Hello,

I was wondering if someone happens to know whats involved when we are making a digital asset in terms of compiling

I am assuming there is some form of compiling with the “Apply” and “Accept” of the “Edit Operator Type Properties” after we “drag and drop” our parameters that we want to use for our Digital Asset.

The digital asset I've created basically has about 1500 - 2000 paramters including the elements like “labels” and “separators”

To make this asset I've created Keyboard/Mouse Macros to save the drudgery of manually dragging and dropping them one by one. ( I can't transfer with the “From Nodes” tab of the Create Parameters column because it starts erroring out with some sort of “QT” message indicating the “list” is too long. Also this method doesn't include your works layout that uses seperators and labels - you have to manually do this anyway.

Creating the initial asset in this way goes rather quickly even with 1500-2000 parameters.

However, this asset is being used in another digital asset multiple times.

Because of the naming convention with Houdini, when I start transferring the parameters of the same sequential digital assets, my original expressions “get broken”, because my parameters are now being tagged with suffixes like “granulate_tx” of the first asset becomes “granulate_tx2” of the second asset.

This is not the issues, because what I am doing is duplicating the original asset with the same content and layout but ( another reason I made “Typing Macros” ) but with different parameter names and their corresponding expressions.

However, even though I am using the “same” digital asset with a “different” name in the new asset that is containing multiple assets I am starting to run into problems as I transfer more and more parameters.

Even as I start in on into transferring into the second “same” asset I notice things start to slow down.

By the time I have come to finishing this second asset it now takes about 2 seconds for example of waiting for a dragged and dropped “Seperator” parameter to finish “registering” as having done so. If I don't wait then subsequent drag and drops of parameters don't occur.

In writting my macro scripts I noticed this affect happening so I included a global sleep cycle time that I can increase those times as I see the delays getting longer and longer.

The thing is that I am only at 2 digital assets having had their parameters transferred. I am wanting to have 72 of them in this particular asset and further down the road more.

It looks like this plan will not work without an immense amount of time to transfer parameters.

Which is why I am wondering what is involved with making these assets processor and memory wise.

I'm using an i7 processor with 8GB memory on my system and using an SSD.

I was already in the plans for increasing my memory since the prices have come down since I first bought my system last year.

But I also came across this idea of Ram Disks.

If I did create a ram disk I've noticed the install size of Houdini is about 1.54 GB and the files ( hdalc, hiplc ) I am working with only about 11MB.

So in theory it seems even with the memory I have, save setting aside for opearing system a ram disk might help speed up this process of making the digital asset.

I am thinking what I have to do is once I've made the Ram Disk is to install Houdini onto that Ram Disk?

I guess the question would be how does this affect my licensing? I guess I would have to uninstall Houdini from the hard disk and hopefully the licensing process still sees Houdini as being on the same system? ( I really don't understand how the licensing works under the hood which is why I am asking this ).

Any thoughts on all of this is appreciated. Thanks

I have no advice to give on this…but I'd think it might be VERY important to let us know what sort of work you're doing that would require an HDA to have 2000 parameters…and why you'd need 72 of them.

https://www.youtube.com/watch?v=AY7OUQjedfc [youtube.com]

This clip was made to test my CHOP net which is at the heart of my project.

It has 5 different groups of nodes within it that each lends themselves to producing certain types of waveforms that can be easily modified or switched between.

In the clip there are 5 objects each composed of 4 pairs of PathCV points.

Each point has the inherent 9 parameters of Tx, Ty, Tz, Rx, Ry, Rz, Sx, Sy, Sz, Px, Py, and Pz.

The Ts' and Ps of course work torwards moving of the points position. The Rotation and Scale parameters work together to give direction and shape of the curves coming off the points.

Each of those parameters is controlled by its' own ‘waveform’ CHOP net that I made so that the points are able to move independently of each other and in ‘concert’ with each other.

This last part of what I mean by in ‘concert’ with each other is difficult to explain as its more into the realm the artist side of things and what I envision.

Suffice to say that the more controls that I have the more ability I will have to compose and fine tune what I produce.

Yet it's not just the one final project. I believe I would have been able to make a single project like in the video clip months ago, but I'm thinking down the line of being able to make different versions of the same project without having to do additional work like adding or removing nodes and wiring them.

In this clip I only have 5 objects but in my project there will be about 200 of these objects oriented along inside a cube shape room. All of these objects will have their own independent shapes and movement.

Like morphing wallpaper.

So each object has the 8 points of 9 parameters that are each controlled by dedicated CHOP net, hence 72 per object.

That CHOP net was easily made into a Digital Asset and so the Object has 72 of those Digital Assets.

In thinking about this further I really don't need to ‘promote’ all those 72 Digital Assets up to the object level.

I think I will just make the ‘Object’ Digital Asset with those 72 other assets and leave it at that; Dropping them into the scene as I need them and keeping the assets in their state of “allow editing of contents”.

So much work just to save the one step of diving into the asset and accessing the other assets?

Perhaps I should have thought of that before posting.

looks cool
from a usability standpoint I can't imagine anyone wanting to have to adjust all those parameters to create new shapes + animations

have you thought about
building a set of shapes with the CVs and storing them as base shape presets (or whatever) and a set of animations/transitions.
then add some duplication options - mirror, mirror offset, mirror direction, duplicate, copy to points etc etc
then you could, for each object, have an interface that allowed the user to create animated transitions between different settings.
then you could add fewer objects in total (because they can do more)
and maybe even use instancing…

Yes Michael, I could do that…and I think some parts of my project I will incorporate those ideas.

I think I just needed to verbalize what I was doing a little bit so I could stand back and re-evaluate.

So with your ideas and others that have occured to me since I first posted on this topic, I think I'm on my way simplifying my approach much more.

Thanks for taking the time to read and respond. It's great to have a forum like this and throw things out to mull over.

np
looking forward to seeing the results