NEW Raylectron v2, a photorealistic SketchUp render plugin.
-
@unknownuser said:
"...certain bias"? Strictly raytrace, no photon, no path tracer.
I guess Pete meant that youwere not unbiased, not the rendering method.
-
@gaieus said:
@unknownuser said:
"...certain bias"? Strictly raytrace, no photon, no path tracer.
I guess Pete meant that youwere not unbiased, not the rendering method.
Indeed!
-
Michael, the file doesn't have to be perfect, even better that its the same that you used. I think its only fair to offer a chance to the other renderers to show what they can do also by experiences users 
-
One question:
If you set your camera looking in Xray through the corner of a volume, do you set the number of planes to 2 or 1? Or does the camera "see" both planes as one plane? -
@mitcorb said:
One question:
If you set your camera looking in Xray through the corner of a volume, do you set the number of planes to 2 or 1? Or does the camera "see" both planes as one plane?It all depend on how many planes your wall is made of. If your camera is outside, pointing in, how many planes does it have to go through before it's actually inside? That's the number you put in.
-
@mike lucey said:
:lol: Michael, the file doesn't have to be perfect, even better that its the same that you used. I think its only fair to offer a chance to the other renderers to show what they can do also by experiences users
The other users forced me into it by saying that the others are better. So I did a compare. But the model remain off limits until we release it on our site as a complete model specifically for Raylectron. It will include just about all the feature of the software.
Sorry about that Mike, I don't want to sound rude or anything, but if anyone wants to make a comparison, they can in fact use any models from the 3D warehouse, which is where this one is from anyway!
Or better yet, why not make a new topic with a complex model that we can all download and post our rendered result along with the settings we used, so that everyone else can try it, to make sure no one cheated. This way, it will not be only a Raylectron thing, but a true comparison of all renders, at least those willing to participate. What do you think?
-
I dont think the test was biased, although he obviously is more experienced with raylectron! To me twilight looked the best as it has a nice balance between the 2 (crisp/realism/warmth)
Cant do video can it? -
@unknownuser said:
Or better yet, why not make a new topic with a complex model that we can all download and post our rendered result along with the settings we used, so that everyone else can try it, to make sure no one cheated. This way, it will not be only a Raylectron thing, but a true comparison of all renders, at least those willing to participate. What do you think?
I like the idea! Maybe a standard interior scene, rendered day time with light coming from a window and night time with artificial lights. Then each one can post the steps taken to achieve the result so we can not only see the end result but also how easy or complicated was to achieve it. This could also serve as a way to measure render times.
I made a quick search at 3D Warehouse and I came up with this: http://sketchup.google.com/3dwarehouse/cldetails?mid=7e192f09383ff157eb01393c3387d97d which are highly rated and maybe one of those rooms with some modifications could serve as the test scene.
-
@notareal said:
@unknownuser said:
Or better yet, why not make a new topic with a complex model that we can all download and post our rendered result along with the settings we used, so that everyone else can try it, to make sure no one cheated. This way, it will not be only a Raylectron thing, but a true comparison of all renders, at least those willing to participate. What do you think?
There is one, popular challenge, based on famous H.Jensen's frosted glass dragon at http://forums.cgsociety.org/showthread.php?f=2&t=902691&page=1&pp=15
Keep in mind that it's not a all around challenge, but it does address to one particular area that can make rendering of a photo realistic scene impossible. Probably most of the renders are already updated from the time of challenge, but it's still a good test to try. Good news is that you only need two (or three materials if emitter plane is counted): a diffuse and a rough glass (no SSS).Interesting challenge, I'd love to see an experienced user of Raylectron attempt it, please post results.
-
@unknownuser said:
Or better yet, why not make a new topic with a complex model that we can all download and post our rendered result along with the settings we used, so that everyone else can try it, to make sure no one cheated. This way, it will not be only a Raylectron thing, but a true comparison of all renders, at least those willing to participate. What do you think?
There is one, popular challenge, based on famous H.Jensen's frosted glass dragon at http://forums.cgsociety.org/showthread.php?f=2&t=902691&page=1&pp=15
Keep in mind that it's not a all around challenge, but it does address to one particular area that can make rendering of a photo realistic scene impossible. Probably most of the renders are already updated from the time of challenge, but it's still a good test to try. Good news is that you only need two (or three materials if emitter plane is counted): a diffuse and a rough glass (no SSS).@mrmikeesq said:
I dont think the test was biased, although he obviously is more experienced with raylectron! To me twilight looked the best as it has a nice balance between the 2 (crisp/realism/warmth)
Cant do video can it?Honestly it did look crap comparing what Twilight can do; grainy shadows, AA issues, odd texture jumping (could it be some modeling issue, like skewed textures that TWL tries to fix?),... I'd use "easy" settings or maybe alternative. Not to mention Maxwell, well have to say that I have not tested SU integrated version, but what I seen it should not be that difficult to use.
If I where a author of those compared renders, I'd be fairly upset.
Inexperience is not a excuse, if scene is not provided.
-
I have installed the trial version of Raylectron, basically to compare it with Vray for Sketchup, the rendering program I am accustomed (and happy) with. Is there any tutorial or manual for Raylectron? And if anyone here has any experience with both Raylectron and Vray, how would they compare?
-
Hello All
FYI
Test made with about 200 K and complex architectural project file.
Lots of sub-groups and components.
i7 unlocked cpu + 8 mb Win 7 64 4 threads used
Raylectron V2 1hr. plus and still working.
Kerkythea 32 bit under 20 min to finish.
Artlantis 1.2.6 under 20 min to finish
Twilight same as Kerky.
Photon map with anti-alias.
(Atl uses radiosity)dtr
-
@odafos said:
I have installed the trial version of Raylectron, basically to compare it with Vray for Sketchup, the rendering program I am accustomed (and happy) with. Is there any tutorial or manual for Raylectron? And if anyone here has any experience with both Raylectron and Vray, how would they compare?
We've lost our tutorial on our site, so we are rebuilding it now.
See our forum gallery for renders we and people have produced. Only you can be the judge.
-
@dtrarch said:
Hello All
FYI
Test made with about 200 K and complex architectural project file.
Lots of sub-groups and components.
i7 unlocked cpu + 8 mb Win 7 64 4 threads used
Raylectron V2 1hr. plus and still working.
Kerkythea 32 bit under 20 min to finish.
Artlantis 1.2.6 under 20 min to finish
Twilight same as Kerky.
Photon map with anti-alias.
(Atl uses radiosity)dtr
Raylectron never stops. It's up to you to stop it when the quality is to your satisfaction.
-
@dtrarch said:
Kerkythea 32 bit under 20 min to finish.
Twilight same as Kerky.
Try Kerkythea Echo Boost, it will go even faster than Twilight! And if you run on 64bit, you will get much more speed. Make a test and come with the results !
-
@unknownuser said:
@dtrarch said:
Kerkythea 32 bit under 20 min to finish.
Twilight same as Kerky.
Try Kerkythea Echo Boost, it will go even faster than Twilight! And if you run on 64bit, you will get much more speed. Make a test and come with the results !
The Intel Embree kernel is faster than all of them and runs only on the CPU. It renders the Imperial Crown in about 20 secs.
-
@unknownuser said:
@unknownuser said:
@dtrarch said:
Kerkythea 32 bit under 20 min to finish.
Twilight same as Kerky.
Try Kerkythea Echo Boost, it will go even faster than Twilight! And if you run on 64bit, you will get much more speed. Make a test and come with the results !
The Intel Embree kernel is faster than all of them and runs only on the CPU. It renders the Imperial Crown in about 20 secs.
Is this what Raylectron is using?
I had a quick look and it looks interesting:
http://software.intel.com/en-us/articles/embree-photo-realistic-ray-tracing-kernels/ -
@zoom123 said:
Is this what Raylectron is using?
I had a quick look and it looks interesting:
http://software.intel.com/en-us/articles/embree-photo-realistic-ray-tracing-kernels/No, but we are investigating the possibility and legality. It is very interesting!
-
@unknownuser said:
@unknownuser said:
@dtrarch said:
Kerkythea 32 bit under 20 min to finish.
Twilight same as Kerky.
Try Kerkythea Echo Boost, it will go even faster than Twilight! And if you run on 64bit, you will get much more speed. Make a test and come with the results !
The Intel Embree kernel is faster than all of them and runs only on the CPU. It renders the Imperial Crown in about 20 secs.
Give any modern CPU based renderer a four socket workstation (four independent xeon CPUs, 40 physical cores) , that was used in my understanding in said Embree demo, and you will get a pretty fast renderer. Anyhow Embree is a nice demonstration about how CPU (=intel) based renderer can compete against GPU (=nividia)
And it was not 20 sec, if you where referring the intel video, "a prefect image takes about a minute to compute". No doubt Embree is a impressive collection of ray tracing kernels.
The original Austrian Imperial Crown by Martin Lubich http://www.loramel.net/showroom/3d/
Advertisement