Italian flying saucer video

[FatPhil]

3d graphics cards appeared on the consumer market in 1996. People were using Autocad to make models. And video capture hardware was also ubiquitious. So the tools were definitely there, even if you didn't have access to Silicon Graphics hardware—which the right kinds of students did have.

2003 would not have been an "early release" of Blender, either.

I can confirm that Autodesk published a proper 3D renderer called "3D Studio" back in 1992, as I remember once having the task of quickly knocking up some shoes order to persuade Clarks to buy grahics workstations from us. It wasn't integrated with their AutoCAD suite, as it was written by a different company, but file formats overlapped so you could share assets between the two.

 

[NorCal Dave]

2003 would not have been an "early release" of Blender, either.

Yes:

Blender was initially developed as an in-house application by the Dutch animation studio NeoGeo (no relation to the video game brand), and was officially launched on January 2, 1994.[13]​ Version 1.00 was released in January 1995,

https://en.wikipedia.org/wiki/Blender_(software)

I haven't played with it, but does this video look out of the realm of possibility for the early '00s? If anything, the cleaner version looks a bit more CGIed to my eye. The later versions with the tracking errors or bars at the bottom look like an attempt to make the original appear as VHS or other analog recording.

 

[Max Phalange]

This one should be the best quality, uncropped segment you can find.

I've combined it with the lower quality tv segments at the start and end:

Source: https://www.youtube.com/watch?v=F0bQMBTUff0

Thanks for this, it was very handy to demonstrate how poor the motion tracking really is.

 

[John J.]

Can you expound as to why? Other than CGI, I'm not sure how you get he morphing appearing/disappearing lumps. That would be hard to do with a physical prop. (I suppose you could do it by stop motion or cell animation, but CGI seems easier.

Ah, yes, I'm glad you asked me that.
The answer is of courseI hadn't really thought about the shape changing bits.

I still feel the overall motion of the "craft" resembles that of something suspended on a line, but have to agree that the craft itself seems to have some moving features. Probably easier to do with CGI than with e.g. RC model aircraft servos.

 

[qsek]

Thanks for this, it was very handy to demonstrate how poor the motion tracking really is.

This looks more like proper stabilizing. Well done and thanks for the effort.

But i would like to point out that your previously linked stabilized video (on page 1) was indeed a glitched effect resulting from double frames before the stabilizing was applied.

So just to be sure that in this one no similar oversights were made, did you check for double frames caused by mismatching FPS before you applied the stabilizing?
Can i ask what software you used and what filter/plugin?
And would you be so kind to render and link a higher quality version of this?

 

[JMartJr]

Did anybody have experience with some of those earlier CGI/VFX programs, and do you recall if motion tracking to a handheld video was difficult to do? That might fit in with the janky tracking, if so...

 

[Fin]

Here's an example of what people were making in Blender in 2004. I'm confident it could achieve what we are seeing in the UFO video in 2003 (assuming the UFO is CGI and the rest of the scene is a real video)

Source: https://www.youtube.com/watch?v=Cu53H8jsZ54

 

[jdog]

Here's an example of what people were making in Blender in 2004. I'm confident it could achieve what we are seeing in the UFO video in 2003 (assuming the UFO is CGI and the rest of the scene is a real video)

Source: https://www.youtube.com/watch?v=Cu53H8jsZ54

Yeah, I took a class in 3D Studio Max around then -- before I learned how expensive it was to license, though students got it free. (Max was released in 1996.) Maya 3D (released in 1998) was also available for free trials. Adobe After Effects had been out for about 10 years.

Designing a reflective metal object and animating it along a path in flat lighting wouldn't have been that hard. With no interaction between the animated object and the environment, all you really need to do is drop the animation with an alpha channel over the base clip, then add an adjustment layer over both to help the animation blend in. (Notably the saucer doesn't go behind anything, and the shiny metallic bottom doesn't reflect any changes as it passes over the dark green trees or the bright sand.)

 

[JMartJr]

Here's an example of what people were making in Blender in 2004. I'm confident it could achieve what we are seeing in the UFO video in 2003 (assuming the UFO is CGI and the rest of the scene is a real video)

Source: https://www.youtube.com/watch?v=Cu53H8jsZ54

This bit at 5:10 is particularly relevant!

Source: https://youtu.be/Cu53H8jsZ54?t=310

 

[purpleivan]

It is not clear, if he got this "group of students" from Hangar18b who did their own research i suppose, or he read something on the internet about it. As far i have been digging there are conflating informations in various old forums and sites. But both the tv report and Chiumiento speak about a anonymous source.


I think its not helpful to this discussion when i am going to try to explain impossible physics. Im not a physicist.

And im also no VFX Artist nor i am in the industry. But i am interested in the field and everything digital/mathematical/optical related, which brings me to the reason i came here in the first place.

In my view this witness account does not make the accompanying video any more credible but ... more interesting and worth a second look.
The video itself has a couple of red flags. No source. No sound. It "looks cgi". Extremely lucky to capture the whole sequence and get get a display like that.
But what is has is detail. It has a reasonable resolution. It is a real location. It is at least from 2005 since then Paola Harris got a copy and made it public in the US (link1, link2), if not from 2003 if you believe Chiumiento and the tv report. Its not a fuzzy blob or grainy dot. There are moving parts seemingly with function, pertusions, color changes. A particular motion path.
If its is CGI then it has hollywood like lighting, motion tracking (sometimes with bad background reference, zoom changes, partial out of frame movements etc), a motion blur to match with a weird digital image recombination of a camcorder of that time. I bet no VFX artist would have touched that with a pole.
That is a lot of stuff for a hoax from that time just to try to decredit an Italian Ufologist.

Star Wars: Episode II came out 2002 and if you watch the unedited version it is obvious that it is CGI, and they had money, a controlled environment and a VFX expert team.
There are a lot of ufo videos with this detail but most of them are newer and have been debunked. There is always something that was overlooked since most of them are from amatuers. And they dont have have someone coming out after 16 years and say, that it happened.

So instead of focusing on details that are unknown or search for evidence in a witness report, i was hoping to spark more curiousity and to see someone with some VFX knowledge take a look at the data we currently have and can analyse, meaning the video in the best quality available.

What would be the cost of creating such a video in 2003?
Can something be said about that unusual motion blur that is not only on object but on parts on the background as well?
Apart from "standing out" or "looking like cgi" are there some numbers or is there something in the lighting that can be identified as definitely fake?
Are there some practical tricks to help with tracking or lighting?
That is what im interested in.

I'd disagree about the video having Hollywood production like quality for the time period, for one simple reason. The quality of the video is so low that it's impossible to tell if it has such high standards of lighting, motion tracking etc.

The entire scene (including the object) is of very poor quality and the object is airbourne throughout, which means that it would be possible to create movement and size in view (for the changes in zoom) manually, through animation of the object sufficient to appear to track correctly with the camera. Even if this was motion tracked (plus some animation on top, e.g. for the object zipping away at the end), then this is something that has been a part of (or available with plugins) for 3D art tools (e.g. 3DS Max and Maya), prior to 2003. Alternatively, it could even be done without motion tracking, instead animating an object manual, to provide the kind of match the camera movement and zoom that was desired. A bit fiddly to do, but far from impossible, as the object is floating in the air at all times, with little frame of reference regarding size, distance etc. Tracking doesn't need to be correct, just appear close enough to that (in a very low quality video) to make people think that it is.

I'm not saying that a complete amateur would likely have put this together, but a professional VFX artist (i.e. for film or TV production) would not be required, simply someone with some experience using 3D art and animation tools, used for many purposes.

A lot of artists from advertising, computer games development (my background) or other lines of work, could have put something together like this, even as early as 2003 and wouldn't have required a huge amount of work.

The argument that CGI in movies of that period were obvious, doesn't really apply, as movies have clear, high quality video, presented to the viewer, that the CGI needed to match (as best it can), in order to be believable. However, in the case of this video, the quality is so low that CGI added to a separate background video, would not be anything like as apparent as that in a Hollywood movie.

The object and it's movement in the video, avoid a lot of issues that would otherwise complicate the work of someone faking this with CG. It's a simple rounded object (so harder to perceive angle and perspective inacuracies), that never passes behind anything in the scene, nor comes into contact with anything (e.g. landing on the ground). It's also high enough from the ground that a shadow would not be visible, nor comes close enough to anything to require some kind of reflection of it. In addition to that, the video quality is so low that matching lighting/reflections to a background scene video, could be done at a very basic level. Just enough to provide the apparent influence of the general colour of the the sky and ground, plus a reasonable match to the sun light source direction, intensity and diffusion.

Someone attempting to make a video like this could have started with a reasonable quality video (for the time period) of the background scene, then added the CGI of the object, before doing some minor post processing to "muddy the waters" visually (blurs, colour adjustment etc.), then reduce the resolution of the video before heavily compressing it, to arrive at the video that we are discussing.