Vertigo is the hardest track in the game Trackmania Sunrise in platform mode (the tracks are insanely difficult and full of traps, and you have to reach the finish line without respawning)
I managed to complete this track in one shot, so i decided to make a video.
I had already completed this video a few month ago in 2D, but when i heard that the latest trackmania game supported native stereoscopic video (.avi) output. I decided to re-render the video in 3D.
Warning !!! the H264 version requires a good PC, and the use of the right software to play it. Not all players and not all codecs are able to open it it the moment.
You'll need a multithreaded H264 codec, and for some reason 3dtv.at's player crashes.
So if you don't know how complex codecs work, just use the Xvid version, it'll work.
Stereoscopic 3D configuration : Over/Under, right image on top : 2x 1280x720 (2x 720p)
watched the xvid version on my 19' crt+shutterglasses. the only 3d I noticed was the logo and sometimes when the track was very close but could have been my mind playing tricks on me. I appreciate the effort you put into this and I enjoyed watching this video, make more, but in the next one set the stereo separation much much higher.
by the way very cool clips.. how did you render them? with the replay? how did you do exactly? i would like to try it too..
if i give you access to a special mtbs ftp could you upload all 3 the clips there? i could add them to the gallery so you dont need to share them on other sites or even worse bittorrent.
whats the total size?
let me know
igor
Adam Savage: "I reject your reality and substitute it for my own."
Jamie Hyneman: "It's really cool, but really unusual."
@yuriythebest : the video was part of trackmania forever beta tests, the goal was to find a workflow for 3d video production and provide 3d videos watchable by anyone.
The trackmania gaming community produces lots of videos, so you should expect to see some more 3d video coming soon... (or at least as sson as i will have enough time to write a complete in-depth tutorial for them)
My firsts tests used a separation about twice higher than in the version you downloaded. But a lot of testers found it too strong and to difficult to view. All of them were absolute beginners in 3D. Also, the game developper played the video on a big tv. The separation became too high and gave him headaches.
So i became quickly obvious that when making videos for everyone you have to go easy on separation, otherwise most people would have an unpleasant experience. S-3d already has a terrible reputation of giving headaches, etc... I don't want to make this worse.
With the reduced separation , all the people who saw the video quickly got the depth effect and enjoyed it.
@sharky
i used quite high bitrates to get a good image quality, (because the image size is quite big : it's HD !).
The Xvid version uses a 5Mbps bitrate, about 180MB and the H264 (master version) uses an even higher 8Mbps, about 250MB.
The anaglyph one is about the same as the xvid version.
The process of making a 3d video in trackmania goes like this :
-play a track
-record the replay by clicking the "save replay" button at the end of a race (or you can also press the "S" key if you can't make it to the finish line)
-go to replay editor, setup your cameras, time, etc.. it uses a key system similar to most 3d and cinema tools. Except you don't have the kind of fine tuning ability you could get from these software. Don't add text or images in the replay, thei're not supported in 3d, you'll do that during editing.
-now you need to setup the S-3d parameters : turn off all external S-3D drivers and use the ingame engine (hope you still got some anaglyph glasses), and add convergence and separation keys. You should use a constant separation value for your whole video, but move the convergence point manually as close to the action as possible (especially when zooming in), but avoid any pop-up effects.
-render your scenes using the "shoot" function, you can either do it inside the replay editor or straight from the main menu, just activate the 3d rendering option, you can choose either anaglyph Red/Cyan mode or Left/Right dual images side by side mode.
The in-game 3d engine has some trouble with most post process FXs, i already told the devs about it during the beta but they didn't have the time to correct these bugs before release. So when i rendered the video, i had to disable all FXs and use the graphics card antialiasing instead of the game's.
If you see that an effect is only applied to one eye, or that only one eye is rendered -> you have an unsupported FX still activated..
chose your codec, and the game will render every frame to an .avi file. Just make sure your file doesn't get bigger than 2GB or the file will get corrupt (it's an other bug from the game). So don't render uncompressed frames.
Then you can edit the video in your favorite movie editor.
Anaglyph works just like a normal video.
Dual images is harder, since you have to make sure you work on the images separately. So that compositing, FX, and transition are applied correctly.
I didn't use any dedicated 3d video editing software or plugin, i did everything manually in my usual video editing software (Vegas). I should test some plugins next time i make a video.
When i made vertigo, i first did the whole editing on a 2D flat version (my original video was made month ago).
Then i used anaglyph, which was very easy, (just replacing the sequences) and allowed me to see exactly at what depth i was placing my text (i did everything by hand), it was my 1st 3D editing experience so didn't get how to make the compositing look right in anaglyph.
Finally when doing the dual images i rendered the two eyes separately, because i got some painful experience getting my compositing right in anaglyph.
I only reassembled the two eyes into one video for final compression using an avisynth script.
It was quite difficult the first time since i had to learn everything from scratch, from getting the render out of the game to the final file format.
I even found that the matroska container (.mkv) can carry the two eye streams separately in one file with tags on each stream to identify them -> which would make it ideal for 3D video distribution. Unfortunately, the only media player able to detect the 3D info contained in the file is VLC.
But vlc doesn't have any 3d output module, so it opens the two streams into two separate windows and keeps the streams synchronized, so it should work on dual output systems (dual projectors), but it wont work on any other 3d display system... unless someone makes a plugin.
Nice video. I like it very much, although I could use some higher seperation, too. But I think it might be better like this. The xvid version is already very fast and it is possible, my system (or my brain) is a little to slow for it sometimes...
I think making 3D videos is a little too complicated for me, but I'd love to watch some more, soon
Keep up the good work!!!
Play Nations at WAR with this code to get 5.000$ as a Starterbonus:
ayqz1u0s http://mtbs3d.com/naw/" onclick="window.open(this.href);return false;
AMD x2 4200+ 2gb Dualchannel
GF 7900gs for old CRT with Elsa Revelator SG's
currently 94.24 Forceware and 94.24 Stereo with XP sp2!
Nice. I watched it with Stereoscopic player and shutter glasses. I found the depth subtle but I think it might be because the camera was far away from everything most of the time. The depth would be more noticeable if the camera was closer to the car and the track (by using a bumper view or an inside view).
CPU : Intel i7-7700K RAM : 32 GB ram Video card : GeForce GTX 980 Ti OS : Windows 10 Display : Samsung UN40JU7500 Curved 40-Inch UHD TV with shutter glasses HMD : Oculus Rift
Tril wrote:Nice. I watched it with Stereoscopic player and shutter glasses. I found the depth subtle but I think it might be because the camera was far away from everything most of the time. The depth would be more noticeable if the camera was closer to the car and the track (by using a bumper view or an inside view).
watched it in the same solution and same results. As I've said before higher separation is key. It will look like a toy car track but still awesome.
The "ideal" stereo separation value depends on the size of the screen.
I'm trying to calculate this value on my video.
When i play trackmania i mersured the maximum separation on my screen, to correspond with the distance between my eyes.
My monitor is a 22" wide screen
I get this value with a separation of 0.8 (trackmania value). (but i can't use over 0.6, or i get headaches)
I shot and rendered the video with a value of 0.2 (trackmania value)
So this means that in order to get the ideal maximum separation, you'd need roughly a 70" or maybe 80" wide screen... That's big !
Maybe next time i make a 3D video i should try and make these calculations before and make multiple versions ? (one for 3d beginners and one for advanced 3d users)
What's your screen size ? (if it's 4:3, you should substract about 1 inche to include the letterboxing effect)
I bet most of you use 19" 4:3 CRT monitors ?
Actually Tril is spot on the "problem" here. The camera is too far away and then the stereoeffect (as in real life) is very low. It's another question if you shoot the replay with the view from the car (or playing the game when you have the view as seen from the car). I have not tinkered too much with the editor settings but i think you can do a lot with camera settings in trackmania to accomplish a really great effect! Since you've probably recorded your run it shouldn't be any problem to edit the cameraviews. Will just take a "little" time.
Really great run. Well done on a very difficult track.
nice logo.. 
so would you like to upload the files to mtbs? are you confident with ftp's? if yes i set up one for you. 
