Resurrected graphics switch

Hi guys, check out if you haven't seen the progress made with the database yet.

Just thinking ahead a bit now - next patch will be trade, some small changes, and performance/SEO stuff to prepare for launch. After that, I'd like to work on a patch that will introduce a Resurrected graphics skin switch for the site.

In order for that to happen, there will be a ton of graphics and artwork stuff to be done. Pulling the item inventory graphics from the game files has proven simple enough, but that's only items. The skin switch mechanic itself is already coded, so that's already a lot of the work out of the way.

However there are other things will require a bit of teamwork, and I'd like to plan ahead and pick your brains about how we're going to achieve this. I know a lot of people here are very technical indeed so hoping for some Insight on the methods used.

First of all, I want to maintain both the scale and the perspective of every graphic and icon that already exists on the site:

• I know Resurrected is offering much higher fidelity artwork and models, which are presentable at a larger size, but I'd like them to remain at the same scale as the original D2 ones here on d2io. This also ties into future plans for a tetris inventory (which requires items to meet those itype 'boxes occupied' dimensions that you're all familiar with. Whilst the graphics will seem small for some, it will keep the continuity of design on the site and has practical reasons too as mentioned re: the tetris inventory for grail tracker, build planner etc.
• The original game had sprites that could be rendered from the same perspective and direction using an extraction tool. Resurrected, on the other hand, is using full 3D environment with 3D models, instead of sprite sheets. Therefore we need to find out a way to consistently 'photograph' these models - so first we'd need to find a way to extract the 3D models, get them rendered in 3D with a program, setup consistent lighting/lighting angle, and from the same camera altitude/angle as the current graphics on d2io take a picture, then
  Mask it for transparency. This applies to things like Monsters and NPCs for example.
• Using that same method, I'd like for any equippable item to have a block on the entry's page with either 1. A 3D model preview of how the item looks when equipped on the different character models, or 2. Have in-game screenshots of characters wearing said item. I'd also add an extra block with a much larger and higher quality item artwork for those items that have it.

My main concern currently is how we're going to extract the models and photograph our own 'sprites' from 3D renders. It would be cool also if I could make inroads into on-page 3D rendering of items, if we can convert/render the files using javascript or something like that.

What are your guys' thoughts on my plans for this, and what advice can you give me in facing the technical challenges ahead?

Teebling wrote: 3 years ago

What are your guys' thoughts on my plans for this, and what advice can you give me in facing the technical challenges ahead?

From me? Absolutely nothing! This is way over my payroll. My brain melts every time I read convos like this...
I'm just a stupid guy that can Copy/Paste things around and has a little bit of knowledge about the game. That's it. That's all I have to offer.

Sorry!

Ok, so did some research into this tonight with mild success.

Was able to get a render within a program and was able to export to .obj files with png textures
I'm pretty clueless about 3d models outside of those for 3d printing, so not sure how to go about posing the models other than a t-pose, nor do I know whether or not things can be fixed like the positioning of the book in Deckard's hand.

As far as getting proper renders, after converting the files to a more common format, it shouldn't terribly difficult to create a little OpenGL program with a consistent light source, camera placement, place the model, and render a png (unsure about the transparency).

Here's the step I took to render the models:

1) Download Noesis: https://richwhitehouse.com/index.php?co ... ojects.php
2) Download granny2.dll https://github.com/Kyle-Cerniglia/EvE-3 ... er-0.5.1.1
3) Place granny2.dll in the root directory of Noesis (wherever you extract it to)
4) Download a GR2 formating python script for Noesis: https://forum.xentax.com/viewtopic.php? ... &start=283 (You have to sign up on their forum to download)
5) Open the python file and edit the first set of constants to have these values:
Note, found these settings in the same forum posts, might be different ones that work better, no clue

6) Put the python script into {noesis_root}/plugins/python

From there you can open noesis.exe (Make sure you don't use the 64 bit version because the granny2.dll was just the 32 bit version. Navigate to the D2R files extract from CASC storage. Double click a model files, then cancel the dialogue box that opens up. Wait about 5 seconds, it should say loading in the bottom left corner.

That's it

I have no idea if noesis is scriptable to mass export the models, but pretty sure we'd have to export. There are libraries to use the GR2 files, but that's beyond my capabilities in C, no idea what it would take to do.


As far as your plans, the first two bullets seem pretty doable. The third I'm not sure how to go about doing a render with the character model also holding the item model. Going the route of screenshots seems easier, but we would also have to wait a few weeks until the game is released. Could make more consistent images by running your character into a corner prior to taking the screenshot. This would help automate things by being able to

Key off of landmarks near the character to determine pixel ratio, then auto crop/scale the images. Just a thought. Hopefully theres an easy way to do it with the model files

Sounds plausible Noemard and thanks for looking into this. If we can get them posed, lit, and photographed from the correct angle then that would do the trick. See these renders by wowhead for example:

Oh, yeah, that would be a whole lot of work for every npc and monster. Having never used any posing software idk what they are capable of,

I wonder if there is a way to setup a scene for lighting/camera perspective with the transparent background and save that for multiple people to load along with the model files. Then it'd just be up to the poser (assuming that's what you'd call someone that poses models) to make a natural stance.

Although it looks like in the examples you posted above the camera angle changes a little (second

Akara image looks to be a little higher than the rest).

Once the models are posed it might be possible to script a 0 degrees, 45 degree, 90 degree, an 180 degree rotation to speed things up.

To be honest - what kind of "improvement" would that mean to the site?
I appreciate the feature but's its something I will look at a few times and then won't come back to.

It's a nice gimmick in my opinion but not an improvement

If I'm allowed to say:
The site need more lists and shortcuts.
e.g. Overviews as you can find them in the

Arreat Summit.

Just my 2 cents
Keep up the good work!

@Noemard @Teebling

Playing model animations in Noesis can be used to pose them/ get diff animations looping


Settings in the Noesis plugin that I used
```
MULTIFILE = 1
SMART_DETECTION = 1
ANIMATION_TRACK = 7
ANIMATION_MODE = 1
SKELETON_LOAD = 0
MERGE_SCENE = 0
```

ANIMATION_TRACK 7 was the idle animation. I had to play each one to find the idle one.

https://imgur.com/a/xohFoCL

Akara idle (also track 7) matches what wowhead has
https://imgur.com/a/9s3Itkd

Gheed idle (track 7)
https://imgur.com/a/RMbWuCY

Woah nice, didn't realize it could load the animation.

Stormlash wrote: 3 years ago

Teebling wrote: 3 years ago

What are your guys' thoughts on my plans for this, and what advice can you give me in facing the technical challenges ahead?

From me? Absolutely nothing! This is way over my payroll. My brain melts every time I read convos like this...
I'm just a stupid guy that can Copy/Paste things around and has a little bit of knowledge about the game. That's it. That's all I have to offer.

Sorry!

Here I be looking up to you as some kind of skilled god as I copy and paste. Your game is strong. You say your a stupid guy, I viewed myself as a schmuck copy and pastying my way to a name around here heheh! If only there were enough entries left to contend your superiority! Praise be

[Edit] My fingers are available to help push buttons if anyone has directions.