[Tea Monster]

Just to the maintainers of the pack. Polymer IS capable of coloured specular maps. Please don't vandalise the look of the pack just to save a few Mb. Did any of you maintain copies of the original files?

[Mark]

If noting else, you might be able to download an older version of the HRP from the website? Or I'm sure NightFright has it backed up somewhere. I might even have it on some backups of old projects. I guess it depends on how long ago somebody messed with the textures.

[Roma Loom]

There were quite a few color specmaps in hrp, around 3 or 5 iirc...

[Tea Monster]

Not any more

Steeevie had at least one up there too.

[LeoD]

Tea Monster, on 02 April 2016 - 07:13 AM, said:

Just to the maintainers of the pack. Polymer IS capable of coloured specular maps. Please don't vandalise the look of the pack just to save a few Mb. Did any of you maintain copies of the original files?

Any repository contribution since 2009-05-17 can be restored. Please name a specific file that has been "vandalized" so the change can be tracked down. IIRC quite a number of specmaps has always been greyscale.

[Tea Monster]

Check Roma's and Steeevies contributions.

[Roma Loom]

Mine were always grey 8 bit PNG since I haven't bothered and most of them textures had usual metallic surface. Also my goal was quantity over quality. Shame on me :/

[TON]

Hendricks266, on 03 December 2015 - 11:25 PM, said:

Assuming Roma still has all the source models from his Polymerized textures (which I believe to be the case), how much work would it take to get them into shape for PBR?

Tea Monster, on 03 December 2015 - 11:37 PM, said:

For PBR, You take all the AO out of the diffuse texture, and create a 'metalness' map or a gloss map, depending on what sort of material you are using. Currently Polymer supports gloss maps, but I don't think anyone uses them. The AO is baked separately and added to the diffuse as an overlay before saving out to texture, so, in theory, creating the diffuse wouldn't be that difficult - as long as he still has his work files, and if he wants to go through and spend the time to re-jig them all.

TL:DR - you'd have to ask him.

I'd like PBR, but I think optimising is probably more important at the moment.

The main reason that people want to mod something else other than EDuke is that the engine dosen't have so many basic modern features. Optimising Polymer is great, but it dosen't give EDuke multiplayer, particles, shaders, etc. etc. ect. I realise that's way beyond what icecoldduke is doing at the moment, but just putting that out there. I would personally rather the effort go into the whole package of modernising EDuke, rather than just race to put PBR in there. JMHO.

Roma Loom do you have it?

[Tea Monster]

I asked Steeevie and he told me that the fire hydrant had a tinted spec map. I'm not sure about his other models.

[Steveeeie]

Looking now at my source and my hydrant was grey scale, I remember now the colour ones causing a weird Pearlescent effect so I made it grey scale.

[Roma Loom]

TON, on 04 April 2016 - 03:37 AM, said:

Roma Loom do you have it?

Yep

[LeoD]

Tea Monster, on 02 April 2016 - 04:14 PM, said:

Check Roma's and Steeevies contributions.

Roma Loom, on 03 April 2016 - 01:29 PM, said:

Mine were always grey 8 bit PNG since I haven't bothered and most of them textures had usual metallic surface. Also my goal was quantity over quality. Shame on me :/

Steveeeie, on 04 April 2016 - 10:55 AM, said:

Looking now at my source and my hydrant was grey scale, I remember now the colour ones causing a weird Pearlescent effect so I made it grey scale.

Case closed or more files to deal with?

[Tea Monster]

Sigh. That looks like it.

[OpenMaw]

The thing to remember when you're dealing with color specs is you also have to take into account brightness/what the material is. If it's a highly metallic surface you don't want to do a 255 0 0 RGB color for the spec. You'd want something that's faintly, just a hint of, red. If it's a duller specular, you'd want a darker, more saturated color.

[Mark]

A few people have asked me for the fem10 replacement model I made using the Elexis model. I had another request yesterday so I thought I would include it here. I never finished it or tested it in the game. I was going to tweak positions of the girls and add a few frames of animation but I never got around to doing it. Hopefully one of the people that requested it can make use of it.

edit: Sorry, for some reason I can't attach the 10.7MB zip file. I choose it and click to attach, and when it finishes uploading I get a message saying no file attached. So, I guess no model for now.

[1337DirtAlliance]

Mark., on 10 April 2016 - 05:09 AM, said:

A few people have asked me for the fem10 replacement model I made using the Elexis model. I had another request yesterday so I thought I would include it here. I never finished it or tested it in the game. I was going to tweak positions of the girls and add a few frames of animation but I never got around to doing it. Hopefully one of the people that requested it can make use of it.

edit: Sorry, for some reason I can't attach the 10.7MB zip file. I choose it and click to attach, and when it finishes uploading I get a message saying no file attached. So, I guess no model for now.

Try using google drive or drop box.

[Tea Monster]

Just a heads up. Very few if any of the older HRP models have a specpower/specfactor line in their defs. This is why they all look like Christmas tree ornaments in Polymer.

Even if they don't have specular maps, you can still define a general level of specularity for the model. It will be the same for the entire surface of the model, but you can control how shiny it is, even without a specular map. Same for the glossiness of the model.

As an example, adding specpower 0.2 specfactor 0.2 to one of the characters, even without any specular maps, made them look a lot more human and less ridiculous.

[Hendricks266]

Hey Plagman, is there a reason why the defaults for these values isn't 0?

[Plagman]

No particular reason, it's not like the default value was ever intended to be used. It seems weird that someone would add an asset without setting the most basic material properties for it.

[Mark]

Maybe a lot of the models were created by a Polymost user?? Just guessing since I have no idea who or when most of those problematic HRP models were made.

[Tea Monster]

Most of the models were made before polymer came out. For whatever reason, nobody added the spec lines in the defs for the old models.

The majority were constructed before we had a renderer with next gen features. They don't have normal or spec maps.

This is why I think its foolish to create the NG renderer to work with the old assets.

[icecoldduke]

Tea Monster, on 16 June 2016 - 05:57 AM, said:

Most of the models were made before polymer came out. For whatever reason, nobody added the spec lines in the defs for the old models.

The majority were constructed before we had a renderer with next gen features. They don't have normal or spec maps.

This is why I think its foolish to create the NG renderer to work with the old assets.

I agree. If there is bandwidth in the community to make new HRP models, that would be the best case scenario.

[LeoD]

Tea Monster, on 16 June 2016 - 05:57 AM, said:

Most of the models were made before polymer came out. For whatever reason, nobody added the spec lines in the defs for the old models.

Apparently no one of those who knew about the technical details told the HRP maintainers ...
So should I go ahead and add "specpower 0.2 specfactor 0.2" to the Polymost model definitions? Opinions, anyone?

Tea Monster, on 16 June 2016 - 05:57 AM, said:

This is why I think its foolish to create the NG renderer to work with the old assets.

icecoldduke, on 16 June 2016 - 06:43 AM, said:

I agree. If there is bandwidth in the community to make new HRP models, that would be the best case scenario.

In reality I expect some Advanced PolymerNG++forte renderer to be ready before even half the HRP is remade, unfortunately.

[Mark]

Instead of adding all those def lines would it be better to change Eduke's default values from whatever they are to the .2 and .2 or whatever is agreed upon? that way the default is a usable value if no def line is present, and you can still over-ride the default with your own line as we do now. But then we run the risk of vastly differing opinions on what default looks best.

[LeoD]

Mark., on 19 June 2016 - 06:25 PM, said:

Instead of adding all those def lines would it be better to change Eduke's default values from whatever they are to the .2 and .2 or whatever is agreed upon? that way the default is a usable value if no def line is present, and you can still over-ride the default with your own line as we do now.

Yeah, that makes more sense. (Although it might irritate people when upgrading EDuke32 -> cfg option.)

Mark., on 19 June 2016 - 06:25 PM, said:

But then we run the risk of vastly differing opinions on what default looks best.

If we can agree that some non-zero default within a certain range looks better, that would still be a good starting point.

[Mark]

yep.

[Plagman]

I'm not sure if changing the default from under's people existing content would be the right approach here, and doing a blanket change of all the HRP's models without looking at before/after doesn't seem right either.

[Danukem]

Plagman, on 20 June 2016 - 11:46 AM, said:

I'm not sure if changing the default from under's people existing content would be the right approach here,

I'm usually among the first to complain about this sort of thing, but in this case I don't think it's a problem. If someone released stuff without bothering to specify basic properties for their assets, then they are going to have to deal with the default values and the possibility they could change. Changing those is way less intrusive than other things I have seen in this community: for example, when huge changes were made to shading, making several years worth of maps almost unplayable with the default values. In this case we are talking about a relatively rare kind of content in which the content providers failed to do their job properly.

[LeoD]

Trooper Dan, on 20 June 2016 - 12:05 PM, said:

I'm usually among the first to complain about this sort of thing, but in this case I don't think it's a problem. If someone released stuff without bothering to specify basic properties for their assets, then they are going to have to deal with the default values and the possibility they could change. Changing those is way less intrusive than other things I have seen in this community: for example, when huge changes were made to shading, making several years worth of maps almost unplayable with the default values.

Yeah, and that might have been less bad if the all the important information for mappers and players hadn't been buried in some techno-babble thread which only the hardcore nerds are able/willing to follow.

Trooper Dan, on 20 June 2016 - 12:05 PM, said:

In this case we are talking about a relatively rare kind of content in which the content providers failed to do their job properly.

At least in the props department there are quite a number of pre-Polymer models affected. I can come up with some update-pack style zip with "specpower 0.2 specfactor 0.2" added to the Polymost model definitions, so the followers of this thread can check it out. Although a patched EDuke32 executable would be nice, too.

[LeoD]

Trooper Dan, on 22 June 2016 - 08:42 PM, said:

Yes, but would changing the default affect them in a bad way?

Not much tested yet: brainless (almost) addition of specular defaults (0.20) to all models without specific entry.
 duke3d_hrp_test_spec_defaults.zip (33.62K)
Number of downloads: 206