CGTalk > Software > Autodesk 3ds max
Login register
Thread Closed share thread « Previous Thread | Next Thread »  
 
Thread Tools Search this Thread Display Modes
Old 02-09-2013, 03:36 AM   #1
Want2be3D
Veteran
Omar
London, United Kingdom
 
Join Date: Apr 2008
Posts: 39
Camera Map (WSM), Projection and Unwrap UVW modifier issue with footage...

Hi everyone, how are you all? I dont know if this is the right place to post this but I was wondering if someone can help me in regards to the Camera Map (WSM) and Unwrap UVW modifiers.

Here I am working with some homemade footage I shot quickly with an actor only 10secs long with it looking like this:

http://www.vfxtalk.com/attachment.php?attachmentid=13128&d=1359253374

...and managed to track itís head with a simple piece of head geometry that I made using FaceGen using reference images of the actor which went quite well after a few attempts of readjusting the geometry and retracking. Itís not perfect but I think it will do :

http://www.vfxtalk.com/attachment.php?attachmentid=13129&d=1359253676

Iíve exported the shot to a maxscript which I can open in 3ds max itself, but before doing so, I set the necessaries like setting the correct framerate and resolution in the render setup. I run the script and everything loaded up (i.e. the head with movements) and then loaded the footage (image sequence) into the background of the camera viewport. Everything seems to match.

Now, from learning how to Camera Map through tutorials around, I seem to have successfully mapped the footage as a texture to the geometry with all the UV coordinates matching on every frame. This was accomplished using Camera Map (WSM):

http://www.vfxtalk.com/attachment.php?attachmentid=13130&d=1359255758

This is what I sort of wanted to accomplish originally and thought i was nearly there, however this is what I get stuck. I want to take the geometry with the animated texture and render it out as an animated UV Map. This is usually done by unwrapping the model with coordinates using Unwrap UVW modifier. BUT...when I apply this, it never ever stack itself ontop of the Camera Map (WSM) modifier which is proving to give me quite a headace now and even when I apply it with the normal CAmera Map modifier (which does stack ontop of it) I dont get the footage flattened either as a texture in the UV editor.

I believe Iíve gotten really close but everytime I do so, I can never get the texture to be rendered out or even get viewable when flattening the model.

This is where I am at the moment, I dont know if I am going through the correct steps or if I am missing something once again, but this is what I am trying to somewhat achieve:

http://vimeo.com/41770090

How would I go about this may I ask? Apologize if you may find this a tad long
 
Old 02-10-2013, 02:25 PM   #2
Want2be3D
Veteran
Omar
London, United Kingdom
 
Join Date: Apr 2008
Posts: 39
Anyone at all may i ask? Im sorry if it was abit long or if wasnt clear on what i was saying. Please let me know.

Thanks!
 
Old 02-11-2013, 04:23 PM   #3
PiXeL_MoNKeY
Expert
 
PiXeL_MoNKeY's Avatar
Eric Craft
Dallas, USA
 
Join Date: Mar 2002
Posts: 5,668
WSM will always sit on top of the stack. You could create a Mesher Object and reference the data with the WSM and add standard modifiers on top of the Mesher.

The way you will want to do this is have the CameraMap WSM map to say Channel 2, and the Unwrap Modifier map to Channel 1. You can then do a RTT to Channel 1, which will bake the CameraMap data to the desired unwrap data.

-Eric
__________________
"The Evil Monkey hiding in your closet."
 
Old 02-12-2013, 08:34 AM   #4
Swahn
Expert
 
Swahn's Avatar
Daniel Swahn
Sweden
 
Join Date: May 2005
Posts: 344
Something that might help you is the use of the Camera Map Per Pixel map.
If I understand correctly you want to render out a custom unwrap for later editing?
Well, you should be able to create Camera Map PP, load the footage into it, create a new UV channel and unwrap on your tracked mesh, and set the Camera Map PP to use the same channel.

Once you do a Render To Texture your texture should go well into the new UV channel.
Is that what you seek?
 
Old 02-15-2013, 02:49 AM   #5
Want2be3D
Veteran
Omar
London, United Kingdom
 
Join Date: Apr 2008
Posts: 39
Quote:
Originally Posted by PiXeL_MoNKeY
WSM will always sit on top of the stack. You could create a Mesher Object and reference the data with the WSM and add standard modifiers on top of the Mesher.


Yeah, I am aware that a World-Space Modifier would always be ontop of the stack. The problem is that when I use the normal Camera Map modifier, the talent's head in the footage no longer matches the tracked head geometry except for the first frame only.

And is it okay if you can elaborate abit on when you talk about creating a 'Mesher Object' and referencing the data with the WSM? What proceedure is this called usually?

Quote:
Originally Posted by PiXeL_MoNKeY
The way you will want to do this is have the CameraMap WSM map to say Channel 2, and the Unwrap Modifier map to Channel 1. You can then do a RTT to Channel 1, which will bake the CameraMap data to the desired unwrap data.


Quote:
Originally Posted by Swahn
Something that might help you is the use of the Camera Map Per Pixel map.
If I understand correctly you want to render out a custom unwrap for later editing?
Well, you should be able to create Camera Map PP, load the footage into it, create a new UV channel and unwrap on your tracked mesh, and set the Camera Map PP to use the same channel.

Once you do a Render To Texture your texture should go well into the new UV channel.
Is that what you seek?


Yes!!! I believe this is what I seek indeed. However, I've never quite heard of the term Camera Map Per Pixel map. Isnt this technically the same as Camera Mapping?

To further answer your quesyion, I do want to render out a custom unwrap of the talent's face/head for later editing (I assume you mean painting or adding stuff to the animated face texture right?) and then putting it back onto talent's face/head in the footage. If that is the case, how do I go about it exactly may i ask? I cant quite figure out a workflow for it.

And yes, Render To Texture , although I am aware of this, isnt image quality or resolution a problem with this matter? This is something I assume from what I've read and heard so far as it relates to texture baking.

Please correct me if I am wrong! greatest apologize if sounding quite noobish, I highly appreciate your reply here! I look forward hearing back.

P.S. (belated) Happy "Commercialised Love" Day!
 
Old 02-15-2013, 08:51 AM   #6
Swahn
Expert
 
Swahn's Avatar
Daniel Swahn
Sweden
 
Join Date: May 2005
Posts: 344
Quote:
Originally Posted by Want2be3D

Yes!!! I believe this is what I seek indeed. However, I've never quite heard of the term Camera Map Per Pixel map. Isnt this technically the same as Camera Mapping?


The principle is the same, but the Camera Map PP will give you a 1:1 pixel accuracy compared to the original footage, rather than just creating an UV space projecting from a camera. You can compare projecting using a modifier and Camera Map PP respectively and you'll notice how the Camera Map PP will give you 1:1 accuracy, something that will come in handy if you seek to maintain as much of the original look as possible when rendering out to texture.

Quote:
Originally Posted by Want2be3D

To further answer your quesyion, I do want to render out a custom unwrap of the talent's face/head for later editing (I assume you mean painting or adding stuff to the animated face texture right?) and then putting it back onto talent's face/head in the footage. If that is the case, how do I go about it exactly may i ask? I cant quite figure out a workflow for it.


I would create a suitable UV space, which is centred around the area which you want to add detail to. By doing this the area you want to add detail too will remain as static as possible in the rendered texture. It will be easier to work with.

Quote:
Originally Posted by Want2be3D
And yes, Render To Texture , although I am aware of this, isnt image quality or resolution a problem with this matter? This is something I assume from what I've read and heard so far as it relates to texture baking.


Image quality is not a problem as long as you render out the texture in a high enough quality. There is virtually no limit of how large you render out the texture, better render out too large than too small so to speak. I think it's a good idea to render out a 2:1 image aspect texture (width twice the size of the height) as an UV map of a face is usually more wide than tall. But if you are using Vray you should be aware that there seem to be an old bug with the Render To Texture that will always output square textures. In such case it would make more sense to render out the raw texture with Scanline or MentalRay, and render the final material with Vray later on.
 
Old 02-15-2013, 04:36 PM   #7
PiXeL_MoNKeY
Expert
 
PiXeL_MoNKeY's Avatar
Eric Craft
Dallas, USA
 
Join Date: Mar 2002
Posts: 5,668
Quote:
Originally Posted by Want2be3D
And is it okay if you can elaborate abit on when you talk about creating a 'Mesher Object' and referencing the data with the WSM? What proceedure is this called usually?
Think of it as an in scene Xref procedure. Typically the Mesher is designed to reference particle systems to allow you to apply standard 3ds max modifiers to them. Basically it is a Dynamic/Procedural snapshot per frame of the referenced object in it's current World Space state. Read more on the Mesher here. Once you have your Head Mapped with the Camera Map WSM, you create a Mesher Object, Pick the Head as the referenced object, you would then hide the head, now apply any standard modifiers you want on top of the Mesher object.

-Eric
__________________
"The Evil Monkey hiding in your closet."
 
Old 02-20-2013, 02:36 AM   #8
Want2be3D
Veteran
Omar
London, United Kingdom
 
Join Date: Apr 2008
Posts: 39
Quote:
Originally Posted by Swahn
The principle is the same, but the Camera Map PP will give you a 1:1 pixel accuracy compared to the original footage, rather than just creating an UV space projecting from a camera. You can compare projecting using a modifier and Camera Map PP respectively and you'll notice how the Camera Map PP will give you 1:1 accuracy, something that will come in handy if you seek to maintain as much of the original look as possible when rendering out to texture.


I see then, so it seems this is something I shouldnt really at all then? I should've guessed it from its name 'Per Pixel' I suppose, hahaha But I wonder how it would be possible to compare with a modifier as I dont think there is one that would even allow me to get a UV Map of the footage (which was what I was struggling with at the start), hence why Camera Map PP is the way forth.

Quote:
Originally Posted by Swahn
I would create a suitable UV space, which is centred around the area which you want to add detail to. By doing this the area you want to add detail too will remain as static as possible in the rendered texture. It will be easier to work with.


Ahhh this is interesting, because the areas that I want to add detail to are usually moving noticably or subtly, also, as in the example screenshots in my first post, I originally wanted to do the whole face. In addition, in the video I linked previously:

http://vimeo.com/41770090

...from 19 secs. during the breakdown, the painted patches which helps cover the seams of the makeup show that it is moving along with the original footage in the UV map. I really find it hard to believe that it was hand animated!

Quote:
Originally Posted by Swahn
Image quality is not a problem as long as you render out the texture in a high enough quality. There is virtually no limit of how large you render out the texture, better render out too large than too small so to speak. I think it's a good idea to render out a 2:1 image aspect texture (width twice the size of the height) as an UV map of a face is usually more wide than tall. But if you are using Vray you should be aware that there seem to be an old bug with the Render To Texture that will always output square textures. In such case it would make more sense to render out the raw texture with Scanline or MentalRay, and render the final material with Vray later on.


I think I understand you here, and also that I do agree with you about rendering out a 2:1 image aspect texture as a UV map, this would give an artist more room to play and also not to worry about image degradation at the same time. I think this is something I shouldnt worry about either as my worry came from when viewing the textures from the viewport when doing projection or mapping and having to add segments (like on a Plane) inorder to get the best viewing quality, however it turns out from my knowlege nowthat this only applies to the Viewport only and not the actual render. An example I saw someone just clicking the render button and everything pretty much came out crisp

At the moment I will be using Scanline and Mental Ray...however I would like to ask also, could I just use the Scanline render to render out my face texture and the later edited face texture without having to worry about image quality (like as discussed on Camera Map PP)??

Quote:
Originally Posted by PiXeL_MoNKeY
Think of it as an in scene Xref procedure. Typically the Mesher is designed to reference particle systems to allow you to apply standard 3ds max modifiers to them. Basically it is a Dynamic/Procedural snapshot per frame of the referenced object in it's current World Space state. Read more on the Mesher here. Once you have your Head Mapped with the Camera Map WSM, you create a Mesher Object, Pick the Head as the referenced object, you would then hide the head, now apply any standard modifiers you want on top of the Mesher object.


Thanks Eric for that! Iw ill definitely look into it and try it out too! If I have any problems, will it be alright to also come back to you?
 
Old 05-13-2013, 02:56 AM   #9
Want2be3D
Veteran
Omar
London, United Kingdom
 
Join Date: Apr 2008
Posts: 39
Hi guys! How are you all??? Long time no hear!

I do sincerely aplogize for not updating you lot with my progress,I have been incredibly busy with university work that I had some put this on pause however not long after I have made my last post here I had somesort of success in achieving what I was after with 'Render to Texture'. Now I am back on it

I am going to get back into practise with it, however the main problem I came across (merely a simple one) was everytime I made a render I could only do one frame at a time, now we all know this is not practical at all, and I was wondering, how do I render a whole sequence through using Render to Texture???

Hope all is well and hope to hear back!

P.S. Will post results soon!
 
Old 05-13-2013, 02:00 PM   #10
PiXeL_MoNKeY
Expert
 
PiXeL_MoNKeY's Avatar
Eric Craft
Dallas, USA
 
Join Date: Mar 2002
Posts: 5,668
Quote:
Originally Posted by Want2be3D
I am going to get back into practise with it, however the main problem I came across (merely a simple one) was everytime I made a render I could only do one frame at a time, now we all know this is not practical at all, and I was wondering, how do I render a whole sequence through using Render to Texture???
Just like you do any animation in 3ds max, you need to change the render setup to a sequence instead of single frame. Press the Setup button in the Render To Texture > General Settings > Render Settings section of the dialog and make sure it is changed to a frame sequence.

-Eric
__________________
"The Evil Monkey hiding in your closet."
 
Old 05-18-2013, 04:05 AM   #11
Want2be3D
Veteran
Omar
London, United Kingdom
 
Join Date: Apr 2008
Posts: 39
Right, I think I got it, however I was not able to do it as a movie file only as a TGA sequence (even though i setted it up for it to render as a video file, the Render to Texture only ever outputs a sequence).

Anyways, I think I am coming along the right tracks, I made the image sequence to a movie file anyways and this is what I got so far:

https://mega.co.nz/#!IAhWAKID!MORvW...-S9r1UMWMYuTdjw

I think I might redo this again but I would cap the eye holes so I could see the eyes and also try and somehow relax the vertices around the chin/neck area as they seem to be overlapping.

Let me know what you think.

P.S. If you are unable to play it, it means you would need the Lagarith Lossless codec, otherwise you might just get away with it with the latest version of VLC Player.

Look forward to hear wat ya'll have to say (nothing exciting of course)
 
Old 05-20-2013, 12:50 AM   #12
Want2be3D
Veteran
Omar
London, United Kingdom
 
Join Date: Apr 2008
Posts: 39
Anyone may I ask?

Did anyone get a chance or manage to download or view the clip???
 
Old 05-21-2013, 01:39 AM   #13
Want2be3D
Veteran
Omar
London, United Kingdom
 
Join Date: Apr 2008
Posts: 39
Hi again everyonme, how are you all doing?

I do apologize for the posts, but I just would've like some feedback.

But anyways, here I present to you an improved render of the mapping, here I have capped the eyes of the geometry head (therefore they would be visible in the output) and also I have extended the edges of the head so more of the face is shown (to the point that abit of the background of the footage is revelaed).

Also to help your viewing, I have encoded it to a Quicktime H.264 and the filesize should be releatively small, so you should have no problems viewing it now.

http://depositfiles.com/files/rd1dk6y3k

Thanks and I look forward hearing back!
 
Old 06-08-2013, 04:06 AM   #14
Want2be3D
Veteran
Omar
London, United Kingdom
 
Join Date: Apr 2008
Posts: 39
No one?


Anyone at all may i ask???

:(

If there is any problems with trying to view the video, please let me know!
 
Old 06-08-2013, 04:06 AM   #15
CGTalk Moderation
Expert
CGTalk Forum Leader
 
Join Date: Sep 2003
Posts: 1,066,478
Thread automatically closed

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.
__________________
CGTalk Policy/Legalities
Note that as CGTalk Members, you agree to the terms and conditions of using this website.
 
Thread Closed share thread


Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
CGSociety
Society of Digital Artists
www.cgsociety.org

Powered by vBulletin
Copyright ©2000 - 2006,
Jelsoft Enterprises Ltd.
Minimize Ads
Forum Jump
Miscellaneous

All times are GMT. The time now is 08:50 PM.


Powered by vBulletin
Copyright ©2000 - 2016, Jelsoft Enterprises Ltd.