Forum

That's a pretty good summation of why we wrote something from scratch, very well observed. To add to it and clarify from the Eigenlabs point of view;

*We made the decision to develop EigenD in 2004/2005 - Max was not even remotely capable of supporting what the Eigenharp does then and there was no indication it would be able to. If there had I been would have been on a plane to California instead of kicking off a difficult and expensive software project. I'm not stupid and I hate the 'not invented here' syndrome. I have paid heavily enough in the past for it and watch for it assiduously. This has not been and is still not one of those moments.

*There do exist Open Source variants on Max, but Max isn't one of them. Open source is the only guarantee that the instrument might still be usable, viable and still developed in 50 years. I am personally passionate about this and the idea that any component of what makes it playable would be unusable because some third party died and took the rights to code with them is and will continue to be unacceptable. You will note that we even chose an open source cross platform framework (JUCE) for our GUI tools with a view that over time they might be viable in that sphere in the future.

*Max is expensive and for Pico owners would raise the cost of ownership to an unacceptable level. This is not that complicated an equation. If we make a cheaper Eigenharp in the future this picture just gets more stark.

*I'm seriously unconvinced that you can do what EigenD does in Max. Happy to be proved wrong, go for it, but having spent some time coding in EigenD for the first time this year, I've realised that it's a hard world dealing with that amount of data, in real time and making audio from it. And Max, as far as I'm aware, has never handled anything like that. And yes, you can decimate the data so it can cope, but that's really not the point - the reason we create that amount of data has a serious purpose and throwing it away just to run some piece of software that costs a load of money and doesn't give you anything spectacular in return is pretty stupid.

* And lastly, Max is old. Lots of the stuff in it is MIDI centric, and that is simply not where we're headed. It's a lot more mature than EigenD for sure, but it's formed from a different set of objectives than the ones we have and meets a different set of needs. We have a different vision of the future, one revolving around highly expressive performance rather than offline composition, academic musical pursuits and twiddling with noise. Those things are interesting in their own right, just not to me and as I have found over the last two years, not to the vast majority of Eigenharp players either.

We have invited other developers to share in that journey towards a seriously expressive electronic future, and a proprietary development environment that might disappear sometime at the whim of others does not figure strongly on that road right now. I'm very happy to see someone interfacing the Eigenharp to Max in some way, I'm sure there could be some interesting things come from it. However, we have near zero customer interest in such a thing, so it's not something that we can give any bandwidth to at the moment.

cheers

John


PS; to add to NothanUmber's OSC comment - we have substantial plans with OSC and any volunteers to help us with that would be greatly appreciated. A test OSC input would be a great little project...

John, I don't know when you last looked at Max but it's certainly not for offline composition....are you thinking of CSound?

Max is a state-of-the-art system that includes real-time DSP (audio rates) and real-time video processing. People are building synths (subtractive, additive, granular, physical modeling and so forth) as well as effects and graphics processing. It has had network (udp,tcp) + OSC support for years. People from the top music centers (CCRMA, CNMAT among others) are working with it....trust me, they are not futzing with old MIDI stuff, they are pushing the limits.

There's a reason it got integrated into Ableton Live almost 2 years ago. There's a significant and vibrant community building instruments and effects processing in Max For Live with new stuff showing up daily. Some of it is astonishingly cool. People are also interfacing it to other physical devices.

The Max Runtime System is available free so if you've created an application with Max, people can use it without having to purchase their own Max license.

It is actively supported and a new version will be out in the next month or so and they're reducing the price significantly to try and get it into more hands.

I cannot imagine that a system designed to process high quality audio and video in real-time could not handle the data rate from an Eigenharp. The only reason I mentioned decimating the data was so that it could be connected to external MIDI devices (and the point is that there's no need for EigenD to do it, Max can).

Finally, while I don't know this, I would not be surprised if you could sell a lot of Picos into the Ableton Live market, if it integrated nicely (i.e, via Max). That market is hungry for interaction and physical devices for performance and the Pico is not that expensive compared with other controllers and physical devices that people are buying to use with Ableton.

I'm not suggesting for one moment that Eigenlabs get involved in Max development, I realize resources are limited. But helping to kickstart things for people who can contribute Max stuff would not be a bad idea. Once you have a base system, the growing Max community will take over.

If I can get the OSC thing to work, I'll go from there, and I'm excited to give it a try (thanks barone)....but even that stuff is overkill as really the only thing necessary is to feed the physical parameters into a Max external.

There's a 3-day Max conference in Brooklyn this weekend and I'm hoping to be able to discuss this with others at that time.

Cheers,
D

Just to add to what dhjdhj has said. I like many use ableton, it's used in the studio and also live. I choose to use a pico as one of my controllers, whilst occasionally I feed audio from eigend to live, mostly it is midi. Having a simple lightweight interface to live via max would cut down what I'm asking my CPU to do, remove a couple of windows from my screen (oh for stage on my iPad). The new setups by geert help and thanks for those but as you know already, choosing midi as the transport means we are loosing so much.
Not everyone wants to use something like ableton for live performance, and that is what the eigenharps specialise in, but the majority of people using computers in a live performance to make audio are using ableton. Ableton push max for live in a big way. These people are a large untapped customer base for eigenlabs given the right technology and marketing. I see opportunities for both approaches, and whilst I can see that a solution based on max may limit subscription revenue, the extra revenue from ableton/max users should more than compensate?

Hi dhjdhj,

It would indeed be very interesting for certain people if Max could be make to work to directly take the data streams from the Eigenharp. It's interesting that you mention audio rates. Each data stream on the Eigenharp is 2kHz, which brings the total maximum data rate on the Alpha up to almost 17 simultaneous audio tracks running at 48kHz, consisting out of 399 individual streams (132 keys * 3, 2 strips, 1 breath) that each carry clocking information, contextual information to allow for correlation, data formats, boundaries, ... This needs to be processed in real time, without introducing latency, without losing the clocking and while respecting the data formats and doing the required data conversions. It does indeed not leave much room of inefficiencies, but it might indeed technically be possible in Max, it certainly would be cool. I just wanted you to realize what the actual rates and data relationships are. (note that this still doesn't include the 96kHz microphone input and 48kHz headphone output)

Finally, to add to the excellent summary from NothanNumber, the Eigenharp is intended to be an instrument that is accessible for people that don't want to tinker. With EigenD they can just plug it in, start the software and begin playing. Then gradually discover the depths of the system and if they so wish, customize it. Having software that is specifically tailored for the Eigenharp makes this possible. Most people are totally happy with the factory setups and don't even delve into external MIDI control, they just want to play it as a traditional instrument, accepting the boundaries and functionalities that we have provided with the shipping factory setups.

Take care,

Geert

" I have no idea what EigenD does for me that I couldn't do with existing environments. "

I, on the other hand, *do*

It is simply not valid to say, in effect, that one has ignored something such that one is unaware of what it can do and that, therefore, it can't do anything.

So, let me say that I'm quite Live and MAX phobic. I've had a (brief) run at them but they are not interesting to me musically and quite opaque to me without a level of effort I see as unjustified. If they had been the required environment for an Eigenharp, I would not own one. This probably feels familiar to you as it is simply the inverse of you own position. It is nothing more.

This is *not* a critique of people who do create in that context nor am I, in any way, arguing against *including* the ability for people who have an established interest in Max and or Live to work in that idiom. Personally, I'm looking forward to see what they do as well. To be clear, I'll absolutely enjoy the outcome but have no plans to suffer through creating such an outcome... I'm a good audience.

Finally, my experience of the Eigen-system is that both contexts (the controller and the SW environment) are excitingly new in a way that argues against exclusively using the HW as a controller for older (if esoteric) solutions. Looking back, it is clear I would have been deeply underwhelmed if it had been just another controller.

m

Not sure I understand what a datastream represents --- is it the case that data is being sent out continuously for each key, even if you're not touching it? If so, then it seems to me that it would make sense to block unnecessary data (i.e, nothing happened) at driver levels. It's my understanding that the upcoming release of Max 6 will have an option (for a small extra payment) to compile the patchers, leading to even faster response times.
It would indeed be very interesting for certain people if Max could be make to work to directly take the data streams from the Eigenharp....



Actually, this is one of the key issues for me. I agree completely that it should be useful for people who don't want to tinker. I am not suggesting that everyone who gets an Eigenharp would touch the Max code. It should be feasible to build a sophisticated Max environment, that at the top level can simply accept commands (using Eigenhap keystroke sequences if that's desirable, but lots of other ways) to change the behavior (e.g., similar functionality to keygroups/splits, layering, scale definition and so forth). Max already has built-in support for playing sequences, drum loops, samples, managing VSTs and so forth that can also just be triggered remotely.
Finally, to add to the excellent summary from NothanNumber, the Eigenharp is intended to be an instrument that is accessible for people that don't want to tinker.


This is the other key issue for me. The problem from my perspective is that EigenD wants to be in charge. Whereas, I view the Eigenharp as something to be integrated into an existing environment. I think that is true for many people. I have had 45 years of training in keyboards and guitars....I'm not about to drop those instruments and dedicate myself to the Eigenharp. I want to extend my expressibility where it makes sense
With EigenD they can just plug it in, start the software and begin playing. Then gradually discover the depths of the system and if they so wish, customize it.


This takes me back to the earlier point --- I do not know what EigenD is able to do for me what I cannot do in other environments. Some examples that show me where I'm misunderstanding EigenD capability would be terrific.
I realize I'm hammering away on this and maybe people are getting fed up with me. I apologize for that but understand that I am a PASSIONATE believer in the capability of the instrument itself....just not so much EigenD right now
Having software that is specifically tailored for the Eigenharp makes this possible. Most people are totally happy with the factory setups and don't even delve into external MIDI control, they just want to play it as a traditional instrument, accepting the boundaries and functionalities that we have provided with the shipping factory setups.

Forgot one reply, here's a very basic example David.

Imagine the Eigenharp just outputting raw data (Max isn't open-source so it's not even an option to be the standard 'brain', gratis is irrelevant here) ... how would anyone that just wants to play an instrument pick the Eigenharp up? Learn about whatever data format it outputs, learn how to transform and correlate that data, figure out how to make it work with musical notes, ... and so on. I vastly prefer the current experience that gives users effortless access to all the expressiveness through EigenD and still allows you to integrate with other environments.

Nobody is expected/required to do anything for me. I absolutely have what I need.

I am sorry you feel your time would be wasted.

I'm done here.

@NothanUmber

That is an interesting suggestion.

Note that I have been "having fun" doing Max processing with the Eigenharp for the last year by essentially just using a MIDI out port from EigenD. I then brought that MIDI data into a Max patcher where I did all my processing. I used some keys for controlling the patch itself, and others for playing, both solo notes as well as triggering chords in a Max sequence.

The problem I ran into was due to pitchbend and aftertouch being channel messages and so unless one was in poly mode, it wasn't possible to add vibrato to multiple notes independently (for example). But poly mode caused a different set of issues for me at the time.

I'd like to be able to do things like floating splits where, (for example) my left hand is playing one sound, my right is playing another, and as I move up and down the keyboard, as long as there is at least one (say) row separating my hands, then the left hand might use one set of MIDI channels and the right hand could use a different set of MIDI channels. To do this, I need to know things like physical key locations.


Now, your suggestion makes me realize that the combination of a scale that maps each key to a unique raw midi note number (from 1 to 120 for the alpha) AND poly mode could be used so that you could always tell which specific keys had aftertouch (or pitchbend) applied. That could then be merged back into a representation that just gives me keynumber + X,Y,Z values at all times which is what I would like to be working from.

I will explore this.

Thanks,
D

/keyboard_1/key/4 siffff "48" 47 2.000000 0.028564 -0.027344 0.050781
/keyboard_1/key/4 siffff "" 0 0.000000 0.000000 0.000000 0.000000

This is interesting output. What's with the empty string and the all zeros? Just wondering.