Adapting the EuroRack audio world for video synthesis

Discussion of modular and standalone video generating/processing techniques and associated hardware.

Moderators: Kent, Joe., analogdigital, infradead, lisa, parasitk, plord, daverj, lizlarsen, nickciontea

Post Reply
User avatar
lizlarsen
Super Deluxe Wiggler
Posts: 2155
Joined: Wed Jan 13, 2010 1:45 pm
Location: Denton, TX

Adapting the EuroRack audio world for video synthesis

Post by lizlarsen » Wed May 04, 2011 9:50 am

EDIT: I've decided to make this thread a sticky, and put a list of all video demos posted in this top post.

Audio Oscillators with the LZX Visionary
Roland V-Synth XT as a Video Oscillator (waveforms)
What the Malekko Wiard Anti-Oscillator looks like and what the LZX VWG sounds like
Moddemix VS. LZX
Cwejman SPH-2 demo as a video module
LZX + Doepfer A137-2 Wave Multiplier
A-136 Solarization Tests
Colors of Shadows
WMD - Synchrodyne - LZX VIDEO

There was a request recently for a thread concerning techniques for using and adapting modules typically designed for audio synthesis for the purpose of video synthesis, particularly in the world of EuroRack (although this thread doesn't need to stop there.)

So chime up with your impressions and advice on techniques with using your favorite audio devices with your video synthesizer!

Here's a new demo by Johnny Woods to start us off, which coveres WMD Gamma Wave Source, QMMG, Wiard Oscillator, and Toppobrillo Triple Wave Folder.

[video][/video]

I will be collecting "field reports" from this thread to create a database with information on things like which audio oscillators sync to video clocks, etc. For a future resource page on my website.
Last edited by lizlarsen on Mon Mar 26, 2012 2:19 pm, edited 3 times in total.

kjellb
Common Wiggler
Posts: 173
Joined: Wed Dec 08, 2010 12:00 am
Contact:

Post by kjellb » Thu May 05, 2011 4:08 am

We can save time and money by sharing which audio modules work for video. My experience is that you can do a lot with standard modules and even with audio mixers and effect pedals. As for oscillators and syncing, the terminology used by the LZX system does not seem to be the standard used in the industry. As far as I can see what Lars calls "field rate" equals Vertical Drive (VD) pulse, and "line rate" equals Horizontal Drive (HD). Please correct me if I am wrong, but this is the way it looks from the tutorial in the post. The A-111 VCO2 has good locking capabilities. This module will lock to VD which will give you locked horizontal bars. To produce vertical bars you should get a dedicated video oscillator from either LZX or from the announced upcoming Jones system. From my experience the audio oscillator that goes pretty high is the Livewire AFG. But this has a much smaller window for locking than the A-111 (see the tutorial above and notice how you much approach a certain frequency before the oscillator locks). The Livewire AFG will lock to HD and produce vertical bars (but not like a dedicated high frequency video oscillator). Mix one VD locked oscillator and one HD locked oscillator, add some CVs and you start to engage the eroticism of analogue video shaping. For wave shaping I used the A-136 (which works fine) a lot, but now I prefer the Toppobrillo Triple Wavefolder.

User avatar
MrDys
Super Deluxe Wiggler
Posts: 1911
Joined: Sun Mar 23, 2008 12:40 pm
Location: Pittsburgh

Post by MrDys » Thu May 05, 2011 8:23 am

I want to try to get some video up (hopefully this weekend...?) of the audio modules I use (A137-2, A196, A126).

But...has anyone tried the Piston Honda as a waveform source yet? I'm thinking that custom ROMs might provide some unique opportunities for "drawing" waveforms. I guess I'm looking for how well it syncs, etc...?
darenager wrote:I wonder how many people with Maths don't know it can file a tax return, or that Plague Bearer can indeed give bears the plague
suitandtieguy wrote:STG IS ALL ABOUT THE PLUR.

User avatar
lizlarsen
Super Deluxe Wiggler
Posts: 2155
Joined: Wed Jan 13, 2010 1:45 pm
Location: Denton, TX

Post by lizlarsen » Thu May 05, 2011 9:28 am

Kjell, thanks for the awesome post. :)
As for oscillators and syncing, the terminology used by the LZX system does not seem to be the standard used in the industry. As far as I can see what Lars calls "field rate" equals Vertical Drive (VD) pulse, and "line rate" equals Horizontal Drive (HD).
Here is some reasoning/philosophy behind our choice in terminology...

-- Horizontal / Vertical drive, which are video sync signals, are active low (negative trigger) logic signals, which is the opposite of what is typically dealt with in the synthesizer world, which are active high (positive trigger) logic signals. In LZX terms, Line rate clock and Field rate clock, are inverted versions of HD and VD, meaning there is a positive pulse at the beginning of the horizontal, and vertical intervals. One reason for the naming change is to avoid confusion with inverted and non-inverted sync signals.

-- From a beginner's standpoint, most users we assume are coming at this system with no previous experience of working with broadcast gear and video sync generators. However, most users are at least comfortable and understand the concept of "Frame rate" and also of a "clock/bpm clock" signal used for audio. We felt a natural extension to this understanding of "Frame rate" would be to promote an understanding of vertical and horizontal sync signals as "Field rate clock" and "Line rate clock", since these terms are intuitively clear in that they are dealing with timing/clocks and with what timing they are dealing with.

-- We felt "horizontal drive" and "vertical drive" were also confusing terminology for a beginner when working with sync'ing oscillators, since the "Horizontal Drive" signal creates vertical bars, and the "Vertical Drive" signal creates horizontal bars.

User avatar
lizlarsen
Super Deluxe Wiggler
Posts: 2155
Joined: Wed Jan 13, 2010 1:45 pm
Location: Denton, TX

Post by lizlarsen » Thu May 05, 2011 9:59 am

Piston Honda as a waveform source yet?
Scott has a system and he's been experimenting, hopefully he'll chime in with some notes!
I'm thinking that custom ROMs might provide some unique opportunities for "drawing" waveforms.
A wavetable based video oscillator would have tremendous potential! I'd really love to see a dual wavetable video oscillator, designed specifically for sync'ing one side to Field rate and one side to Line rate, and some cross-modulation between wavetables specifically for drawing "shapes" as you say.

I know the new Zorlon Cannon revision will have two sections, designed to be used this way to generate pseudo-random pixel fields/sequences.
I guess I'm looking for how well it syncs, etc...?
The problem with digital oscillators (or any digital module that uses a DSP engine or microprocessor for a core timing component) is that the crystal clock which drives the microprocessor will be running asynchronous to video clock. So while it may sync... you will have jitter. What is needed to circumvent this is future consideration taken for having a connection (either back panel or front panel) for insertion of an external clock signal which then gets PLL'ed (multiplied) to drive the microprocessor clock directly. This is why the TipTop Z-DSP is intriguing, because it has this feature, which makes it possible to drive the entire DSP engine off of a phase-locked video clock.

User avatar
daverj
Vintage Video Wiggler
Posts: 8508
Joined: Fri Jun 19, 2009 11:09 am

Post by daverj » Thu May 05, 2011 1:24 pm

"VD" Vertical Drive and "HD" Horizontal Drive are standard EIA sync signals from sync generators. Those are 4 volt peak to peak negative going, into 75 ohm terminations.

But "vertical sync" and "horizontal sync" are timing concepts rather than signal standards. Vertical sync usually refers to the whole complex sync signal during the "vertical interval" (the vertical sync pulse and any serration pulses), while horizontal sync usually refers to the sync signals happening during the "horizontal interval" (the horizontal sync, burst, front porch and back porch).

Yes it can be a bit confusing to noobs that oscillators locked to vertical sync make horizontal bars, and oscillators locked to horizontal sync make vertical bars. But it's more about the concept. The video image scans from top to bottom (vertical) and from side to side (horizontal). The sync signals are the video sync signals, and are conceptually correct when thinking about the image itself (vertical sync = vertical movement on screen, horizontal sync = horizontal movement on screen). "Vertical rate" are signals (like oscillators) running at multiples of vertical sync (50Hz or 60Hz) while "horizontal rate" are signals running at multiple of horizontal sync (15 KHz).

Personally I think it's better to educate new people about the video image rather than mask the concepts using non-standard terms. Otherwise they'll still be confused when they have to deal with equipment from other companies. (such as cameras that need H sync and V sync) Besides, once you're teaching them the difference between "frame" which they probably are familiar with, and "field" which they probably are not familiar with, you might as well teach them the rest of it.

On my upcoming genlock module the H and V pulses for triggering oscillators are marked with "H" and "V" and a graphic showing positive pulses.

Image

User avatar
johnnywoods
Super Deluxe Wiggler
Posts: 1229
Joined: Thu Apr 22, 2010 1:10 pm
Location: Santa Fe

Post by johnnywoods » Thu May 05, 2011 1:43 pm

Could we move the discussion of sync nomenclature to it's own thread to keep this one on topic? Thanks!

Back to the original topic, I'm interested to see what you can get out of that PLL. What have you been using it for?

User avatar
lizlarsen
Super Deluxe Wiggler
Posts: 2155
Joined: Wed Jan 13, 2010 1:45 pm
Location: Denton, TX

Post by lizlarsen » Thu May 05, 2011 1:45 pm

Hey Dave,
Personally I think it's better to educate new people about the video image rather than mask the concepts using non-standard terms. Otherwise they'll still be confused when they have to deal with equipment from other companies. (such as cameras that need H sync and V sync)
I agree that education is paramount in all this. I'm not claiming our way of labeling the outputs on the Video Sync Generator is the proper way, it's just what made sense to us at the time, and I was explaining our thought process as to why.

I would hold (and correct me if I'm wrong) that the "Field rate clock" and "Line rate clock" labels are still technically accurate descriptions for these signals, and ones that do not work against an educational effort on the nature of these signals. However I honestly really like the way you're approaching the labelling with the H & V and pulse symbols -- and I am not inflexible, and happy to mark our VSG module panel like this for future revisions if that's what the community would prefer. After all, it does mean the same thing, and the only difference is labeling on a single module.

User avatar
lizlarsen
Super Deluxe Wiggler
Posts: 2155
Joined: Wed Jan 13, 2010 1:45 pm
Location: Denton, TX

Post by lizlarsen » Thu May 05, 2011 1:56 pm

Sorry Johnny!!

I'd be interested to see what someone could get out of the Doepfer PLL module. I am still a newbie when it comes to various PLL schemes, but the basic idea is that you can lock two sources in phase/cycle with each other (phase-locked loop) by voltage controlling one of them (a VCO) with the low-pass filtered version of the phase difference between the two of them. The 4046 PLL chip incorporates several of these functions (phase comparators, a VCO, a low pass filter, etc.) into a single chip.

In order to do frequency multiplication with it, you need to divide down the output of the faster VCO to a division that matches the frequency of the slower one, using clock division of some sort.

So, in the clocking a microcontroller sense, you use a PLL multiplied version of Line rate (HD! :mrgreen:) instead of an asynchronous crystal oscillator. So for example, if you want your microcontroller to run at 16.128MHz, you would multiply line rate by 1024 by setting up the microcontroller to send a CLK/1024 signal back into the phase comparator of the PLL.

User avatar
MrDys
Super Deluxe Wiggler
Posts: 1911
Joined: Sun Mar 23, 2008 12:40 pm
Location: Pittsburgh

Post by MrDys » Sun May 08, 2011 7:49 pm

Along the lines of what I was asking above about the Piston Honda....

I tried out my V-Synth XT as a waveform source this weekend. It works great. Two oscillators that can be FMed, ring modded, mixed, or hard sync'd together. Each oscillator can be a variety of analog-modeled waveforms, or I can load PCM single-cycle samples and use those. I can't wait to get some time to go through the Adventure Kid Waveform set.

I tried to trick the V-Synth into giving a stable image by setting one osc to be an analog shape and the other to be an external input (line rate clock signal) and hard sync'ing the two. It would lock, somewhat stably, but it wouldn't show the analog shape, just a single small bar, which I think was how it wanted to show the line rate. I couldn't figure out how to get it to hard sync to the signal without also mixing that signal into the output. I think I need to explore a little more here. (The only downside of actually getting this working is that I won't be able to use the PCM waveforms. It only supports hard-syncing with its VA-oscillators. Oh well.)

Once you have the oscs set up (detuning the oscs and FMing them seemed to have the most useful results), they can be run through filters/waveshapers and then finally a multi-effects section with reverbs/delays/chorus. The filters in the V-Synth give some pretty fantastic results. There are these sideband filters which give some amazing results. Say you have a big triangle bar going, the sideband filters will chop it up like the IBM logo. Modulate the bandwidth, and the size of those chops will grow/contract and do a kind of 80s video transition.

Everything can be automated with individual LFOs, and there are also 4 16-step sequence modulators (which can be stepped or smooth) that can be assigned to any number of parameters. There's lots of modulating fun in a single patch.

I need to read a little further in the manual, but if I can get the V-Synth to play two different patches simultaneously, and I can hard-pan each to L and R, then I have two more video oscillators to play with in a TVF. Fun stuff.
darenager wrote:I wonder how many people with Maths don't know it can file a tax return, or that Plague Bearer can indeed give bears the plague
suitandtieguy wrote:STG IS ALL ABOUT THE PLUR.

User avatar
lizlarsen
Super Deluxe Wiggler
Posts: 2155
Joined: Wed Jan 13, 2010 1:45 pm
Location: Denton, TX

Post by lizlarsen » Sun May 08, 2011 8:48 pm

That sounds fascinating!

Getting the v-synth to sync up, jitter-free to the sync from an external videl clock may be pretty impossible, BUT... I wonder, there may be some tricks you can do to get synchronized waveforms out of a computer/DAW...

For example, if you had a recording of video sync signals sampled as a loop and base all your other samples/waveforms off of this loop as a master timing source... and you send this loop out of your DAW directly to the external sync input of the VSG. And then use the other DAW outputs to output waveforms sync'ed or playing in multiples of that same timing. The VSG should be fairly tolerant as far as the voltage range it can lock onto, so it should "sync" to this false video without issue, as long as the timing and sync pulses are correct for interlaced video.

Hmm, not sure if it would work, but it's worth a shot!!

User avatar
lizlarsen
Super Deluxe Wiggler
Posts: 2155
Joined: Wed Jan 13, 2010 1:45 pm
Location: Denton, TX

Post by lizlarsen » Sun May 08, 2011 8:49 pm

One way to get a video sync sampled loop would be to just record a black/blank output of the CVE's video outputs into the DAW as audio. I think it would work, anyway.

User avatar
johnnywoods
Super Deluxe Wiggler
Posts: 1229
Joined: Thu Apr 22, 2010 1:10 pm
Location: Santa Fe

Post by johnnywoods » Sun May 08, 2011 11:15 pm

Very curious to see some results, MrDys. In my (very limited) experience, I've gotten great results from digital oscillators as fm sources. I'd love to see some of the possibilities you described below, especially what is happening with the FX sections.

User avatar
ocd synthnerd
Common Wiggler
Posts: 206
Joined: Tue Oct 05, 2010 6:28 pm
Location: Oregon

Post by ocd synthnerd » Wed May 11, 2011 5:32 am

MrDys wrote:I want to try to get some video up (hopefully this weekend...?) of the audio modules I use (A137-2, A196, A126).

But...has anyone tried the Piston Honda as a waveform source yet? I'm thinking that custom ROMs might provide some unique opportunities for "drawing" waveforms. I guess I'm looking for how well it syncs, etc...?
I have video I need to edit but the piston honda is amazing as a video oscillator. It syncs to field rate pretty good. A little jitter. The hertz donut is also great and syncs to field rate equally well. They both make very interesting video oscillators. They are quite beautiful. Modulating yon and hither with a ph-4 or your favorite lfo while the Piston Honda is in field sync is very cool looking. Lot's of variety if you have the waveform rom expander loaded. It takes steady hands but playing around with yon and hither can generate all kinds of stuff. I got a squigley line sync out of one of the roms adjusted just right. Haven't been able to reproduce it but I have it on video. The Hertz donut will ghetto line sync but it produces just a few to five rough vertical bars and never loses the field lines scrolling unless you field rate sync the modulating oscillator and line rate sync the primary oscillator but you get fewer, less vivid, vertical, jagged, pixilated bars.


A big surprise was the anti oscillator. It syncs to field rate and line rate through the 1 volt per octave jack. Line rate sync produces jagged jittery bars that have a lot of texture. You need to play with it a bit and mayhem out seems to produce more bars than triangle but it works pretty damn good.

Dmf-2 is a big winner just using it as dual sine oscillators, wont sync but all the cwejman stuff you can tune the frequency pretty easily to get the bars to stop scrolling and stay as still as sync almost. Plus it's an amazing filter or vco and filter. Does a lot.

MMf-6 this filter has a ton of tricks and works as a very interesting filter or very interesting wave form generator. You have to see this one. Like it's made to do this almost. The mmf-6 will do something similar to the piston honda, generating squigly line rate bars but no line sync.

vco-6 wont sync using the sync jack but syncs to field rate perfectly patched into fm2(expo fm) triangle and sine wave sync perfect to field rate, the other waveforms you kind of see the shape of the different waveforms moving along the bottom, side to side but with no scrolling

D-lfo another big hit, syncs to field rate patched into any of the two sync jacks for each lfo. Perfectly stable field sync with sine and triangle waves. RM/AM output can be put into field rate sync for either the left side or right side or both at the same time. Syncing one side and letting one scroll is a great effect when patching out of the RM/AM output. This is an amazingly awesome and complex dual lfo that works as a great lower frequency double oscillator with all kinds of tricks, has to be seen.

Envelator and Maths also have a lot of video functionality. Not a big surprise.

Cwejman S1 I love you. Every part of this beast is video awesome.


I'll eventually get video posted but wanted to share my findings.
I suddenly have a much bigger video synth system without getting any new modules.

A cve, a vsg, one vwg, and a triple fader and key generator along with the right audio modules that you probably already own would make for a very powerful little video synth set up on the fairly cheep. Having everything is ideal but not necessary.(I don't really believe this I need everything though, I'm full of shit and hooked on eurocrack, now video eurocrack but I do feel like 3 vwg's is almost enough now for me with the great tricks my audio vco's can pull, i'd like another vwg and a vbm yesterday though )

But if someone wants to just test the video synth waters the two core LZX modules (cve and vsg) an anti and the two Harvestman oscillators(or your choice of a couple full featured vco's can do a lot of cool stuff.

User avatar
lizlarsen
Super Deluxe Wiggler
Posts: 2155
Joined: Wed Jan 13, 2010 1:45 pm
Location: Denton, TX

Post by lizlarsen » Wed May 11, 2011 9:38 am

Fantastic post!! :hyper:
I'm going to start putting all this "VCO evaluation" info into a spreadsheet.
If you feel up to a project at some point, it'd be great to get some screenshots of each VCO. Probably the best way to visualize an audio rate waveform for screenshots is to use it to pulse-width modulate a line-sync'ed VWG.

User avatar
MrDys
Super Deluxe Wiggler
Posts: 1911
Joined: Sun Mar 23, 2008 12:40 pm
Location: Pittsburgh

Post by MrDys » Sun May 15, 2011 10:49 pm

I managed to get my video capture device (just a USB TV thing) working finally. Here's what the V-Synth XT analog-modeled waveforms look like:

[video][/video]

I'll do something more interesting (what the different filters do, modulation, etc.) when I have some more time. Oh, and if you notice the waveform jump at certain points, that's just when the sequencer has looped back to the beginning and re-triggers the note. I need to find a better way to get it to just sound all by itself...
darenager wrote:I wonder how many people with Maths don't know it can file a tax return, or that Plague Bearer can indeed give bears the plague
suitandtieguy wrote:STG IS ALL ABOUT THE PLUR.

User avatar
ocd synthnerd
Common Wiggler
Posts: 206
Joined: Tue Oct 05, 2010 6:28 pm
Location: Oregon

Post by ocd synthnerd » Sat Jun 11, 2011 7:59 pm

[video][/video]

This is a link to a demo of some eurorack modules and a few Lzx
core modules to show you can start with a small LZX system and get some results.

This is also about a neat trick I learned that the Anti oscillator can do, as well as some filtering demos done with audio filters.

The audio is the sound of 2 and 3 VWG's.

This is a demo there is no artistic merit to anything you will watch or listen to.

User avatar
johnnywoods
Super Deluxe Wiggler
Posts: 1229
Joined: Thu Apr 22, 2010 1:10 pm
Location: Santa Fe

Post by johnnywoods » Sun Jun 12, 2011 11:02 am

very cool ocd! lots of interesting tips in there. :tu: :tu: :tu:

User avatar
lizlarsen
Super Deluxe Wiggler
Posts: 2155
Joined: Wed Jan 13, 2010 1:45 pm
Location: Denton, TX

Post by lizlarsen » Sun Jun 12, 2011 11:35 am

I really love the audio track, the very precise sync circuit of the VWG has a very interesting sonic quality.

User avatar
Cat-A-Tonic
Lobsters love Muff
Posts: 3694
Joined: Mon Feb 04, 2008 12:03 am
Location: Yokohama, Japan

Post by Cat-A-Tonic » Fri Jul 08, 2011 1:44 am

How useful would the TipTop Z3000mkII be for video with its 30kHz at the upper range?
It seems faster than most audio oscillators, but it doesn't do the 2mHz of the LZX.

A similar line of thought is A-188-1 BBD clock out.
Pretty high frequency I think, but mHz? :hmm:

User avatar
lizlarsen
Super Deluxe Wiggler
Posts: 2155
Joined: Wed Jan 13, 2010 1:45 pm
Location: Denton, TX

Post by lizlarsen » Fri Jul 08, 2011 9:35 am

How useful would the TipTop Z3000mkII be for video with its 30kHz at the upper range?
Perfectly useful for a Y-axis modulation source. See the chart here:
http://www.lzxindustries.net/video-sync-guide

If it can sync to Line rate/HSync then it will be useful in some cases, but at 30KHz you're not getting even close to the full depth of the X-axis range. (Divide the frequency by 15.734KHz to get a rough approximation of how many vertical bars you could create.)

User avatar
Veqtor
Super Deluxe Wiggler
Posts: 2193
Joined: Fri Nov 14, 2008 6:25 pm
Location: Stockholm, Sweden

Post by Veqtor » Sun Jul 10, 2011 12:11 pm

I've had some strange behavior when polarizing high-frequency signals with LFO's when using the makenoise modemix, anyone w. a lzx system care to experiment? I get these funky "noisy" ringing high-freq content, might be interesting in a video context.

User avatar
lizlarsen
Super Deluxe Wiggler
Posts: 2155
Joined: Wed Jan 13, 2010 1:45 pm
Location: Denton, TX

Post by lizlarsen » Mon Jul 11, 2011 5:47 pm

Ringmod or amplitude modulation is definitely really interesting and useful as a video technique. I don't have a modemix to experiment with, but Tony has the two LZX core modules, maybe we can get him to do more experiments with them. He's mainly interested in vector display stuff, and is waiting on the video ramps module to get really started in that.

User avatar
numan7
the most autistic amongst us
Posts: 6207
Joined: Wed Nov 25, 2009 9:09 pm

Post by numan7 » Mon Jul 11, 2011 7:44 pm

:yay: what a fun thread (that makes me want to do some experimenting)! :tu:

cheers,
7n7
"knowledge has to be organized so it can be taught, and it has to be reduced to information so it can be organized do you follow that? in other words this leads you to assume that organization is an inherent property of knowledge itself, and that disorder and chaos are simply irrelevant forces that threaten it from outside. in fact it's exactly the opposite. order is simply a thin, perilous condition we try to impose on the basic reality of chaos..." --william gaddis

User avatar
johnnywoods
Super Deluxe Wiggler
Posts: 1229
Joined: Thu Apr 22, 2010 1:10 pm
Location: Santa Fe

Post by johnnywoods » Mon Jul 11, 2011 10:02 pm

oh it works!

I'd tried using the moddemix for video before, but for some reason, it's very very picky about working with low voltages (ie: it doesn't work!). Using the prototype voltage bridge to boost two oscillators up to 5v, it worked brilliantly! Well into line rate!!! I made a quick little video. It's uploading to my youtube now.
I'll post once it's done uploading.

Post Reply

Return to “Video Synthesis”