Jump to content

bac9

Members
  • Posts

    470
  • Joined

  • Last visited

Everything posted by bac9

  1. B9 update is in the works and mostly depends on the free time of Taverius who has to look over new configs. Some changes: - Reworked HL part set, everything is now using just one texture instead of four - New HL parts (proper tail without a cargo door, new 2.5m/3.75m adapters, round versions of some parts. - New MK2 bicoupler (short one, very useful for some designs) - New set of massive station-building parts, somewhat similar to LLL designs (all using one texture, around 20 part types, including engines and 4-point docking ports)
  2. First screenshots look interesting, although I'm noticing a lot of smoothing issues on every model showcased so far. Is that a Blender viewport issue or are smoothing groups indeed not set up properly?
  3. Absolutely great work, huge improvement over stock textures! The only thing I'd recommend changing is red captions on the engines: the serif font is very out of place and deep red on dark gray is not a good looking combination in general. I realize KW originals had something like this too, but I think it was one thing not worth reproducing from the otherwise brilliant quality KW texture work.
  4. Great work here, glad to see someone with the patience to make more IVAs
  5. No point to make it as there is no fixed standard thickness for wing parts and it will be horribly clipped into surfaces 99% of the time. Using separate per-surface airbrake parts is better considering the variety of parts around.
  6. There is no point in adding parts like those as they will only provide vertical gear attachment from one fixed angle around the hull (not to mention it's not the best idea to add extremely specialized decorative parts). It's entirely possible to make nice looking "bulges" with 1.25m cylinders and existing endpieces, which I did on Strugatsky sample craft, I think. It's also impossible to make a "bulge" part with integrated wheels at the moment: as KSP has no proper mirroring, so only way to make that work is using two separate manually mirrored part types (in which case you won't even be able to perfectly align the wheels). SABRE intake is facing roughly seven degrees down from your direction, which makes it better for high-altitude gliding where you can't ever stay level if flying aligned with your velocity vector. Other than that, nope, stock SSTO is all about intake air numbers.
  7. I'll look into it, but sometimes it's better to use as little unique detail as possible if you'll have dozens of parts showing the same area of the texture next to each other. One part with a cool burnt edge looks nice, ten parts with the same repeated edge aren't. By the way, every single part in this ~50-part addition is using just one texture atlas, so it would be a very lightweight update.
  8. If you mean adding thermal protection plating and other stuff to those big parts, nope, won't happen. They aren't meant to be used for atmospheric craft at all. No, there is KW Rocketry for that, zero need to make another mod filling the same niche. This is largely a physics issue and not something that can be solved on the side of a part model or a config, so I can't really help here.
  9. There are example crafts included with the mod and every single one of them was built by Taverius. I don't have anything stock+B9 based that could be shared on top of already vast selection of excellent designs Taverius created.
  10. Something to dock your spaceplanes to. I think I've posted these many times before. Few old albums: http://imgur.com/a/uSost#0 VTOL http://imgur.com/a/jlo0h#0 Mothership from the previous page, two of three sections (it's too large to assemble in VAB in one piece) http://imgur.com/a/pSGsm#0 Whatever that is
  11. (for scale - those are four 3.75m tanks slotted inside the hollow section on the last screenshot)
  12. That can be one of the reasons, yep. You can always bump weight limit on the parts so see if it helps, but it's not going to solve some situations like those.
  13. Correct. In addition to that, wheels actually aren't behaving like wheels in most games (including KSP). Wheels are essentially one ray shooting downwards and expecting the contact on a certain distance, not a simulated circle or a cylinder at all - which is why wheels never work properly when tilted and should always be installed strictly perpendicular to the runway if you don't want unexpected jumps.
  14. There is no stability parameter at all, and trying to improve it through means of adding multiple wheel axes per part is not possible either. No release estimate at the moment.
  15. Not really, the detail texture is a simple overlay that leaves the original texture resolution completely obvious and visible (not to mention that current implementation of detail texturing is introducing very noticeable tiling). It works well for stuff like closeups of concrete walls in FPS games, but not for masked clouds. Silhouette detail is the issue here, not the frequency of detail in the filled areas.
  16. And what's the point in that when it will be exactly as stable as every single wheel I have already added?
  17. As I've said multiple times, not going to happen, KSP can't support multiple wheel axes per part.
  18. I think there is a misunderstanding as to what noise means too. It's not this: Noise algorithms are used everywhere to generate extremely complex patterns like those: So it's a great way to render clouds, there is nothing about them requiring a predefined source texture that much.
  19. There are ways to draw volumetric clouds without billboards, which would be the best way if you want to get planetary-scale coverage while still getting decent detail and as little overdraw as possible. One example of that approach: http://www.simul.co.uk/truesky (used in ARMA3 and few other products), which probably relies on some low-resolution coverage mask and then draws very neat volumetric clumps using random noise filters. I recall plugin for clouds with a similar approach being available for Unity too. P.S.: I think that's the one I was thinking about: http://forum.unity3d.com/threads/115825-RELEASE-Nuaj-the-3D-clouds-simulator The documentation offers more indepth look into how it works and how it's configured, I believe. http://www.nuaj.net/Download/Nuaj'%20-%20User%20Documentation.pdf Webplayer demo: http://www.nuaj.net/WebPlayer/ Notice how it's completely procedural and therefore allows you to seamlessly make transitions from clear to overcast weather without any heavy source textures involved.
  20. Would be nice if someone will pick that up and make an optimized version with per-craft instead of per-part cubemap rendering.
  21. I don't recommend using non-square textures, to be honest. They are pretty badly managed by the GPU and in worst cases can be automatically converted to a square of a nearest size, thus adding dead weight without actually using it for anything useful. It would be especially a problem with enormous 8k textures many like to use with this mod. I'd recommend either to find a way to use your triplanar with a 1:1 textures, or store multiple texture types in one file. On other subject, I think the enormous size of planetary bodies and desire of everyone to, for some reason, have completely unique clouds on every latitude and longitude (creating planet-wide texture atlas instead of a noise-based unrecognizably repeating pattern) makes the current implemementation a bit of a dead end in terms of looks. Clouds will never look good from low orbit with the current approach if raw texture will continue to be used. The problem can be solved, though, if there will be a per-pixel shader that generates visual effect (bunch of cloudy noise, for example) using the texture only as the source mask, never showing the size of actual texels. Then, as the next logical step, it's possible to ditch the source texture altogether (saving enormous amount of memory), generating the mask from few noise filters like PQS system does. Many games use a similar approach. P.S.: I can't recommend using particle systems for volumetric clouds at all, that approach gives insane overdraw, lighting problems, and clipping problems.
  22. That's a pretty cheaty goal, to be honest. What's the challenge in not having to use fuel?
  23. This is pretty cool to hear, I'd love to integrate something like that into B9 to prevent cheating with radial intake spam or intake stacking.
  24. I'll look into releasing an update in the coming weeks. After all, there is still a pile of unreleased parts over here.
×
×
  • Create New...