Jump to content
3DCoat Forums

Blender Applink


haikalle
 Share

Recommended Posts

  • Advanced Member
Yes! I've had the same issue since forever. I always just set it to -1. But it's not just 3dCoat it's from other programs too. So I think it's less of an Applink issue and more of a blender issue. I was looking for an option in blender so that the default normal map value would be negative 1 but I couldn't find one. If you set the value in the applink I would be happy with that!

So a negative value for normal maps? In blender 2.5.x I suppose. Blender 2.49b uses positive normal values! And the funnier is that after saving there and open it in bl 2.5.x everything works fine. But when importing normals directly in 2.5x ... if lot of seams exist... just do some tests. The half islands have positive and the other half negative values. I reported so many times this major bug and the only I have is the offense of lot of blender users. You may remember me from blenderartists, Kay_Eva.

Just have in mind this issue. The funny thing begins when opening the 2.49 file in 2.5 and see that a "camera" value is in place of "tangent" map. ????

As you did, as many other users, use to reply to my complains... This blender 2.5.x is in beta stage, better use 2.49b, as some non existent old stable version of 3dcoat in place of latest builds as well.

I had to say these to you, to Haikalle and to all others that loosing their time with extremely buggy apps. And still keep offending me when I say the truth. That you can't work as an artist like this!

Link to comment
Share on other sites

  • Advanced Member

Of course, I never mentioned that haikalle offended me. You Kay_Eva offended me me once, in blenderartists.org but its OK.

I just said that you're loosing your time being in the middle of two buggy apps. Especially blender. Don't misunderstand me, you were already talking about positive and negative values of normal maps, in blender 2.5.x. Well, do some tests first. ITS BROKEN!!!

Try 2.49b instead, it works and positive values is what you have to use. Just save and open in 2.5, it works now. ( In bl 2.44 normals were negative BTW)

Kay_Eva, sorry for this, you offended me when you tried to remind me that all these versions are betas and I should avoid working with them. Have you realized what about you were talking? There aren't any other versions except betas!

So , no more offenses, you know that I'm right saying that an artist can't work like this. Don't try to fix this problem with normals (positive or negative), this will change in the future. Its not working anyway.

There's a kind of ancient temple in the gallery 3dc room (mine) . Half of normal maps are wrong. Its a blender 2.5.x render.

A simple test is to use lot of seams-islands in the middle of areas that nobody likes to use them this way. OK its a test anyway.

I did same tests using zbrush, the same wrong results on blender 2.5x

All "tangent" mode of course.

Link to comment
Share on other sites

  • Advanced Member

I personally recode the blender scripts myself when the API changes and I need them before the script author has time to do the recode themselves. But I happen to be a bit code savvy, and can totally sympathize with artists who arent able to adapt scripts to Apps with changing APIs. But I have to say that I absolutely, personally, enjoy my move to Blender. Even though I still use other 3D apps often enough, I still end up going back to Blender for my "general" 3D app of choice, but thats just me. No disrespect to other 3D suites, they all have their strengths(Max/Maya/XSI)/LW). That being said, I use ZBrush and 3DCoat for my hi-res-sculpting/texture-intensive work of course.

and, to the point of this thread, THIS SCRIPT ROCKS!!!! Makes the Blender 3DC workflow sooooooo fluid, thanks big time for it!

Link to comment
Share on other sites

  • Applink Developer

Thanks guys. Really glad to hear that.

Right now 3d-coat->blender workflow dosen't work at the moment. There is new features coming into blender API to make it better so I will wait until they are coded into API.

Meanwhile blender->3d-coat->blender workflow is working.

Link to comment
Share on other sites

  • Reputable Contributor

Hey haikalle, this is a off topic, but you aren't accepting PM's and I know of no other way to communicate with you....Just who assigned you charge over my bug reports on Mantis? Undo whatever you did. I took the time to not only post the issues, but to screen record it and send Andrew the work file. I would imagine Javis, Chris, Phil or anyone else would be pretty upset if you did the same to one of theirs.

I had them assigned to Andrew, so if he wants to remove or close a report, that's his call. In the future, please contact me first if you have a problem with one of mine.

Link to comment
Share on other sites

  • Applink Developer

Hi! I think that you are talking about this issue:

http://www.3d-coat.com/mantis/view.php?id=148

I closed it because it included several bugs in one report. It would be easier for Andrew if one bug report would include one bug. Andrew said that some of those bugs

are already fixed, so it makes that issue even more confused. It would be better to separate them because then we can discuss about a specific bug.

I have tried to keep mantis system clean. I have closed some other assigned issues because they were fixed or something else. I think that in the future it would be better to change the system that the person who reports the issue can't assing it. It has to be assigned by someone else. It's better to assign bugs to Andrew if they are confirmed by two or more users. I have to admit that I have assigned my issues to Andrew too, but this maybe is not the best way to do it...

I will open your thread again, but I hope that you have time to separate them at some point. Sorry about this.

Link to comment
Share on other sites

  • Reputable Contributor

Hi! I think that you are talking about this issue:

http://www.3d-coat.com/mantis/view.php?id=148

I closed it because it included several bugs in one report. It would be easier for Andrew if one bug report would include one bug. Andrew said that some of those bugs

are already fixed, so it makes that issue even more confused. It would be better to separate them because then we can discuss about a specific bug.

I have tried to keep mantis system clean. I have closed some other assigned issues because they were fixed or something else. I think that in the future it would be better to change the system that the person who reports the issue can't assing it. It has to be assigned by someone else. It's better to assign bugs to Andrew if they are confirmed by two or more users. I have to admit that I have assigned my issues to Andrew too, but this maybe is not the best way to do it...

I will open your thread again, but I hope that you have time to separate them at some point. Sorry about this.

Actually, in that case, since they were all inter-related, and I had already spent a lot of time logging other bugs (I had been ambushed with about a dozen bugs or so in the same day...it took the better part of the morning to try and log them all), I decided to keep it in one video and one work file. I understand where you are coming from, in a general sense, but I wouldn't presume to close one of yours or other 3DC contributors/administrators. I have a presentation to make later this month and it's critical that those bugs get fixed beforehand, so as not to crop up in the process and reflect very poorly upon the software.

I asked Andrew what version he wanted to use for the presentation and he stated .13 was the most stable. Not if these issues are swept under the rug.

Link to comment
Share on other sites

  • Applink Developer

Ok. And I can test these issues too and put some feedback to report. Andrew said that he has already fixed some bugs. I think it would be better to check all

these bugs and see if some issues are fixed and put some feedback if they are or not.

Link to comment
Share on other sites

  • Reputable Contributor

Ok. And I can test these issues too and put some feedback to report. Andrew said that he has already fixed some bugs. I think it would be better to check all

these bugs and see if some issues are fixed and put some feedback if they are or not.

Good idea..I'm rechecking them now.

EDIT: I re-checked, and if you can test the UV's on your end, for a good 5 minutes or so, see if they don't occur for you as well. Thanks.

Link to comment
Share on other sites

  • Advanced Member

Sorry I'm posting this here, haikalle, I could pm you about this.

This is also related with 3dc and blender. It seems that resulting normal maps from baking voxel relief come as positive values in blender. Normal maps from depth painting in PPP come negative!!!

With or without an applink use. Now I'm really confused. Is this related with the number of islands or the way these islands are unwrapped, I really don't know. From the first time I was using 3dcoat I had this problem. Sometimes these work sometimes not.

Link to comment
Share on other sites

  • Member

Hi guys,

Just saw this thread and saw that people are having a few issues with blender normal maps(inverted relief or visible seams).

I haven't tried the applink yet but have extensive experience with using 3dcoat and blender in conjunction with each other.

The point I want to make is on the how blender treats normal maps compared to Maya and Max/LW(3dcoats natively supported export options).

If you're exporting max/LW style normal maps from 3d coat(changable from the preferences) you will need to invert the red channel of those images in photoshop.

If exporting maya normal maps you will need to invert both the green and red channels.

I'm not sure if this applink already does this and peoples issues are unrelated. I've been doing this manually for a wide range of models for a while, and acheived perfect results

every time.

Hope this helps with development.

Link to comment
Share on other sites

  • Advanced Member

@Haikalle

@kay_Eva

@Heath_3d

You have so wrong about normal maps values in blender that my mouth is open.

HOW MANY times should I say that "do some tests first!!!!" ?

I have to see negative values working from blender 2.44

As for inverting red channel of LW-max style exporting, well I keep reading this here from time to time, lets finish this myth, its happening in your dreams my friends.

All this confusion... there's a reason for this. If you use lot of islands (more than ~ 15) then blender has a great BUG (especially 2.5x). Some of the islands will be inverted.

So, haikalle, please, POSITIVE VALUES ON NORMAL MAPS.

See the test I'm posting, I used ugly seams on a 1024 x 1024 normal map. Please do your tests if you like.

As you may see, positive values gave a nice result and seams are almost invisible in this ugly case.

I checked both bl 2.49b and bl 2.56 (same results)]

Please check it, under different OS, here OSX builds.

I also checked zbrush normal map exports, exactly the same results.

The link to download the blend file and the maps

http://www.mediafire.com/file/1fqitpc7jilx2cc/cubetestBlend.zip

Link to comment
Share on other sites

  • Member

@michalis

Sorry to open your mouth michalis.

In regards to doing tests, it goes without saying. My point is purely based on using these two apps together in a production environment.

In regards to this blender bug you mentioned, could you point me to a thread, bug report or any other URL regarding it? It would be

genuinely appreciated.

In regards to the inverted channel "myth", here's a quote right out of the blender manual. Warranted it's talking about the normal maps generated by blender,

but it stands to reason that blender would generate normal maps it could then use.

"Blender's normal map baking tool is different from other applications, such as xNormal and Maya.

Both of these are widely used in the games industry and are already production proven (meaning the way they bake their normal maps is more commonly used and can be regarded as a standard).

To convert a Blender normal map to an xNormal, you need to invert the Red and Green channels."

I'm not wanting to start an argument or anything, it's just that my experiences with blender normal maps matches what the manual tells me to expect.

That said, I opened your file in blender 2.49b and it seemed to work off the bat, so kudos for finding a method that works for you.

Link to comment
Share on other sites

  • Advanced Member

I tested what's happening with 3dcoat-max/LW normal maps. As I can see, it works without any inversion. Positive values.

If you have any tests, please post some, we have to make it clear. We could start a blender dedicated thread. What do you say?

Link to comment
Share on other sites

  • Member

Well, I'll be a monkeys uncle.

I did a couple more tests and it appears that max/LW export works fine..... Sort of.

On most of the work we've done there are many materials on individual objects. In this case I noticed all but one material

use normal maps with the color orientations as I described above. The one that doesn't, uses the max standard by the looks of things.

I'd say that at some point someones decided to bring blender normals into line with Max and Lightwave but there's some legacy issues/bugs.

This would explain why some people have been having problems while others haven't.

As a disclaimer, I only tested this on blender 2.49 and 2.54 so am not sure as to the consistency of the issue.

Also, I tried splitting an object up so that it's individual faces were different UV shells, but didn't notice any problems occurring with

more than 15 UV shells.

BTW our workflow here has people working on different object parts and then joining them together in blender.

Hope this helps in some way with dev.

cheers

Link to comment
Share on other sites

  • Advanced Member

Well, I'll be a monkeys uncle.

I did a couple more tests and it appears that max/LW export works fine..... Sort of.

On most of the work we've done there are many materials on individual objects. In this case I noticed all but one material

use normal maps with the color orientations as I described above. The one that doesn't, uses the max standard by the looks of things.

I'd say that at some point someones decided to bring blender normals into line with Max and Lightwave but there's some legacy issues/bugs.

This would explain why some people have been having problems while others haven't.

As a disclaimer, I only tested this on blender 2.49 and 2.54 so am not sure as to the consistency of the issue.

Also, I tried splitting an object up so that it's individual faces were different UV shells, but didn't notice any problems occurring with

more than 15 UV shells.

BTW our workflow here has people working on different object parts and then joining them together in blender.

Hope this helps in some way with dev.

cheers

My monkey is an uncle too. :D

Thanks for testing.

Link to comment
Share on other sites

  • Advanced Member

All newer blender builds from graphicall have a new bumps support algorithm.

Some interesting results this way as displacement maps can be used as nor or bump maps. Similar to modod501 new displ methods.

http://www.3d-coat.com/forum/index.php?showtopic=6894&view=findpost&p=60300

Blender also supports direct realtime preview bump painting as well

http://blenderartists.org/forum/showthread.php?t=207874

Link to comment
Share on other sites

  • Advanced Member

And here the explanation of all my confusion.

Newer blender builds aren't compatible with older blender projects!!!

I tested r35474 (3dc applink included) using a groboto test mesh, painted depth in 3dcoat and asked for 'maya compatible' normal maps export. No seams or inverted normal are visible.

Its fine so far, what I'll do with my older blender projects is another matter. Even a simple obj append from there ends with a complete nor-maps mess.

Link to comment
Share on other sites

  • 1 month later...
  • Member

Hi Haikalle

I made some tests. Sometimes it's work and anothers time not. I'm obliged to restart 3DC each time.

For my tests, Once I used a simple cube = doesn't work at all

Then I used a sphere = Blender ->3DC (I take a subdivision about 8k). So I practice some test with color and depth.

Thx

Jean

Link to comment
Share on other sites

  • Applink Developer

Hi!

I did quick test and it seams that old bug has woke up. You can try this if this helps. When you bring a mesh into 3d-coat.

Go to uv room and press unwrap and then apply uv-set. There is this issue that one of the uv islands is very small. That would explain bad

resolution. Can you check if there is one uv island very small.

Link to comment
Share on other sites

  • Carlosan changed the title to Blender 2.83 Applink

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...