HK UMP .45

[frell]

Im breaking this out of my other thread

High poly
http://img443.imageshack.us/img443/9348/newpic.png

Low poly wire
http://img156.imageshack.us/img156/5634/wirenew.png




I did a quick flatten map bake before uving just to catch any errors. Obviously some floating geometry didnt export right but itll all be fixed

Im also unaware of the texture standard for fps weapons, i know 2048 is good for the texture but what about normals? Ive baked out both:

http://img18.imageshack.us/img18/8864/bake.png

http://img9.imageshack.us/img9/6131/bake2.png

Crits before uving?

[StefanH]

I'd like a bigger wire which is not see-through and not triangulated to better judge the topology.

Having built alot of guns at work with similar polybudgets i can say already that there is potential for better poly usage. For a firstperson weapon you want to chamfer the edges of the sights so they appear less blocky when you aim.

Also your edges are still very hard. Looks almost like you are not using a normal map. You want the nice curved edges that catch nice highlights.

[frell]

I havent been making guns very long but on every gun i seem to under shoot my poly limit by a good 1-2k

Heres the wire
http://img101.imageshack.us/img101/4444/wrie.jpg

[Kitteh]

Curve the top of the stock! It'll be right in the player's face in first person view.

Also, you've got a lot of unneeded segments on your low poly, like above the trigger. And your flash hider is still wrong.

[frell]

Whats wrong with the flash hider? I modeled it the same as it is in racers reference he showed me...

And im very used to modeling box like and keeping everything equal, what should i do with the un-needed segments?

And will do to the stock.

[StefanH]

also why did you put so many polies in the trigger? that's an area a player is very likely to never see or just a fraction of a second on reload!

[frell]

Oh for some reason thats how it rendered out, theres maybe only 12 polys on the trigger.

[an aggressive napkin]

Looks alright, The highpoly should really have softer edges to get the normal's detail onto the lowpoly perfectly. And this is plastic so they should be softer anyway

[CGvanHoudt]

I don't get why this scene is 4100 tris, also I feel like you have some technical errors in your low poly mesh, you might want to take a look at that.

[frell]

Starting the texturing, heres my base color and the final bake in marmo:

http://img704.imageshack.us/img704/240/texwip.png

Heres the normals and the uv mapping
http://img820.imageshack.us/img820/5456/umpnormals.jpg

[Firebert]

There are larger things at play here than your poly count. You need to really sit down, take the time to address these areas, and not rush through things. I am only writing this because, simply put, you still do not seem to grasp the idea of making a quality asset. Most everyone here on this thread and the last have given you sound technical advice, which you should follow, but before you do any of that, you should address the points I've made below.



Become overly self critical - You have a willingness to make art for games. That much is clear. What you do not have is the ability to self-gauge your own work by being self critical. I'll repeat it again as I stated in the previous thread; you have to detach yourself from the work you complete as your own and view it from the perspective of someone else. Simply put, no employer will want to hire you if you cannot critique your own work with the utmost scrutiny.
Observe, create, re-observe, tweak - Reference is the key to making a successful asset such as this. Sometimes you can do guess work, so long as it is logical guess work, but for assets such as weapons or items that need to feel very life like, you need to really, really, really pay attention to your reference. You have it right there in front of you, so you have no excuses for it to be wrong. This goes hand in hand with being self-critical.
Learn to do it the RIGHT way - You can't get it "close", call that part finished, and move onto the next. Why? Because you can't. That's all there is to it. Learn to do it the right way, no matter how long it takes you. You will have many failures, but you will also see better results in the end. This is only going to work if you make use of the two points above. Topology, poly distribution, UV mapping, texturing; these are all pieces to a much larger puzzle that takes endless practice to get right.
Slow Down - You are moving way too fast. Slow down, and take note of the point above. You will get faster with time, you just aren't there yet, and that is okay. If you take it down a bit, you'll be able to complete the previous point with a greater understanding. THEN you can speed it up a bit.
Lose the ego, drop the excuses, and listen to critique. - There is an evident sense of pride at play with your posts with repeated excuses as to why something was made the way was. If you don't kill your own ego with self criticism, then you will have a greater sense of failure and frustration later on down the road. Excuses are meant for those that refuse to account for, or accept, their own failures. LISTEN to what people are telling you. Most everyone here knows what they are doing or are willing to learn. You seem like you just want to get it done however you feel it should be done, and the industry just doesn't work like that. A fine example of this is how you just completely ignored most everyone's critique and dived right into a bad UV map and an OS bake.


Try working on these ideas first before working on all the technical crit you have received.

This may seem a bit harsh, but I promise you that if you really try to work on these things, you will see yourself grow much more as an artist. The unfortunate side of this post is that I really feel you will completely ignore it.

[frell]

When i accept critique i try see if i can work it into the existing model, if i cannot then i try to not make the same mistake again on my next.

At the end of my other thread i fixed what racer suggested, and i didn't really get any critiques after that other than the edges shouldn't be so hard. So i told myself "ok, im going to try not to do that again".

In this thread someone said the flash hider was still wrong, but i don't see how it is because its been changed twice, both using reference. You can't say im ignoring critique, because my uvs look alot better and i have changed alot of stuff on this model due to critique. On my ak i got so much critique that i decided to just ditch the model and try to make another one alot better, and to assure that i posted a wip thread, which i don't normally do. Would it help if i posted WIP's of my uving from now on? I don't really see people doing that so i didnt do it. Ive also been accepting critique for the ump for 1-2 days and changing as much as i can. I like to keep working but still leave room for critique changes, because im trying to learn as much as i can and the only way i can do that is by practicing ALOT. And i tend to finish a gun in about 3 days.

Ill have to agree with your third bullet the most. I have a tendency to do that and ill admit it. But ever since i started getting into weapon modeling a month ago i've been trying to push myself to make it the right way. And you're right, i feel alot better about it in the end.

Why don't you tell me whats specifically wrong with it?

[mheyman]

Can you post an actual straight on side view of your gun, one that matches the angle of your reference you posted. It's easier to critique errors in size and shape when you post up a side view instead of slightly off angle perspective views.

http://img295.imageshack.us/img295/825/umpm.jpg

[bbob]

Its pretty darn important to be able to go back and fix things, such as the sharpness of edges, before you bake the whole thing down.

[frell]

Sure. The barrel is different than the reference because of a suggestion by racer, which showed reference of a different barrel variant.

bbob, it really (for me) depends on the edge. If it was maybe 2 or 3 main edges that were too sharp i'd probably fix them. But im assuming that they're meaning the overall edges of the model, in which case ill try not to repeat.

http://img139.imageshack.us/img139/2526/gungund.png

[Racer445]

re-read my post bro, you missed a lot of stuff

make sure you view a LOT of references. every angle, not just the side. with weapons we have this luxury so make the best of it

[mheyman]

I actually meant a shot of the low poly (or high poly, doesnt matter to me) shaded. That makes it easier to compare how the light is affecting the surface. I can already tell from the wire frame that there are some subtle details that may have been left out. Like at the bottom of the clip and the handle (I don't know gun terminology sorry).

[frell]

I was using http://www.pixagogo.com/2778744214 as references. Before i model i usually go through and find the most complex shapes and break down how ill make them, then i begin modeling side by side with reference.

[Firebert]

When i accept critique i try see if i can work it into the existing model, if i cannot then i try to not make the same mistake again on my next.

I like to keep working but still leave room for critique changes, because im trying to learn as much as i can and the only way i can do that is by practicing ALOT.


*insert face palm smiley*

Do you not see how contradictory those two statements are?

I've already taken my time to help you out, and Racer gave you an excellent post with really great reference and pointers, which as he said, you overlooked a lot. So why should I go back and tell you what's wrong if it is overwhelmingly apparent that someone else already has AND that you can't critique yourself enough to make it right?

[frell]

Alright, well right now id like to know more about where i stand because you guys sure are good at making me feel like i might be falling down the wrong flight of stairs :poly122:. Here are all the guns i have made since i started making guns, which was about a month ago.

http://fc07.deviantart.net/fs70/f/2010/149/2/4/Colt_1911_Pract__by_Frell262.png
http://fc08.deviantart.net/fs70/f/2010/150/8/d/Lever_action_game_rifle_by_Frell262.jpg

http://fc05.deviantart.net/fs70/f/2010/158/f/0/Thompson_SMG_by_Frell262.png
http://img37.imageshack.us/img37/9047/newnew.png

And heres my current gun. Im having some trouble with the specular because im really only used to metal.
http://img143.imageshack.us/img143/7427/tdxchvjkn.png

How is my progression so far, and am i behind / ahead for my 5th gun?

[PixelMasher]

there is definitely progress, but by not going back and fixing things you are not readily applying critiques to your art work and that is a key thing in a production environment. there is no "I wont do it on the next one"

When you are making game assets for a living, if it needs to be fixed, fix it. Thats the main thing I think people are trying to convey to you and something you need to get used to, I know its annoying to go back and fix/redo things but the end result after some polishing is usually 2-5x better than the 1st pass.

when you are starting out there is the urge to produce a ton of art really quickly but if you slow down a bit you might have fewer pieces but they will be higher quality. qaulity trumps quantity every time when it comes to portfolios, I would rather hire someone with 3-5 awesome peices than someone who has 20 so so pieces, and thats what people are trying to help you out with here.

if you only apply critiques to you next model there isnt really any point in in depth crits like racer's where he lists a ton of inaccuracies, so you will find less and less people crit your work. ass pats dont help you improve, applying critiques and listening to feedback do and is a critical skill companies look for.

that said, your UMP is lookin nice, the UVs are better than the last one but could still benefit from a major overhaul/tightening up. the time it would take to re-do them and get back to where you are now wouldnt be too bad, you already have a decent highpoly and thats half the work on this piece done already, shoot for the high quality result in all aspects of it man!

[frell]

Hmm. I think what i've done is jumped into uving and the low poly way to fast, as firebert said. I know you're getting tired of hearing this but on my next model ill have a couple more days open for crits and fixes on the high poly before i start removing loops and making the low poly.

Just for future knowledge, if i collapse the Unwrap UVW modifier and then add new geometry to the model via attach, will it mess anything up?

[mheyman]

Yeah I still think you completely ignored some curvature in the handle/grip (where you hold the gun near the trigger) Right now it just looks like a block.

I don't really understand what you're looking for? Praise? If you're on the right track? I don't see the logic in ignoring critique until the next project.

Let's see what you have now: 5 incrementally better guns, all with flaws.

What you could have: 5 very nice polished assets, critiques were addressed, probably less flaws.

Do you see what I'm saying here? What I see happening from here. You apply some critique to your next project, you get more suggestions, and then ignore those until the next gun.

[frell]

Yea i see what you guys mean. Im going to try not to do that anymore by waiting longer for critiques and then fixing them before i go any further.

I figured the handle gets covered up by a hand 99.9% of the time :s

[Taylor Hood]

Well, I can understand that there may be technical errors but that latest one looks pretty kickass.

[Racer445]

where is the hinge and latch and all the nice shapes from the iconic UMP stock? view more refs and dont rush!

http://pix.pixagogo.com/S5ODTr%21fWlbmTEEcG52ENtlqd076Z%21P5Nza%21VnbltT9V aj00Tynkg5sw82kWpj2V4JFesLzGL6yQJTwInom3kJ42%21sZJ C-nvr%21/00021.jpg

How is my progression so far, and am i behind / ahead for my 5th gun?

who cares? people are all different in the speed that they progress. slow down and make a few really nice items before trying to speed up anything. technique over speed!

[EarthQuake]

who cares? people are all different in the speed that they progress. slow down and make a few really nice items before trying to speed up anything. technique over speed!

I just want to sort of say something in addition to this, rushing isn't the same as speed. I'm not saying you're rushing, but there is a very very important difference here. In a studio environment, rushing to get the job done and making lots of mistakes is going to cost you a lot more time when your art director gives you a huge list of revisions that you need to fix. Slowing down and doing things methodically and carefully, and not rush, is virtually always going to be FASTER.

[Racer445]

+1

[frell]

Anyone want to help me pick my next gun? :P

[bbob]

You need to capture the roughish surface of the polymer (plastic bits), which is should be way less shiny than the metal. Also, do some small scratches and wear on exposed edges, especially on the picatinny rails, as they would probably have seen a rotation of attatchments. This should be accentuated on the specular, even if the paint is intact, because it gets polished before it wears off.

Reference on the difference of materials
http://photos.imageevent.com/smglee/ss2008/huge/DSCN0077.JPG

Ref for worn edges on guns:
http://www.oldwestgunsmith.com/guns/3425rclose.JPG
http://gunpics.net/uk/no4mk1/No4mk1_010.JPG
http://gunpics.net/uk/no4mk1/No4mk1_013.JPG
http://gunpics.net/uk/no4mk1/No4mk1_018.JPG


Oh and for the love of everything you cherish, do not hesitate to go back and change stuff, it is a crucial part of being a professional digital artist of any kind. No matter how closely you are paying attention to detail and accuracy, there will always be something you miss, and your superior will most likely spot it and demand it changed. So please, for your own good, change that awful attitude at once.

EDIT: You are calling that result finished after two days of work? I could understand if you were struggling to get it done, but considering the time you put into this "quality", you are definitely selling yourself short.

[Stromberg90]

Anyone want to help me pick my next gun? :P

Here is a suggestion: HK UMP .45

Do as the other says and get this one done properly, if you start another thread people wont reply cause they are tired of you not listening.

[frell]

http://img18.imageshack.us/img18/6967/umpr.png

I sorta do feel like remaking the stock right and then fixing up the uvs.

So since im probably going to do that, what else should i fix in the high poly while im at it?

[Taylor Hood]

Yeah, I've noticed you go from Modeling to Texturing really quickly. Stay on one gun and make it perfect, mate. Look at all the crits you're getting! =]

[frell]

Id also like to re-ask if attaching new geometry to the low poly mesh will mess up any uvs i already have? Assuming i collapsed the Unwrap UVW modifier.

[EarthQuake]

Frell: You're showing some great ambition, but hopefully i can give you a little advice so you can avoid needing to start over entirely on your next asset. What you want to do is break it down into stages.

EDIT: 0. Gather reference, study, plan, and think about all of the shapes and how you will go about constructing them.

1. Highpoly, post multiple wips of this, get feed back and implement it(possibly multiple rounds of feedback) before you even think about starting your low. Very important to listen to suggestions on what will/wont bake well.

2. Lowpoly, post multiple wips of this, get feed back and implement it(possibly multiple rounds of feedback) before you even think about starting your low. Pay very close attention to suggestions relation to polygon usage, and anything that has to do with making edits so that you'll get a good bake. You may still want to tweak your highpoly a little bit during this stage, to make sure things will line up as accurately as possible. Sometimes you'll want to nudge over a piece on your high to be able to simplify your low.

3. Now its time to do a TEST bake. This is not your final bake, you will want to use temporary uvs. Here i like to have "proper" uvs unwraped, however it simply is not the final packing for me, just make a copy of these and do an auto-pack, its good enough for a test bake. With our test bake we want to make sure that all of the details are projecting accurately(minimum skewing or wavy lines) that we have enough lowpoly geometry to support the shape, that our highpoly details arent too fine to read properly(ie that we have enough pixels to represent all of our details). And of course that there arent any smoothing errors or other technical problems(very important with tangent normals).

This is the last time you will ever want to touch your highpoly mesh, make any final tweaks now. Post for feedback and implement anything else still unresolved, this is pretty much the no-going-back point. Any model changes need to happen now.

4. Pack your final uvs, post for feedback and do not do anything else until you are sure you have a good layout.

5. Do another testbake, just rendering normals, to double check everything you saw in #3(that everything bakes well). You dont have to post this stage, but it is an important stage, just incase you need to make any last minute uv changes.

6. Do your final bakes, normals, AO, color(assign materials with RGB colors to your high and bake this to create a mask for texturing) anything else. Post these results as well.

--- Texturing

7. Create your base materials, dont do any detail work, scratches, wear etc. Just work on getting the materials to look like "wood" or "metal". Post and take feedback.

8. Work on unique details, scratches, wear, etc. Post and take feedback.

9. Do any final touches, little details, set up a nice scene or something, take some big ol' screenshots, size them down to get some good AA, post them on polycount and pat yourself on the back.


In the future you can post less wips etc, but if you are honestly interested in maximizing the wonderful resource that polycount is, you'll follow these steps to the T. =)

[bbob]

Id also like to re-ask if attaching new geometry to the low poly mesh will mess up any uvs i already have? Assuming i collapsed the Unwrap UVW modifier.


it wont.

[Grimm_Wrecking]

In this thread someone said the flash hider was still wrong, but i don't see how it is because its been changed twice, both using reference.

You're muzzle break/flash hider is wrong[period]

Muzzle Break/Flash Hider WtF Ever you wanna call it #1 (http://pixagogo.com/Photos/Albums/Photo.aspx?id=S4!sepfydtFxQyLsoBIE9746lJH08Lr-xxBj!3YHUSih9kpjH!JnjBALKHxbkN!Wxo)

Muzzle Break/Flash Hider WtF Ever you wanna call it #2 (http://pixagogo.com/Photos/Albums/Photo.aspx?id=S4JdC0p2jeAn2dUuYlKHIhX7u60lsGzw5f4h 7PBtRUpGCVV7!5XT0NDZvtlHEOm8lV)

Muzzle Break/Flash Hider WtF Ever you wanna call it #3 (http://pixagogo.com/Photos/Albums/Photo.aspx?id=S4J-6q8mbr7yHG6ZU4ulrq21GCm3iaSNTzJ1htfs4ByqAwccuIAf4J BhWlIPuyLwGT)

It seems like you changed it because what other people said without actually slowing down and examining what you were about to model. Or when you did you didn't take the time to error check against reference. Or if you did both of those things...you just missed it which we all do from time to time.

http://dl.dropbox.com/u/2323296/PC%20Folder/ump_po.png
Thats just a basic silhoette of how it should look. Its not 100% accurate it was quick in PS just to show you how it should look.
Its the same dia before the flare and taper as it is after, because there is another sharp taper where the end of the polymer is that brings it down to that size.

Edit: Also it's just out of scale length wise. (muzzle break/flash hider)

Raw 8" Barrel #1 (http://dl.dropbox.com/u/2323296/PC%20Folder/ump458inchus-2_935_detail.jpg)

Raw 8" Barrel #2 (http://dl.dropbox.com/u/2323296/PC%20Folder/ump458inchus_283_detail.jpg)

Others pointed out other stuff, and yeah seconding bbob about time & Stromberg about sticking to the 45.

Personal Time Comparison:
I rebuilt at least twice every piece of the HP of the M4 when I was actively working on it before posting, and then rebuilt it again (at least once) after people told me what was off, and I'll probably have to rebuild the damn thing again. Its frustrating yes, but you only learn if you listen, apply and try. And if you think 2 days is a while currently that damn M4 is @ at 158 working hours for just a HP showoff (and for a bake I'll have to spend probably a day loosening edges.)

[frell]

Alright i fixed the barrel and redid the stock. I wont be smoothing the inner part of the stock though for the same reason as before (normal problems).

Now i will be figuring out how to add the stock button thingy and ill add the hinge.


I just saw the missing loop cut on the inside of the barrel and it has been fixed.http://img706.imageshack.us/img706/1975/fixtd.png

[EarthQuake]

What normals issue on the stock? Getting some wavy lines or something? What you have now is going to look much worse than some wavy edges, and you can just add a couple more edges to round it off to help the wavyness.

[frell]

But its also not seen by the player =\. To get it smoothed it really wouldnt have THAT much of an effect and id have to add more polys around the hole.

[EarthQuake]

Yes, but what you're doing now looks bad, baking from high to low is all about the illusion of having a much more detailed model than you really have, when you leave parts of your mesh as lame faceted lowpoly stuff, it ruins that illusion. The "player wont see it" argument is equally good for any normal issues you would have.

The player likely will never see the barrel, and definitely will never see the huge whole the barrel resides in, but that hasn't stopped you from adding a lot of detail there. Keep it consistent, dont slack off on one part just because you have a convenient excuse to.

And if you ever plan on having this in a portfolio or something, you need to consider other angles than FPV... FPV is most important for a weapon like this, but unless you plan on croping the stock out of all your presentation images, it should look good. Simple as that, make it look good.

[frell]

Alright added a couple more polys to that bridge in the middle and it smoothed quite well. You can also see the buttons for the stock off to the left, i was going to have the purple one as floating geometry and the blue one in as the button, you can also barely see the hinge on the other side of the gun.

http://img202.imageshack.us/img202/3329/wiper.png

[EarthQuake]

As has been said quite a few times, your edges are still too hard. =)

[frell]

http://img257.imageshack.us/img257/4544/wiperg.png

Better?

I gotta start getting the distance between loop cuts right, i usually slide them almost to the very edge of the poly without realizing it.

[EarthQuake]

Thats a little better, i think you can go more tho. I like to exaggerate my edge bevels a bit, it helps to make things more interesting. Especially with a gun like this, that has a softer polymer material.

A sort of rule of thumb that i use, if your edges are so tight that there is noticable aliasing when viewing the entire gun, or most of the gun, your edges are toooo tight. This also helps with small details/floaters/screws, if your floater turns into a mess of aliased noise, your details are much too fine and sharp.

[frell]

So crits on the buttons? The purple will be baked and the blue will be geometry. They should come out fine. I had a little clash of references problem with the beginning of the stock, so i went with my side image mostly.

Heres the hinge, a few minutes ago it was a cylinder smoothed and pushed halfway into the seam between the stock and the gun and i was about to upload it when i took a second look at the reference and fixed it to match. Another little push but now im even happier with it :)

http://img708.imageshack.us/img708/6925/hingesd.png

[an aggressive napkin]

Not saying that my model is perfect or anything, but this is a metal weapon, and look at how "loose" my edges are.
http://langdavis.com/images/mac11_hp.jpg

To get a good effect, you really must smooth out your lines. plus, I have never seen plastic with that hard of edges

[samgriffiths]

I think some of your edges may be a tiny bit sharp and won't show so well on your normal map, specifically the butt.

[frell]

working on a scope for the gun, heres the wip so far:

http://img341.imageshack.us/img341/7801/picch.jpg

Gotta add screws on the side etc and then ill fix up some edges on the gun.

[Rumkugel]

If this should be a eotech holosight you might wanna gather more ref :P

[frell]

This was the only one i had:
http://www.gunshopfinder.com/Eotech/eotech552_2.jpg

Its also not done (dont know if you read that or not)

Could you please tell me whats wrong with it? Its probably something to do with the battery pack :S

[frell]

Here was a test render just to get a preview of it on the gun
http://img24.imageshack.us/img24/7670/scopeon.jpg
Still trying to figure out how im going to make that battery pack.. hmm...

[Pedro Amorim]

the front part of the scope is nothing like in the ref.
change it :)

[frell]

You're funny :P

[Kitteh]

You're funny :P

No, he's right

[DKK]

Still trying to figure out how im going to make that battery pack.. hmm...

I think what frell means, is that hes still working on the front part, you can see in the last shots its actually just a box instead of the wrong geometry from before.

Also I gotta say big ups man, You're improving a lot, and you should be proud that you're taking all the advice into consideration, that's probably the most important part of Polycount, and it's something some people neglect to realize. People may bust your balls, but it's because they want you to get better, all you can do is say thanks, and come back tomorrow and ask for more.

[frell]

wtf is wrong with it? Thats what i spent hours on going back and forth making it match the ref more and more.

http://img695.imageshack.us/img695/1538/watcf.png

The only problem i see is the....

Wait, by front do you mean the battery or where the buttons are?I havent finished that. Im still wondering how im going to model the battery.

Thanks DKK, ive been getting more determined to make this gun kickass :) And im not even done yet, im going to completely re organize the UV's so they cant breathe for shit.

[Grimm_Wrecking]

EoTech - Lots (http://pixagogo.com/2722774636)

If you just go by the way the eotech rests on the rails the front would be the battery pack and the rear would be the display/controls. Also you're missing the details from the side, you're probably getting to it, but don't forget.

[frell]

Im not done at all, and thanks for the refs.

Im guessing i should have this as a separate asset? 1024x1024 should do fine. That was my original plan.

[Moosey_G]

I believe they were referring to this.

http://img824.imageshack.us/img824/6426/takeadvicedamnyou.png

[frell]

Oh, after i found out the name of the sight i went looking at more refs and in one of them the battery only came out that far and i thought it looked better. All their sights look similar so it shouldnt be a big problem. I cant find the ref again so i probably will end up extending it.

[frell]

http://www.gunshopfinder.com/Eotech/eotech552_2.jpg
Should those bolts on the side (big ones) be floating? They're on the complete opposite side of the gun in FPS view and when scoped in it wouldnt be visible at all.

[Moosey_G]

I'd make them floaters. I usually make most of bolts floaters.

[Racer445]

The first PC Workshop modeled this sight, perhaps look through those threads and you can find out how other people modeled theirs.

[frell]

Who would have thought a battery pack could be so frustrating.

Thanks racer, i just reviewed about 10 threads. I noticed that they dont use floating geometry very much, is there a problem with this? My top screws on that back panel are floaters along with the text. The ridges above the buttons and the arrows will also be baked into the mesh. Of course i tried to not use it too much because the sight is right in the players face when aiming.

[frell]

http://img197.imageshack.us/img197/7336/87533595.png

Here it is so far, ill probably loosen that very front edge to be the smooth and round like its supposed it. But then ill probably call it done because the player really wont be that close to this side of the scope. In MW2 i never even knew that there was a huge battery pack in the front until i modeled it :P

[mheyman]

Are you going to do those sweet screws on the sides like in the reference?

[frell]

Are you going to do those sweet screws on the sides like in the reference?
Forgot about those, yes ill do them they're not a problem.

Did a quick fix on the back

http://img697.imageshack.us/img697/873/newgu.png

[frell]

Pretty content with calling the scope done now. Im very tired so if theres any main features i missed let me know

http://img517.imageshack.us/img517/2714/scopedun.jpg

[Squiggers]

Had about 3 threads on Crymod at once with EoTech sights going, that one got interesting. :P

I do agree with what others are saying - perhaps slow down a little bit, if only to review your work before you go onto the next stage. Not that getting an asset out in a little over.. i think 2-3 days (?) is a bad thing mind, you just need to be really, really careful not to rush it.

There aren't any key features you've missed I'd say - are you using any floating geometry, by the way?

[Grimm_Wrecking]

There's nothing wrong with using floaters, in fact they're an essential part of production. Most of the guys doing the workshops were probably doing it for a solid high to show / working on complex modeling workflows.
Practicing complex mesh modeling now saves time later.
My opinion on that anyway.

[frell]

Had about 3 threads on Crymod at once with EoTech sights going, that one got interesting. :P

I do agree with what others are saying - perhaps slow down a little bit, if only to review your work before you go onto the next stage. Not that getting an asset out in a little over.. i think 2-3 days (?) is a bad thing mind, you just need to be really, really careful not to rush it.

There aren't any key features you've missed I'd say - are you using any floating geometry, by the way?
Ive been fixing crits for the last 2 days :P

Yes im using floaters. All the bolts accept the bottoms on the front, the ridges above the buttons along with the arrows and text.

[Squiggers]

Ive been fixing crits for the last 2 days :P

Yes im using floaters. All the bolts accept the bottoms on the front, the ridges above the buttons along with the arrows and text.

It was more in general I meant. ;) Either way, my personal take is that it matters whether it looks good, the geometry is well distributed, and the texture is up to scratch - if it doesn't exactly match a reference.. well, its not really that much of a big deal in my opinion.

Only the anally retentive gamer will pick up on a bolt being in the wrong place, for example. :poly124:

Cheers for the clarification, was just interested where you were using it on the model.

[frell]

Heres the ugly test bake
http://img37.imageshack.us/img37/6835/uglybake.png

Front panel cape out great but I never learned how to fix up those weird edges where the normal map gets cut off (right image). I guess ill have to fix it in PS?

[Pedro Amorim]

try and make the edges of the highpoly less hard on the outter part of the scope. or else you wont benifit from the normal mapping.

[Millenia]

Tweak your cage so that it hugs the mesh closer

[frell]

Tweak your cage so that it hugs the mesh closer
Didn't know i could do that :O How exactly would i do that?

Before i read that I tried disabling the TS on the inner and outer shells and it looks pretty good.

http://img805.imageshack.us/img805/3596/seedis.png

[EarthQuake]

Just add more geometry to your lowpoly, its too "jagged" right now, especially for something that will be very visible and very important in FPV, and even possibly zoomed in on in an iron-sight animation. You've just got much too little geometry to smooth out those curves right now, and fixing that will also fix any normal issues you're having.

[ZacD]

It might also be a good idea to model the holes in the side for those larger screws, and there's some detail on that area that's missing in the normal map.

How does the sight look on a gun in a fps view? that can really help you see where you need to add more detail.

[frell]

It might also be a good idea to model the holes in the side for those larger screws, and there's some detail on that area that's missing in the normal map.

How does the sight look on a gun in a fps view? that can really help you see where you need to add more detail.
Those bolts will never be seen by the player, because they go into the side rather than out. That, along with them being on the opposite side of the gun made me decide to just float them.

Heres the final bake, you were right EQ. I chamfered all those edges and it baked fine with TS turned on on the high poly. I also tried to make my uvs better. Just a note, the only empty space is at the top, some faces were baked straight on and are the same color as the backgorund.

http://img337.imageshack.us/img337/4849/finalnf.png

http://img156.imageshack.us/img156/7805/sightnormals.jpg

[frell]

Looking great. The texture isn't done and this was just a text render, this also isn't the fixed up gun.

http://img10.imageshack.us/img10/2876/scopewip.png

[EarthQuake]

Those bolts will never be seen by the player, because they go into the side rather than out. That, along with them being on the opposite side of the gun made me decide to just float them.

Heres the final bake, you were right EQ. I chamfered all those edges and it baked fine with TS turned on on the high poly. I also tried to make my uvs better. Just a note, the only empty space is at the top, some faces were baked straight on and are the same color as the backgorund.

http://img337.imageshack.us/img337/4849/finalnf.png

http://img156.imageshack.us/img156/7805/sightnormals.jpg

This UV layout is a lot better. There are still quite a few things you could do to improve it, however they are all relatively minor. Look into how you're placing the smaller bits, as i think you can make that more efficiant, and scale up everything by a small %. Something to pay attention to specifically, where you have some long thing peices but you have one horizontal and the other vertical, right next to each other like this:

___ |

and instead you could stack them like this:

___
___

or

| |

These sort of things can go pretty far in maxing out your layout. One thing i like to do is scale my layout so that it stretches farther than you would think is possible to pack, and then try to really fit some of the smaller stuff in the cracks and be creative with how you lay it all out.

But yeah, much better layout than before.

[frell]

I see what you mean.

Still working on the texture, marmo has a hard time getting the light on it right:
http://img705.imageshack.us/img705/8458/sadfg.png

[Firebert]

marmo has a hard time getting the light on it right:


have you tried using one of the different light setups provided by 3PS in your toolbag install?

toolbag_install\toolbag\base\sky

there's 5 different setups available, or you can create your own

see here from the Toolbag thread. Really buried in there now that it has become a monster....

http://www.polycount.com/forum/showpost.php?p=921855&postcount=337

it also looks like you aren't rotating your light at all... well you may be, but this is probably because you might be using the same light setup?
hold shift, LMB to rotate -or- shiftRMB to spin it. hold down ctrl to rotate your model as opposed to alt to move the camera. this can be useful for getting a different lighting angle... sometimes useful.

and Johny made some more skies you can use here.

http://www.polycount.com/forum/showpost.php?p=921916&postcount=338

OH, and to load a new sky, load toolbag, go to the Sky tab, select Open Environment, and then go directory your skies are stored.

[frell]

I rotate the light the best i can in some views. I mainly find the sunny environment to be the best, but sometimes it doesn't light everything it could.

Thanks for the link to EQ's skies.

Heres another picture, id say the scope has just about everything done to it i can do. The color scheme is really just so bland on the actual ones. I added in the tan for personal preference.

http://img156.imageshack.us/img156/9651/ghnb.png

[frell]

Heres the final bake on the ump:

http://img704.imageshack.us/img704/4537/newump.png

Heres the normals. I really started to get cornered towards the end and couldn't find anything to move to fill the gap at the top and take advantage of space :S.
http://img228.imageshack.us/img228/8788/umpnnormals.jpg

Hopefully thanks to everyone that helped me fix up this gun i wont make the same mistakes on my next gun. Im glad i got the sight done because it can be a great asset to customize my guns with, and came out surprisingly well.

Im getting cornered by a team project so im going to toss this in my game art folder and texture it in my spare time.

[EricV]

you're stillnot getting what I was talking about about with your unwrapping.

sure you can get away with this on a pretty easy going mod, but I cant imagine any art director or lead at a professional studio would be ok with this.

To give you an idea of how much work is involved, it will sometimes take me a whole day to unwrap a single asset. What you've shown looks like it was done in less than hour

[odium]

That unwrap is pretty dire mate. You should be focusing on that before texturing. For some reason you keep jumping the gun (no pun intended) and rushing into stuff before the stuff before are nowhere near as good as they can be. You are wasting like a collective 40% of your UV space which is going to result in some poor details ingame, not to mention everything is split up like its just come right out of a "click here to unwrap" program.

Don't forget you don't NEED to use 1:1 textures, you can use for example a 2048x1024 (2:1).

Also, model those screws on the side of the sight, they look poor flat like that.

[Mr. Bean]

UV layout should be as evenly distributed as possible...

[odium]

That as well. Usually what I do is unwrap everything even if it wont fit on the unwrap, then move all the pieces together to form a square, then scale up or down as needed. Sometimes you can get away with giving a few areas more pixel density (Like iron sight areas) so its not a set rule. But at the moment, the weapon has quite a lot of really random breaks in the unwrap, and even more random scaling.

DO NOT just say "oh i'll move onto a new gun, get it right next time". Just fix what you started. A good unwrap makes a huge world of difference and you may even be able to get out 2x the detail with the same resolution textures, just from making a better unwrap.

[brandoom]

Heres what you should be striving for. This is a UV map made by Rens - he is a saint.

http://dl.dropbox.com/u/1559778/bulroguvs.jpg

[StefanH]

listen to what they tell you about the unwrap. It's definetly very important for getting hired.

still i just wanted to give you a headsup. I think you are improving very fast and doing good progress. Just try to get really down to the nuts and bolts of unwrapping and a little bit better lowpoly modelling.

[frell]

I didn't re-unwrap the uv, which is why its probably still messy. The pieces were sloppy and didn't fit together that well. My sight was unwrapped and then packed by hand and it looks alot better compared to the gun which i don't have time to re-unwrap, but it could be better.

Odium, 2 things. 1, no one will see the bolts on that side of the sight so i would prefer not to spend time fooling with it. And actually, they're there i just had the rays too low to get the full indent so i might just re do the normal map. And I have been fixing stuff constantly on the last 2 pages of this thread instead of putting it off until the next gun. I have a request for another gun that i need to get to work on, and hopefully thanks to this thread i wont rush anymore pieces and i can unwrap it much better to pack.

Ill keep working on them though, thanks guys.

[ZacD]

Heres what you should be striving for. This is a UV map made by Rens - he is a saint.

http://dl.dropbox.com/u/1559778/bulroguvs.jpg

if you are using a 1024 texture you might want a little more padding than that, I believe Rens said that was for a 2048.

[Moosey_G]

Yeah don't forget you still want room for padding. It won't be fun doing it again because everything bled over.

[frell]

Yeah don't forget you still want room for padding. It won't be fun doing it again because everything bled over.
Thats what i worry about. Ill put them real close and then worry that they might touch each other so i move them apart, and then in the end it looks really spaced out.

[ZacD]

anything more than 3 pixels is too much.

[Kitteh]

You also made the mistake of mapping both sides of a lot of parts that are perfectly symmetrical - wasting UV space and texturing time. Like the stock and magazine, to name a couple.

[Millenia]

You also made the mistake of mapping both sides of a lot of parts that are perfectly symmetrical - wasting UV space and texturing time. Like the stock and magazine, to name a couple.

Yeah but he's using object space normals

[frell]

Im trying to learn to use Tangent normals. I actually did my first test bake for the UMP with tangent space and it looked good except for a few smoothing errors. And the quality of the map looked really bad too.