3D Cliffs Of Dover Pictures

[Les]

Another video. Again, best watched in expanded or full-screen view on Youtube.

The aim was to see if I could re-do the Black Death Benchmark Track in 3D. I think it turned out alright all things considered.

Tried to write a list of instructions here explaining how to make these kinds of video's, but it got too long and complicated.

I think the most important thing I can say, and the biggest lesson I've learned from this test is, if you want to make a video using the method I've used here, do it with a track you've recorded yourself specifically for that purpose.

Due to the nature of the manual 3D capturing method, and more so the limits of the games replay system, there are some kinds of shots you must avoid altogether in order to end up with a useable image at all. Beyond that, 3D video has it's own unique set of pre-requisites when it comes to shot composition and content. I don't know what all those do's and dont's are, but I have a better idea after making these test video's.

So, hopefully, if I do another one it will be less of a test and more of an attempt at ending up with a more finished result. But in the meantime -

http://www.youtube.com/watch?v=mGHY7VYLY6U

[Wolf_Rider]

Thank you


that brought a tear to an old bloke's eye - cheers

[Les]

With the help of Superman providing the link to the excellent stereo image viewer and converter here - http://stereo.jpn.org/eng/stphmkr/ and Stefem for explaining how to take stereo screenshots using the nVidia 3D capabilities, I've been able to compare nVidia's semi-automated 3D results to those you can get using the manual screencap & edit approach.

My results aren't definitive, and with some tweaking of the nVidia 3D parameters, maybe these results could be different, but...

Cliffs Of Dover and nVidia 3D are clearly having some issues.

Manual

nVidia

Manual

nVidia

Manual

nVidia

Manual

nVidia

Manual

nVidia

Manual

nVidia

Manual

nVidia

Manual

nVidia

Manual

nVidia

Manual

nVidia

Manual

nVidia


These nVidia screenshots were taken in PNS format, using what looked at the time to be approximately the same amount of camera offset as the PNG FRAPS captures, but the 3D effectiveness is quite different. This might be a result of adjusting some other nVidia settings incorrectly, or it might not be, I don't know. I'm assuming there must be some way of adjusting the nVidia 3D parameters so that the basic 3D effect is more pronounced, because if there isn't then what's the point?

I don't think the bugged out water and shadows on the nVidia shots is the result of an end-user setting though.

I was hoping to see if I could use the nVidia 3D, then run the game in slow motion and take repeated 3D screenshots that I could then compile in a video-editor as a moving image sequence. That would have meant not having to use the game's buggy replay system and manually resetting the camera position for every shot. But I guess there ain't no easy way out

Maybe some of the other nVidia 3D users can confirm these graphics bugs I'm seeing and perhaps suggest some solutions.

[Stefem]

There's not yet a dedicated profile for CoD so convergence isn't set at the best, anyway, users can create custom convergence profile if they want and the 3D depth can be adjusted in the NVIDIA control panel.
Shadows are bugged in 3D but I don't know if it's a driver problem or must be corrected on the engine side.

[irR4tiOn4L]

Many thanks for the 3d content, however this is not the correct depth/seperation. Its similar to what would often happen in IL2, which is that the cockpit didnt seem to be rendered at the correct depth forcing you into a 'toy story' type 3d.

Basically, your perspective are way too far apart and make everything look small - including itty bitty cute toy planes.

This is an entertaining effect, be sure, but thats not how it should look! Doing this makes depth perception in the distance (Where you really need it) suffer and makes it very hard to shift from looking in and out of the cockpit.

[Les]

Originally Posted by irR4tiOn4L

Many thanks for the 3d content, however this is not the correct depth/seperation. Its similar to what would often happen in IL2, which is that the cockpit didnt seem to be rendered at the correct depth forcing you into a 'toy story' type 3d.

Basically, your perspective are way too far apart and make everything look small - including itty bitty cute toy planes.

This is an entertaining effect, be sure, but thats not how it should look! Doing this makes depth perception in the distance (Where you really need it) suffer and makes it very hard to shift from looking in and out of the cockpit.

While I did push the 3D separation about as far as I could in order to make the 3D effect as obvious as possible, I'm not sure I agree with you about what the 'correct' depth/separation should be. I have my doubts that the kind of depth separation you're describing is actually possible to create on something as small as even the largest computer monitor. I'm not even sure it happens in real-life, where there are also limits to the distance at which we can perceive things stereoscopically. Do you have any pictures to show what you mean?

So looking at the clod pictures , I can't tell the difference at all.

[JG26_EZ]

Originally Posted by winny

The photo that was linked in one of the earlier posts that has the picture of the dinosaur, best describes what irR4tiOn4L is trying to say (I think). If you look closely, you can see that the nose moves less than the rear section of the head. This is the depth perception problem that I believe is being discussed above.

So, I think that it's not just as easy as taking an image from two different angles when you've got such depth as miles of background visible. My suggestion would be, to crop the spit away from the background, and have the spitfire's amount of adjustment different than the background scenery's adjustment. (Basically, if you were 5 feet in front of a spitfire looking back at it, and you moved your head from side to side, you'd see more unseen areas of the spitfire than you'd see of the background.)

If ya get me..

(I'd set it up, but it is 100% impossible for me to see 3D with those types of images, and my eye doctor will back me up on that one.)

[Ailantd]

Having researched 3D stereoscopy myself I can tell you that for a realistic effect two things are needed: First, the shift between the two virtual cameras should be the same in game that in real between the two eyes. With only this you have a "realistic" representation of depth in game, BUT an unrealistic feeling of it, because... the fov and monitor size. For a realistic and natural feeling you also need to fit the fov to a realistic value AND use a surface where that realistic fov actually fits your real field of view. If you do both you will have a complete being there feeling with a nice and beliable depth effect. Of course depth is also more noticeable in close objects that in very far ones. That´s normal: you can accurately tell the distance between two close object in cm, but hardly you will do the same between two aircraft even in kilometers without using other info like apparent size.

The main probem with nvidia screenshots is the shadow. It seems that shadow map is not recalculed for the second screenshot ( don´t know if thsis is driver problem or it needs to be implementd by devs, but I would bet for the second one), so shadow projection in the second image fails.

[Ailantd]

Originally Posted by JG26_EZ

The photo that was linked in one of the earlier posts that has the picture of the dinosaur, best describes what irR4tiOn4L is trying to say (I think). If you look closely, you can see that the nose moves less than the rear section of the head. This is the depth perception problem that I believe is being discussed above.

So, I think that it's not just as easy as taking an image from two different angles when you've got such depth as miles of background visible. My suggestion would be, to crop the spit away from the background, and have the spitfire's amount of adjustment different than the background scenery's adjustment. (Basically, if you were 5 feet in front of a spitfire looking back at it, and you moved your head from side to side, you'd see more unseen areas of the spitfire than you'd see of the background.)

If ya get me..

(I'd set it up, but it is 100% impossible for me to see 3D with those types of images, and my eye doctor will back me up on that one.)

Problem with that dinosaur is that 3d effect is achieved moving the dino head, and no simulating the two eyes point of view. The 3d effect is similar, but the way the model moves is not. As you noted in real stereoscopic the closer an object is to the eyes, the more shift it has between the two images.