It is currently Fri Oct 20, 2017 9:11 am

All times are UTC - 5 hours




Post new topic Reply to topic  [ 107 posts ]  Go to page Previous  1 ... 4, 5, 6, 7, 8  Next
Author Message
 Post subject: Re: 3DSimED 3.1i
PostPosted: Thu Jul 13, 2017 2:32 pm 
Offline
US Pits Crew
US Pits Crew

Joined: Thu Sep 01, 2005 9:58 am
Posts: 1273
Location: Barnet, London
Thanks for this, I hadn't noticed and I think I know what has gone wrong - I shall try to fix for the next release.

Cheers!

Dave N.

br555 wrote:
Last one for a while, I promise. Forgive me if this is modeling ignorance speaking, but I'm curious so here goes:

I'm not sure if FH3 is doing something especially strange for steering wheels in particular, but the normals (I'm guessing it's a normals issue but I'm no modeler) always appear the polar opposite of what they should be. Even a steering wheel and steering column fused together on the same submesh using the same material (as in the case of the '95 Cobra R pictured) are lit very differently. Has anyone noticed this? I haven't seen this phenomenon on any other type of part so far, just steering wheels. :confuse:

[edit: it's definitely the imported normals, if I recalculate the normals the part looks fine]

Thanks,
- Ben


Top
 Profile  
 
 Post subject: Re: 3DSimED 3.1i
PostPosted: Thu Jul 13, 2017 2:34 pm 
Offline
US Pits Crew
US Pits Crew

Joined: Thu Sep 01, 2005 9:58 am
Posts: 1273
Location: Barnet, London
The limit shouldn't be that small but it is there, yes. Thankfully, it is fairly easy to raise and should be in the next release.

Cheers

Dave N.

br555 wrote:
Possible Texture Browser bug:

I haven't nailed down the exact number yet, but if you do a batch conversion and you select more than around 250 (probably 256) textures, it will ignore/not process any textures beyond that number - Almost as if there is a fixed-size buffer or 8-bit unsigned integer counter variable lurking somewhere.

Also it would be nice if the output folder defaulted to the input folder, or at least remembered the last output folder selected. But I'm nitpicking - it's by far the best texture converter for FH3 work IMO.

Thanks,
- Ben


Top
 Profile  
 
 Post subject: Re: 3DSimED 3.1i
PostPosted: Thu Jul 13, 2017 2:35 pm 
Offline
US Pits Crew
US Pits Crew

Joined: Thu Sep 01, 2005 9:58 am
Posts: 1273
Location: Barnet, London
I think I know what is going here so it should be fixed for the next release.

Cheers!

Dave N.

br555 wrote:
Another potentially more serious issue relates to material names changing and/or losing information when exporting various ways...

- Maserati A6 when entire imported FH3 model is exported to FBX via plugin, the color swatch portion of the material name for many materials changes to #FFFFFFFF or #00000000 and the original color is lost. This does not happen if you do an old-school export to GMT, so the issue is either the FBX export or plugin exports in general but I haven't tried any other plugin exports. See attached (ignore the uppercase and loss of "#", that's a result of ZModeler)

- (no picture) The '53 Fiat 8V Supersonic model when exported to FBX using plugin has a different but possibly related material naming problem. It's material names consist of only the color swatch/texture name portion by itself (e.g. "rgba#8a8b8eff" or "bmw_stitch_001_diff_cd6276c9-fdb1-4dfc-a148-1389bceadb3a") instead of the the usual [material name]_[color swatch/texture name] combination we usually see. Also curious is that when texture names do appear, they are the long "hashed" Forza filename version instead of the shortened names we usually get. [edit: Oops, ignore the bit about shortened filenames, I was doing this myself. Also I forgot to mention that so far the Fiat 8V is the only car that has experienced this particular issue]

- (no picture) If you isolate any FH3 part to a new window and then export that to FBX, you also get just the color swatch portion of the material name. This issue does not come up if you export the isolated part to GMT for example.

Thanks,
- Ben


Top
 Profile  
 
 Post subject: Re: 3DSimED 3.1i
PostPosted: Thu Jul 13, 2017 2:37 pm 
Offline
US Pits Crew
US Pits Crew

Joined: Thu Sep 01, 2005 9:58 am
Posts: 1273
Location: Barnet, London
I think this because of a custom shader, they're the key to how Forza manages to render, not much I can do with this.

Cheers

Dave N.

br555 wrote:
Possible FH3 material-assignment bug on '53 Maserati A6GCS:
- Heatsink on kickpanel is assigned material screw_044_#010202FF
- Heatsink UVs appear to want an opaque, finely-tiled texture (see rendered screenshot, tiled @ 8x) instead of a 1:1 transparent
- Heatsink shares this material with the screws that do require a 1:1 transparent mapping, which has two implications -
a. It makes it impossible to separate the heatsink from the screws and give them both a material/texture mapping that works for each
b. It leads me to believe this is perhaps the wrong material assignment rather than FH3 using the screw texture in some novel way (i.e. custom shader)

Thanks,
- Ben


Top
 Profile  
 
 Post subject: Re: 3DSimED 3.1i
PostPosted: Thu Jul 13, 2017 3:01 pm 
Offline
Tirechanger
Tirechanger

Joined: Tue Oct 04, 2016 3:44 pm
Posts: 16
Dave Noonan wrote:
I think this because of a custom shader, they're the key to how Forza manages to render, not much I can do with this.

Cheers

Dave N.

br555 wrote:
Possible FH3 material-assignment bug on '53 Maserati A6GCS:
- Heatsink on kickpanel is assigned material screw_044_#010202FF
- Heatsink UVs appear to want an opaque, finely-tiled texture (see rendered screenshot, tiled @ 8x) instead of a 1:1 transparent
- Heatsink shares this material with the screws that do require a 1:1 transparent mapping, which has two implications -
a. It makes it impossible to separate the heatsink from the screws and give them both a material/texture mapping that works for each
b. It leads me to believe this is perhaps the wrong material assignment rather than FH3 using the screw texture in some novel way (i.e. custom shader)

Thanks,
- Ben

Fair enough. While I don't necessarily need two identical (as far as we can tell) materials defined for the same part, it would be nice to at least split the UV maps when the Forzabin data gives you any indication that it might be using that material in more than one way (different Forza shader name for example). That way I can at least manually create a new material and re-assign one of the UV maps.


Top
 Profile  
 
 Post subject: Re: 3DSimED 3.1i
PostPosted: Thu Jul 13, 2017 3:44 pm 
Offline
Tirechanger
Tirechanger

Joined: Tue Oct 04, 2016 3:44 pm
Posts: 16
Dave Noonan wrote:
The limit shouldn't be that small but it is there, yes. Thankfully, it is fairly easy to raise and should be in the next release.

Cheers

Dave N.

br555 wrote:
Possible Texture Browser bug:

I haven't nailed down the exact number yet, but if you do a batch conversion and you select more than around 250 (probably 256) textures, it will ignore/not process any textures beyond that number - Almost as if there is a fixed-size buffer or 8-bit unsigned integer counter variable lurking somewhere.

Also it would be nice if the output folder defaulted to the input folder, or at least remembered the last output folder selected. But I'm nitpicking - it's by far the best texture converter for FH3 work IMO.

Thanks,
- Ben

I was recalling the number from many weeks ago, but must have got it wrong - I'm sure it's the limit you're referring to, thanks. It would also be nice if it didn't fail silently - Maybe inform the user if they select more items than the buffer can handle. Sorry if I sound pedantic but I really do appreciate the attention being paid to these small issues.


Top
 Profile  
 
 Post subject: Re: 3DSimED 3.1i
PostPosted: Thu Jul 13, 2017 4:29 pm 
Offline
Tirechanger
Tirechanger

Joined: Tue Oct 04, 2016 3:44 pm
Posts: 16
Another trivial FH3 import tweak - The current wheel diameter scaling is slightly too small because it's based on the rim mesh bounds instead of the tire bead mating surface. For example, the mesh bounds/lip-to-lip measurement of a "17 in." rim will be closer to 18". Fortunately most rim lips are around the same size so it should be fudge-able by just adding the extra inch

Likewise the rim width measurement should technically not include the lip/flange either, but I haven't compared those yet as they look very close already


Top
 Profile  
 
 Post subject: Re: 3DSimED 3.1i
PostPosted: Thu Jul 13, 2017 9:10 pm 
Offline
Tirechanger
Tirechanger

Joined: Tue Oct 04, 2016 3:44 pm
Posts: 16
br555 wrote:
Fair enough. While I don't necessarily need two identical (as far as we can tell) materials defined for the same part, it would be nice to at least split the UV maps when the Forzabin data gives you any indication that it might be using that material in more than one way (different Forza shader name for example). That way I can at least manually create a new material and re-assign one of the UV maps.

I may have an idea of what's happening here. The attached image shows a pedal that is clearly expecting two different UV mappings/materials:
The UV mapping in the center of the pedal (marked in red) is expecting a 1:1 mapping of the "airvents" texture and the remaining UV mappings (marked in green) require a tiled texture and clearly "airvents" is not it. If you look to the right of the scene, you'll see in the modelbin data what appears to be a second reference to a material of the same name (i.e. leather_embossed_alpha) but with a different texture which happens to be "pyramid_005" which looks like a no-brainer match given the (few) pictures of Jalpa pedals I've found online.

3DSimEd is currently merging both of these UV maps into the "airvents" texture/material, which isn't ideal. Now I'll admit that when faced with more than one choice of material, it's impossible to determine which UVs belongs to what material, but I'm hoping that it is possible to keep the UV maps from getting co-mingled in such a situation.


Attachments:
JalpaPedals.png
JalpaPedals.png [ 537.88 KiB | Viewed 387 times ]
Top
 Profile  
 
 Post subject: Re: 3DSimED 3.1i
PostPosted: Thu Jul 13, 2017 10:51 pm 
Offline
Tirechanger
Tirechanger

Joined: Tue Oct 04, 2016 3:44 pm
Posts: 16
Here's another example where a single submesh has two UVs (marked green and red) merged into a single material, where Forza would be creating two materials. They have the same material "name", color and even the same texture set but one is intended to be transparent (the embossed script) and the other opaque (the vent).

Seeing the same material name/texture set being listed more than once in the modelbin tells me that Forza intends to setup multiple materials each using that same texture set in some way. We may have no idea how they might differ, but if we don't setup more than one material then our UVs end up getting mushed into the same material which creates the conundrum pictured. Hope this helps.


Attachments:
95CobraRDash.png
95CobraRDash.png [ 661.47 KiB | Viewed 384 times ]
95CobraRDash2a.png
95CobraRDash2a.png [ 625.45 KiB | Viewed 384 times ]
Top
 Profile  
 
 Post subject: Re: 3DSimED 3.1i
PostPosted: Thu Jul 20, 2017 1:38 pm 
Offline
Tirechanger
Tirechanger

Joined: Thu Jun 01, 2017 1:26 pm
Posts: 18
br555 wrote:
Possible FH3 material-assignment bug on '53 Maserati A6GCS:
- Heatsink on kickpanel is assigned material screw_044_#010202FF
- Heatsink UVs appear to want an opaque, finely-tiled texture (see rendered screenshot, tiled @ 8x) instead of a 1:1 transparent
- Heatsink shares this material with the screws that do require a 1:1 transparent mapping, which has two implications -
a. It makes it impossible to separate the heatsink from the screws and give them both a material/texture mapping that works for each
b. It leads me to believe this is perhaps the wrong material assignment rather than FH3 using the screw texture in some novel way (i.e. custom shader)

Thanks,
- Ben

Unfortunately this has already been caused by forza itself: https://puu.sh/wOET3/fced374cee.jpg


Top
 Profile  
 
 Post subject: Re: 3DSimED 3.1i
PostPosted: Thu Jul 20, 2017 2:50 pm 
Offline
Tirechanger
Tirechanger

Joined: Tue Oct 04, 2016 3:44 pm
Posts: 16
Tomatosauce288 wrote:
br555 wrote:
Possible FH3 material-assignment bug on '53 Maserati A6GCS:
- Heatsink on kickpanel is assigned material screw_044_#010202FF
- Heatsink UVs appear to want an opaque, finely-tiled texture (see rendered screenshot, tiled @ 8x) instead of a 1:1 transparent
- Heatsink shares this material with the screws that do require a 1:1 transparent mapping, which has two implications -
a. It makes it impossible to separate the heatsink from the screws and give them both a material/texture mapping that works for each
b. It leads me to believe this is perhaps the wrong material assignment rather than FH3 using the screw texture in some novel way (i.e. custom shader)

Thanks,
- Ben

Unfortunately this has already been caused by forza itself: https://puu.sh/wOET3/fced374cee.jpg

Nice catch, thanks! I did try to peek at the model in FH3 showroom but forgot you could open the doors :dunce:
Well I guess even the Forza modelers are not perfect


Top
 Profile  
 
 Post subject: Re: 3DSimED 3.1i
PostPosted: Fri Jul 21, 2017 2:12 pm 
Offline
Tirechanger
Tirechanger

Joined: Thu Jun 01, 2017 1:26 pm
Posts: 18
br555 wrote:
Tomatosauce288 wrote:
br555 wrote:
Possible FH3 material-assignment bug on '53 Maserati A6GCS:
- Heatsink on kickpanel is assigned material screw_044_#010202FF
- Heatsink UVs appear to want an opaque, finely-tiled texture (see rendered screenshot, tiled @ 8x) instead of a 1:1 transparent
- Heatsink shares this material with the screws that do require a 1:1 transparent mapping, which has two implications -
a. It makes it impossible to separate the heatsink from the screws and give them both a material/texture mapping that works for each
b. It leads me to believe this is perhaps the wrong material assignment rather than FH3 using the screw texture in some novel way (i.e. custom shader)

Thanks,
- Ben

Unfortunately this has already been caused by forza itself: https://puu.sh/wOET3/fced374cee.jpg

Nice catch, thanks! I did try to peek at the model in FH3 showroom but forgot you could open the doors :dunce:
Well I guess even the Forza modelers are not perfect

Haha the forza models are so bugged xD. Since turn ten just recycles the model from previous forzas and just adds new parts to the old model. Which results in holes and polybugs etc.


Top
 Profile  
 
 Post subject: Re: 3DSimED 3.1i
PostPosted: Mon Jul 24, 2017 2:51 pm 
Offline
Tirechanger
Tirechanger

Joined: Thu Jun 01, 2017 1:26 pm
Posts: 18
encountered another bug: https://puu.sh/wS1C3/528ed8605f.png. That also happens when I import higher LOD levels: https://puu.sh/wS1Dp/903c9becb9.png


Top
 Profile  
 
 Post subject: Re: 3DSimED 3.1i
PostPosted: Mon Jul 24, 2017 3:05 pm 
Offline
Tirechanger
Tirechanger

Joined: Thu Jun 01, 2017 1:26 pm
Posts: 18
almost the same bug on the amg gt: https://puu.sh/wS2hU/7fd9039448.png and it doesn't even appear on the gtr: https://puu.sh/wS2kp/2b92cc13d1.png

SLS AMG: https://puu.sh/wS2tv/d249bd2ec2.png

and alot of other mercedeces too like the SLK55: https://puu.sh/wS2BH/867ddd11f1.png


Top
 Profile  
 
 Post subject: Re: 3DSimED 3.1i
PostPosted: Mon Jul 24, 2017 3:29 pm 
Offline
Tirechanger
Tirechanger

Joined: Thu Jun 01, 2017 1:26 pm
Posts: 18
ok most of these hood badge bugs are just texture bugs i guess: https://puu.sh/wS3nF/749525ac33.png


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 107 posts ]  Go to page Previous  1 ... 4, 5, 6, 7, 8  Next

All times are UTC - 5 hours


Who is online

Users browsing this forum: Google [Bot] and 3 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group