I’m running 13.6 on the MacBook I’m currently testing on. I have an older one which I’ve yet to get to.
@thetechnobear has sent me a build from the build server yesterday that works. So it seems the new build chain makes problems with old systems - but fortunately the release builds will work.
1 Like
I was thinking about this a little more on my run today. So I here’s my 2c:
- I think I agree with the “clutter” comment regarding the Eigenharp module, and often the user might not be interested in using the strips at all. However, I also think that more control over the strip behaviour than described above is desirable. So would it make sense to split this out into it’s own module? Maybe move the whole set of non-key inputs to it’s own module (breath, strips and pedals?).
- As @NothanUmber mentioned, it’s common to use both absolute and relative strips, and the output voltages should, IMHO, differ. ±5v for bipolar relative, 0-10v for unipolar absolute.
- I also think the user should have control over the return to zero, and if that’s disabled, be able to specify
if a new touch instantaneously jumps, or allows for a “pickup” of the value (more desirable in the case of a relative output). - It’s not clear to me that we can achieve this using the standard VCV modules?
Aside from that, the new release definitely solves the crash I was seeing. I do note that the mode key on the bottom row changing between voices seem to have four states? The behaviour of the octave switches is also a little odd if pressed, but I guess that’s down to the limitations of the free modules?
so you cannot really put other outputs (strips etc) into a ‘separate module’, as the main. eigenharp processing is done within this module … its not just a ui 
that said, I am considering the role of ‘expander’ modules, though these have a one sample delay on them.
all that said, we are looking at possible redesign of the modules, and this is becoming a moot point, as there will be room for abs/rel and gate for strips.
so really its a non-issue, just a part of it all evolving.
You’ve stated that MetaMorph is open source, but I’m not seeing it on your github, and I’m not finding any links on this site. Maybe you haven’t made it public yet?
FWIW it’s technically possible to implement outputs on extender (or other kinds of companion) modules.
The processing doesn’t necessarily have to happen in the same module that has the output jacks. There are issues like the 1-sample delay to deal with, but a few samples offset is relatively unimportant for most control-rate information and it’s certainly possible to implement synchronization across cooperating modules where that’s important. Of course, doing all that is a lor more work than doing it all in one module, so maybe not worth the effort, but again Rack UI isn’t a comfortable place for a giant module. All interesting tradeoffs, and I’m having the same process on the plugin I’m currently working on for the EaganMatrix.
I don’t release source code until product is finished/ released.
Yes, I already mentioned (in the post above yours) you can have expanders modules which can be used for many different uses eg more in/out and params.
I’ve been considering their role, if any.
However, I’ve noticed they can be a bit confusing to end users.
also you still end up with what is essentially a large module , it’s just now parts of it are kind of optional.
Technically the delay is not really a big issue, I know how to deal with the oddities it throws up.
But yeah, I’m still considering if they have a role, particularly with split and layers, which I think is the most obvious use case - and in the last update I re-wrote the code of these to allow for this possibility.
Posting here to express interest in EA testing but I cannot DM you yet since I am a new forum user. We’ll see if this posting comment allows me to send DMs.
1 Like
Welcome! DMs should work now.
2 Likes
sorry, I’ve been quiet on this.
Ive kind of put this EA on pause, as I’ve been hoping to release this properly.
(as Ive not heard of any issues from those that have used it)
however, someone is helping me to get the UI looking a bit nicer,
which Id like to have complete before the release.
so its, going to take a little longer.
I’ll post here, with news once I have a timeline… or a release 
3 Likes
Hey. Im new around here but not elsewhere. I just got an Eigenharp Pico this week and getting it up it and running. Is there availability to test out the VCV module still?
hi @midst, welcome to polyexpression
sorry, as above, at the moment Im pausing on the early access, whilst I figure out a release schedule fro Meta Morph.
watch this space
1 Like
Thanks for reaching out. Im really looking forward to this endeavor.
I had an EWI unboxing dream last night, the label on the box said
‘PLAY MORE NOTES
WITH MORE ACCURACY’
I can’t wait. lol
1 Like
Can’t wait for the release! I appreciate the effort that was put into EigenD and the tools but they were actively interfering with me using the Eigenharp the way I wanted, just too slow and difficult to use. I think Rack does it much better and Meta Morph looks like a great fit.
yeah, my hope for Meta Morph is that it will open up the Eigenharp for many…
whilst vcv is not ‘simple’ , it is very well understood and documented.
so once you understand the basics of how Meta Morph ‘exposes’ the Eigenharp, you are then able to leverage alot of pre-existing knowledge on vcv.
also as I mentioned at the beginning… the protocol between Meta Morph is also documented.
the means that others can develop vcv modules that use the meta morph data streams, to add specific eigenharp handling.
in this sense, I like the way modular system decouple development projects… a 3rd party developer does not need to know/see/understand the Meta Morph code base, to be able to create something new for the Eigenharp… they can focus on ‘normal’ vcv development, which has established ‘developer support’.
anyway… we are making progress, hopefully release is not far off !
1 Like
calling testers !
does anyone in the early access group have time to do a bit of testing?
I’m close to having a release candidate, based on some much improved panel designs
however, the change means , I need to go thru every module, checking every jack and parameters, as its possible jacks or parameter functions could have been flipped.
Im pretty confident its all ok, from my surface testing, but Im only going to release
I will, of course, be doing this myself, but having a second set of ‘eyes’ would be useful, in case I overlook something.
General Release
public release is very close…
assuming I don’t find anything problematic doing this round of testing - which ofc, could happen!
functionality will be, as per, the early access.
final tasks are …
a) full testing, mainly regression testing
b) move documentation to wiki
c) tidying up
(c) is a grab bag of small tasks, whilst small can inevitably eat up quite a bit of time.
(fortunately, I did a lot of prep for the early access release, so lightened the load for public release!)
I will initially release directly, rather than via the VCV store - so I do not have to figure that side out, and in case there are any ‘blockers’ on that side of distribution.
anyway, we are close… as I think I mentioned earlier, I actually have other projects I need to turn my attention too, so its important I ‘wrap this up’, and get it out of the door
1 Like
I’m here
I can go thru all the Macs I have and Eigenharps…
1 Like
thanks… I’ll send it over later
1 Like
Question : Strip behaviour
in the new version of Meta Morph, we will have TWO strip values
Strip Abs (Absolute)
Strip Rel (Relative)
(we also have Strip gate, but behaviour of that is obvious)
I need to decide on what each of these ‘means’ and its ranges
Strip Absolute
currently, Strip Absolute is a bi-polar signal going from -5v to 5v where 0v = centre of strip.
however, I could make it unipolar 0 to 10v, this feels a bit more ‘intuitive’.
the disadvantage of this, is a common use-case would be to send Strip Abs into Scaler:Global Pitch Bend and for that you want a bi-polar value cv input.
however, see below… because perhaps we use Strip Relative for this purpose.
notes:
- its easy enough to ‘convert’ from one to the other… so no functional gains really
- im not going to make this ‘optional’, as it complicates many things.
Strip Relative
ok, this could have two meaning
A) output that is relative to the position you initially touches the strip
so touch it EXACTLY in the centre of the strip its the same as Strip ABS.
but rather, the ideas is touch it anywhere, then you get a voltage thats relative to that.
so would make pitch bends much easier… since you dont have to be exactly centered.
ironically, this means that perhaps Strip Absolute should be 0…10v since, we could use THIS
note: I think might clip the change at -5v to 5v… for consistency.
note 2: I think this is the EigenD behaviour.
B) change
here the strip is reporting the change / movement… so its kind of like velocity.
so move slowly down you might get 0.5v, but move faster you’ll get 2v - dont move you have 0v.
I personally don’t see this as useful… as (a) though I can think of some ideas where it’d be fun
note: obviously can derive this value if you used Strip ABS with a module that calc’s differential… so perhaps this is better.
overall, I think my tendency is:
change Strip Abs to uni-polar - 0 to 10v, absolute position
have Strip Relative as relative to where you start the touch, which mean this will be the ‘suggested’ input you use for Scaler Global Bend.
EigenLite notes:
changing Strip ABS to unipolar means I will also change EigenLite to make it uni-polar.
Ive considered adding Strip Ref to eigenlite, but honestly, its pretty easy to calculate, so I think for now its easier to leave interface ‘as is’ with just strip abs.
Conceptually, I think a unipolar signal makes more sense for the absolute value.
I think the pitch bend is far more likely to be a Strip Relative. Thinking out loud, what if you wanted to bend into a note? Ie trigger the envelope when pitch a few cents off and resolve to the note itself? I think you’re triggering the note from the key anyway, so would still use a Strip Relative input?
I think this conclusion is sensible!
1 Like