These are all analogue video encoders. There are no digital formats involved. The output is HD component analogue video. In NTSC/PAL modes, composite video outputs will be active. In all modes, component video outputs will be active.
TBC2 & Chromagnon, in addition to their other functions, have tbc/scaler inputs. These inputs are a digital time base correction function for frame synchronization and upscaling/downscaling. The output format of the scaler is always the same as the sync generator mode. So these video inputs take the external video source, and convert that to whatever the current sync generator mode is.
If you want to upgrade your Visual Cortex based system, you can do it a few ways : a) Just run all the new stuff at NTSC/PAL and don’t worry about it. Everything stays compatible. b) Start a new system with Chromagnon or TBC2 and Gen3 modules. Use the Chromagnon or TBC2 to upscale SD System Output to your new HD System Input. In this case cross patching between the HD and SD systems won’t really work out, or becomes a mental puzzle to figure out what you can use or what you cant. There aren’t any rules to break here though – so crosspatching between formats may be a cool glitch world to explore for some of you. If you aren’t doing any cross patching (using the Chromagnon as an independent upscaler / post amplifier / mastering proc) then this should work just fine with minimal headache. Separate cases would make a split workflow easier. c) replace any NTSC/PAL only modules (like Visual Cortex) with their Gen3 equivalents, so that your entire rig is “15 mode compatible.” Most Expedition modules will work fine in either system – anything without sync IO on the rear is going to be compatible without a thought.
That was longer than I wanted, but let me know if I can clear anything up.
Since Sync Generation is no longer a modular function, and internal sync is included in all IO modules, and the ESG has no input decoder to derive sync, I agree with @sean that the ESG is more an Encoder than a Sync Gen, so I vote for “OUT-3”.
There’s a “Triple Key Generator” coming, it’s just arranged as a a triple RGB proc/quantizer/solarizer w/high gain controls – the other (FKG-3) is an RGB switcher/fader w/keying input. Having switching and generating on separate modules has numerous advantages, as does having the keying control on the module doing the switching.
The 12HP limit on FKG-3 is very important, since it is designed to be used as “n-level priority mixer/colorizer/alpha compositor” in chained multiples – an RGB compositing bus, or infinite downstream keyer.
ESG: Have been interested in integrating a second output encoder into my system, but maybe realistically just want a more stripped-down, 4hp one (hoping maybe revised Cadet includes a 4hp component encoder??).
For power and sync related form factor reasons, the smallest module with any sort of video input or output will be 12HP. You can budget on that though – 12HP per IO channel. ESG-3 is a massive triple PCB stack of floor-to-ceiling SMT circuitry all designed so you can still fit it < 40mm mounting depth (with sync cables attached!) So it’s really a question of “mounting depth or HP width”. 12HP is the compromise that gets us the biggest amount of PCB area without prohibitive mounting depths, so it just solves a lot of problems right out the gate to not try to cut corners to make it smaller.
SMX: This is the only one where I feel a little bit that the price perhaps exceeds the functionality…?
In Gen3, any time you see an Attenuverter – it is an active control for 2x discrete VCAs that are controlling the signal path. No video signal actually runs through the pot you’re turning at all – it’s generating a nicely filtered and buffered control source that drives the complementary positive and negative VCAs. This is massively better for the signals and just feels better. True nulling with center detent, etc.
So with SMX-3 you’re really getting 18 discrete Video VCAs! And that’s why it costs what it does (and is similar in scale to these other modules on the inside.)
Some of the future modules will likely be less $ per HP than these 6. Some of them will only require dual PCB stacks instead of triple (aka half the non-power supply related circuitry) so we can lower the price on those closer to $299 or $329, even though they will still be 12HP.
We can also do some 8HP modules, but those would need to be modules not requiring sync.
Fair enough. And sorta figured this was the case — a lot of behind-the-scenes stuff making it work really nicely. Guess I was just thinking more on the level of what does it do, vs how well does it do it. I might also just be a little more of Geo Metro than a BMW kind of guy (or, in reality, I am a bike guy without a car at all).
But 100% appreciate the thought and care put into all this!
And, indeed, I think it makes a lot of sense to leave the down-and-dirty solutions to the 3rd party DIY developers and have LZX stuff be the top-shelf versions.
Well to be fair, the original VBM offered the same basic function – 9 attenuverters, with some extra IO points, for $599. Color Chords is a little less expensive, at $299 – but that’s because there are no multipliers in the signal path. Color Chords can’t also double as a triple RGB mixer (since you don’t have independent inputs) or as a YPbPr/RGB or RGB/YPbPr Colorspace Conversion Matrix (because you don’t have an attenuverter on each point in the matrix.)
So we feel like it’s about on par with what we’ve been offering at the same price-per-function, despite the performance upgrade. If you already have a VBM and/or Color Chords, multiple Passage modules, etc, and aren’t concerned with the performance upgrades then you’re right! You are good!
For us, it’s important that there’s consistency – it’s part of the reason to do a Gen3 core set. We can’t treat the signal in a hi-fi way on one module and a lo-fi way on the next. There’s definitely a big bandwidth increase in the fully active circuitry compared to older modules, and that’s only going to help with HD workflows.
But yes, if you wanted to, you can make a matrix mixer with some pots, jacks, and just three op-amps! It will soften your signals a lot, be a little noisier, mess with impedance – but if it makes you happy that’s awesome, and it would be cheap to build. LZX format is all about a wide range of this. Nobody has to get “gen 3 module certified” or anything! It’s just that for LZX modular, we’re aiming for robust, practical, super solid build, consistent signal expectations/ranges, etc throughout the core set. We’re not aiming for top shelf (that would be more like $3999 RF grade rack modules with digital control systems) – we’re aiming more for “pro gear for working artists.” We don’t need $50 switches… but are going to use the $2 ones over the $0.25 ones, for example.
I think of it as analogous to a mixing bus, I guess? For example – even if you’re feeding in all kinds of mess and noise and hum and weird musical instruments into your audio mix, you usually want to mix that all down using something more clinical and transparent. Making the mix dirtier again is usually a mastering process rather than something you want to be inherited by your summing bus. So FKG-3 and SMX-3 offer what is supposed to be a clean and transparent mix/switch of the signals.
I just want to add some long winded rambling to this thread, related to what I said above.
So with “LZX video synthesis” you really have a few options:
Stick to the instruments like Chromagnon, and expand/patch between “voices.” Like a studio full of monosynths you are mixing.
Use an instrument like Chromagnon as a “base station” and add modules to it as expander functions. For example, adding a DWO to Chromagnon is very useful!
Go “pure modular” and roll your own instrument/studio/88 band colorizer, etc. In this case you’d want to start with ESG-3.
#1 This approach is a great place to start because you’ve got something “complete” without having to buy multiple modules or a case. It might also be the best answer for integrating into a workflow with video mixers and glitch boxes you already own and use. Also you may just prefer working with “complete voices” rather than “separate functions” – a ton of folks do.
#2 This is probably the closest to what I’d consider a “modern EuroRack approach”, where you often have complete voices living alongside weird utility modules in the same environment. If you are putting Chromagnon inside your EuroRack case, this is probably where you fall. In this case, you can rely on the fact that Chromagnon is a complete workflow, and just add to it the modules that are especially exciting to you.
#3 This is old school. This is more like what modular meant in 2006 when I was introduced to it (not that long ago) – it’s about dividing modules by function with “design/customize/tune your own instrument” as an end goal. It may involve some knowledge of physics and signal processing, some research – but that is all part of the fun and challenge of it. The functions should be navigable by and understandable by someone with that background. For many of you, thinking in modular function blocks consumes you! It is just a way of thinking that also relates to DIY ethos, etc. For many of you, this kind of granularity just gets in the way of making your art, which is perfectly valid as well.
“All of the above!” is valid too.
At LZX we’d like to cover this range of uses in a few instruments and a couple dozen modules, and then just do a really good job at supporting, extending and maintaining them. We hope many others will make compatible modules/instruments and lean on us as a foundation that makes their experiments possible.
Yes, Memory Palace Mk2 will follow Chromagnon’s release and be our second Automata instrument. It’s become necessary for supply chain reasons that we combine our hardware for the digital/FPGA heart of all our products to use the same core assembly, so this update will occur before Memory Palace’s next production batch in early 2022.
Will Memory Palace Mk2 have different functionality than Mk1 or is this just a move to standardize the brain behind these offerings?
A move to standardize the FPGA brain PCBA used across multiple products, and update Memory Palace to Gen3 power expectations. This won’t have any effect on future development – firmware will be running on the exact same platform (7Z010), so both versions of the hardware will run the same firmware image and have the same functions.
Usually we completely overhaul things when we revise something, but in the case of Memory Palace we don’t feel the need to make any big changes. So this is more of a form factor update so we can keep it in production and keep developing for it!
Some hardware context may be important – right now we have separate board designs to host the FPGA and video encoders/decoders on TBC2, Chromagnon, and Memory Palace. These are very expensive assemblies, costing $100 - $150 each (our component cost!) to make. (NOTE: This is the cost of one of many assemblies, not the cost to make a unit. That is much higher.)
With minimum order quantities on production batch sizes come into play, this puts us in the position of needing to come up with orders for qty 200-250+ regularly, to get them down to the right cost for our budget. We can usually do that with one product at a time, but this ends up creating a rolling target where one project is always overstocked and the others are understocked. Since all three of these devices use the same part (7Z010), we can revise our core assembly to be shared across these products (really, they will all be getting the new board we designed for Chromagnon).
So it is a vital part of making the supply chain less wobbly, and future proofed.
ooooh does that mean an enclosure like the one for chromagnon may be made for memory palace? Even if it is just for the new version that would be pretty great!
Yes, the Memory Palace Mk2 update will use the same 52HP enclosure and rear power/sync/digital boards as Chromagnon. The Automata instruments following it will all do the same. So there are two shared boards (power and FPGA) and each instrument either has one or two specialized boards on top of those. Chromagnon has a big analog computer board and a big analog control surface board, in addition to the two shared boards, for example. Memory Palace would keep it’s existing control board, with a few tweaks for new parts we’re using now, and of course where the power entry/sync comes in.
The three-letter abbreviations don’t need to be initialisms nor acronyms. They can just be abbreviations that most succinctly convey the semantics of the full name. Commonly used words like “generator” can just be dropped for brevity and clarity. For example:
But I would leave Affine Ramp Transform as ART, because ART.
Also, I don’t see the need for the “-3” at the end of everything. Unless you’re making single channel versions of these, which I don’t think you are. If the only form of these is the triple-channel version, then the “-3” is superflous and just takes up panel space.
The number helps the brain see the acronym as a model number rather than a word that’s supposed to be read, was the conclusion we came to there. The -3 is generational, for “third generation.” Colloquially I imagine we will omit the number when talking about the modules in forum threads.
The three-letter abbreviations don’t need to be initialisms nor acronyms. They can just be abbreviations that most succinctly convey the semantics of the full name.
I agree, but part of it is how it sounds when you speak it, for me. I like all the D’s and G’s just from the rhythm of the phonemes they give in speech. “LZX ESG” and “LZX DWO” has better meter than “LZX ESY” and “LZX WBO” for example.
I think maybe “no rules except it’s 3 characters” might be the best approach – coming up with a ruleset for naming consistently is going to be difficult. Three characters is short enough where even a single character being different is enough of a difference. (And we are all nerds enough to have it all memorized within a few minutes of getting ourselves excited, I presume.)
FKG and ESG are set in stone simply from the perspective of lineage – they are concatenations of previous product names they are descended from. So we have TVFKG → FKG and CVE+VSG → ESG. Since there’s an existing precedent in the community for these names, it’s what makes sense.
DWO, DSG, SMX, we may have some wiggle room to reconsider. I like DWO all right because D rhymes with V, and it makes me think VCO.
I like the names & just want to add that I’m really stoked. It really seems like y’all have taken some deliberate time & effort with this new Gen3 move to do a top down reorg/reimagine of everything.
It’s like LZX has been cruising along for years & years, just making the next module or instrument, fully entrenched in the cycle… And now y’all took a week off (lol can you imagine, LZX take a week off), dropped acid, and looked at the whole thing from atop your intelligent design crane.
So basically, hats off to the crew. Y’all are amazing & it seems to me the future is brighter than ever.
Couldn’t be more thrilled about the “modular” lineup. The instruments are cool, but the building blocks are where my heart’s at.
Can you speak more on the dual discrete VCA’s on all the attenu’s? For instance on the SMX - just stack an LFO or envelope along with whatever video signal into A1 & the LFO will open the VCA’s allowing the video signal to pass? Is that the idea?
As a “voltage control absolutely everything freakazoid” it’s so easy to fall into “where’s the CV ins?!” when looking at these panels. What’s the response like?
