Any 3D renders
I originally posted a Porsche 959 animation in "BF1 Eagle Rendered Movie." thread as a test. Since then I worked a little and wanted to post this back, but the title of that thread was "BF1", so I think I better to start a new thread to post this in.
Please post in all other 3D renders here to share (even not LFS related)
Rendered in 3ds max with Mental Ray. Glare[lume], Volume light, fog were used. 720 x 576 pixels 25fps, using Microsoft Mpeg4 to create this avi file.
MR_03.avi
7.22mb
Also, some still renders of the car.
Please post in all other 3D renders here to share (even not LFS related)
Rendered in 3ds max with Mental Ray. Glare[lume], Volume light, fog were used. 720 x 576 pixels 25fps, using Microsoft Mpeg4 to create this avi file.
MR_03.avi
7.22mb
Also, some still renders of the car.
Last edited by samyip, .
I am afraid Photoshop made up car is not allowed. If you read the original "Custom Car Contest" rules...
http://www.lfsforum.net/showthread.php?p=86950#post86950
Have you try the new modifier 'Sweep' in max? It is a handy new tool which can build surface from any spline shape very simple. You draw out a path, you can even have another path come out of it (like a juction) then apply the sweep with a custom shape, and it will build the road automatically, just like lofting, you then can fine tune with edit poly to your desire track.
For texturing, that should assigned mapping co-ordinate automatically. You can use blend material with vertix colour map for the blending mask, which you can paint directly on the actual object. The two materials to make up the blend can be e.g. smooth track surface and cracked track surface, This technique can help breaks the regular tiling patterns from normal texture mappings.
Last edited by samyip, .
Is 'Simulator D' a commercial product? Do you have their website? I could'nt find them in the search. Love to know more about it. Does it work on all 3ds max versions? Unfortunately the tutorial is not in English...
LOL That man is trying to controlling the car very hard on the roof, pulling left and right as the driver was missing on that car
(well all the cars infact...)Hi yamakawa, I remember I saw these avi before, very nice work. You used reelmotion for these with 3ds max? But reelmotion has gone now, so I have not tried it, is it similar to the Reactor car in 3ds max?
Oh, I have noticed in your Race.avi, the blue car intersets with the red car.
Oh, I have noticed in your Race.avi, the blue car intersets with the red car.
Not bad Ian, look forward to see that finished.
Ian,
To make all poly to same smooth group sometimes will have a too rounded effect. it may be better to adjust the spinner next to auto smooth to get the smooth blending you want (as jpg attached) The number is the degree angle between 2 faces that will blend.
I think you can try checking the sort by material in Turbosmooth/meshsmooth to avoid the material mapping screwed up. (see attached jpg), I agree the LFS default tyres are too angular, need to do something if you are going to render a high res image, you don't want it to spoil your wonderful image
To make all poly to same smooth group sometimes will have a too rounded effect. it may be better to adjust the spinner next to auto smooth to get the smooth blending you want (as jpg attached) The number is the degree angle between 2 faces that will blend.
I think you can try checking the sort by material in Turbosmooth/meshsmooth to avoid the material mapping screwed up. (see attached jpg), I agree the LFS default tyres are too angular, need to do something if you are going to render a high res image, you don't want it to spoil your wonderful image
The pixelated result is normally seen on compressed file, that is strange if it shown up on your uncompressed file, because an uncompressed frame should look exactly like your single still image, and I don't recall seeing any of your still images posted were pixelated.
The Bink videos don't seems to be compressing as the best quality as you described. The videos seems pixelated in places (i.e. the driver seat has square pixels in some frames and the floor shading has the ribbing)
Don't get me wrong here, I am not criticizing your animation, it is very nice. But I don't think Bink has showing the best of your work here yet.
It is just that I don't think 5.76 MB justify for the quality of 512x384@24FPS and 589824 Bytes/Second seen here. Compare to mine which was done at 720 x 576 @ 25FPS and 6000kb/s is only 2.33mb
Don't get me wrong here, I am not criticizing your animation, it is very nice. But I don't think Bink has showing the best of your work here yet.
It is just that I don't think 5.76 MB justify for the quality of 512x384@24FPS and 589824 Bytes/Second seen here. Compare to mine which was done at 720 x 576 @ 25FPS and 6000kb/s is only 2.33mb
Well, of course if only making a still image of a race car on a section of race track isn't difficult, because you only need to model the track and background to what your camera can see, you don't have to worry to much about anything beyond the camera view. And because it is a still image, some of your backgrounds can be 'cardboard' props, no need to go full 3D, but of course you can if you insist to...
You can get quite a lot of track textures from the LFS folder already. Not bad for a starting point.
XCNuse is quite right about the lighting of a scene like that, very important to get it right, also you may need to consider the atmosphere, i.e. haze, mist etc. They are all the little details to make your still image to look better.
LOL Ian, you have got ghosts on the hill watching the race
You can get quite a lot of track textures from the LFS folder already. Not bad for a starting point.
XCNuse is quite right about the lighting of a scene like that, very important to get it right, also you may need to consider the atmosphere, i.e. haze, mist etc. They are all the little details to make your still image to look better.
LOL Ian, you have got ghosts on the hill watching the race
Last edited by samyip, .
http://www.mcb-ltd.co.uk/MCB_3d/959/MR_01.avi
This is a test render of my old modeled car rendered in mental ray. (This car is a very high poly model, many times higher than LFS models. But it wasn't completed, many part were missing, i.e. the rear brake) A sequence of taga still images import to an editing programm and used Microsoft mpeg4 to create this 6 seconds 720 x 576 25fs animation, bit rate set at 6000kb/second.
But for speed, this animation did not render with fields, which will play smoother if did so.
file size: 2.33m
This is a test render of my old modeled car rendered in mental ray. (This car is a very high poly model, many times higher than LFS models. But it wasn't completed, many part were missing, i.e. the rear brake
) A sequence of taga still images import to an editing programm and used Microsoft mpeg4 to create this 6 seconds 720 x 576 25fs animation, bit rate set at 6000kb/second. But for speed, this animation did not render with fields, which will play smoother if did so.
file size: 2.33m
Last edited by samyip, .
If using skylight and default scanline renderer, it will take ages to render.
But I've noticed if you switch to Mental Ray renderer, it renders much quicker. Probably Skylight in Max isn't really for scanline renderer to use.
Bob,
That certainly doesn't look like a full HDTV quality. (may be the file size would be too big to be attach to this forum if it was in HDTV size)
The jerks probably caused by the slowest of refreshing each frame, the codec's quality setting may be too high which it has not compressed to run smoothly.
DivX or Mpeg are not bad for compressing and at the same time hold good enough quality. I personally perfer AVI/mpeg format because they can be played on a larger range of media.
Bink is quite often seen in Games, which it captures the real time 3D (not pre-rendered) animations. Anyway, I could be wrong.
That certainly doesn't look like a full HDTV quality. (may be the file size would be too big to be attach to this forum if it was in HDTV size)
The jerks probably caused by the slowest of refreshing each frame, the codec's quality setting may be too high which it has not compressed to run smoothly.
DivX or Mpeg are not bad for compressing and at the same time hold good enough quality. I personally perfer AVI/mpeg format because they can be played on a larger range of media.
Bink is quite often seen in Games, which it captures the real time 3D (not pre-rendered) animations. Anyway, I could be wrong.
BTW, Compression in video codec is different issue.
The quality lost in video compression has got nothing to do with your scene set up.
What I mentioned on the above post was pure raw renders(image has not compressed), As I said before, I render animation in a sequence of still images(uncompressed) and then import them to a video editing software to stitch them together.
I think a lot of people have sharp eye too, but when you watch a movie, you don't pause every frame to view for a minute, then forward to the next frame and do the same, do you? you don't break the flow of the motion.
What I meant was if you render an image in 4000 x 3000 pixels, of course you can see all the details, but try resampling that image in photoshop to 720 x 576 pixels, will you still see those details? Nope! they became a blob in a larger pixel. What I was trying to say is you can find the right visual balance to fit the TV/video resolutions which means you don't have to waste any unneccesary time for some details which are not going to hold anyway.
Also, capturing stills from TV/video will never give you a good quality to print, most TV/video are interlaced (which means two images, e.g. frame 1 and frame 2 combined together with lines between them), again, they are only 720 x 576 resoultions, will that be good for wallpaper?
I have done many animations at work which time is always against you depending on client's budget as well. I have done exactly what you said before, create every 3d objects in highest resolutions (even the objects are only in the background) with loads of fine details, and took weeks to render, but the result is alot of the work I put to the finest details do not show up in the 720 x 576 resolutions.
By learning from the experience, you will know how to plan you animation before hand such as will the object be seen in close up or not? (close up will require higher res modeling) if not, you could cut the modeling time by half, and if some objects can be texture map rather then real models or bake the GI textures to objects could save alot of time, which the end result still be as good as the original GI scene. There are a lot of tricks can help you cut the time and still render in good result.
It may be I am in the trade where delivering the good is more important than not delivered at all, where time and quality got to be compromised. Who don't want to have a large render farm and ultra powerful machines to do 3D? When you don't want to or can't wait for the slow renders, you will have to seek alternatives. This is only my experience, of course you don't have to agree.
What I meant was if you render an image in 4000 x 3000 pixels, of course you can see all the details, but try resampling that image in photoshop to 720 x 576 pixels, will you still see those details? Nope! they became a blob in a larger pixel. What I was trying to say is you can find the right visual balance to fit the TV/video resolutions which means you don't have to waste any unneccesary time for some details which are not going to hold anyway.
Also, capturing stills from TV/video will never give you a good quality to print, most TV/video are interlaced (which means two images, e.g. frame 1 and frame 2 combined together with lines between them), again, they are only 720 x 576 resoultions, will that be good for wallpaper?
I have done many animations at work which time is always against you depending on client's budget as well. I have done exactly what you said before, create every 3d objects in highest resolutions (even the objects are only in the background) with loads of fine details, and took weeks to render, but the result is alot of the work I put to the finest details do not show up in the 720 x 576 resolutions.
By learning from the experience, you will know how to plan you animation before hand such as will the object be seen in close up or not? (close up will require higher res modeling) if not, you could cut the modeling time by half, and if some objects can be texture map rather then real models or bake the GI textures to objects could save alot of time, which the end result still be as good as the original GI scene. There are a lot of tricks can help you cut the time and still render in good result.
It may be I am in the trade where delivering the good is more important than not delivered at all, where time and quality got to be compromised. Who don't want to have a large render farm and ultra powerful machines to do 3D? When you don't want to or can't wait for the slow renders, you will have to seek alternatives. This is only my experience, of course you don't have to agree.
What XCNuse said about the rendering is quite right. For a still image, you normally render it to over 1000s pixels and all the details are crisp and sharp, because the image is still, any imperfection will shown if it is not set to the highest quality. But for video, a standard TV/video resolution is no more than 720 x 576 (except HDTV) therefore you don't have to set your renderer to the highest sampling rate, because those extra quality will not be seen in this resolution. Also because it is moving, each frame of render will not stay in your vision for more than 1/30 of a second.
By lowering the quality sampling, you can speed up each frame's rendering time.
By lowering the quality sampling, you can speed up each frame's rendering time.
http://www.lfsforum.net/showthread.php?p=122186#post122186
I posted a simple animation of my car in 3ds max on a different thread before. This one took about 3-4 hours to render, I done them in sequential Taga files and then import them in Adobe Premiere to create an avi
I posted a simple animation of my car in 3ds max on a different thread before. This one took about 3-4 hours to render, I done them in sequential Taga files and then import them in Adobe Premiere to create an avi
On my colour corrected screen at work, these images look a bit grey/under exposed.
Same kind of thing happened on my home monitor, I rendered an image that is looking bright but when I took it to view at work it looks under exposed...
Nice wheel BTW. (oh, how do I pump up your tyre? )
Same kind of thing happened on my home monitor, I rendered an image that is looking bright but when I took it to view at work it looks under exposed...
Nice wheel BTW. (oh, how do I pump up your tyre?
)Ian, have you try the MultiRes modifier? it can reduce you objects polygons to a specific poly counts you want. Sometimes it works well, I used it on LOD set up when I have made a Hi-res model. (for example, Hi-res model when is in foreground, change to lower res at middle and reduce to lowest res in background) But I am not suggesting this is how people should model a low poly car, this is just an alternative to reduce your model's poly in a quick way. The result could be different to what you wanted.
This 2 pictures are not great, but they only took about 3mins to render. Together with Glare and mist applied. If I saved the Final Gather map file to be reused on future rendering, the render time will cut down further.
Have you try the Mental Ray? it can be quite fast to render if you lower the Final Gather samples and increase the filter. Just find a right balance of speed and a reasonable quality.
I think what Orion meant was the car in your render is too over exposed on the bonnet. If you tweak the lighting/direction of the light, the shading on the bonnet will improve.
Mato, very nice mod of the FZR, I like it
look very sexy.
The only thing is I think after you meshsmooth the bonnet, the corners of the bonnet have gone too rounded, may be increase the weighting on the corner vertexes in meshsmooth, or refine your corners before applying meshsmooth could help.
look very sexy.The only thing is I think after you meshsmooth the bonnet, the corners of the bonnet have gone too rounded, may be increase the weighting on the corner vertexes in meshsmooth, or refine your corners before applying meshsmooth could help.
sgt.flippy, I don't think that is only apply to the M coupe.
To confuse you more, take a look at this photo. (sorry XCNuse, that will be the last photo I will attach in your thread )
To confuse you more, take a look at this photo. (sorry XCNuse, that will be the last photo I will attach in your thread
)FGED GREDG RDFGDR GSFDG