Wanted to tackle a Assault Riffle in my Hard Surface Practice aswell as to have a gun in my portfolio, my overall plan was to model each part individually. Thus in the future I can kitbash some guns (watching the Red Storm Zbrush R7 video they stated it best to make it detail now so in the future it just a matter of re-topology). I found this gun online and it varied in weapon parts, hope to have the high poly done by 20th and the low by end of month. Texturing a environment so time is split
Lower Receiver: 2 days (setting up the proportions for the rest of the gun, then it just a matter off slapping on the finished pieces)
Stock: 1 day and half
Magazine: 1 hour
That with 5 hour work day, I have great collection of references ill share at the end if anyone wants (just Google searches) of each part from almost all angles. Ill do a exploded render at the end. Trying to think of a good plan to layout the texture/uv to also be re-useable as I do plan on doing a quick ndo or I may just model all the minor details in and do a bake. *Upper Noveske Receiver next then the grip and a scope*
Progress with 4 days work, managed to knock out the Upper Receiver Left and Right Side took me a bit longer than expected due to the lack of reference for the certain version I was going for. I may also revisit the bolt carrier depending how much it seen I only did a small section. All parts I used has been updated
My plans for the RDS attachments is a canted design:
Alright, there really not much to look at in terms on non wireframe as I dont add any supports till the very end (easier to model/make changes with less geo). That being said it rather flat and lifeless haha:
Once I finnish the High ill add my support and toss it for a full critique, with a exploded view to see the parts individually
I began adding the scopes/holo and finishing off all the attachments I would be adding to the gun, I am missing some bolts but didn't feel like making them and replacing them when I add my supports so doing it at the end. I should get this done next week in the next 4 day spread. So far I have 10 days into the gun I took pretty much a full day throughout looking up reference/analyzing how it actually works
Time Span: Magnifer/QD Pivot: Took half a day EOTech: 1 half Day, didnt expect it to be so big Iron Sights/Clean up any problem areas: 2 days
Iron Sights Front and Back *71 A.R.M.S
EOTECH 512 with EOLAD 2 Laser Sight
LaRue Tactical QD Pivot Mount for canter sights Aimpoint 3x Magnifier most of the details will be done in nDo
Detailed item list found at my first post. Wire for now ill post some nice sexy render next time when i finished it on Dec 1st. I have determined my barrel/and lw bar rail already, may add a handgrip at the front aswell
Looking good, although the lower reciever has a few problems (missing a few details).
I'm doing the same reference as I said in the other thread, so it will be cool to see how both of ours turn out.
What do you mean missing a few details could you elaborate, there a couple of things like the small holes/the front squar'sh pattern that will all be done through nDo or Substance Painter. Didn't want to deal with the extra poly for something that can be done through a normal map
Im also going fully accurate with it and doing it separated, so you may get yours done quicker than mine. The high poly should be done Dec 1st more than likely; than I have to think of a way to uv/low poly to use it with as a kitbash set
Thank you for the input, you may not be able to see it but I do have the slove and the subtle change with the trigger guard. It may just be due to the fact of the reference snapshots I took, ill do a more in depth breakdown for the high poly renders
Awesome breakdown I wasnt going to in-debth with the bolts/screws though but man so sexy looking
Hey Alex, i think you make a good job, but can you explain the function of the Sight?
I think that here are missing some details like a feather or something.
It a backup iron sight incase your EoTech runs out of battery or gets damaged during combat. Most of the time it flipped over on its side, the back one is the only one I will have up
I am missing 2 bolts on the front and the spring, then I am missing 1 bolt/spring on the right *It easier to just make a set of each type of bolt then spread them around the model then having to make one high poly then go back in and replace each bolt with the new high poly version/same with doing it when I make the low (this way I can just do it once)*
Any little bump and effect will all be done in nDO, so not sure what your mean by feather or detail I am missing.
Took me a total of 16 days 2 of which included adding my support loops and setting up some keyshot renders. Along with all the time spent looking up reference and chatting on skype with Tim and on facebook w/ Demian. I went as accurate as the reference and help I got aswell as feedback from everyone else thanks alot, I will do a better breakdown during the low poly on each individual parts
Edges look super tight in a lot of places along with areas of bad supporting edges making it look almost like it's not even sub-divided. These won't bake down very well to the LP. The spike fore grip is looking super flimsy, the real one looks like it has some thicker construction to it. Not by much but enough to make a difference I feel. The magpul rail covers also look much too narrow. It would even be worth slightly exaggerating them since when viewed in FPV they're going to get super noisy, especially in their current narrowed state.
Mag Release Button/Charging Handle/Groove & graved detail: I will do all that during the low poly inside of nDo the same way I did with the jigsaw, if any detail doesn't come out right in the normal I will just model it on the high. Trying to cut back on the minor stuff and capture most of the model with geo
@Pedro/TheLonePenguin
As for the edge being tight it something that keyshot seems to do making them sharper, there actually semi-loose the same amount of distance I did with my Jigsaw and I had no problems baking that
@LP
Thanks I will make the rail pannels/dust cover less narrower by beefing them up a bit more. I will also extrude the hand-guard vents more thicker shouldn't be too hard to fix these issues
I also don't agree with bad supporting edges most of them are capturing the shape and I don't see to many visible errors of bad topology.
This is looking ace man, keep it up I'll keep my eye on this for the low poly afterwards.
What kind of final result are you looking for? a clean bake or some mid-core damage afterwards?
Small update while I prepare my final UV and analyze my shells to save as much space as possible by mirroring any parts that needs it. *Side not imma push the rail pannels a bit more and exagerate further)
Gun with all parts is 66k poly
Gun without the lw rail/rail panels (biggest poly) = 22k
LW Rail = 16k & Rail Panels = 25k (16 of them)
Since I am taking closeup shots of each parts on its own aswell as the inside detail the poly is a little high, aswell as it being a folio piece. This is my test bake to gauge any problems, have a few minor errors/detail I missed I will take care of; next update will be all the ndo detail/bake done and ready for texture (aswell as new portfolio renders of parts on there own)
Yeah the rail panels are the biggest poly crunchers since I also did the inside, but without any rail panels it ends up at 33k with all the inside being done as-well. I could delete the inside for a game res version/game as a whole and cut it back say to 24k and maintain the high res detail of all the parts you will see. But since I am basically taking it apart I kept it pretty high, I think the only big thing I would really need to do is repack my uv/re-bake to get a good game res version
Anyways just finished the final UV's and going to do a pack/re-bake and toss it into ndo
Finished baking and ndo each individual piece, moving on to texturing next week inside of Substance Designer/Painter. It all in 1 4k map, the magnifier was all ndo work and the pattern on magazine/grip is a stippled hand dotted pattern
I hate to say it after watching you put these sheets together, but you misspelled "Toolbag" at the bottom of each image, and "panels" in the last image.
Also I think the noise/stipple pattern on the rail panels is a bit too strong. It just reads as too much noise to me on the final gun.
Really nice detailed model though, must have taken a long time!
Fixed thanks also reduce the noise a bit still a lil more as it seen at a farther distance (lastly it took me 2 weeks to make the high and 10 days to low poly/uv&bake/ndo), changed Magpul Pistol Grip to Magpul MOE Grip, and added a indention of fire selection knob/moved charging handle in more/setup a better render for portfolio that being developed.
The magnifier probably should have had the insets you did in NDo modeled in. Those grooves are quite pronounced in FP view and because they look a bit cheap being flat like that with generated normals
I like this project. I just wonder if the Magpul Foregrip is actually based on the real Magpul model? In my freetime I work on a similar project and I head to deal with the AFG, too and mine has some small differences.
Thanks everyone, trying to see a layout to go with my portfolio layout for the breakdown. As for the AFG I used a side view image of a Magpul AFG from a gun store website only thing I didnt do was model in the bolt hole on the side, my uvs where to split up to ndo on that spot since it had a series of hard edges
I could go back in and model 2 holes on the side to add some depth it will be covered in black from AO so wasnt to sure if I needed it. I did model the holes for the LarueQT Pivot. As for the groves they where subtle in the reference so I just opted to ndo it as it does still capture the silhouette and visible on the fps views (plus wanted to practice adding grove like that in ndo as it was a little tricky
Same reference but without the support attachment, depending on the lighting and angle the groves can be seen as less pronounced, image I just linked which is what I used the bigger grove are pretty flat while the smaller ones are more pronounced. Guess it how picky one is for such detail, but definitely something I kept in mind I was more linient this time around then my jigsaw in trying to not fake to many deep indents
Thanks Pedro, I also didn't do the text. Split my UV there till I noticed it after it was baked
Lastly got a wip portfolio up: http://www.alexrodriguez3d.com/ so I dont have to link everyone to my artstation still early WIP as I work should have it done by GDC
But if this is an fp weapon and even if the grooves aren't that pronounced, when they are right in your face you would notice the lack of geometry in that area. I would have modeled that if I had as you. Your call though.
Started my first material pass to get the rough color scheme in order aswell as start to hone in on my overall materials. Haven't added the dirt/decay as I want to get a good foundation first to go into Painter and start a new pass.
Struggling on the metal such as the overall metal on Upper/Lower Receiver/LW Rail and Bolt Metal. Got a good start on the plastic surfaces such as my magazine/stock, will continue to do passes with hopes if getting close by end of the month.
Thank Gazu ill add a edge and get that taken care of, got around towards doing my first texture pass in Substance Designer. Will be going into painter now and overlaying noise filters/painting in decay and wanted some feedback on improvements I could do on the base materials before proceeding.
Hey, general critique that I wouldn't expect you to change for this model, but something to keep in mind for feature assets.
It seems like you've modeled some areas in your lowpoly like you would your high, by that I mean you have these separate meshes in the lowpoly, which creates a jagged, lowpoly silhouette that wouldn't be visible had you merged those elements into a solid, watertight mesh.
Those two areas are the most obvious. It looks like you modeled them as if they could be animated/removed. If you plan on animating those bits that makes sense, if not, its generally very inefficient, and again, looks worse due to the blockyness. The flap in the stock is especially confusing, as that area would never be seen from FPV, so you could simplify the model significantly there.
Then, there are some areas like the scope which have relatively low geometry use, areas that will be up close in FPS are where you should spend your effort, not on the unseen stuff.
Overall great job, highpoly is especially good, but you could improve in a few areas here and there with your lowpoly modeling.
Edit: I see you've modeled internals for a lot of parts. Sort of cool to see this done, but its really, really inefficient. There's so much waste in terms of both polygon usage and texture usage here, this isn't something you would ever do in production so, in terms of portfolio work, it sends mixed signals. Yes you can make cool sub-d work, but do you understand what it takes to make production art? That might linger in my head after viewing this piece if I was someone reviewing your folio for a job. You end up with a 65K model that is missing detail in key places, when you probably could have had the same detail or more with 20-30K by prioritizing what would be visible in the end result.
Replies
Stock: 1 day and half
Magazine: 1 hour
That with 5 hour work day, I have great collection of references ill share at the end if anyone wants (just Google searches) of each part from almost all angles. Ill do a exploded render at the end. Trying to think of a good plan to layout the texture/uv to also be re-useable as I do plan on doing a quick ndo or I may just model all the minor details in and do a bake. *Upper Noveske Receiver next then the grip and a scope*
My plans for the RDS attachments is a canted design:
Once I finnish the High ill add my support and toss it for a full critique, with a exploded view to see the parts individually
Time Span:
Magnifer/QD Pivot: Took half a day
EOTech: 1 half Day, didnt expect it to be so big
Iron Sights/Clean up any problem areas: 2 days
Iron Sights Front and Back *71 A.R.M.S
EOTECH 512 with EOLAD 2 Laser Sight
LaRue Tactical QD Pivot Mount for canter sights
Aimpoint 3x Magnifier most of the details will be done in nDo
Detailed item list found at my first post. Wire for now ill post some nice sexy render next time when i finished it on Dec 1st. I have determined my barrel/and lw bar rail already, may add a handgrip at the front aswell
I'm doing the same reference as I said in the other thread, so it will be cool to see how both of ours turn out.
Im also going fully accurate with it and doing it separated, so you may get yours done quicker than mine. The high poly should be done Dec 1st more than likely; than I have to think of a way to uv/low poly to use it with as a kitbash set
Also your trigger guard (again no good with names), yours smoothly fades in with the main frame of the receiver where there should be some spacing.
Awesome breakdown I wasnt going to in-debth with the bolts/screws though but man so sexy looking
I think that here are missing some details like a feather or something.
It a backup iron sight incase your EoTech runs out of battery or gets damaged during combat. Most of the time it flipped over on its side, the back one is the only one I will have up
I am missing 2 bolts on the front and the spring, then I am missing 1 bolt/spring on the right *It easier to just make a set of each type of bolt then spread them around the model then having to make one high poly then go back in and replace each bolt with the new high poly version/same with doing it when I make the low (this way I can just do it once)*
Any little bump and effect will all be done in nDO, so not sure what your mean by feather or detail I am missing.
Could you explain what you mean
nothing youll really see in closeup.
The reference link you posted http://www.spikestactical.com/images/DSC_0424.JPG
Yours however looks pretty thin and flimsy, and the top holes seem to be to elongated as apposed to having circular/rounded off ends to the holes.
http://i.gyazo.com/d2120eacde027c520714e43cce31a662.jpg
http://i.gyazo.com/6f07989599cb85e8e8c78f758f68e2f0.jpg
Your mag release button is not done. Right now it's just a flat shape.
The dust cover is also unfinished. Looks too thin as well.
The charging handle doesn't have those nice groove in it. Neither that cool chamfer on top. Plus everything a bit too hard edged.
These are just somethings i caught at first glance.
Mag Release Button/Charging Handle/Groove & graved detail: I will do all that during the low poly inside of nDo the same way I did with the jigsaw, if any detail doesn't come out right in the normal I will just model it on the high. Trying to cut back on the minor stuff and capture most of the model with geo
@Pedro/TheLonePenguin
As for the edge being tight it something that keyshot seems to do making them sharper, there actually semi-loose the same amount of distance I did with my Jigsaw and I had no problems baking that
@LP
Thanks I will make the rail pannels/dust cover less narrower by beefing them up a bit more. I will also extrude the hand-guard vents more thicker shouldn't be too hard to fix these issues
I also don't agree with bad supporting edges most of them are capturing the shape and I don't see to many visible errors of bad topology.
Here some marmoset render and maya viewports
Its missing its micro detail all of which will be ndo save on poly, I may model the bolt hole though
I did that one which supports the Aimpoint Mag 3x the best, it just missing it grooves
What kind of final result are you looking for? a clean bake or some mid-core damage afterwards?
Keeping enough topology to have my bake looks like its the high poly, then use ndo to add all the micro detail to boost up the model/accuracy
Gun with all parts is 66k poly
Gun without the lw rail/rail panels (biggest poly) = 22k
LW Rail = 16k & Rail Panels = 25k (16 of them)
Since I am taking closeup shots of each parts on its own aswell as the inside detail the poly is a little high, aswell as it being a folio piece. This is my test bake to gauge any problems, have a few minor errors/detail I missed I will take care of; next update will be all the ndo detail/bake done and ready for texture (aswell as new portfolio renders of parts on there own)
Low Poly
Wires1
Wires2
Anyways just finished the final UV's and going to do a pack/re-bake and toss it into ndo
Also I think the noise/stipple pattern on the rail panels is a bit too strong. It just reads as too much noise to me on the final gun.
Really nice detailed model though, must have taken a long time!
Breakdowns
Full
Also setup new renders for my High Poly as I fixed a few issues while I made the Low. More shots on my artstation: http://www.artstation.com/artist/alexrodriguez
Anyway, good job : )
I could go back in and model 2 holes on the side to add some depth it will be covered in black from AO so wasnt to sure if I needed it. I did model the holes for the LarueQT Pivot. As for the groves they where subtle in the reference so I just opted to ndo it as it does still capture the silhouette and visible on the fps views (plus wanted to practice adding grove like that in ndo as it was a little tricky
http://i.gyazo.com/1a7bbeea98177ce262389d4f012f2108.png
Same reference but without the support attachment, depending on the lighting and angle the groves can be seen as less pronounced, image I just linked which is what I used the bigger grove are pretty flat while the smaller ones are more pronounced. Guess it how picky one is for such detail, but definitely something I kept in mind I was more linient this time around then my jigsaw in trying to not fake to many deep indents
Thanks Pedro, I also didn't do the text. Split my UV there till I noticed it after it was baked
Lastly got a wip portfolio up: http://www.alexrodriguez3d.com/ so I dont have to link everyone to my artstation still early WIP as I work should have it done by GDC
Started my first material pass to get the rough color scheme in order aswell as start to hone in on my overall materials. Haven't added the dirt/decay as I want to get a good foundation first to go into Painter and start a new pass.
Struggling on the metal such as the overall metal on Upper/Lower Receiver/LW Rail and Bolt Metal. Got a good start on the plastic surfaces such as my magazine/stock, will continue to do passes with hopes if getting close by end of the month.
Tweak the materials further and started to refine the base metal for the gun a-lot more been struggling trying to get the right base value before moving onto painter/considering using DDO aswell for some of the noise generators for a few minor areas. All work has been designer thus far, found this nice lighting condition of the material I am trying to replicate:
http://wallpaperest.com/wallpapers/weapons-ar15-sbr_330935.jpg
http://wallpaperest.com/wallpapers/weapons-ar15-sbr_360296.jpg
Brown color scheme: http://i221.photobucket.com/albums/dd136/jblomenberg16/March%202010%20AR%20and%20M1%20Out%20Door%20Pictures/DSC06772.jpg
*Crits and suggestion are welcomed*
One thing that is to blocky for first person :
But anyway, very good job
Thanks.
Materials I am referecing:
Metal: http://wallpaperest.com/wallpapers/weapons-ar15-sbr_330935.jpg
Metal: http://wallpaperest.com/wallpapers/weapons-ar15-sbr_360296.jpg
Plastic coating/dark earth parts: http://www.airsoftoutletnw.com/media/catalog/product/cache/1/image/9df78eab33525d08d6e5fb8d27136e95/m/a/magpul_pts_gbb_ubr_stock_dark_earth.jpg
http://images.rockwellarms.com/ProductImages/Magpul/MAG210FDE-alt1.jpg
Color Scheme: https://www.exilemachine.net/shop/images/exm1/IMG_3750.JPGhttp://farm5.static.flickr.com/4012/4442265378_a752bfa33c_z.jpg
It seems like you've modeled some areas in your lowpoly like you would your high, by that I mean you have these separate meshes in the lowpoly, which creates a jagged, lowpoly silhouette that wouldn't be visible had you merged those elements into a solid, watertight mesh.
Those two areas are the most obvious. It looks like you modeled them as if they could be animated/removed. If you plan on animating those bits that makes sense, if not, its generally very inefficient, and again, looks worse due to the blockyness. The flap in the stock is especially confusing, as that area would never be seen from FPV, so you could simplify the model significantly there.
Then, there are some areas like the scope which have relatively low geometry use, areas that will be up close in FPS are where you should spend your effort, not on the unseen stuff.
Overall great job, highpoly is especially good, but you could improve in a few areas here and there with your lowpoly modeling.
Edit: I see you've modeled internals for a lot of parts. Sort of cool to see this done, but its really, really inefficient. There's so much waste in terms of both polygon usage and texture usage here, this isn't something you would ever do in production so, in terms of portfolio work, it sends mixed signals. Yes you can make cool sub-d work, but do you understand what it takes to make production art? That might linger in my head after viewing this piece if I was someone reviewing your folio for a job. You end up with a 65K model that is missing detail in key places, when you probably could have had the same detail or more with 20-30K by prioritizing what would be visible in the end result.