Need help with modeling weapons

MediumSolid · Sep 2015

Hi guys, this is my first thread here in polycount. So don't be too mean to me. :poly124:

I have in plan in the next 80-100 days to model 120 realistic high poly weapons, I reckon each weapon can be model withing 3-4 hours depending on how much detail it has. I have some skill, I'm not the best modeler but I think I can reach that goal before the set date. But here's where it gets tricky, I don't have many HD references, and I've been googling and searching for them, but it's so hard to get my hands on them. For example, AK-12 assault rifle which is fairly new weapon doesn't even have that many up-close HD photos on the internet, I've even searched in russian sites. So could anyone tell me where I can look for diagrams, pictures and even videos for weapons in general? How do you guys get your weapons so precisely done and how much does it take you to model them and does everyone bother with the details or do some modelers just bluntly model them? How do professionals get their hands on preferences for their models, is there a special site for that? I'm very interested to see what you have to write.
Here's a list of all the weapons I have in mind to model with extreme detail:

Assault rifles
ACE-52
AEK-971
AK-5
AK-12
AK-74M
AN-94
AUG-A3
AS-VAL
CZ-805
FAMAS
FN FAL
FN-F2000
G3A3
G-36
L-85
M16A3
M16A4
M416
QBZ-95
SAR-21
SCAR-L
TAR -21

Carbines
A-91
ACE-21 CQB
ACE 52 CQB
ACW-R
AKS-74U
AKU-12
G-36 C
G-53
M-4
MX-4 STORM
SCAR-H
SG-553
SKS

Grenades and missile lanuncher
DEVICE DM
GM-94
HK-69
Panzerfaust 3
QLB-06 QLZ-87B
QLZ-87
RG-6
RGS-50

Handguns and revolvers
Taurus raging bull
Berreta 92FS
Berreta 93R
CZ-75
Desert Eagle
FN FNX45
FN-57
Glock-17 G4
M-9
M-1911
Mateba Model 6
MP-412 Rex
MP-443
MP-446
P-226
QSZ-92
Walter P99

Machine guns
IMI NEGEV
LSAT
M27 IAR
M-60
M-240B
M-249
M-G4
MG-36
Mk-48
Negev NG7
PKP
PKP Pecheneg
RPK-74M
Type 88LMG
U-100

Shotguns
870 MCS
AA-12
Bandaevsky RB-12
Bekas-Auto
Benelli M4 Super 90
DAO-12
Hawk-12
Hawksemi-automatic police shotgun
Mossberg 500
MTs-225
QBS-09
RMB-93
SAIGA-12K
Shorty_12G
SPAS-12
USAS-12
UTS-15

SMG's
CBJ-MS
CZ-3A1
Kriss Vector
M5K
MAC-10
MP7
P-90
PP-19
PP-90M1
PP-91
PP-93
PP-2000
QCW-05
SR-2 Veresk
TMP
UMP-9
UMP-45

Sniper rifles
ACE-53 SV
Cheytac Intervention
JNG-90
L-96
M12-12.7x108mm Black Spear
M39
M40A5
M82
M92 Black Arrow
M98B
m417
MK11 Mod0
Orsis T-5000
OSV-96
QBU-10 12.7 sniper rifle
QBU-88 (Type 88)
Remington XM2010 rifle
RFB
Scout Elite
SIG-50
[*]SV-98
SVD
SVDK
SVU-AS
Tikka T3 Tactical
VSS vintorez
VSSK Vychlop

Joost · Sep 2015

I'm curious what kind of "extreme quality" you're expecting out of 3-4 hours of work?

FourtyNights · Sep 2015

I give you an advise. Why don't you just pick up your... let's say 10-15 (or less) favourite weapons from that list and model them. 120 is overkill. Even ten weapons is a lot, and I'm sure it's versatile enough for your portfolio. And you'll have more time to work on them to achieve better quality.

Not to mention that making the low polies with UVs, bakes and textures is kind of mandatory to make them worth for your portfolio, if you're ever gonna work in the game industry.

Doxturtle · Sep 2015

it seems like you are going to be hard pushed to get this many weapons done so quickly

MediumSolid · Sep 2015

Well something like this. But yeah.... I also feel like I went a little overboard or too enthusiastic with 3-4 hour requirement per model, it would be more like 6-8 hours.

Visceral · Sep 2015

Looking good @MediumSolid if that was done in 3-4 hours i would be really really impressed. Just dont push yourself too hard to meet the requirements. Its good to challenge yourself but you might get discouraged if you fail.

Ill post all the pages with refs i know of.
http://forum.guns.ru/forum/36/466341.html

X-One · Sep 2015

Do you plan on texturing all of these weapons?

Honestly, if you're modeling using traditional Sub-D techniques, there is almost no chance at all of getting 120 guns done in 100 days. (Unless you're incredible, and even then, I'd doubt it)

I have no idea how good you are, but my suggestion is to start with one you like, and that you can find good reference for. That should, at a minimum, give you a gauge on how long each one would take, under most ideal circumstances.

MediumSolid · Sep 2015

Visceral wrote: »

Looking good @MediumSolid if that was done in 3-4 hours i would be really really impressed. Just dont push yourself too hard to meet the requirements. Its good to challenge yourself but you might get discouraged if you fail.

Ill post all the pages with refs i know of.
http://forum.guns.ru/forum/36/466341.html

Thanks man, that's the stuff I"m talking about. Sadly this wasn't done in 3-4 hours. It actually took me 8-10 hours. But this was done couple of months ago. Now that I've seen where I have made my mistakes I can model things, especialy hard surface stuff even faster. Because most of the basic weapon shapes you can get with boolean operations which Blender handles pretty well.

WarrenM · Sep 2015

Seriously, scale that list waaaaay back. Start slow.

MediumSolid · Sep 2015

Here's a couple more. The first one would be ACE-53 and second one Taurus Raging Bull 480.

Anth · Sep 2015

My advice is to take it step by step, just try and produce your favourite weapon and a really high quality and then take it from there! But everyone has their own ways of doing things

MediumSolid · Sep 2015

That's all fine and dandy. But what about aquiring reference photos for these weapons. Am I really doomed by googling endlessly for pictures of a weapon that has probably never even been snapped with a camera, at least not up close with HD one?

Sir Apple · Sep 2015

A single accurate, well modeled weapon holds a lot more weight and importance then 100 poorly modeled, sub-par quality weapons. Especially in terms of learning and portfolio.

I guess I'm just confused as to why you want to do this.

BradMyers82 · Sep 2015

yea, all you need is 2 or 3 really awesome weapons and your set. I have to say though, some of the stuff you said you modeled in 8-10 hours I think turned out really well for that amount of time. It would be cool to see what you could do on a complicated piece that you spent your time on.

redux · Sep 2015

I think you could manage to do this by the time you have set. After a while you would know how a weapon is made. Just make shure you take some time on the first ones. The weapons you have now seems a bit wide, and the edges seems to be a bit hard. Remember to uv, texture and shade your weapons, or else you would end up with alot of unfinnished weapons.

Amsterdam Hilton Hotel · Sep 2015

I dunno why everyone's poo-pooing the goal here, even if you don't nearly finish you'll get a lot of experience and familiarity with weapon platforms this way. Follow ur dreamz

MediumSolid wrote: »

But here's where it gets tricky, I don't have many HD references, and I've been googling and searching for them, but it's so hard to get my hands on them. For example, AK-12 assault rifle which is fairly new weapon doesn't even have that many up-close HD photos on the internet, I've even searched in russian sites. So could anyone tell me where I can look for diagrams, pictures and even videos for weapons in general?

Newer weapons that don't do the military-industrial convention circuit will always have less ref but in general it really is just googling. Usually you can find better results if you search for specific parts ("ak47 charging handle") or manufacturer. Always check auction sites if the gun is old enough to be on them because those guys take great pics.

RedRogueXIII · Sep 2015

Cool project! I'm working on something similar, a list of about 20 guns between some mods, and it's taking me an average of 3 days to make the high poly with some extra variant parts, although I have been focusing on the simpler ones, so when I get around to something complex like a lmg, it'll likely take 5 days or so. That would put it at 2-3 months to do them all but I don't think that would be enough time still. I like to model as much detail in as I can, I try not to use floating geometry, but it'll help save a few hours when I do. If you omit details that could be normal mapped in later then it'll also save a lot more time.

For 8-10 hours those weapon models are very impressive results! If you intend to bake a low poly version though I'd recommend exaggerating the edges so that the highlights are preserved at smaller resolutions. I definitely have to give Blender's modeling tools a good look now!

From the weapon list it looks like the Battlefield 4 list with some extras. I've got refs on quite a bit of your list, but the problem is how to transfer all that data. Here's a list of reference image sites I use; http://wiki.polycount.com/wiki/Prop_Reference quality varies and for searching for a specific subject in particular Google image search is faster and good enough, the rarer stuff you have to search specialty sites like auction sites.

Love to see how this challenge goes, good luck!

MediumSolid · Sep 2015

RedRogueXIII wrote: »

Cool project! I'm working on something similar, a list of about 20 guns between some mods, and it's taking me an average of 3 days to make the high poly with some extra variant parts, although I have been focusing on the simpler ones, so when I get around to something complex like a lmg, it'll likely take 5 days or so. That would put it at 2-3 months to do them all but I don't think that would be enough time still. I like to model as much detail in as I can, I try not to use floating geometry, but it'll help save a few hours when I do. If you omit details that could be normal mapped in later then it'll also save a lot more time.

For 8-10 hours those weapon models are very impressive results! If you intend to bake a low poly version though I'd recommend exaggerating the edges so that the highlights are preserved at smaller resolutions. I definitely have to give Blender's modeling tools a good look now!

From the weapon list it looks like the Battlefield 4 list with some extras. I've got refs on quite a bit of your list, but the problem is how to transfer all that data. Here's a list of reference image sites I use; http://wiki.polycount.com/wiki/Prop_Reference quality varies and for searching for a specific subject in particular Google image search is faster and good enough, the rarer stuff you have to search specialty sites like auction sites.

Love to see how this challenge goes, good luck!

Yeah you're right, I don't like floating geometry neither. I think that's cheating except if you don't have time to embed more details into your surface normally because of the strict deadline. Thanks for the links man, you're really helpful. And yes, these are BF4 weapons with many additional ones in it. @Amsterdam Hilton Hotel Thanks for the advice and words of encouragement, I also don't see what's so wrong with the deadline I have set for me? Sure, I might not finish them before time but at least I'll have some finished weapons, a lot of experience and good idea of what I'm capable off. @RedRogueXIII, LMG's are a pain in the ass and this IMI Negev took me couple of days to make, it really was hard to model it even with HD references.

Joopson · Sep 2015

MediumSolid wrote: »

Yeah you're right, I don't like floating geometry neither. I think that's cheating except if you don't have time to embed more details into your surface normally because of the strict deadline.

Floating geometry is absolutely not cheating. It saves time and delivers equal results, for all situations it should practically be used in. It's working smart.

No use pushing yourself to solve a very complicated puzzle that doesn't need to be solved. Unless, of course, you want to solve it.

donofdon · Sep 2015

Yeah man, use all the tricks available. If you are texturing later, remember you can batter out simple shapes with the likes of NDo.
You're mental for attempting this. The models posted above look fine, (first one looked a bit fat though). Do you have to be 100% accurate with each?
Ref. Try Pinterest, airsoft sites, reddit/r/gunporn, search for Shot Show images and forum entries. You have to suck it up and spend the time internet searching. Eventually things like the underside of a LMG bullet reload lid will come up on an obscure forum.

MediumSolid · Sep 2015

Joopson wrote: »

Floating geometry is absolutely not cheating. It saves time and delivers equal results, for all situations it should practically be used in. It's working smart.

No use pushing yourself to solve a very complicated puzzle that doesn't need to be solved. Unless, of course, you want to solve it.

Well not for all situations. I doubt it could be used in films where close ups are common. Floating geometry is almost exclusively used in map baking, but then again you could also achieve the same result with modern softwares for texturing like PSC in conjuction with dDo and nDo2.

Bek · Sep 2015

MediumSolid wrote: »

But what about aquiring reference photos for these weapons. Am I really doomed by googling endlessly for pictures of a weapon that has probably never even been snapped with a camera, at least not up close with HD one?

Some things will necessarily be difficult because they are new and not released extensively to the public (AK12 world.guns.ru having the most recent info I know of). Others will be even more so as they appear to be fictional (AKU12, which from a quick google seems to be an invention of the Battlefield 4 team).

I would also recommend creating a generic Kalashnikov kit as you'll be able to re-use it in tonnes of stuff on that list. You can probably do a similar thing with the AR platform. Another example is the VSS Vintorez and the AS val are literally the same gun but with a different stock (and typically a different mag, 10rnd for the former, 20rnd for the latter).

If it were me I'd also be working in real-world units and perhaps knocking out all the ammunition first so you can keep barrel diameters consistent. I'd probably work on all the ARs one after the other; all the AK's one after the other and so on. And from looking at the pic you posted; if you intend to bake a normal map from these later on your edges are far, far too sharp.

You've also listed the SCAR-H as a carbine and the SCAR-L as an assault rifle. Those are two others you can do at the same time due to their sharing of parts.

Amsterdam Hilton Hotel wrote: »

I dunno why everyone's poo-pooing the goal here, even if you don't nearly finish you'll get a lot of experience and familiarity with weapon platforms this way. Follow ur dreamz

Eh; I think it's coming from the perspective of what's needed to get a viable portfolio for work. If you're not interested in that then sure; model 120 highpoly guns. Otherwise, pick some of your favorites and produce the best textured lowpoly you can.

MediumSolid · Sep 2015

I don't get it why y'all complaining about AEK-971 being too wide? If it's the magazine that makes you think that the weapon is wide then I must inform you that that's how it is in real life. I'm aiming for extreme detail and precision when doing my weapons/models. And it's not too wide, am I right guys... guys?

MediumSolid · Sep 2015

Bek wrote: »

Some things will necessarily be difficult because they are new and not released extensively to the public (AK12 world.guns.ru having the most recent info I know of). Others will be even more so as they appear to be fictional (AKU12, which from a quick google seems to be an invention of the Battlefield 4 team).

I would also recommend creating a generic Kalashnikov kit as you'll be able to re-use it in tonnes of stuff on that list. You can probably do a similar thing with the AR platform. Another example is the VSS Vintorez and the AS val are literally the same gun but with a different stock (and typically a different mag, 10rnd for the former, 20rnd for the latter).

If it were me I'd also be working in real-world units and perhaps knocking out all the ammunition first so you can keep barrel diameters consistent. I'd probably work on all the ARs one after the other; all the AK's one after the other and so on. And from looking at the pic you posted; if you intend to bake a normal map from these later on your edges are far, far too sharp.

Eh; I think it's coming from the perspective of what's needed to get a viable portfolio for work. If you're not interested in that then sure; model 120 highpoly guns. Otherwise, pick some of your favorites and produce the best textured lowpoly you can.

Yes, I'm a aware that some of the carbines listed are actually just fictional and scaled down versions of assault rifles. And yes, many weapons of the same brand are similar to one another just like MP443 and MP446 and many ak weapons have similar parts like magazines, rear sights etc. Not to mention Berreta 92FS and 93R.

Zieg_Reborn · Sep 2015

Even scaled back your estimates are going to get you very poor results. If you aren't extremely confident in your ability to produce high quality models (as in you have done this before, and quite a lot), I would recommend you start with 1 and get it to an acceptable quality level before even tackling a second one.

You'll get better with each one, but you will get no benefit by just rushing through things without a solid understanding of the fundamentals.

Edit: Also I haven't read (perhaps I missed it) as to whether or not you plan on texturing these, and if it is planned for film or games... portfolio or learning... etc. It might give is a better idea on how to critique/advise you on your progress.

MediumSolid · Sep 2015

@Zieg_Reborn, no you haven't missed it. I never mentioned for what would I need this and what would I do with this ammount of weapons once I make them. I would make low poly versions of them and of course texture them. For the high poly weapons I would only put basic metallic, wooden, plastic/polymer materials on them.

Need help with modeling weapons

Replies