Voxel Section Editor III 1.34: The search continues!

Banshee

The search for the perfect normals continues. Someone has given me a mech voxel some years ago and I don't even remember the reason... if it was rather a bug with any voxel tool or to help with the normalizer, but the fact is that it's pratically impossible to get perfect normalization results with it because... well the voxel itself is awesome, but it has all the situations that would kill my normalizer.

Few minutes after the release of the VXLSE III 1.33, I was aware that this mech wasn't normalizing properly. After some analysis, I had a kick ass idea to fix some of the issues with it and, at the same time, it would simplify a lot the first part of the normalization process, which is to build a kind of 3d matrix that would be used to calculate the normals. The idea worked very well and the Mech's normalization is almost perfect, except for a little detail on its face.

So, Voxel Section Editor III 1.34 still features the AutoNormals 5.1 from 1.33 (as Auto Normals) and it comes with this new, cooler and faster Auto Normals 6.0 (as New Auto Normals). It's up to you to try and choose your favourite. If one option becomes far too much popular than the other one, the worse one will be removed on future versions of the program.

So, download the latest VXLSE III at the left menu of the site and enjoy the new normalizers!

Dupl3xxx · Posted: Sat Feb 17, 2007 1:24 am Post subject:

I say no more #Tongue

(sorry Banshee)

_________________
This is a signature

Banshee · Posted: Sat Feb 17, 2007 2:17 am Post subject:

Yea, 5.1 is better in this specific case. I'm glad I haven't throwed this one away. Thanks for the testing.

Edit: Actually, I'm not sure, really. 6.0 is more smooth and more natural...

MadHQ · Commander **Joined**: 07 Nov 2003

Well i have finally decided to see what you got going on with this new normalizer stuff as i see you have been spending alot of time with it. And i will say alot of it is very impressive. but i tested out 6.0 and, some things just not right.

wasp_001.gif

Description:

Filesize:

402.5 KB

Viewed:

1965 Time(s)

_________________
MadHQ's Graveyard - Click here!
(Permissions) - (F.A.Q.)

Dupl3xxx · Posted: Sat Feb 17, 2007 12:10 pm Post subject:

I have found the weaknes of AN 6! When two "parts" go over eachother, it gives it one normal Confused

And some small detail (1-4voxels) looks not that good Sad

But the AN 5.1 is tho most impressive #Tongue

Thank you for all of the work, you make life a lot easyer Smile

_________________
This is a signature

ORCACommander · Posted: Sat Feb 17, 2007 12:17 pm Post subject:

damn I wonder what gotten banshe so motivated?

_________________

Dupl3xxx · Posted: Sat Feb 17, 2007 12:55 pm Post subject:

I don't know, but whatever it is, find some more of it Laughing

_________________
This is a signature

Banshee · Posted: Sat Feb 17, 2007 8:42 pm Post subject:

ORCACommander wrote:

damn I wonder what gotten banshe so motivated?

Your answer is in the VXLSE III 1.33 release news.

Banshee wrote:

Keep up with error reports and showing samples where the normalizer doesn't work. I'm interested on it and if I manage to sort all problems, I'm considering the option of writing some scientific article about it... (but if you know any article about normalizing voxels, please show me).

Dupl3xxx wrote:

I have found the weaknes of AN 6! When two "parts" go over eachother, it gives it one normal Confused And some small detail (1-4voxels) looks not that good

I've found one weakness with the influence area function. I've fixed it on VXLSE III 1.35, that will come with both 5.2 and 6.1. 6.1 is considerably better than 6.0. The difference between 5.1 and 5.2 is nearly insignificant.

I'm packing the program now and I'll release it asap.