Modelling and Rendering the TMP Enterprise

[Science Officer]

I've taken a decent step forward with it tonight. My problems turned out to be a combination of 2, 3 and 4.

On 2, I've tweaked the camera position for ref 46 again and sure enough the specular is moving into the correct place on the saucer.

On 3, the following image helps determine where the floor is in relation to the model.

http://www.cloudster.com/Sets&Vehicles/STMPEnterprise/STMPent44.jpg

This helps the radiosity, as illumination from the studio light will bounce off the floor and onto the saucer.

On 4, I wasn't being brave enough with the contrast on the specular channel. On diffuse, I've tweaked the Phong falloff for the lighter shades. Now Cloudster images look a better match as well as the movie snapshot of the saucer bottom with spotlight.

One problem it does highlight is how basic my texturing is. I'll need to put in a lot more effort to break up that darned Aztec! The Cloudster images show a lot of different shades and some large scale patchiness.

Cheers,

S.O.

 

[Science Officer]

Hi Folks,

Things are progressing. Below are some comparison images:

The first two sets use Cloudster images as references. The third set uses a HD scene from the film. The fourth shows the diffuse/specular inversion. It's a bit like some reference images of the NCC-1701A and like the orange saucer image on the previous page.

There are obvious differences. I still haven't got the radiosity correct on the Cloudster images, so parts of the saucer are darker than they should be.

On all images you can see that parts of the Aztec pattern have a higher contrast and others are lower. This is especially obvious in the ring of panels below "NCC-1701". So I've got yet more work on the texturing - add more shades and inconsistency.

Cheers,

S.O.

 

[wjaspers]

Very nice S.O.

When I look at the cloudster pics, and see the bottom dome, it looks like they had the light just a bit further down than you have, could brighten up the darker areas.

Yeah, the breakup, when you study the model, there is much more breakup than you would have guessed in the beginning
Nice job that Olsen did, didn't he
Can you now imagine looking at other artist renders of the Ent, and go searching for their breakup? and findout there is none, just that aztec pattern every one uses.

Don't know anymore how I did it, so no tips from me, but you'll get there, no doubt.

 

[Vance]

You might add a smidge more gray to the aztec pattern, but you're so close I'm thinking you may want to add the few other color elements first and see how it goes from there. I would hate for you to eternally 'correct' a beautiful feature that isn't even consistant within TMP itself!

The only other thing I might recommend is to add some detail to your light source itself (the sensor dome), since the pale light, for some reason, really stands out there. I'm not sure what the best thing to add would be - maybe just a textured recess to give that area some physical presense.

 

[Science Officer]

Hi Wil and Vance,

Wil, concerning the lighting angle, I think it's pretty much correct in 35 and 46. The geometry of my sensor dome is a little parabolic compared to the model and that would throw the specular highlight out. The shadows cast by the sensor array and phaser turrets look correct.

My thinking is that as I increase the brightness of the floor, I'll have to slightly reduce the studio light intensity. That should provide more illumination for the back of the saucer and the specular in 46 (at the 10-11 o'clock position) will decrease slightly.

I'm hoping then that the environment in which the Cloudster images were taken will no longer be a factor. Any remaining differences will be down to texture alone.

Vance, elements of the sensor array including the dome are completely the wrong colour. I just haven't bothered to correct them yet. I agree about the spotlight rectangles - I've got my own lighting rig which is different from the TMP setup (see image in post 32 on page 3). Yet to figure out what to do about those white rectangles. The NCC-1701 decal isn't correct either, but I suspect that may be more than an issue of specular highlights. I think it is slightly transparent and therefore there is a little subsurface scattering and/or reflection from the underlying white hull. On "C -" in the movie snapshot, there may be patches where the decal hasn't quite stuck down as well.

Advancing the texturing is not something I look forward to!

Assuming 4 pearl colours and one coat for each, there would be up to 16 (2^4) shades. I don't think I can go that far, so it will be a case of add a a couple more shades and see how it looks. The other issue is that I'm painting polygons, so there is a limit to the size of detail I can add.

Cheers,

S.O.

 

[wjaspers]

The shadows cast by the sensor array and phaser turrets look correct.

Ah yes, my mistake.

Yet to figure out what to do about those white rectangles. The NCC-1701 decal isn't correct either, but I suspect that may be more than an issue of specular highlights.

For the white rectangles you can use a gradient.

The decal is a noisy very dark bleu/black. So is the specular, noisy.

Assuming 4 pearl colours and one coat for each, there would be up to 16 (2^4) shades. I don't think I can go that far, so it will be a case of add a a couple more shades and see how it looks.

I used 6 different ones, and it looked close enough.

Oh yes, btw, looking good

 

[FalTorPan]

Looks beautiful!

 

[Science Officer]

Hi Folks,

A difficult week and consequently not as much progress as I'd like. I think my computer's on the way out, so I'll have to fork out for a new one.

I've been concentrating on a new texture for the saucer. The old one was too simplistic using 4 shades. The new one uses 11 with specular intensity going from 50% to 100% in steps of 5%.

I'm looking at the Cloudster images and trying to replicate the details. I'm doing this by storing the numbers in an Excel spreadsheet and using conditional formatting to colour the cells. This is easier than trying to make judgements in C4D.

It is slow work, but it is starting to pay off. I've only managed to texture 4 1/2 of the 16 slices. Most shades I'll have to create myself as the Cloudster images only show details for some parts of the saucer.

Just jope I can get this done and backed up before the computer dies. Decided I'll never buy Dell again - are HP any good?

Regards,

S.O.

 

[Captain Robert April]

My uncle swears by 'em...

 

[largo]

Just jope I can get this done and backed up before the computer dies. Decided I'll never buy Dell again - are HP any good?

never wait to back up your stuff. back up now, if you value your work.

i've usually had the best luck with PowerSpec and the like. (never heard of them? that's the point.) they're all commodity items these days, and the less you're paying for a brand name and all the advertising it takes to support it, the more often you can afford to replace the damned thing.

 

[AnyStar]

that's coming along beautifully! i'm rather envious right now

btw, this was typed on a HP. aside from a power supply, well, lets just call it a "malfunciton"... its never given me any trouble at all!

 

[Science Officer]

Well the PC seems to be reliable at the moment. I suspect the problem was caused by a Windows Update preventing the PC getting past the Welcome screen. So I've disabled Automatic Updates to see if the reliability continues.

I don't think my Inspiron 530 is a good PC for CGI work. I used to get a nasty ticking noise coming from the hard disk that also locked up the OS. I disabled Vista's PreFetch and Fast Searching a while back and only recently upgraded it from 2 to 4GB. Those stopped the disk from being thrashed.

But I am tempted to go with a bigger PC with more memory and 4 cores, just to get some of my render times down.

 

[Psion]

Just jope I can get this done and backed up before the computer dies. Decided I'll never buy Dell again - are HP any good?

I've had nothing but trouble with HP laptops, but those problems have always involved shoddy cooling solutions and you're less likely to encounter those with a desktop. I have three Dell laptops in my house, however, and those work flawlessly.

I'd recommend putting one of your own together, if you're up to the challenge. You can then select the best case, motherboard, CPU, graphics card, drives, RAM, etc. that you can afford and build the machine to your specs. I did that with my Linux box two summers ago, and with the exception of a failed hard drive, that computer has served me flawlessly.

 

[Science Officer]

Hi Folks,

The Dell is back up and running thanks to a reinstall of the OS. So far so good.

I've made a little more progress on the new texture with 9 of the 16 saucer segments textured. Images below...

Before I start tinkering with the colours, I aim to get texturing completed on the saucer bottom, sensor array and rim. But it's a slow process teasing the information from reference images.

Talking about reference images - I've noticed that the Cloudster site (the only reference for TMP/pre-TWOK images) is no longer! I don't know if this is temporary or not, so keep a close eye on your B&W image files if you've got them.

Cheers,

S.O.

 

[Science Officer]

Phew! I've now completed the 16 segments of the saucer bar a few bits and pieces. Even though I haven't put in any varying hues for specular highlights, it's working nicely. The diffuse/specular inversion is there as well as a slight purple halo surrounding the highlights. This halo is lacking compared to the movie and I need to tease it out a little bit more.

However, I'll work on the rim next.

Below are two images using the same lighting and different cameras.

Cheers,

S.O.

P.S. Anybody know if the Cloudster website is gone for good?

 

[publiusr]

Have you thought about maybe an Ariel class shuttlecarrier, something with a bit wider saucer hull? I love some of the ST-II era ships in the new SOTL.

 

[sojourner]

^I think he is very focused on the TMP Enterprise as a major project right now.

 

[Science Officer]

She tasks me. She tasks me and I shall have her! I'll chase her 'round the moons of Nibia and 'round the Antares Maelstrom and 'round Perdition's flames before I give her up!

Ahem, Cough! I am pretty fixated on the TMP Enterprise!

The only problem is time - this project is moving along very, very slowly. I'll be old and grey before it's finished.

I'd just like to complete modelling and texturing it, so I don't use any of Dennis Bailey's original v1 mesh. I just see too many problems with the LW port into C4D, and the best way to resolve them is remodelling. But I'm not brilliant at it, so it takes time. Also it's a good feeling to know it's something you've done and can call it all your own work.

But never say never - it would be nice to try something different!

Cheers,

S.O.

 

[wjaspers]

The only problem is time - this project is moving along very, very slowly. I'll be old and grey before it's finished.

No you won't don't forget what you are doing now, the rest of the ship will benefit from.

Modeling will go smooth, the things you are doing now are the thoughest.

And should I see it wrong, I'll buy some coloring for your hair

 

[Science Officer]

Hi Wil,

The shoulders slumped a little last week trying to do the texturing. Trying to pick the details out of the reference photos takes longer than I thought.

The bottom of the rim is progressing, I'm two thirds complete, with the final third requiring refinement. The sensor array may have to wait as it uses a triangular texture and I'll have to put together some masks for that.

I've moved the shading on a little by enabling reflections and giving them the same properties as the specular highlights. It's a little rough and ready, but as the image below shows, it adds more life to the render.

I'm close now to looking at the hues. If all goes according to plan, I'll have something to show in that area a fortnight from now.

Cheers,

S.O.