Sorry to take you down another rabbit hole! I keep hijacking your threads.
Thank you, I think I’m finally seeing how this would work, multing the signal into the Steps CV input to modulate the base frequency. Or otherwise plugging some version of the input signal into Steps CV. E.g. running it through a gamma corrector like Syntonie VU005.
Exciting stuff, thank you again!
Yeah, gamma, log, expo, anything slope processor related, would be an interesting thing to patch into the feedback path.
Right, Arch or DSG3 would work there… but I like the Syntonie waveshaper because it’s got a variable control.
Right, Arch or DSG3 would work there… but I like the Syntonie waveshaper because it’s got a variable control.
Yes, all good options – but variable control definitely in the feedback case. Another Stairs in the feedback path is always an option too. Give that some quadrature LFO modulation and watch it go.
I think you mean Baja modulation 
@creatorlars Regarding Matte… having slept on this, I realized that you are right, as usual. Voltage range of 0 to 1 makes way more sense for a dedicated color picker. It gives the full range of mechanical precision. I was planning to use 4x @Fox Foxing Hour Access and 3x FKG3 to roll my own EAB colorizer. But Access isn’t optimized for that, since its output range is -1 to +1. One can only use the mechanical range from 12:00 to 5:00 for legit RGB values. It’s better to have the full mechanical range of 7:00 to 5:00. If you want to subtract colors, that’s what the SMX3 input attenuverters are for.
Color picking is a critical feature for me. Video synthesis artwork has always suffered from limited color palettes simply due to the design / limitations of the tools. Any time an artist uncritically ends up with RGB primaries, or even worse, secondaries, it makes the angels weep. The precision of Matte should help with that problem.
Hopefully you can make the price point competitive with some of the options out there for static voltage generation. I think that’s doable despite the more expensive power assembly and whatever high end components. E.g. 2x Access = $318 … and obviously that’s much more than just a static voltage generator, so it’s not a direct comparison. But you see where I’m going with this. An 8HP Matte should land significantly under $300 considering that much more baroque 12HP Gen3 modules are $399 standard.
Thanks
Yes, I would expect $199 at the most. Matte is intentionally a lower cost system builder module.
The first five Gen3 modules all have triple stack PCB assemblies (control board, core board, power supply board.) Matte and the other new ones here aside from Color Wheel are double stack PCB assemblies (just control board + power supply board.)
Triple stack modules I would expect to range $30 - $40 per HP.
Dual stack modules would be more like $20 - $30 per HP.
We may shift some pricing soon due to costs being crazy right now, but that would be a percentage overall change rather than module specific.
Will we see a triple filter soon? I’m imagining an RGB Curtain with independent filters per color channel.
I could see a triple 8HP VC differentiator (high pass filter, fixed to video rate response) being very handy. A good sculpting tool for triple keys as well.
I think Fox is working on a triple RGB filter. I’m not sure if that is triple or single cutoff – there are definitely advantages to both approaches. Filters will shift the image around, so if you want to filter RGB without creating color alignment shift, you need three filters calibrated to perform identically with single controls.
We do have a new dual filter/texture source planned for 12HP (previously teased as DWF3.) This one is actually based on our OTA VCO core on DWO3, so it’s a new design, but will also serve as a War of the Ants update (“noise” will be one of the filter range options.)
A luma enhancer, but with RGB IO option is something I’d like to do for sure – that would be the most direct follow up to Curtain, with separate main parameters for cutoff and depth. It would focus specifically on brightening/darkening the edges of an RGB source, kind of a temporal Dodge/Burn tool.
Some of the most exciting FPGA based stuff we haven’t attempted yet involves spatial and temporal filters – something that’s not possible without frame buffers and memory.
I am working on a filter, yes!
Not to steal from the thread but my filter is a single state variable filter with High- and Low-Pass outputs in 4HP based on an available example in the LT1228 datasheet. For RGB, three modules are needed.
Three separate filters used in parallel will not share the same voltage-to-current source, true. Alignment shift can be observed but using different pairs of matched transistors on each module has proven to be quite predictable.
@creatorlars, side question, would it be possible to add a dip switch to the Matte to make the “A+B” out A-B?
Additionally could switches be added to the mixer on the sum/dist module so that one or two of the RGB inputs to make them negative? What I’m thinking is basically like the mixer section of bridge, expanded to the the ARGB format of the sum/dist circuit?
Sum/Dist I think needs to stay very straightforward, as it’s purpose is more about signal routing for parallel signal paths, where you often need to mult a source multiple places, and then have a straightforward way to get a unity sum of the results from multiple processors.
For Matte, if you have an A-B option there, it is possible for the middle column to contain colors that are below black (for example, if A is black and B is white, A-B would just clip at black, rendering it identical to A.) This isn’t necessarily a bad thing, but it does mean that it introduces a rule to remember if you’re using it specifically as a color selector.
I think in both cases, what you probably want is Proc! That’s a very efficient way to do a variable subtracting/differential mix – and it also doubles as a general purpose triple gain control or triple bias generator. SMX3 is also a good option for variable additive/subtractive mixes.
I see the appeal of a fully patchable differential amp though, maybe there’s a Diff/Multiplier in the same format as Sum/Dist at some point?
Not to steal from the thread but my filter is a single state variable filter with High- and Low-Pass outputs in 4HP based on an available example in the LT1228 datasheet. For RGB, three modules are needed.
Awesome! This is what the Triple Video Multimode Filter was based on, too. I had to trim all the inputs to get matched response curve, but it worked nicely. Curtain is a VC differentiator rather than multimode, but also uses an LT1228.
Yes, it is certainly possible to dial it in. I guess my point is that I could see good reasons for both topologies as valid function blocks / module concepts (triple filter vs three filters.)
Word, that makes sense. I have multiple passages. Big fan of the mighty proc!
You know I want that. Actually I want all arithmetic operations. In combo with Matte you’ve got basic algebra. The division operation would be a gain amplifier since you’re only feeding it values from zero to one.
For that matter I hope there’s a plan for updated analog logic. I do have 3x Arch but I know I’m going to need more.
Thanks!
Thanks for the replies guys! I guess looking for a reasonably priced third Curtain is my best bet for now; being able to independently shift each color around and mess with the alignment of each channel is exactly what I’m after.
This is what DSG-3 is.
I’m a big user of bandpass filters. I like the LP and HP filters too, but I find BP most useful in my video feedback patches.
@rempesm Not entirely… DSG3 doesn’t have all of the logic functions of Arch. It’s missing the Intersection and Clip modes.
Plus it would be great to have more compositing modes. Lars has written about this, I’m sure it will come in some form.
dang that sounds sick!