3D Master
Rear Admiral
CGI could be better than model work, except that the CGI guys, under contract by producers and such, don't make it better.
CGI would allow one to do things a model wouldn't be allowed; like making ships meet each other NOT both straight up; meeting in every angle possible. Similarly the ships could move at every angle.
The problem with CGI is the problem that SF SFX have been plaguing since Star Wars; stylish, brightly lit, flashy, and thus extremely 2D pictures. They look like cartoons.
The reason why model work thus looks better, is even with the flashy overlit stuff that makes it look very flat, it's still an actual, solid model, an object, being filmed.
CGI by contrast isn't an actual model, and so when they make it look brightly light, and flashy, shadows become either non-existent or exist only lightly, often with conflicting lightsources: it becomes extremely flat, even more so than wrongly lit models.
You can see this extremely well in Star Trek versus the new effects versions and later TOS mvoies. The original Star Trek was before the Star Wars flashy convention; they're budget was small, so they wanted to get as much out of it as possible; they lit and filmed the model to highlight that it is an actual solid object; you can see it physically there. Watch the later TOS movies still with models, and you'll notice they are much less physically there; looking very flat and 2D - completely wrong lighting. Then check out the remastered new SFX, and you notice it is even flatter than the original effects and the TOS movie effects.
The problem is thus with the lighting and the conventional SFX for space ships to be brightly lit and visible - NOT with CGI versus model work itself.
Modeler's on the net, aren't limited to that, so to illustrate, I'll put a couple of pictures here; and I'd like you to say which are models and which are CGI. Now, don't reason about it; especially if you know the models and look for differences in the CGI mesh used, you could reason it out, so you should just look at them a few moments, and then decide, model or CGI:
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
CGI would allow one to do things a model wouldn't be allowed; like making ships meet each other NOT both straight up; meeting in every angle possible. Similarly the ships could move at every angle.
The problem with CGI is the problem that SF SFX have been plaguing since Star Wars; stylish, brightly lit, flashy, and thus extremely 2D pictures. They look like cartoons.
The reason why model work thus looks better, is even with the flashy overlit stuff that makes it look very flat, it's still an actual, solid model, an object, being filmed.
CGI by contrast isn't an actual model, and so when they make it look brightly light, and flashy, shadows become either non-existent or exist only lightly, often with conflicting lightsources: it becomes extremely flat, even more so than wrongly lit models.
You can see this extremely well in Star Trek versus the new effects versions and later TOS mvoies. The original Star Trek was before the Star Wars flashy convention; they're budget was small, so they wanted to get as much out of it as possible; they lit and filmed the model to highlight that it is an actual solid object; you can see it physically there. Watch the later TOS movies still with models, and you'll notice they are much less physically there; looking very flat and 2D - completely wrong lighting. Then check out the remastered new SFX, and you notice it is even flatter than the original effects and the TOS movie effects.
The problem is thus with the lighting and the conventional SFX for space ships to be brightly lit and visible - NOT with CGI versus model work itself.
Modeler's on the net, aren't limited to that, so to illustrate, I'll put a couple of pictures here; and I'd like you to say which are models and which are CGI. Now, don't reason about it; especially if you know the models and look for differences in the CGI mesh used, you could reason it out, so you should just look at them a few moments, and then decide, model or CGI:
1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

Last edited: