Does the presence of CG models in the animation really allow for a bet

[ArchieNov]

Sorry, it seems my title was too long. I don't know how to edit it.

But basically I was asking if the presence of CG models in the animation really allow for a better engineered/more screen accurate toy?

I know this has been mentioned a lot and that the earlier Macross valkyries were more difficult to translate into a working toy because of anime magic or the lack of a CG model.

The reason I ask is because I came across a nicely done compilation video of updated animation scenes for Macross (from the Pachinko game, BluRay extras, etc). And this has a LOT of CG models that can be used as reference material if needed.

So my question is, if someone was to redesign the VF-1 toy series and use the CG models in the videos above as a reference, are they going to be able to come up with a better VF-1 than the Yamato v2?

PS: I thought the video was awesome. Personally, (with the exception of the Minmay songs) it was nice to hear the Robotech BGMs again as these were the ones I heard in my childhood.

Some notable scenes:

8:18 Armored VF 1-S

8:28 - 8:47 Britai VS Global

9:13 - 9:37 Hikaru VS Kamjin

10:52 Wowza!

11:14 - 11:28 The Virgin Road VF remains a mystery. Is it really a VF-1D or a 2 seater VF-1A?

Edited September 10, 2015 by ArchieNov

[no3Ljm]

Yeah. I remember mentioning this CG Pachinko videos before in a different thread. And basing on that, I guess it's doable. And if that happens... Spanner will cry over the announcement of Version 3.

But yeah, I'm interested to see a transformable take on this CG VF-1's.

[no3Ljm]

But basically I was asking if the presence of CG models in the animation really allow for a better engineered/more screen accurate toy?

Just to answer this question, yes, it really does help creating better engineered screen accurate (if they used the actual CG footages) transforming toys. Since they already have 3D representation of the unit itself for 3 modes. Proportion wise they don't need to tweak each mode not like when they still hand-drawn. All parts are already fit with each other even if they just need to fold or rotating. The best example I could think of is the valks from Macross Frontier.

[Bolt]

You might certainly produce a different version off CG's now. But I think Yammato did a great job translating from line art and hand drawn animation. There's no exception when it comes to CG based modeling , look at the fact that Bandai had to do v2's of the all the VF-25's. Those were all based on solid CG references as well as line art..and ultimately it's a debate on preferance.

Oh man ArchieNov..can of worms topic..

[neoexcaliber]

When I went through a making-of video for the first Michael Bay Transformer's movie, the CG team did mention having limitations as all their designs and transformations had to be approved by the merchandising department. I'm more curious as to whether the insistence on having merchandise is removing the magic from anime.

When mechas were hand-drawn instead of CG, I used to see lots of very dynamic and probably unrealistic angles and movements. Sure, the newer stuff does look good when they put effort into it like the final battle in the 2nd Frontier movie but compared to SDF/DYRL and Macross Plus, I feel like they're missing a lot of the 'wow', but yes they do make toys screen-accurate as they're all modelled in 3d.

[ArchieNov]

You might certainly produce a different version off CG's now. But I think Yammato did a great job translating from line art and hand drawn animation. There's no exception when it comes to CG based modeling , look at the fact that Bandai had to do v2's of the all the VF-25's. Those were all based on solid CG references as well as line art..and ultimately it's a debate on preferance.

Oh man ArchieNov..can of worms topic..

Hehe sorry, I didn't realize that some would see this as a can of worms topic.

I actually have a similar opinion as you do. I don't think that having a CG model will necessary lead to a better designed (especially if transforming) toy.

But I guess this depends on the quality of the CG model. I mean, if the CG model itself was designed to transform in the same way as it would in the real world, then I would understand that turning it into a toy would be easy. But if the CG model also used screen magic (e.g. Michael Bay's transformers), then I don't really see how this would help that much. (By the way, though not fully accurate, I'm fairly impressed at what the toy designers accomplished in trying to convert Michael Bay's transformers into toy form)

And we also have several cases where not having a CG model doesn't mean you can't have a good toy (e.g. Anime figurines). A lot of the current ones look pretty dead on accurate.

[Ignacio Ocamica]

When talking about valkyries you can design three different CG models (battroid gerwalk and fighter) and blend them and make them transform while still cheating. If you watch Roys VF-0S CG transformation sequence in Macross Zero you'll see ultra thin takatoku style hip bars that magically disappear into the nosecone/forward fuselage when changing to battroid.

So a CG model doesn't necessarily translate into a better/screen accurate toy.

[SuperSenpai]

Basing the toys off of CG models will ensure that proportions stay the same from one mode to another, but it won't ensure that the transformation itself is mechanically feasible. You can still cheat the transformation process using CG by having parts move from one point in 3D space to another with no apparent mechanism for doing so.

[Scyla]

I think we have to differentiate two things. CGI designs that transform from one mode to another in a burst of polygons like the TF4 Optimus Prime where it is hard to find any elements from the truck in robot mode and designs where the transformation follows a logical transition from one mode to another (aka the transformations in the Macross series).

I think that Kawamori tries to follow the second approach to create a design that has at least a plausible transformation.

In combination with the refined production methods and progress in engineering I think that the combination will produce better representations of the on screen models.

However It think it is actual the other way round. Because the Valkyries in Macross have a plausible transformation it is harder to cheat on screen. For example the proportions of the VF-1 in SDF-1 change from mode to mode to have a sleek jet and a bulky robot. With CG models as a baseline you probably won't try to cheat too much because it will look unrealistic and I think Kawamori is shooting for more realism.

I also believe that this is the reason for the Armored VF-25 to have a bulky robot (with armor) and a sleek jet (without armor).

So my idea is that because the shows has plausible design the resulting toys are also more accurate because the don't cheat that much to begin with.

Of course the show can still cheat in some areas like collision detection and exact placement and size of the parts to get a better look. For example placement and movement of the hips that magically are not limited by the wing backpack, etc. I think this is one of the reasons why the VF VF-25 Valkyries failed so massively because they where to reliant on the 3D model (I posted this somewhere else on the boards). My idea was that they took the model from the show and literally translated it in a toy without taking into account that the animators in the show cheated a bit to get more heroic and lanky proportions for Battroid-Mode. Thus resulting in a design that looks like a Messiah Homunculus. With the Renewal versions Bandai introduced stuff like the hip joint that has an additional transformation step to recreate a better looking Battroid-Mode.

The question that is more interesting to me is if this is actually a good thing. For me as an engineering enthusiast I really like all the mechanisms and designs the toy designers came up over the years to recreate the look of handdrawn Japanese robots as screen accurate as possible. Stuff like pull out joints to allow for more movement or spring-loaded flaps in the forearms to allow for a double jointed elbow are great feats of engineering. Also the ROTF leader Optimus Prime or DotM Leader Sentinel Prime have done an awesome job to recreate a physical impossible transformation and are a better (but also complex and maybe not for kids) toy for it.

On the other hand the VF-25 and VF-0 toys are actually pretty simple. all the limps don't need to transform much because they are all solid blocks attached to the toy resulting in a much more standard (i.e. boring) transformation while the VF-19ADVANCE has many built in mechanism to cover for the anime magic of the YF-19 design. Prime example would be the flip out lower leg plates that cover the gaps that are necessary for Fighter-Mode.

[Sildani]

Basing the toys off of CG models will ensure that proportions stay the same from one mode to another, but it won't ensure that the transformation itself is mechanically feasible. You can still cheat the transformation process using CG by having parts move from one point in 3D space to another with no apparent mechanism for doing so.

Seeing as how the 25 uses electromagnetic actuators to transform and move the various things, mechanisms aren't even supposed to be present. Since the toys have to use mechanical means, they're automatically less anime accurate, which should irritate some people. We'll never have a perfectly accurate 25 until everything slides and locks into place magnetically!

[Valkyrie addict]

AFAIK the updated VF-1 CG design on the Pachinko videos are based on the Yammie V2 already.

[ArchieNov]

Seeing as how the 25 uses electromagnetic actuators to transform and move the various things, mechanisms aren't even supposed to be present. Since the toys have to use mechanical means, they're automatically less anime accurate, which should irritate some people. We'll never have a perfectly accurate 25 until everything slides and locks into place magnetically!

Oh I didn't know about that. So for some parts, there's nothing holding them together except for magnetism?

[SuperSenpai]

Seeing as how the 25 uses electromagnetic actuators to transform and move the various things, mechanisms aren't even supposed to be present. Since the toys have to use mechanical means, they're automatically less anime accurate, which should irritate some people. We'll never have a perfectly accurate 25 until everything slides and locks into place magnetically!

And this is where the tension between so-called "anime accuracy" and mechanically feasible perfect transformation comes into play. Anime (or line art) accuracy is all well and good as far as basic proportions and aesthetic details are concerned, but I don't think it's something that should constrain the design of a mechanically feasible PT toy as far as the transformation process. Animators and mecha designers are generally artists, not engineers. They make things that look cool, not things that actually work. Guys like Kawamori do a good job of making sure that the transformation works in broad strokes, but they don't bother with the detailed nuances of exactly how each piece moves and fits into the others using exact measurements. Thus they are free to invoke "electromagnetic actuators" or whatever other magical in-universe explanation as to why this part suddenly shows up over there with nothing attached to it.

In order for a CG model to translate exactly into a PT toy design, you'd have to design the CG model to be mechanically feasible in the first place -- meaning, with all the connections, hinges, and joints to make PT work. You'd also have to animate such that all the parts move as they would in the physical object. But at that point, you'd essentially be making a animated 3D CAD model for the toy. I suspect that the reason why this is not done is because actually designing all the mechanical elements would probably take too much time and expense for the animators, and besides isn't that what the toy designers are paid to do?

[Reïvaj]

AFAIK the updated VF-1 CG design on the Pachinko videos are based on the Yammie V2 already.

That's what I thought.