The current shovel (found under the handheld weapons/klappspaten) is about 3000 tris for it's GEOM0 (first person perspective) so sure, feel free to optimize it but you can also afford to throw a few more tris at it.
Most important thing will be to actually get decent LODs for the GEOM1 (third person perspective).
For example the klappspaten is only ~850 tris for the LOD0 in GEOM1. So you'd have to optimize quite a bit for the 3rd person perspective to be on par with the current standard.
_________________________________________
Edit: That being said, the optimizations for GEOM1 would be pretty easy, and you would easily save huge tri-counts by:
Reducing complexity of the rounded part at the bottom of the handle.
Reducing number of smoothing sides on the handle/grip (The purple part basically in pic #4)
Changing the majority of the raises/bumps on the shovel face to normal maps, except maybe the middle one which gives it some shape.
I understand that. I figured it easier to make a higher-tri model and make it less, rather than the other way round. I can remove a lot of tris on the handle and finer details (bolts 'n such).
ZondaX15 wrote:I understand that. I figured it easier to make a higher-tri model and make it less, rather than the other way round. I can remove a lot of tris on the handle and finer details (bolts 'n such).
For sure! I thought you were asking for advice, maybe I misread your post.
Hey, I am fairly new to modeling too, if you want to check out my flak 88 thread I've asked a bunch of questions and gotten tons of answers, itll give ya some idea what to do
I've seen your Flak. It's a wonderful model. Considerably more complicated than mine . I've already read through your thread a few times. Keep up the good work, man
Alright, so when I make the three different parts become one, it always destroys the UVW map of the thing which I am attaching everything to. If that makes sense.
Edit: Got it fixed. Just had to redo some of the UV map.
Last edited by ZondaX15 on 2012-10-02 03:52, edited 1 time in total.
This looks like a version of the main shovel MD made for opfor, ie, his normal shovel but with a wooden handle. Don't think it ever got ingame thou for some reason, not sure why.
Got any refs for it?
[quote=""'[USF"]Doc.Pock;1820551']commentary in the pic. COLOR LINKED...
[/quote]
The bit marked in red on pock's image above needs to be removed, if this is a low poly model that is meant to go ingame like it looks like to me. It will not only zfight like crazy but is also not needed, since a normal map can do just the same job perfectly well with saving both tris and UV space and not zfight.
No point on working on UVs until the model is done otherwise you will just end up doing them over and over.
Right now I would work on fixing the smoothing of your model as that needs quite a lot of work tbh.
[quote="Ratface""]That's a lot nicer looking than my flak UV will be, that thing is going to be nucking futz when I start it -.-[/quote]
I hope not as that kind of UV wouldn't come close to acceptable tbh.
Like a noob (which I clearly am), I was working of only one ref. Mainly because I couldn't find any other photos of the same model of shovel.
I've had a go at the smoothing, but I can't really see where I went wrong. Mainly because I have nothing really to compare it to. Aside from other models, but that doesn't really help me. I'll have a better crack at it tomorrow.
Out of curiosity, what is it about the UV that isn't acceptable?
Ye, from memory that looks like pretty much the same shovel MD made but all his pics on it are dead so can't say for sure
EDIT: one of the refs MD worked off for the wooden handle:
ZondaX15 wrote:Out of curiosity, what is it about the UV that isn't acceptable?
Mainly the packing side is the worst bit as your aware of, so much wasted space which is really unacceptable since even with a high rez texturing, taking up quite a few mbs, it will still look pretty low quality. Need to maximise the mount you get out of a texture for its cost and a good UV map means you can have a much higher quality result, for less mbs, or at the very least the same mbs, but I suspect with better packing, you could get a better result with a 512 texture, than you could with your current pack with a 1024 texture, possibly even a 2048.
. I've already read through your thread a few times. Keep up the good work, man 
