Eurorack Video Synthesis

Greetings OD users! I am a long time lurker awaiting my ER301. I am new to eurorack. Currently I am using a 6U system to process video and sequence motion graphics. I have not found much by way of longstanding or collated information in this space and have been wanting to collect my thoughts and present them to other eurorack users. If this is of interest and within the guidelines of the general discussion forum, I would like to post notes from my daily work here in this thread. This comprises use of both eurorack audio and video synthesis tools in effort to produce visual effects that tightly coordinate with audio treatment and composition. At the very least, I am interested in helping other users create their own audio visuals as it has become a fulfilling aspect of my work flow.


Very cool man, definitely interested in what you’re doing. Always loved visuals, if they are done well and it completely changes your experience. Some of the stuff people can do these days is pretty amazing (all the mapping type stuff).

I went to Shambhala music festival a couple of years ago and that completely blew my mind with respect to visuals. By far the best epxerience I have every had when it comes to visuals and music (all this in an old growth BC forest - unreal).

Anyways, I’ll be book marking this thread. Cheers!

Ypsi Kid

1 Like

Welcome and thank you for the introduction. :star_struck:

I know almost nothing about video synthesis, but I have to say, after watching the tutorial video, that Memory Palace is a wonderful creation.

1 Like

I am currently involved in a short film production that relies on a scene shot from the perspective of a surveillance drone. Drone footage will provide the background and a motion graphic HUD will occupy the foreground. The actual design of the foreground is pending and I am beginning the effect development process.

The production of this scene can be visualized in layers, from BG to FG:

  • External Video
  • HUD Layer A, a point of reference for subsequent layers (HUD actual)
  • HUD Layer B, an animated layer that tracks the transformation of layer A (HUD blinkys)
  • HUD Layer C, an animated layer that tracks the subject (in the External Video)

Although the source files and specifics for the HUD are not available, producing iterative proof of concept reduces midproject frustration. I am starting with test reel of potential file type and key method combinations.

(Break to notes from session:)

The Memory Palace and I are still getting acquainted and often times I find keying surprising. Tonight, @creatorlars prompted me to experiment with dither on and off for matte .pngs. I settled on a white image on black via chroma key. Thresholding starts from home row with stills loaded, key set to chroma, and threshold set to single. Width, center, and softness can be tuned together for best results

Following this conclusion, I exported a series of mock HUD layers and external video for playback. During this time I encountered some resistance from the signal chain. This concluded a great fact finding session that will reduce my hours spent thrashing during numerous takes. Out with a single take improv patch with notes.

The board state of the instrument following my half baked work lent itself to transformation of ‘Layer B’ or null layer tracking effects. Without a story this sort of utility is underwhelming, but with direction of audio it begins to breathe. Quadtradure from the Batumi drives a metronome of visual effects on the Memory Palace. The coupling of Prismatic Rays as well as a ramp output combine to inform the logature fill, the scanning horizontal bars, the modulated flicker, and the gradient overlay. It is remarkable to me the amount of hidden utility the Prismatic Rays provide. On their own they lack the ability to create meaningful and varied scene elements. Applied as a stylization tool they are elegant, simple and deep.

1 Like

I saw the prototype of Rick’s Structure at Knobcon '18 and it looked killer. Have you seen this?

Anyway, keeping an eye on this thread as video synthesis is a fascinating unexplored world of eurocrack…


Reapproaching proof of concept from last night and listening to a Drew Neumann compilation from the 90s. I am eager to confirm that a multi pass treatment from last night’s proposed effect will maintain reasonable fidelity of the external video.

(Break to notes from session:)

Committing to analog effects when possible involves planning around the constraints of my system. Within the context of a multipass treatment of external video, my system would like to embed the external video to the null layer first. This allows me to avoid sync complications in lieu of a time based corrector or an external video device that can receive sync. I am thinking about the quality of edges of keyed assets and the artifact introduced during the resampling process. I captured some footage in a few formats and set up an A/B of luma and chroma keying results. In this case, luma keying has the function and aesthetic quality I would like. External video definitely displays considerable artifact and will require further format, codec, and bit depth consideration. That being said surveillance footage has a processed vibe that the resampling artifacts approximate. In short, these hours quickly create a roadmap that improves productivity and awareness of the practical challenges associated with this aspect of the project.

@NeilParfitt yes! The structure is interesting indeed and still in final beta (I think.) The venerable Bill Wiatroski has one and is posting output in concert with Memory Palace in the ‘LZX Video Synth’ faebook.


Do you have any plans of using the ER-301 or any other OD modules for video synthesis? Looking forward to some tips.

I’m new to video synthesis. I just got a Visual Cortex and a Bridge a few weeks ago, and been experimenting with what I got.

The ER-301 works well passing through and generating vertical ramps, with its own oscillator units, as well as with single cycle samples, synced to the Visual Cortex (approx. 60 kHz) using a 0.06 threshold. However, like most (all?) audio modules, it seems unable to reach video frequencies high enough to generate horizontal ramps, or to pass them through. After all, the outputs are limited to 96kHz.

Have you considered the ER-101 for controlling the Memory Palace?

1 Like

Work got the better of me this week! Intermission patch->


Following the concept art style from last night, mocked up another spread of panels and sent a favorite off to the printer.

@JohnWynberg Sorry for the late reply! To be honest my initial interest in the 301 beyond it being a badass instrument was primarily motivated by the story behind it. Currently I work out of a 6U case and am always hard up for real estate. That being said the more I have dove the wiki the more excited I am about carving out a home for the 301 in my rig. (A little context before I address your question more specifically:) I work in multimedia and small scale fabrication. Prior to stumbling into video synthesis I was shackled to my workstation for nearly every creative portion of post processing. I have always had a soft spot for the motion graphic treatments of the late 70s and was always left wanting by my digital approximations of such. I found LZX industries and the rest is romance, kind of. Without an exposure to the concepts of control voltage prior, I have had to spend quite a bit of time educating myself and conceptualizing Eurorack in the context of what I must accomplish with reasonable professionalism for work.

My current best practices involve leveraging the Memory Palace as a subject or scene generator, capable of receiving keyframe type data via CV. The visual cortex serves as a master compositor for this information, and can provide a loop in point for external video, ie. a motion graphic pass. Frames provides a nearly perfect analog of a curves based animation suite. The remaining HP can be budgeted for any number of things that will significantly change the timbre of the instrument so to speak. Currently I use that space for effects and image processing, Shapechanger and PR in combination can cover down on an assortment of graphic novel type particle effects. It’s true, there’s a bandwidth issue in some use cases for eurorack audio in regards to video synthesis, but the deeper I get into Larsen’s modules the more I find the following to be true: Lars has plugged a variety of holes that allow Swiss army type access to video. One must essentially conceptualize their need and then carefully work backward to a system that will best accomplish such. (I mean to say that my initial understanding of a video instrument on a conceptual level was roughly ‘can make sick videos.’ Now that I have spent a couple hundred hours using it, I think of my particular video instrument as a cut scene or short sequence generator that can very quickly mill out alternate looks and colorizations per scene very quickly.

This is getting rambly, but in closing I must add that the audio modules I use for animation leave behind opportunities for sound effects and score composition that not only occurs with the video but also from the video. I am hoping that the integration of an HP efficient sequencer and/or CV generator will be the link between my incoming 301 and the video weapon. I am eager to speak with anyone willing to hear my amateur thoughts on the matter, as I am still young in Eurorack.

A2 16x20", Gloss, ‘Peace in Every Step!’ Printable.pdf

Till next time!
PS. I have an extra test print if any one is into this sort of thing. Free to a good home!

Previous post got me thinking about the arrival of the 301 and I couldn’t help but get me some modular grid time. Here is the system as I see it, the core cards that make the deck happen. I am interested in devoting the remaining 26HP to experimenting with audio modules that will allow me a functional 301-based microsystem for short sound snippets to accompany motion graphic sequences or for sound effects. I have been lurking the microsystems thread for some time, and of course everything looks like the most elegant thing that will definitely solve all of my problems as soon as I buy it etc. Sigh… anyways, in the spirit of limiting my GAS I have this complete and <30min use time BSO set that the budding video synthesist might find interesting. I’m not too terribly keen on cash but maybe someone out there has some unused audio modules that might suit my use case and would be up for trade?

Either way, I would greatly appreciate any consideration Ye Old Lords might have on filling this 26HP hole with ER301 augmenting audio goodness.

Thanks for your lengthy replies @Yumibill
I assume you want portability —that explains the Liquid TV, so as not to been dependant on external monitors. In that’s the case, you might want to consider an audio interface module to go with the ER-301. MakeNoise’s Rosie is my favourite, although with the ER-301 almost anything seems possible, so I wouldn’t be too surprised if someone just plugs their headphones or speakers right from the ER-301.
As for oscillators, one of my favourites (and in my opinion most versatile) is Intellijel’s Shapeshifter. It’s exactly 26HP, and syncs well with video, though it can only generate synced vertical ramps once the sync from Visual Cortex has been multiplied x5 through Bridge. However, the ER-301 can perform like a very good oscillator. The available oscillator units are limited, but there is still a lot you can do with them… plus single cycle waveforms through a player unit. So more than audio modules, which the ER-301 can emulate well, I would rather suggest getting more CV generators.

Here is a screenshot of my current 12+3U system that I’m using for audio and video. (Braids and Clouds are last night additions from another case, to see how they perform in this one).

Looking forward to see how your case develops after the arrival of your ER-301.

1 Like

I have always thought Control Forge might fit for a devastating bottom row of bigguns. (Mp+301+CF :fist_left:)

1 Like

Got some news today that my employment situation might be changing. Which is not here nor there but at least I can put some more effort into my personal projects. I have been thinking a lot lately about the formative influences from my childhood and how they’ve contributed to my perspective and have wanted to give back in some way. I am a big into games and was thinking about maybe putting together some sculpts and a rule set for a little tabletop shindig.

Anyways, this thread has slowly become not only about video synthesis but about creative process and utilizing tools such as my Eurorack instrument within the context of industrial throughput. In the spirit of such, here are some thoughts I have been condensing prior to playtesting this concept.

Miniatures are standardized clear acrylic 1/35 scale models on a modeling base. Unit
type can be determined by visual characteristics of each model. Unit altitude can be
visualized by the height from which the model is suspended over its base. The bottom of
the base has three slots on its upward face, and two slots on its downward face. Prior to
play, each player identifies which models will represent his units by affixing a bit to the
underside of three model bases. Six total models are placed on the gameboard by each player.
Three of which represent playable, minimum stat proxy pieces. The other three represent fully
equipped components and are the crux of the game. These models are deployed on the
gameboard without the opposing player’s knowledge of which models will represent the players
controllable units. A single objective marker is affixed to the unit carrying the objective. Slots on the

upward face are used to tally damage for multiple HP units. At the beginning of play, two
die are rolled to determine the grid coordinates of the win condition. Each turn, a drift die
is rolled to reposition the win condition. Each player will take a turn as the attacking and
defending player attempt to deliver the objective to the win condition location.

3 Unit Types:
Industrial, ‘Objective Transport Platform’
3HP, 1DEF, 1ATK, 1SPD, 1ALT, Zero Cost Carry
(5pLV) [pLV = A balanced accumulation of stats and skills]

  • Builds:
  • ’Cargo’: Model is placed in Cargo box to indicate a +1HP shield. While Cargo is
    active, Surveillance drones cannot identify unit or objectives within unit. Successful
    attacks apply to Cargo before HP. Prior to unit deployment, Cargo may be exchanged for
    a card.

Service ‘Multipurpose Manned Utility Platform’
2HP, 1 DEF, 1RATKx2, 2SPD, 2ALT

  • Builds:
  • ’Attack-Carry’: +1HP, 1ATKx2, Can -1SPD to Carry.
  • ’Attack-Identify’: -1HP, -1RATKx1, 2SPD, 2ALT, Can Identify.

Surveillance ‘Monitoring and Targeting Platform’
1HP, 2DEF, 1RATKx1, 3SPD, 3ALT, Zero Cost Identify

  • Model may cast Identify at the beginning of it’s movement phase.

Combat System:
Player announces target, defender rolls based on DEF to establish to hit requirement.
Attacker must roll to hit at x above hit requirement. Attacker may reroll once for every
point of ALT above defender. If attacker fails, equal ALT attackers or higher may roll to
counter at the same hit requirement.

For instance. Player 1’s surveillance drone attacks P2’s industrial drone. P2 rolls a 12.
P1s surveillance drone must roll a 16 or greater to hit, and is 2ALT higher than P2. P1
thus can roll 3 total times to hit. P1 fails three rolls allowing P2 to counter on a roll of 16
or greater.

RATK = Ranged attack, roll to hit a model located one square away edge to edge.
Ignore unit altitude.
Carry = Cargo, objectives, and other units may be carried by this unit. Carried units are
declared by the player then removed from the board until acted upon. Carried units
cannot carry.
Identify = Spend one attack to identify, roll to hit a model within a three square radius.
Successful hits expose model base.

Any update on this topic @Yumibill? Your writing style and creative process seem super interesting, wondering where you’ve got?