Micro Point v.s. Ray tracing movie

Become a member of the CGSociety

Connect, Share, and Learn with our Large Growing CG Art Community. It's Free!

REPLY TO THREAD
 
Thread Tools Search this Thread Display Modes
Old 03 March 2014   #1
Micro Point v.s. Ray tracing movie

Today, Uploaded MicroPoint v.s. Ray tracing movie.
Please watch Full HD resolution environment.
 
Old 03 March 2014   #2
Cool but not enough info. I mean what are we looking at here? Is this a new plugin, a render engine for point cloud data, a new white paper, what?

While the point based rendering is obviously smoother there's no indication of the render time differences between each side. Also what optimizations are you using for each side, what are the limitations? Lighting, shadows (shadowmaps should not be a problem, but how about sharp shadows), what happens when you place a refractive surface in front of your pointcloud, or a reflective surface nearby etc?
__________________
The Third Party | Homepage | My Reel
"You need to know what you're doing before you start, and to start because you need what you're doing."
 
Old 03 March 2014   #3
Please access MicroPoint web site.
http://micropoint.jimdo.com/
Papers are
http://micropoint.jimdo.com/paper/
You can download English version paper PDF.
Samples are
http://micropoint.jimdo.com/sample/
You can download this movie's bunny-far object data.

Thank you.
 
Old 03 March 2014   #4
As I understand it, this is a brand new rendering technique to replace ray tracing.

There's more info about it on the MicroPoint website...
 
Old 03 March 2014   #5
neat stuff

Hope the research materializes into pro render engines. We've all come to just expect a certain amount of noise in our renders and hope no one notices.
 
Old 03 March 2014   #6
- If the comparison you present is at least a bit fair, you could have gotten the paper easily published at any major graphics conference. Why did you choose some small, possibly fradulent transportation conference?

- Why did nobody use your method in the two years it exists? http://forums.cgsociety.org/showthread.php?t=1065959

Sorry, I've seen hundreds of computer graphics papers, and this is 100% gibberish (poor translation aside)
 
Old 03 March 2014   #7
Sorry, Please wait.

[QUOTE=KeyCZ]- If the comparison you present is at least a bit fair, you could have gotten the paper easily published at any major graphics conference. Why did you choose some small, possibly fradulent transportation conference?

Exactly. Currently in preparation. For a new rendering technique, it becomes quite a long paper completely, time it takes to writing than originally planned. Please be patient.
 
Old 03 March 2014   #8
Originally Posted by tacwon: Exactly. Currently in preparation. For a new rendering technique, it becomes quite a long paper completely, time it takes to writing than originally planned. Please be patient.


Ok, but are you aware that if your original paper went through peer review, you cannot publish it anywhere else, even if translated?
 
Old 03 March 2014   #9
Boom! Headshot!
 
Old 03 March 2014   #10
Originally Posted by KeyCZ: Ok, but are you aware that if your original paper went through peer review, you cannot publish it anywhere else, even if translated?


Paper we are preparing is not intended the translation of the original Japanese papers. In addition to more than one paper, which was announced in Japanese already, we plan and detailed implementation and mathematical theory of aliasing is added.
 
Old 03 March 2014   #11
Question

I noticed the ray traced image is much more sharper. It looks like no filter was used.
While the micro point example looks like its being blurred a bit.

If you applied the same filtering to the raytraced one this could help reducing the noise considerably imho.
 
Old 03 March 2014   #12
[QUOTE=CHRiTTeR]I noticed the ray traced image is much more sharper. It looks like no filter was used.
While the micro point example looks like its being blurred a bit.

In this reason data compression when upload a movie to Youtube.
The rendering result of the original,
you can be downloaded from
http://www.sist.ac.jp/~iigura/MagicPoint/
I wish you well.
 
Old 03 March 2014   #13
Originally Posted by tacwon: [QUOTE=CHRiTTeR]I noticed the ray traced image is much more sharper. It looks like no filter was used.
While the micro point example looks like its being blurred a bit.

In this reason data compression when upload a movie to Youtube.
The rendering result of the original,
you can be downloaded from
http://www.sist.ac.jp/~iigura/MagicPoint/
I wish you well.


Hi, i checked the .psd file and the micro point one does look blurrier than the ray traced one...
Is there no image of a noiseless ray traced version, so we can compare better?

Here is a version with the blending mode set to 'difference' in photoshop between the raytaced one with 8192 samples and the micro point one.

As you can see the results are not verry close

Asside from that, how long did it take to render the micro point image?

Theres not much usefull information (that i can read) on the site.

Last edited by CHRiTTeR : 03 March 2014 at 12:04 AM.
 
Old 03 March 2014   #14
[Hi, i checked the .psd file and the micro point one does look blurrier than the ray traced one...
Is there no image of a noiseless ray traced version, so we can compare better?

Stochastic sampling to be used in ray tracing is a technique for obscure by converting to noise aliasing. So, there is no ray tracing image without noise theoretically. bunny-fur is a remarkable sample indelible noise.

Asside from that, how long did it take to render the micro point image?
The movie creation and MicroPoint Ray Tracing (4096 samples / pixel), we took roughly four days. However, MicroPoint, CPU (1 core) implementation, in order to deal with large amounts of data, which is the implementation of out of core, consuming most of the time is I / O access. (Cycles of blender) Ray Tracing, is running on more than 1000 core GPU implementation.
It is not me on the Web because the condition does not compare the difference too.
 
Old 03 March 2014   #15
It looks to me like the method is not tracking sub pixel coverage, as it converts it to alpha instantly? Therefore, wouldn't a pixel that is covered by two polygons at exactly 50% each always look the same, regardless of whether those two polygons overlap each other (50% of the pixel covered by polygons) or not (100% of the pixel covered by polygons)?
 
reply share thread



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 01:01 AM.


Powered by vBulletin
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.