My only concern is that the poly count on those barrels, fence, and tiny rocks in the background are too high if they're only going to stay in the background.
The only criticism i have is that the background looks odd to me.maybe try lowering it down so the horizon line isn't so high. It could be that the dirt advancing towards the camera is too blurry when it is up close.
honestly, the background is ruining it imo. I thought you'd just found some pic on Google Images and stuck it back there at first. . . dunno if it's the perspective or the lighting, but something doesn't jive with the foreground. Otherwise it's fantastic!
The station looks like next-gen game enviroments should! But yeah, that background looks like you just out a picture on a plane, fix that up a bit. 10/10 otherwise, great work, deserves to be in a profile.
I think as has been said some optermisation could be done.
The tubes as StJoris said could be optermised.
I also think the Barrels and the fence have an unessarily high poly density in them. Optermise them.
[ QUOTE ]
looks to high poly for realtime, and too low poly for cinematic
[/ QUOTE ]
Most definately not the case. It may be too high poly for 5 years ago, but this is easily rendered in any modern engine. Texture memory and draw calls are the big performance killers these days, 10,000 or so polys isnt going to really be noticable at all. Environments are very cheap to render, they dont need to be skined or animated and thats really what kills your performance as far as characters go, transforming all the verts.
Some stuff like the barrels are a tad high but wanted those to look super duper smooth for when I rendered it. If I was to take this into a game I would really knock a bunch of that down.
Changed the sky a bit. There was a little DOF that was blurring the sky out.
[ QUOTE ]
[ QUOTE ]
looks to high poly for realtime, and too low poly for cinematic
[/ QUOTE ]
Hey hawken, I would have to disagree with you on this one.
SO what is ok for next gen polygons, lets say for a tunnel style FPS?
[/ QUOTE ]
When you can save polys, you should, because you can use those freed polys to make other things. If he shaved 80-100 polys out of those barrels he could have A LOT of transparency mapped dead grass and weeds, more rocks and more clutter. I don't see the point in saying a 2k poly barrel is ok when that means the rest of the scene will lack detail.
Why blow your whole poly budget on a barrel when you don't have to? Next gen poly limits do not mean everyone gets to make 200k toothbrushes if the rest of the bathroom is devoid of detail. However you do bump into time constraints, if it will take you a day to optimize the scene and that is time you don't have, let 'er go as is. Since this is looks like its a just for fun project, I wouldn't touch the barrels.
Looks great, I love the detail you put into the pumps, really amazing work.
Very cool. The background doesn't actually bother me, but there's something about the boarded windows on the right that I don't quite buy for some reason, I don't know if it's because they're stacked so thick or because of the textures or what, but it looks kind of odd.
the barrels are really the only thing in the scene that screams out as wasteful, hardly enough to say that it is anywhere close to too highres to run in a game. but yeah, the barrels are a bit excessive.
I think there are quite a few things that would need to be optimized if this was game art, since its not, we can all shut up, unless of course its effecting his render times which I bet its not =P
Things to add to the list of optimization:
Fence boards
Hanging light sockets
Cords going to and from the overhead lights
Siding on the building (add a piece of trim on the edges and use a 4 board texture, alternate the UV's on each board to break up the tile)
Beveled edges on most of the broken concrete
Telephone Pole
Wire receiver on the roof
Water spout by the garage
Window boards ("could" be done with single planes, at the bare min the backsides could be deleted)
I don't mean to bash the work, just pointing out there could be a lot more waste that would cause the scene to chug.
I think Vig pretty much nailed it as far as optimization goes. What's holding those boards up? Are they nailed, screwed, glued or what? Can't really tell from your render. Maybe the camera is too far away, but right now the boards look like they're magically stuck together. It's just nitpicking really. Also, what's your lighting setup like? I suspect a directional light of some sort, but I'd be interestid to know what else you've got in there. Keep rockin'.
Shouldn't take long to lower those at all. Keeped the modifier stack on most of those higher poly objects. Was going to plop it into roboblitz after seeing a hotel someone on here was putting in it that looked neat. Once i get some free time. The plywood on the windows are being held up by super glue too.
Vig you missed the mesh smoothed rebar
Using Vray as the renderer. Just have a direction light in there and a lightprobe I shot in the desert for the skylight
[ QUOTE ]
Why blow your whole poly budget on a barrel when you don't have to? Next gen poly limits do not mean everyone gets to make 200k toothbrushes if the rest of the bathroom is devoid of detail.
[/ QUOTE ]
I know I'm a bit blunt but this is the point I'm trying to make. Next gen doesn't mean we can all get sloppy, EVERYTHING should be optimized ALL the time.
Sterling work btw, I too thought it may be a photo on first look.
Looks great... My only comments would be to agree with Hawken and Vig as well as add;
the hanging rod confused me... it reads as a support beam at first glance... but upon further examination is clearly just a broken hangy poll... i would suggest reataching it or lieing it on the ground or something... it draws too much attention to itself and looks like a floating support beam. IMO
also it's confusing to me why you would board up a wood wall... i would probably do something to make it look like you were clearly boaring up glass windows... the amount of wood you have seems excessive... boarding a window is designed to either keep peoeple out or protect the glass... it doesn't look like anyone would care too much about the glass and that much would wouldn't be needed to keep tings out.
the garage doors seem strange to me.. first because they're both the same materials just one is smaller... not uncommon but it looks strange to me... I would expect to see either a different material for the door... a regular door or no door there... the doors don't look like they would open either... and my last ocmment about the door (i promise) why would you nail criss crossing boards to the front of a garage door half-way down it? if the intent is to stop the garage door without using a pad-lock... i'd think they'd want to nail the boards to the top of the garage door to keep it from sliding up seeing as though the door can't really slide to the side.
The background does look like a lowres photo... but in a way that's kind of a complemint ... it doesn't work well with the scene however... neither does the composition of the background really.. i'd suggest removing the mountain and the hill and doing something with the sky backdrop that makes the clouds look more ineresting as a quicky fix.
P.S. I only comment because i like your work. Good job! This is one of the first env pieces i've seen on polycount that has made me say.. "wow, dat's coo"
Replies
Some stuff needs cleaning up on that.
I thought it was a reference photo
awsome.
My only concern is that the poly count on those barrels, fence, and tiny rocks in the background are too high if they're only going to stay in the background.
Other than that, good job!
all in all though great work.
Aside from that, looks fucking great!
-caseyjones
damn impressive mate.
Also - what was your reference?
looks to high poly for realtime, and too low poly for cinematic
[/ QUOTE ]
Hey hawken, I would have to disagree with you on this one.
SO what is ok for next gen polygons, lets say for a tunnel style FPS?
I think as has been said some optermisation could be done.
The tubes as StJoris said could be optermised.
I also think the Barrels and the fence have an unessarily high poly density in them. Optermise them.
Aprt from that kick arse work dude.
John
looks to high poly for realtime, and too low poly for cinematic
[/ QUOTE ]
Most definately not the case. It may be too high poly for 5 years ago, but this is easily rendered in any modern engine. Texture memory and draw calls are the big performance killers these days, 10,000 or so polys isnt going to really be noticable at all. Environments are very cheap to render, they dont need to be skined or animated and thats really what kills your performance as far as characters go, transforming all the verts.
Some stuff like the barrels are a tad high but wanted those to look super duper smooth for when I rendered it. If I was to take this into a game I would really knock a bunch of that down.
Changed the sky a bit. There was a little DOF that was blurring the sky out.
[ QUOTE ]
looks to high poly for realtime, and too low poly for cinematic
[/ QUOTE ]
Hey hawken, I would have to disagree with you on this one.
SO what is ok for next gen polygons, lets say for a tunnel style FPS?
[/ QUOTE ]
When you can save polys, you should, because you can use those freed polys to make other things. If he shaved 80-100 polys out of those barrels he could have A LOT of transparency mapped dead grass and weeds, more rocks and more clutter. I don't see the point in saying a 2k poly barrel is ok when that means the rest of the scene will lack detail.
Why blow your whole poly budget on a barrel when you don't have to? Next gen poly limits do not mean everyone gets to make 200k toothbrushes if the rest of the bathroom is devoid of detail. However you do bump into time constraints, if it will take you a day to optimize the scene and that is time you don't have, let 'er go as is. Since this is looks like its a just for fun project, I wouldn't touch the barrels.
Looks great, I love the detail you put into the pumps, really amazing work.
Things to add to the list of optimization:
Fence boards
Hanging light sockets
Cords going to and from the overhead lights
Siding on the building (add a piece of trim on the edges and use a 4 board texture, alternate the UV's on each board to break up the tile)
Beveled edges on most of the broken concrete
Telephone Pole
Wire receiver on the roof
Water spout by the garage
Window boards ("could" be done with single planes, at the bare min the backsides could be deleted)
I don't mean to bash the work, just pointing out there could be a lot more waste that would cause the scene to chug.
if those were nailed (and I would assume so) you'd have to be a hell of a lot closer to see.. are you expecting massive tent pegs
Vig you missed the mesh smoothed rebar
Using Vray as the renderer. Just have a direction light in there and a lightprobe I shot in the desert for the skylight
Why blow your whole poly budget on a barrel when you don't have to? Next gen poly limits do not mean everyone gets to make 200k toothbrushes if the rest of the bathroom is devoid of detail.
[/ QUOTE ]
I know I'm a bit blunt but this is the point I'm trying to make. Next gen doesn't mean we can all get sloppy, EVERYTHING should be optimized ALL the time.
Sterling work btw, I too thought it may be a photo on first look.
the hanging rod confused me... it reads as a support beam at first glance... but upon further examination is clearly just a broken hangy poll... i would suggest reataching it or lieing it on the ground or something... it draws too much attention to itself and looks like a floating support beam. IMO
also it's confusing to me why you would board up a wood wall... i would probably do something to make it look like you were clearly boaring up glass windows... the amount of wood you have seems excessive... boarding a window is designed to either keep peoeple out or protect the glass... it doesn't look like anyone would care too much about the glass and that much would wouldn't be needed to keep tings out.
the garage doors seem strange to me.. first because they're both the same materials just one is smaller... not uncommon but it looks strange to me... I would expect to see either a different material for the door... a regular door or no door there... the doors don't look like they would open either... and my last ocmment about the door (i promise) why would you nail criss crossing boards to the front of a garage door half-way down it? if the intent is to stop the garage door without using a pad-lock... i'd think they'd want to nail the boards to the top of the garage door to keep it from sliding up seeing as though the door can't really slide to the side.
The background does look like a lowres photo... but in a way that's kind of a complemint
P.S. I only comment because i like your work. Good job! This is one of the first env pieces i've seen on polycount that has made me say.. "wow, dat's coo"
And yes, a little too much poli IMO is in there. Otherwise is great.