Memory Palace Dev Blog July/August 2018

memorypalace
lzx

#21

Yes, we’ve got a couple more components to place! We won’t be turning it on for the first time until tomorrow morning. We have a rule around here that we never fire up a new prototype at the end of a long day! :sweat_smile:


#22

or on fridays… thats the rule we use to deploy systems lol


#23

Happy to report that we have our first prototype booting up and most of the peripherals verified at this point. The first unit is on its way to Ed in Sydney, and we’re building up a second one today so we can work on development in parallel.


#24

@creatorlars wondering if you could elaborate a bit on how the keyer feature/ interface on Memory Palace is intended to be used.


#25

The keyer in Memory Palace serves as a way to create an alpha channel for sources that do not have them (luma/chroma key), as an alpha channel processor, and as a kind of pre-buffer write conditioner (so, luma key based writing creates luma trails effects when masking the input to the buffer.) It can also composite between the input and feedback path. It’s not a patched out key generator module like Doorway and Polar Fringe, or an alternative to them, since it’s tethered directly as an integral part of Memory Palace’s workflow and signal path, which is complex and designed to be a next generation video feedback instrument. Using it as an alpha processor with ramps or patterns as input sources is just as powerful as the luma/chroma modes, since there is a video rate alpha input. You could think of it as a post-processor for analog keys, or as a way to selectively write things into the buffer, or composite feedback into the signal path.


#26

I’m wondering about that DVI output since I first saw the inicital draws of memory palace. I think one of the main problems working live with analog video today is having to use a projector or LED screen that only allows digital input (or maybe it’s the only one generally used and there for it’s the only cable present on stage).

so my question to you Lars is, are you doing latency tests with that output? having something on the modular itself that allows to convert to digital and has low latency would be a huge plus <3


#27

There will always be some latency through a frame buffer in general, and also at the frontend of a projector, but in terms of the IO on this module there shouldn’t be any latency comparatively between the analog video and DVI outputs.


#28

I have a question regarding CV control of the Delay input on Memory Palace: is this input designed for video rate signals, or would it work best with slower frequencies? If you were to send, for instance, a vertical ramp, would the left side of the screen be moving at a slower pace than the right?


#29

All of the modulation/CV inputs on Memory Palace will be either audio rate (so vertical displacement possible) or animation/LFO rate. By its nature, frame delay time modulation won’t respond to frequencies above the actual time delay (the modulation must happen slower.)

The exception is the alpha channel (and its corresponding modulation input, the opacity control.) If you wanted to crossfade smoothly between delayed and non-delayed versions of the image, that’s what the internal alpha compositor/soft keyer are for. You could input a vertical ramp into the alpha input and use that to fade (using the internal compositor) between what’s inside the buffer and the incoming data stream.

There may also be a slitscan mode, that specifically uses internal wipe patterns and a soft key based mask (controlled by an internal LFO) to do time displacement effects.

I can’t wait to talk more, but if I talk too much I won’t have time to implement half the cool features I’d like to before launch. :slight_smile:


#30

tenor

yay things to come


#31

Can’t wait to see it at Knobcon! I am dying to ask some more questions, but I can wait… maybe… lol.


#32

I am ready to start answering some questions, especially if they are specific. :slight_smile:


#33

are there gonna be Easter eggs on my digital video module? :egg::egg::egg::egg::egg:


#34

Of course!! Maybe even a bunny.


#35

will there be the ability to scale the CV inputs via the menu or will that be relegated to attenuation/cadet scaler etc

how many levels deep is the deepest menu currently?

you mentioned somewhere that there may be variable slew on the sliders that could be set via the menu. If that is correct is there any potential for something like a clocked sample and hold? where the clock speed was static and definable via the menu and the sample input is the slider

and if we are dreaming since you have left some mystery in the menu it seems :slight_smile: will there be any way to potentially reference a slider say the delay and scale/sum it with another slider say the opacity?


#36

@creatorlars Be careful what you ask for!! :grinning:


#37

I wasn’t going to ask questions I promise
haha


#38

will there be the ability to scale the CV inputs via the menu or will that be relegated to attenuation/cadet scaler etc

Through the menus. Whether or not it will be a global CV scale select or per channel is one of the issues on our list for RevB hardware (which we are working on now, and will likely be the production release.)

how many levels deep is the deepest menu currently?

Current wireframe is 16 different settings, most of which select between 2-3 options. Currently not at a point where we need to have different pages. Most of the work on the menu and UI tweaking in general will be done in the next few weeks. We’re trying to keep it elegant and simple – the complexity comes from all the different combinations of the 13 slider based controls, the settings menu just defines the context they operate within.

you mentioned somewhere that there may be variable slew on the sliders that could be set via the menu. If that is correct is there any potential for something like a clocked sample and hold? where the clock speed was static and definable via the menu and the sample input is the slider

We could do that, probably as a stretch or post-release goal. But in general, that’s what the trigger input is for. You can assign the trigger input to toggle or gate a switched function (such as any of the 13 pushbuttons or some other internal parameters).

and if we are dreaming since you have left some mystery in the menu it seems will there be any way to potentially reference a slider say the delay and scale/sum it with another slider say the opacity?

Non-linear relationships between the CV inputs and slider controls are possible since they’re sampled separately – but in general what you’re describing (modulation summing and gain control) is something we feel like is best patched with analog control signals outside the module; part of the nature of the modular synth as an instrument. You could mult a CV source to multiple inputs at once and then use attenuators/offsets to define their control range.


#39

Sooooo … Im starting to think how Ill be using Memory Palace in performance/concert. Im asumming there will be some sort of presets save/load stuff…

so when I use presets during a performance, and then I fiddle the effect so much (let say just before the climax/chorus…) I really like to come back to the initial preset via a reload optio, it will be nice to be able to quickly reload just the current page/effect settings not the whole preset.

Im thinking each part/fader of the MP will probably have settings, and just wanted to understand if there will be a reload option for the current settings of a particular fader, in addition to globally reload/load the preset.

A reset option (also individual and global) would be nice, to reload the original/factory settings for that particular fader or tge whole preset.

This reset option is very usefull when learning a new instrument, because it helps you to creatre a known starting point or baseline that you can always come back to start experimenting again


#40

There won’t be any preset save/recall for any of the performance oriented stuff – we placed all of that on the frontpanel so that you can play it like a synth and don’t feel like you have to do any menu diving to use it as an instrument.

The settings are more like configuration options, like blending mode selection (sets the context for how feedback is composited with the incoming signal), midi channel selection, DVE position in the signal path (pre-keyer, post-keyer, etc), video I/O formats, customizing the response/lag of the sliders, etc. There will be a settings profile save/load, but the intention is that it sets up the context for the whole patch/performance rather than being a live thing. I think it will make more sense in context once I post the settings.

I think for what you want (performance based recall that can switch between totally different looks instantly) you will want to use an external MIDI sequencer and manage your presets on the controller side. For me this is very interesting in the context of USB-MIDI control in concert with editing software or a DAW, so you can do parameter automation in time with the timeline.

Think of the settings menu vs Sliders/Buttons as a DSLR camera. On the DSLR camera menu you have lots of access to settings that you may change before or after you start shooting. While you are shooting, you have the exposure wheel, button, viewfinder, etc to interact with.

Or another analogy is that the settings menu is more like the type of strings you pick for your guitar, and how you tune them before you play.