DRD Team

Obviously, md2tool was written by skyjake expressly for Doomsday and yes, you can convert DMDs back to MD2. When doing such a conversion you must specify which LOD level you want to extract from the DMD.

I'm still very much interested in supplimenting jDRP with a set of MODELDEFS for use with GZDoom/Skulltag rather than distributing a seperate model pack.

Too bad that I don't have the time myself to design a better model definition format. The current one really was just a quick means to an end because I wanted to play a few Vavoom WADs with the models in place.

What I'd really like to have is something that can define an animation sequence from a state-label to the next, not something frame based as it is now.

Given that animation is done internally in DOOM as frames as opposed to sequences, I think that its much easier on the engine to keep the models defined per frame. Naturally, that doesn't help the artist when writing model defs though.

I don't know much about ZDoom's internals but if animation is still done by linking disparate frames then if you define models using sequences; it makes the animation process much more difficult, especially when interpolating between frames of different sequences as it means you would have to walk the sequence to find the animation frame you want as the interp target. Not to mention the added complexity of model animations with more/less frames than the number of DOOM sprite frames in the sequence.

Certainly not impossible but definetly more difficult.

We plan to do the same in Doomsday at somepoint in the 1.9.x series as we want to move to full skeletal animation.

Nash wrote:Correct me if I'm wrong though.

Consider yourself corrected.

As has already been said, it can convert to MD2s from DMDs. I did so successfully just now.

One thing to note is that when GZdoom's renderer draws models using unique light values for each model vertex; you'd get far better results than MD2, lighting-wise, using the DMD format, as it stores normals using a much more accurate representation.