Already looked great, but the warmer light added another level to it.
If I may offer a suggestion, older BSDF shaders such as Anisotropic are being phased out in favor of Principled BSDF. In essence Principled is a new 'master shader' added just last year, that's able to be tweaked into being any of the old BSDF shaders (anisotropic, glossy, diffuse, etc.) while also adding PBR settings baked in, plus it already calculates fresnel without needing the specific node. Older shaders still have some uses outside of photorealism, but if you're going for that then Principled is the way to go, not to mention that it's slightly better for performance than the older alternatives.
Unless I have an emmissive material such as LCARS or lights, ever material I do is Principled (well, LCARS are more complex, hybrid materials, part of them are Principled). I can share examples of certain looks if you'd like.
For all my finished Trek fan art, please visit my portfolio
This is the default Principled shader set to emulate the default Anisotropic; with metallic set at 1.0, anisotropic set at 1.0, and ani rotation set at .25 (for some reason, the two shaders 'anisotropic rotation' settings are rotated 90º from one another, so if you have rotation at 0 in one, it's equal to .25 in the other).
So from what you've got on the shader nodes you showed before, I would imagine dragging the same nodes the 'color', 'roughness' and 'anisotropic' inputs would set the same results; noting that you have to add .25 to your rotation value, for a total of 0.425.
For all my finished Trek fan art, please visit my portfolio
Its looking really good sean, and its great to see youre still out here. I havent been on this site since 2003 so I am surprised to see all these vets sill around. I remember your bridges were epic back in the day and just looking at your avatar reminds me they really stood the test of time! looking forward to updates
Thanks, good to see all the familiar faces (avatars?) still here.
Here's the latest render:
That's referencing this:
Current state of the build:
The orange area is duplicated from the group immediately to its left including parts of the radial corridor to keep the light going correctly down the length of the set. The camera in this view is the just about centered in the intersection in front of the turbolift. It's worth noting that my set is rotated 45° relative to the actual Stage 9 set - I started with the radial corridor, so naturally everything flowed from that.
I keep looking back and forth between your render and the screen cap, and I don't think you can get much closer. You really have those materials dialed in and your lighting looks spot on.
Thanks! Believe it or not, this is one of the most practically lit sets ever produced for Star Trek. Virtually all of the lighting comes from those panels. Even in filming, it's apparent, at least in TMP, that very little additional "non-visible" lighting was added, though there is one obvious example in the shot above: there's pretty clearly a shining from spot off-camera in the engineering foyer into the intersection seen behind Kirk's right shoulder. I did duplicate that in my render (you can see it in the screenshot on the right side, actually), though I've moved it since to get the render closer to what you see in the film. I haven't done a full render since, but this is a screenshot of the area render I did to test it (still isn't quite right, BTW):
Fun fact, doing those area renders from a camera with a fisheye lens in Blender is a wee bit.. tricky. Blender doesn't distort the wireframes to match the rendered output, so there's a lot of guesswork involved in getting the area over the right part of the set. I suspect this would be a lot easier if I had a more powerful machine with a dedicated graphics card (I'm working on a mid-2015 MacBook Pro which I wisely maxed out on RAM, but there's no getting around the integrated Intel graphics chip).
Incidentally, having read up on precisely how that widescreen stuff is filmed (mainly looking to get the aspect ratio), I actually need to re-render this. It looks like they did something like the Super 35 (common top) example in this image:
That would explain why I struggled so much to match the angle, but fortunately will be relatively easy to reproduce now having seen that (I can get the camera to the right height and angle and not have excess crap on the bottom by literally cropping it as it appears they did).
In very nearly replicating the precise camera angle from that screencap, I have discovered that the width of the concentric corridor in my set is very likely slightly off, likely by as much as two or even three inches. 😀
I suspected that would be the case as that was the one thing I couldn't reliably extrapolate with the level of accuracy I wanted given the dearth of available reference material. That's the one dimension that I couldn't cleanly extrapolate from the TNG set for which I do have pretty reliable dimensions.
As for how I did that camera setup, the last part was this:
That red dotted line for those not familiar with Blender is the area I'll actually be rendering, while the yellow area is the full camera frame. That gets you the setup I mentioned in the previous post where the actual visible area of the film is cropped so that the bottom portion of what is available from the lens is removed to produce the widescreen ratio (and get our fisheye setup looking relatively correct).
I can't guarantee I've actually matched that camera angle completely, but I can virtually guarantee that the inaccuracy of that part of the set is why I can't.
Thats a fantastic recreation, its just about perfect to the reference picture. Beautifully done! I might recommend a little bit of blurring on your brushed aluminum texture, the lines in it are very sharp. *Or perhaps mixing with a large scale noise to soften the apparent pattern a bit
you might also find fspy useful for using the camera to line up the photo against your set. It's good for arch vis alignment, then you'd just have to guess something like your photos' focal length and such.
Looks stunning! The only thing I could suggest is maybe increasing the poly count on the horizontal wall panels, particularly the light ones; when they meet the bulkheads the segmentation is a bit noticeable.
For all my finished Trek fan art, please visit my portfolio
Looks stunning! The only thing I could suggest is maybe increasing the poly count on the horizontal wall panels, particularly the light ones; when they meet the bulkheads the segmentation is a bit noticeable.
Thanks. I don't expect most people to notice any segmentation unless they're really looking for it - I'm rendering these at a much higher resolution than I expect would ever be common and I only spot it in a couple of very near-distance objects. Either way, it'd be way too much trouble to change that now - I'd have to rebuild all the panels and that's just not going to happen.
I didn't notice the segmentation until Rekkert pointed it out, and I'm usually decent at spotting segmentation. And, even though I know it's there, it still isn't bad. It's not in your face, hence me not noticing it earlier.
Looks spot on from what I can see. You've done a great job here. Just curious, how far are you taking this? Mainly wondering what you have planned to do next.
Looks spot on from what I can see. You've done a great job here. Just curious, how far are you taking this? Mainly wondering what you have planned to do next.
Thanks!
I've been actively debating what to do next, and I think it'll depend a lot on what I can find in the way of reliable reference material. I'm considering the transporter room, sickbay, or possibly something not strictly canon that can branch directly off the corridors and provide some added visual interest. I've also played around with a variant of this set to fit the TFF bridge design scheme and have significant ideas around that as a result. There's also the possibility of an Excelsior variant including a never-before-seen Excelsior engineering.
That said, I also have some ideas floating around relating to a ship concept placed a century or more after the TNG movies (I have a sketch of the exterior already and started building part of that; also have some parts of a bridge concept floating around in my head for it).
All of those ideas sound very cool! As far as references go, I imagine you already have it, but RedGeneral's archive has all of the publicly available building schematics on display.
Post edited by Rekkert on
For all my finished Trek fan art, please visit my portfolio
Posts
If I may offer a suggestion, older BSDF shaders such as Anisotropic are being phased out in favor of Principled BSDF. In essence Principled is a new 'master shader' added just last year, that's able to be tweaked into being any of the old BSDF shaders (anisotropic, glossy, diffuse, etc.) while also adding PBR settings baked in, plus it already calculates fresnel without needing the specific node. Older shaders still have some uses outside of photorealism, but if you're going for that then Principled is the way to go, not to mention that it's slightly better for performance than the older alternatives.
Unless I have an emmissive material such as LCARS or lights, ever material I do is Principled (well, LCARS are more complex, hybrid materials, part of them are Principled). I can share examples of certain looks if you'd like.
So from what you've got on the shader nodes you showed before, I would imagine dragging the same nodes the 'color', 'roughness' and 'anisotropic' inputs would set the same results; noting that you have to add .25 to your rotation value, for a total of 0.425.
Frequent updates at our Discord channel!
Welcome back. And welcome to Blender.
Various Work: U.S.S. Constellation - Matt Jefferies Concept Shuttle
Here's the latest render:
That's referencing this:
Current state of the build:
The orange area is duplicated from the group immediately to its left including parts of the radial corridor to keep the light going correctly down the length of the set. The camera in this view is the just about centered in the intersection in front of the turbolift. It's worth noting that my set is rotated 45° relative to the actual Stage 9 set - I started with the radial corridor, so naturally everything flowed from that.
Fun fact, doing those area renders from a camera with a fisheye lens in Blender is a wee bit.. tricky. Blender doesn't distort the wireframes to match the rendered output, so there's a lot of guesswork involved in getting the area over the right part of the set. I suspect this would be a lot easier if I had a more powerful machine with a dedicated graphics card (I'm working on a mid-2015 MacBook Pro which I wisely maxed out on RAM, but there's no getting around the integrated Intel graphics chip).
https://upload.wikimedia.org/wikipedia/commons/thumb/2/2c/35mm_film_common_formats.svg/440px-35mm_film_common_formats.svg.png
That would explain why I struggled so much to match the angle, but fortunately will be relatively easy to reproduce now having seen that (I can get the camera to the right height and angle and not have excess crap on the bottom by literally cropping it as it appears they did).
I suspected that would be the case as that was the one thing I couldn't reliably extrapolate with the level of accuracy I wanted given the dearth of available reference material. That's the one dimension that I couldn't cleanly extrapolate from the TNG set for which I do have pretty reliable dimensions.
As for how I did that camera setup, the last part was this:
That red dotted line for those not familiar with Blender is the area I'll actually be rendering, while the yellow area is the full camera frame. That gets you the setup I mentioned in the previous post where the actual visible area of the film is cropped so that the bottom portion of what is available from the lens is removed to produce the widescreen ratio (and get our fisheye setup looking relatively correct).
I can't guarantee I've actually matched that camera angle completely, but I can virtually guarantee that the inaccuracy of that part of the set is why I can't.
Current Projects:
Ambassador Class
Thanks for the suggestions, I'll be sure to check them out.
I've been actively debating what to do next, and I think it'll depend a lot on what I can find in the way of reliable reference material. I'm considering the transporter room, sickbay, or possibly something not strictly canon that can branch directly off the corridors and provide some added visual interest. I've also played around with a variant of this set to fit the TFF bridge design scheme and have significant ideas around that as a result. There's also the possibility of an Excelsior variant including a never-before-seen Excelsior engineering.
That said, I also have some ideas floating around relating to a ship concept placed a century or more after the TNG movies (I have a sketch of the exterior already and started building part of that; also have some parts of a bridge concept floating around in my head for it).