition needs a hard cut or a slow fade. It also the descision if a prod needs another five scenes or if it should exit after one minute.

That is applicable to anything from tiny intros on TIC-80 to a short film in UE5. A lot of this is intuitive. At some point a scene becomes booring, so you switch to another or the soundtrack has a slower part, so you turn down the flashing effects. But I think most prods can gain a lot by actively constructing the pacing - over the whole prod and for single scenes.

In film there is usually the disctinction between the "plot pacing", "scene pacing" and "editing pacing". Plot pacing defines how fast you move from one plot-point to another, how many pararell stories are beeing told - how fast you are moving through time. This will be highly different if you are doing a somber drama or an action-packed sci-fi blockbuster (not telling you anything new here...). When working on plot pacing you ask yourself questions like:

- Would the story miss something if this part was missing?
- Does it make sense to move from this point in the story directly to another? Does it need something in between?
- Can I reference another part of the story at this point in time?
- Can I introduce this concept/person/plot-device at this point in time? How does this feel? Overwhelming? Confusing? Thrilling? Somber? Does the feeling change if it would be introduced earlier/later?


One can apply those questions to realtime prods as well. You might not tell a narrative story, but how your content is presented and introduced over time still invokes emotions in the audience. Storyboards are a great way to test things like this - even very simple ones, like cards with text on it describing the content ("now we see a spinning blinking cube", "the majestic chromesphere stands over the monochrome checkerboard like a emperor watching over his kingdom"). Creating such cards for potential scenes in you prod, arranging them, shuffling them around, etc, can be a fun way to play around with the structure of a prod and can help to descide on the overal pacing of the whole prod.

Just like the whole prod, single scenes always have a start, a middle and an end. The "scene pacing" defines how one moves through this scene and how it is perceived. In film, a scene usually "feels" good, if it has a clear purpose in the story, is introduces with an open question, has content which - through action and/or dialoge - explores answers to this question and ends with a conclusion to the question (that can be as simple as "will the cat make the jump?! - oh no there is an explosion! - oh yes! she did it! She survived! - I fucking love this cat!"). To analyse the pacing of a scene, on usually asks questions like:

- Have we learned everything in this scene we need to understand the following ones?
- Do we have enough time in this scene to comprehend what is happening?
- How should one feel at the beginning of this scene and after it ended? Scared? Relieved? Angry? - does the structure of the scene support this?
- How are the characters actions motivated in this scene?


Since scene pacing in film is highly structured and specific, it can be hard to apply the same concepts to demos/intros. But for me the most important aspect is the one about emotional transition during a scene - and this can be applied somewhat to scene prods as well.
While in a traditional story, we might enter a scene in which we are angry at a character, see a revelation of a plot point, and end the scene we feel pitty for that character. Emotional transitions can also be done with purely abstract content. One can use suble effects like color/value-shifts, camera-work as well as the soundtrack to beginn a scene with a chill and relaxed feeling, reveal new elements during the scene and end the scene with a lot of anticipation and hype for the next one. I think the most important question one can ask in regard to scene-pacing in realtime prods is "what purpose does this scene have for the audience? To excite them? To relax them? To wow them?" and build/optimize that scene to deliver that purpose.

And then there is "editing pacing" which is probably the hardest one to grasp, notice and to get right (and probably what most of this thread has been about ;) ). But in my experience, good editing pacing only shines, if plot and scene pacing are solid in the first place.

Editing Pacing simply describes, how the different shot in a scene are laid out. While it is a science in itself to do this right (the links in this thread show this well), I usually ask just the following question to analyse it:

- Is the information presented on the screen shown long enough for me to process everything of relevance in this scene while not beeing shown so long as for my eyes/mind to wander to other places?


This can vary strongly between subjects (personal and content-wise), but if the answer to this question is "no" there will be something off in the editing.

From the point of creating intros/demos, I think the hardest part of getting pacing right is to have the time to play with it before the deadline :). The best way to improve pacing in a prod is usually to try out different versions of cuts, scenes and the prod as a whole. Move things around, remove stuff, make it twice as long, etc. and see what feels different by changing these things. Make three completely different camera-flybys of an object and see which one you like the best - and ask why.

Another thing which can help a lot, is to just make things shorter. Pacing usually suffers because things are shown to long or dragged out for to long. There is a reason for the running gags regarding student films telling a 5min story with 30mins of boring exposition and needless dialogue. Cutting things (and throwing away things) can be really hard (and sometimes hurtfull) especially if one has worked day and night to create it. Of course one wants to show a great effect in all it's glory and to be proud of that achievement. But if the purpose of an effect is to wow the audience and the scene drags on for too long, the editing makes it hard to follow things, it can be very benefitial to just cut it to the bare minimum the audience need to get wowed and to move on from there.

In the end, one of the great things about scene productions, is that they can be anything and thus we can experiment and play until we find something which works and feels good to us. Be it by following established rules and structures, or by disregarding those and doing things in a way which would not make sense in any other context. :)
added on the 2022-04-20 18:15:58 by psykon psykon
Bang on, Psykon!
added on the 2022-04-21 13:04:51 by rp rp
I'm not sure if I can take away much of the discussion. I got really into editing some 10 or so years ago. (My ex started a video editing business, so many of our friends worked in that field). I tried to learn something from books ("In the blink of the eye" was my favorite). But I always had a really hard time applying any of that on to timing demos.

One of the reasons might be that the demo develops in my head while I'm listening to the soundtrack. Without soundtrack, there is no demo. And then the soundtrack sets the pacing and rhythm. Sometimes, I have the opportunity to discuss timing with the musician (something like: 3.5 minutes long, 1min intro, 15s breakdown at 2:45).

But then I'm "forced" to sync to the soundtrack. Not necessarily cutting scenes, but syncing elements on the screen. For me, the effects of a demo are its actors. The soundtrack is their voice.

I just watch a bunch of my older stuff to see if I would "edit" them differently. Although most of the design, typography and effects look dated, I'm unsure what I would change on the timing.

It would an interesting exercise indeed to take the footage of a demo and let several people have a go on editing the material. Similar to a paint-over.

Some demos would probably also benefit from having different people "directing" and "cutting".
added on the 2022-04-23 22:27:32 by pixtur pixtur
Just by accident I stumbled across this again: https://www.youtube.com/watch?v=GF0-wGbRqEs
added on the 2022-04-23 22:36:20 by pixtur pixtur
Quote:
One of the reasons might be that the demo develops in my head while I'm listening to the soundtrack. Without soundtrack, there is no demo.


And that is the reason I'm so astounded that the scores of some ASD-demos were produced AFTER the demo was made. »Lifeforce« for example.

Quote:
It would an interesting exercise indeed to take the footage of a demo and let several people have a go on editing the material. Similar to a paint-over.


Didn't you made a new cut of »The Timeless« by Mercury?
added on the 2022-04-23 23:09:08 by gaspode gaspode
Yes. But I am to timid to mention it here. As far as I can remember cupe didn't like it very much :D

I'm procrastinating by watching some old music videos and just remembered that I did some cut analysis back then to see if they are being cut on the beat. One of the craziest examples for me was All is full of love where the only cut on beat is into the robot sex scene. I remember that I was incredibly irritated to see how well it worked, but me being unable to reproduce it. On the other hand -- Chris Cunningham is a genius.
added on the 2022-04-23 23:35:20 by pixtur pixtur
Honestly, after having seen the Revision compos, I'd much rather focus the scene on scoping and content proportioning - half the demos were 2-3-4 times the length they were able to hold.
added on the 2022-04-24 11:29:00 by Gargaj Gargaj
pixtur, please be as open with that as you want, and share it if you like! the more conversation and exploration the better. Also I appreciate this thread very much, even though I can't contribute much myself. The one thing I can share is that reading "In the Blink of an Eye" by Walter Murch was a super interesting book to me as well.
added on the 2022-04-24 23:03:25 by cupe cupe
ditch the film analogies; doesn't matter whether you're doing a 60min theater dialogue or a 2min aphex twin confusion roller coaster. the underlying principle is the same imho, regardless what you call it. "pacing" is not so bad. "storytelling" is not so bad. how about "drama"? even something that's supposed to be hypnotic, or even highlight a frozen standstill, a mesmerizing force antagonizing the notion of moving forward, could be interpreted in those terms.
i'm too tired and frustrated to add a constructive essay here now (thanks for yours, psykon!), but it's nice seeing a few of you sceners sharing your learnings. i doubt that it's gonna help much.

if you're looking for a way to improve in these aspects of demomaking, it's all out there. study music videos, read books, investigate cinematography, doesn't matter. i get the feeling that those who wanna do so do so indeed, with success. that doesn't make it a shared value of this scene though. so what is this thread about? is this about technique? or is it about values?
added on the 2022-04-25 02:22:07 by jco jco
Let me add a few things now that I'm in a better mood ;)

pixtur already mentioned Chris Cunningham, famed director of memorable music videos. If "All is full of love" is too tame for you, watch Come to Daddy.

Another director whose work has inspired me a lot is Michel Gondry. Unlike Cunningham, in general his works prefer practical effects over digital trickery.
This classic is a masterpiece in audio visual synchronization while not relying on "in your face" beatsync at all: Star Guitar - I still haven't figured out how it works. :D

I don't know who made it, but Eple might have given me the idea for the endless zoom effects in Memorize Your Future.

For no particular reason, let me also share this 4min30 slasher movie: Repentless - It follows the classic formula of "band" cut against "plot" - but it's also loud and has distorted guitars.

This brings me to the thought that many demos appear to follow a classic "electronic dance music" approach, with the music often not serving more purpose than providing a beat grid for the visuals. Some of the most memorable demos of all time deviate from that formula to different degrees. Debris, obviously, but also Second Reality and Desert Dream. Visual Editing, Flow and Rhythm in prods might improve simply by adjusting the development process so both aspects of the demo/intro enable and inspire each other continuously, as opposed to "music" being just functional support of the visuals, or vice versa visuals just supporting the music.

More experimentation and deviation from the established demo formula could make compos more interesting, in particular the PC demo compos, where technology hasn't been the limiting factor for quite a while now.

Go crazy and surprise us, and never stop making demos!
cheers
added on the 2022-04-26 09:57:06 by jco jco
if you checked the directors DVD by Cunningham with his (written) commentary you can also read that most of his music videos (in particularly his famous ones: Come To Daddy, Frozen, etc, apart from All Is Full Of Love) were a 'meh, plan B'-cut due to not having all the intended footage :D

login