I'm trying to figure out what the materials should be. I'm "guessing" ceramic and vinyl are major candidates, which seem to be materials of choice for floors tiles. The main arc structures and dividers seem to be a similar substance which is very glossy. I'm seeing a lot of glossy non-metals. I'm not sure about the perimeter. It looks like it could be somewhat glossy, but I'm leaning toward a more matte finish. Some reference links I'm looking to:
The new Star Trek bridge especially has a nice variety of glossy to matte surfaces. Any help, thoughts, suggestions? Material definition is an area where I get a lot of strong crits over so this needs to be done right.
Thanks
Edit: Admittedly another question that is stalling me now is the workflow to use. I have consistently failed to achieve good material definition using Photoshop and Mudbox to make all the textures. I cannot currently buy Substance Painter whose Beta does not allow you to export materials. However, I love it! I'm thinking I might make the materials there anyway, either hoping to purchase at some point this year, or at least to start with. Maybe looking at the 2d view, seeing the channel setup and getting the practice in may even be the spring board I need to get on the right track?
Update with a question. Can you not lerp constants or roughness? I have a mask separating the leather part of the bench from the bottom, so I set up the same thing in the roughness.
Edit: I got the roughness to work with one value being a bitmap. My new question is how to combine a baked normal map with a normal generated from a height (Unreal 4)?
Update. How is it? Does it stink? Is it awesome? Is it painfully unfinished? Is it close? Do the materials read so far? Is it going in the right direction? Do I look like I know what I'm doing? Is it looking like Balamb Garden? Is it looking fake or phony? Does it look realistic? Cartoony? Any thoughts? Thanks :poly141:
Looks like a good start, Materials look fine to me, shiny polished stone surfaces lol, im guessing you havent added any bumpmaps yet, and the lighting is a work in progress i assume.
Looks like a good start, Materials look fine to me, shiny polished stone surfaces lol, im guessing you havent added any bumpmaps yet, and the lighting is a work in progress i assume.
By bump maps do you mean details normals? Only the leather part of the bench has one. Definitely working lighting.
Definitely get some normals in there, and work on some roughness maps. regardless of how well maintained or polished any of the floors or walls are, they will always have variance in roughness.
Looks fine! In the game the material definition for me was pretty vague lol. Everything looked like it was made of chrome in that area except the floor, but that would look gaudy as heck with modern rendering so maybe it's best you're doing your own thing here.
Questions: How much do any of you deal with seams in the editor? Is it like "once in a while" ... "never" ... "on every piece"? I'm flabbergasted here really.
Thanks. After spending days on a giant technical fiasco, I finally have a creative post for critique!
I redid the bench and trashcan and I'm debating on them. The bench looks fine to me from a distance, while not really helping in terms of color and theme (its not very sci fi) nor is it stellar up close. The new trash can looks nice to me but the old one isn't bad, is recognizable as the old game's design and still could use color adjustments.
Love this! It took me 3 copies of the game to finally finish... had a family of cousins who loved to scratch disks, so many cut scene hangs! Definitely an inspiration to remake an old environment! Keep it up
Update. Some thoughts on color ideas. Once again I'm faced with the thought of how much to change the concept. I seemed to get a very uninterested reaction to my new props so I'm not going in that direction. I want to share some ideas about color. I don't really think the colors in this scene's concept really ...make sense :poly142:. There are a lot of colors, different shades of yellow, orange, blue, green, etc. I prefer simple colors schemes but it would break the original concept. So, I'd like to pass the thought on to Polycount while chewing on the idea.
I agree you should stick with the original colour palette.
How did you go about putting those patterns on the floor? I found myself getting stuck with that bit. As it stands I only have a tiling floor material and was considering a decal but that would probably require a huge texture due to the scale of the scene. It looks like you have found a better solution.
I agree you should stick with the original colour palette.
How did you go about putting those patterns on the floor? I found myself getting stuck with that bit. As it stands I only have a tiling floor material and was considering a decal but that would probably require a huge texture due to the scale of the scene. It looks like you have found a better solution.
Keep up the good work
I might be wrong, but it looks like it's still a paint over.
Not an update but a discussion point on sharing my workflow. The modular pieces have been one of the biggest learning points for me here.
The first pics show how I separated some things. I had a tech thread going to ask for help on circular modules. I got some thoughts without a overall concrete rule as to what to do, so I proceeded as I knew. After trying to fiddle with the pivots to make things easier, I gave up and made all pivots to the objects centers and spent hours tediously lining things up. This is hard because you edit the mesh in one program and have the scene in the other. The walkway is the glue in one giant piece to line the others up to. This is easier said than done because this scene has so many circular piece to line up. The inner circle is divided into four pieces with a flat end modeled in. It would be constructed similarly in real life and making it into four made the placement easier than the others (still hard to line it up with the stairs though). The central pillar/elevator is in three pieces with a groove modeled on the the edge so they can fit together. That and the lack of a front entrance wall can be explained simply by me getting exhausted/discouraged by the bajillion hours of time spent on this whole thing, resulting in me saying, "Forget it, I don't have time, the original game doesn't show them so I don't need them in any portfolio shots". The center wall arcs dividing the pool form the walkway are the big fail so far. Look at the top camera and see how the small semi-circle pieces are not lined up to the exit pathways. This was again caused by the hours of tedious hand placement to get everything together (I will be more careful in 3ds Max before calling modules done from now on). The result was me having to angle the paint arrows towards the exit paths. :poly122: I never really found the solution to the problem. I did my best to limit the issue of these pieces lining up bad and this was the best I could do ... it shouldn't show in the final render shots.
The final pics show an example of me imperfectly (but good enough) lining up things. The walkway lines up to the walls, to the bridge, to the door, to the lower-outer walls, to the upper-outer walls, etc.
The materials are mostly node-based tiles in Unreal with masks painted in Photoshop to reveal the small details (like the lines on the ground). the walkways )both below and above the stairs) are the glue of the scene so they are one big piece both in mesh and texture. The file is bigger but its fine. The marble floor is a tile with the lines and arrows in a separate file made in Photoshop, revealed by the mask I mentioned. This was tedious as well because I had to save it, reimport, check it, go back to PS, move the arrow, repeat, repeat, repeat ...
I hope this thread can be a learning experience for all of us
When I do circular environments, I like to section things up, either 1/4 or 1/8th size pieces for the largest circular structures, and then maybe just cut the smallest stuff in half, then like you, I set the pivots at the center of the circle so I can just copy small pieces around to make texturing easier. Thankfully UE4 can make that kind of modularity look good through world space projections and what not, while not having to rely on vertex blending alone for that variation.
You can setup 4 channel blending ( R G B A )in UE4 as well. That means 4 textures vert blending on eachother. That and layered materials... so much possibility.
You could use decals for the floor paint instead of masks, it will make it much easier to line things up. You could even model out custom decal geometry so things line up perfectly.
Just remember, if you want to be able to duplicate a portion of a circular structure for a larger, more complete version, the full cylinder should have an even number of sides. Then you can section it up.
I also think you should push this further than the source material. There is a variety of things that will help the scene be more believable, such as black streak marks from shoes on that floor. They'd have to be pretty subtle though, but I feel like it could add a lot. In addition to that, any areas where people walk a lot will be faded more than others, anything that people touch a lot, dust on things that aren't used often, small amounts of dust/dirt in corners along the floor....
With that being said, I say this because I am loving your execution of the scene so far. Hope it helps.
When I do circular environments, I like to section things up, either 1/4 or 1/8th size pieces for the largest circular structures, ...
This is one thing I've learned. With the wall on the inner side of the pool, I made it into four pieces, then exported two (front-left and back-left) which could be instanced in Unreal to the right side. This way in keeping it in four pieces (or 8 or 16) it will line up in square shapes, since a circle can evenly divide into four squares. This way you aren't guessing so much with the rotation in Unreal.
You can setup 4 channel blending ( R G B A )in UE4 as well. That means 4 textures vert blending on eachother. That and layered materials... so much possibility.
Not sure if I know the term "world space projections" or not. By 4 channel blending do you mean like having all the masks in one texture though the channels?
Whoops, not world space projections, using world space coordinates for variations though is helpful in a lot of situations.
4 channel blending uses each vertex color channel to blend. Red channel would be one texture, all the way to the alpha. It's in the Unreal documentation.
Thanks. This is literally my first ever game art plant, I think its not bad for a first, but it should look good, not good for a first try so feel free to crit. Granted, I don't know how close up it will be seen. Note that in the second to last one its left at translucent. I couldn't get the roughness to work, then I changed it to mask mode so the roughness shows now, but the mask is super hard, even after painting softness on the edges of the mask map. How can I get the gloss on the right and the mask on the left?
Another first ... Unreal water! I'm looking at this thinking, "It's a lot less blue than I anticipated, but then again in real life if you are up close to a fountain/pool, it isn't blue either ... it's transparent. I will likely adjust the color underneath and wait until the final lighting set up is done to worry about the settings on the water. Any thoughts are still welcome.
As far as the plants go, I'm hearing to add more variety. The thing that makes me uncomfortable about that is the UV layout. I'm nervous about stretching/altering the mesh because of what it will do to the UVs. If I add unique UV pieces (let's say a second and third variety) then this will alter the layout overall, meaning the original may have to be moved and resized ... which could mean everything needing to be repainted. Thoughts?
This is plant attempt #4. The ferns are all the same with slight card variation. The pot and stem are on a different mesh. It's a combo of photo and paint over. Thanks
Plant attempt #4 looks cool
Unfortunately as it stands UE4 doesn't do translucency right.
However there is a card on their roadmap to add a forward rendering pass for translucency which would fix that. https://trello.com/b/gHooNW9I/ue4-roadmap
Thanks. Here's a shot at another type of plant. I took a thin cylinder and surrounded it with leaf cards. I took that stick of leaves into Unreal and instanced it several times to bunch up as a whole shrub.
On a side note, how do it fix textures from lowering quality from a distance?
Edit: Another thing to share about my workflow is that I'm taking this piece by piece, step by step as a continual process, as opposed to finishing one thing and moving on to another. I'm chipping away little by little. there still is very little roughness definition in there. I'm trying to get the major design and construction done first, then I will focus on small details.
Question: How do you (or do you) lightmap such a long shrub section? The shrubs on the outer edge of the scene are a collection of several single sticks with leaf cards that were grouped together. For the inner circle shrubs here, I had to take the group, export it, take it into 3ds, make it one mesh, bring it back into Unreal, then be able to easily duplicate it over with a stable pivot point. However, as you can see with the result of flatten mapping, making a lightmap UV for it doesn't look reasonable.
Also how low poly does this need to be? I kept adding pieces on until it didn't look bare, but its pretty loaded now.
What advice would anyone have about mip gen settings in Unreal 4? Reason being I changed all the settings from the default to "noMipmaps". Now while somethings are looking nice and sharp, other things (trashcan, floor, metal on the TV, etc) are looking way too noisy, like a random extreme Photoshop noise filter was applied to it? How much do you change these settings and to what? As you can see in the older pic, some things such as the orange and blue lines on the central pillar were mipping really bad so I had to change something. Should I have set it to something else? Should I have adjusted distance somehow instead? I've never worked with mipping before at all. Should I simply fix up the maps to smooth out the ones now too noisy (that's my instinct)? Maybe just set up some multiplies of the map to reduce strength? Thoughts?
Edit: note that I'm researching this stuff right now for the first time.
The latest issue I'm addressing now is how the fish normals are clearly reducing in effectiveness after building under certain lighting conditions. I'm attempting to return to the state I was in where the fish looked better but to no avail so far. It seems like depending upon placement in the scene, the fish are getting better or worse normal depth (because of where the sunlight is). Thoughts?
I don't remember Balamb Garden having harsh shadows. I remember a lot of reflective materials, bright lights and lots of reflections from the water. If it is your take on it, that's cool, but it's giving it more mood than it actually has. Keep going man. FFVIII is many people's fave FF, so uhhh no pressure...JK Good luck!
I don't remember Balamb Garden having harsh shadows. I remember a lot of reflective materials, bright lights and lots of reflections from the water. If it is your take on it, that's cool, but it's giving it more mood than it actually has. Keep going man. FFVIII is many people's fave FF, so uhhh no pressure...JK Good luck!
This is the sort of thing I want... I want to make some interesting, new and more extreme lighting that will take it into today. The thought occurred to me early on that if there is, in fact, a little window up above the scene:
then I could make some extreme sunlight come down through there for something more dynamic. The original game lighting, in all due respect, is a bit washed out and even. That doesn't turn heads on to your portfolio piece.
As for the lighting issues I'm working on now, how important is it to be building lightmaps? If I manually move the lights around after building, I'm getting nicer results with the direct lights, while the GI is baked on. Would it make sense to the finalize it like so? In other words, build the lights for the GI, then nudge the sunlight and blue lights in the scene to improve their effects in the scene. Thoughts?
Creative crits on the lighting? The scene in general? If I were to stop now, what crits would you have? Overall thoughts? I am nearing the end. At this point don't assume I'm "going to fix it eventually". Whether I change it or not, I'm always up for crits on anything at any stage of development. Thanks.
as my basis and am now simply tweaking values. From certain angles it disappears in the center pool (fresnel stuff going on in the material, its over my head at this point). The thing is, water is clear, not blue. However, art usually depicts it as blue and viewers expect it to be blue. Often pools have a blue inside so we think blue when we see indoor water. This is despite the PBR setup. I'm not sure what to do. I believe the original game water is extremely fake, but it meets viewer expectations more than mine does, and I'm not a level designer whose confident here in the editor. I am very serious, however, about getting this piece right before I quit. Is it better here, worse, the same?
This is making me wish ff8 had been done in the modern era T_T its awesome to see this bit of the game recreated like this.
As for the water. The blue color comes from it reflecting the sky. Though realistic water is much more complicated than just it being blue.
In real life indoor water feature would be clear. And a the color comes either from underwater lighting or the pool being itself colored blue. You can see in the ps1 screens that they do have blue lights on the floor and what seems to be hidden regressed lighting in the walla of the pool.
Working from a tutorial, I tweaked and tweaked until this happened and I was finally happy with it. I had to ask myself questions about water in real life versus a PSX game. The water here isn't deep so for a somewhat realistic direction, it should look like it isn't deep. I received some good crits on it looking too noisy and such. :thumbup:
Well, its coming to a close soon. It's been quite an experience. Much thanks again. I hope that you might have been able to learn some things from this project as I have. One reason I chose this piece is because it has so many things such as water, plants, metal, marble, vinyl, porcelain, wood, screens with graphics, emissives, modern, futuristic, modules, colored lights, etc. It covered several bases. Also another reason is because this scene is iconic for its source. I've found it hard to choose a good environment that immediately speaks of a story I know and love. My experience here has already given me some ideas about my next piece. Next up will likely be some final shots and narrowing it down to one primary render. :poly009:
Replies
I'm trying to figure out what the materials should be. I'm "guessing" ceramic and vinyl are major candidates, which seem to be materials of choice for floors tiles. The main arc structures and dividers seem to be a similar substance which is very glossy. I'm seeing a lot of glossy non-metals. I'm not sure about the perimeter. It looks like it could be somewhat glossy, but I'm leaning toward a more matte finish. Some reference links I'm looking to:
http://blog.box.com/wp-content/uploads/2013/04/USS_Enterprise_alternate_reality_bridge.jpg
http://gedblog.com/wp-content/uploads/abrams_bridge.jpg
http://i.imgur.com/5aaSM.jpg
http://www.ex-astris-scientia.org/inconsistencies/sets/corridor-enterprised-whentheboughbreaks.jpg
The new Star Trek bridge especially has a nice variety of glossy to matte surfaces. Any help, thoughts, suggestions? Material definition is an area where I get a lot of strong crits over so this needs to be done right.
Thanks
Edit: Admittedly another question that is stalling me now is the workflow to use. I have consistently failed to achieve good material definition using Photoshop and Mudbox to make all the textures. I cannot currently buy Substance Painter whose Beta does not allow you to export materials. However, I love it! I'm thinking I might make the materials there anyway, either hoping to purchase at some point this year, or at least to start with. Maybe looking at the 2d view, seeing the channel setup and getting the practice in may even be the spring board I need to get on the right track?
Update with a question. Can you not lerp constants or roughness? I have a mask separating the leather part of the bench from the bottom, so I set up the same thing in the roughness.
Edit: I got the roughness to work with one value being a bitmap. My new question is how to combine a baked normal map with a normal generated from a height (Unreal 4)?
Update. How is it? Does it stink? Is it awesome? Is it painfully unfinished? Is it close? Do the materials read so far? Is it going in the right direction? Do I look like I know what I'm doing? Is it looking like Balamb Garden? Is it looking fake or phony? Does it look realistic? Cartoony? Any thoughts? Thanks :poly141:
By bump maps do you mean details normals? Only the leather part of the bench has one. Definitely working lighting.
http://www.polycount.com/forum/showthread.php?t=140410
Questions: How much do any of you deal with seams in the editor? Is it like "once in a while" ... "never" ... "on every piece"? I'm flabbergasted here really.
I redid the bench and trashcan and I'm debating on them. The bench looks fine to me from a distance, while not really helping in terms of color and theme (its not very sci fi) nor is it stellar up close. The new trash can looks nice to me but the old one isn't bad, is recognizable as the old game's design and still could use color adjustments.
Quack! figured out that I had the normal compression set to TC_Default instead of TC_normal.
Update. Some thoughts on color ideas. Once again I'm faced with the thought of how much to change the concept. I seemed to get a very uninterested reaction to my new props so I'm not going in that direction. I want to share some ideas about color. I don't really think the colors in this scene's concept really ...make sense :poly142:. There are a lot of colors, different shades of yellow, orange, blue, green, etc. I prefer simple colors schemes but it would break the original concept. So, I'd like to pass the thought on to Polycount while chewing on the idea.
How did you go about putting those patterns on the floor? I found myself getting stuck with that bit. As it stands I only have a tiling floor material and was considering a decal but that would probably require a huge texture due to the scale of the scene. It looks like you have found a better solution.
Keep up the good work
For now all looks good. Keep going.
Not an update but a discussion point on sharing my workflow. The modular pieces have been one of the biggest learning points for me here.
The first pics show how I separated some things. I had a tech thread going to ask for help on circular modules. I got some thoughts without a overall concrete rule as to what to do, so I proceeded as I knew. After trying to fiddle with the pivots to make things easier, I gave up and made all pivots to the objects centers and spent hours tediously lining things up. This is hard because you edit the mesh in one program and have the scene in the other. The walkway is the glue in one giant piece to line the others up to. This is easier said than done because this scene has so many circular piece to line up. The inner circle is divided into four pieces with a flat end modeled in. It would be constructed similarly in real life and making it into four made the placement easier than the others (still hard to line it up with the stairs though). The central pillar/elevator is in three pieces with a groove modeled on the the edge so they can fit together. That and the lack of a front entrance wall can be explained simply by me getting exhausted/discouraged by the bajillion hours of time spent on this whole thing, resulting in me saying, "Forget it, I don't have time, the original game doesn't show them so I don't need them in any portfolio shots". The center wall arcs dividing the pool form the walkway are the big fail so far. Look at the top camera and see how the small semi-circle pieces are not lined up to the exit pathways. This was again caused by the hours of tedious hand placement to get everything together (I will be more careful in 3ds Max before calling modules done from now on). The result was me having to angle the paint arrows towards the exit paths. :poly122: I never really found the solution to the problem. I did my best to limit the issue of these pieces lining up bad and this was the best I could do ... it shouldn't show in the final render shots.
The final pics show an example of me imperfectly (but good enough) lining up things. The walkway lines up to the walls, to the bridge, to the door, to the lower-outer walls, to the upper-outer walls, etc.
The materials are mostly node-based tiles in Unreal with masks painted in Photoshop to reveal the small details (like the lines on the ground). the walkways )both below and above the stairs) are the glue of the scene so they are one big piece both in mesh and texture. The file is bigger but its fine. The marble floor is a tile with the lines and arrows in a separate file made in Photoshop, revealed by the mask I mentioned. This was tedious as well because I had to save it, reimport, check it, go back to PS, move the arrow, repeat, repeat, repeat ...
I hope this thread can be a learning experience for all of us
You can setup 4 channel blending ( R G B A )in UE4 as well. That means 4 textures vert blending on eachother. That and layered materials... so much possibility.
You could use decals for the floor paint instead of masks, it will make it much easier to line things up. You could even model out custom decal geometry so things line up perfectly.
Just remember, if you want to be able to duplicate a portion of a circular structure for a larger, more complete version, the full cylinder should have an even number of sides. Then you can section it up.
I also think you should push this further than the source material. There is a variety of things that will help the scene be more believable, such as black streak marks from shoes on that floor. They'd have to be pretty subtle though, but I feel like it could add a lot. In addition to that, any areas where people walk a lot will be faded more than others, anything that people touch a lot, dust on things that aren't used often, small amounts of dust/dirt in corners along the floor....
With that being said, I say this because I am loving your execution of the scene so far. Hope it helps.
This is one thing I've learned. With the wall on the inner side of the pool, I made it into four pieces, then exported two (front-left and back-left) which could be instanced in Unreal to the right side. This way in keeping it in four pieces (or 8 or 16) it will line up in square shapes, since a circle can evenly divide into four squares. This way you aren't guessing so much with the rotation in Unreal.
Thanks.
Not sure if I know the term "world space projections" or not. By 4 channel blending do you mean like having all the masks in one texture though the channels?
Edit:.
4 channel blending uses each vertex color channel to blend. Red channel would be one texture, all the way to the alpha. It's in the Unreal documentation.
https://docs.unrealengine.com/latest/INT/Engine/UI/LevelEditor/Modes/MeshPaintMode/VertexColor/MaterialSetup/4Way/index.html
Thanks. This is literally my first ever game art plant, I think its not bad for a first, but it should look good, not good for a first try so feel free to crit. Granted, I don't know how close up it will be seen. Note that in the second to last one its left at translucent. I couldn't get the roughness to work, then I changed it to mask mode so the roughness shows now, but the mask is super hard, even after painting softness on the edges of the mask map. How can I get the gloss on the right and the mask on the left?
Another first ... Unreal water! I'm looking at this thinking, "It's a lot less blue than I anticipated, but then again in real life if you are up close to a fountain/pool, it isn't blue either ... it's transparent. I will likely adjust the color underneath and wait until the final lighting set up is done to worry about the settings on the water. Any thoughts are still welcome.
As far as the plants go, I'm hearing to add more variety. The thing that makes me uncomfortable about that is the UV layout. I'm nervous about stretching/altering the mesh because of what it will do to the UVs. If I add unique UV pieces (let's say a second and third variety) then this will alter the layout overall, meaning the original may have to be moved and resized ... which could mean everything needing to be repainted. Thoughts?
This is plant attempt #4. The ferns are all the same with slight card variation. The pot and stem are on a different mesh. It's a combo of photo and paint over. Thanks
Unfortunately as it stands UE4 doesn't do translucency right.
However there is a card on their roadmap to add a forward rendering pass for translucency which would fix that.
https://trello.com/b/gHooNW9I/ue4-roadmap
Loving the progress on this project, keep it up
Thanks. Here's a shot at another type of plant. I took a thin cylinder and surrounded it with leaf cards. I took that stick of leaves into Unreal and instanced it several times to bunch up as a whole shrub.
On a side note, how do it fix textures from lowering quality from a distance?
Edit: Another thing to share about my workflow is that I'm taking this piece by piece, step by step as a continual process, as opposed to finishing one thing and moving on to another. I'm chipping away little by little. there still is very little roughness definition in there. I'm trying to get the major design and construction done first, then I will focus on small details.
Question: How do you (or do you) lightmap such a long shrub section? The shrubs on the outer edge of the scene are a collection of several single sticks with leaf cards that were grouped together. For the inner circle shrubs here, I had to take the group, export it, take it into 3ds, make it one mesh, bring it back into Unreal, then be able to easily duplicate it over with a stable pivot point. However, as you can see with the result of flatten mapping, making a lightmap UV for it doesn't look reasonable.
Also how low poly does this need to be? I kept adding pieces on until it didn't look bare, but its pretty loaded now.
Update. Elevator, lightmaps, plants, some normals, some lighting, a lot of roughness, all kinds of little things here and there, etc. :poly009:
I think the strongest shot in your most recent post is the top right. I feel like you can get the other 3 to that level as well
Thanks.
What advice would anyone have about mip gen settings in Unreal 4? Reason being I changed all the settings from the default to "noMipmaps". Now while somethings are looking nice and sharp, other things (trashcan, floor, metal on the TV, etc) are looking way too noisy, like a random extreme Photoshop noise filter was applied to it? How much do you change these settings and to what? As you can see in the older pic, some things such as the orange and blue lines on the central pillar were mipping really bad so I had to change something. Should I have set it to something else? Should I have adjusted distance somehow instead? I've never worked with mipping before at all. Should I simply fix up the maps to smooth out the ones now too noisy (that's my instinct)? Maybe just set up some multiplies of the map to reduce strength? Thoughts?
Edit: note that I'm researching this stuff right now for the first time.
The latest issue I'm addressing now is how the fish normals are clearly reducing in effectiveness after building under certain lighting conditions. I'm attempting to return to the state I was in where the fish looked better but to no avail so far. It seems like depending upon placement in the scene, the fish are getting better or worse normal depth (because of where the sunlight is). Thoughts?
Thanks. Have you seen Jordan Walker's Bathhouse?
http://eat3d.com/files/imagecache/blog_content/blog_images/jordan_walker_bathhouse.jpg
http://videogam.in/images/0000001/streetfighterii-worldwarrior-japanehonda.png
This is the sort of thing I want... I want to make some interesting, new and more extreme lighting that will take it into today. The thought occurred to me early on that if there is, in fact, a little window up above the scene:
http://img.bhs4.com/9B/3/9B31EFF0B98C52C626FCC261F9A1A4A6EB3A5CF7_large.jpg
then I could make some extreme sunlight come down through there for something more dynamic. The original game lighting, in all due respect, is a bit washed out and even. That doesn't turn heads on to your portfolio piece.
As for the lighting issues I'm working on now, how important is it to be building lightmaps? If I manually move the lights around after building, I'm getting nicer results with the direct lights, while the GI is baked on. Would it make sense to the finalize it like so? In other words, build the lights for the GI, then nudge the sunlight and blue lights in the scene to improve their effects in the scene. Thoughts?
Creative crits on the lighting? The scene in general? If I were to stop now, what crits would you have? Overall thoughts? I am nearing the end. At this point don't assume I'm "going to fix it eventually". Whether I change it or not, I'm always up for crits on anything at any stage of development. Thanks.
edit: Thoughts on the water?
Crits or thoughts on water pt. 2. Should I go opaque on the water? Admittedly I'm using this tut here:
https://www.youtube.com/watch?v=KPmRV1Z9ikY
as my basis and am now simply tweaking values. From certain angles it disappears in the center pool (fresnel stuff going on in the material, its over my head at this point). The thing is, water is clear, not blue. However, art usually depicts it as blue and viewers expect it to be blue. Often pools have a blue inside so we think blue when we see indoor water. This is despite the PBR setup. I'm not sure what to do. I believe the original game water is extremely fake, but it meets viewer expectations more than mine does, and I'm not a level designer whose confident here in the editor. I am very serious, however, about getting this piece right before I quit. Is it better here, worse, the same?
http://img.bhs4.com/E7/D/E7D505E66FF75D1CC615A16C8E27EA96622EAEF2_large.jpg
http://jegged.com/Image/Strategy/Final-Fantasy-8/Walkthrough/FF8-0014-Balamb-Garden-Library.png
As for the water. The blue color comes from it reflecting the sky. Though realistic water is much more complicated than just it being blue.
In real life indoor water feature would be clear. And a the color comes either from underwater lighting or the pool being itself colored blue. You can see in the ps1 screens that they do have blue lights on the floor and what seems to be hidden regressed lighting in the walla of the pool.
New Water! Shader nodes:
http://i.imgur.com/IzWYabD.jpg
Working from a tutorial, I tweaked and tweaked until this happened and I was finally happy with it. I had to ask myself questions about water in real life versus a PSX game. The water here isn't deep so for a somewhat realistic direction, it should look like it isn't deep. I received some good crits on it looking too noisy and such. :thumbup:
Well, its coming to a close soon. It's been quite an experience. Much thanks again. I hope that you might have been able to learn some things from this project as I have. One reason I chose this piece is because it has so many things such as water, plants, metal, marble, vinyl, porcelain, wood, screens with graphics, emissives, modern, futuristic, modules, colored lights, etc. It covered several bases. Also another reason is because this scene is iconic for its source. I've found it hard to choose a good environment that immediately speaks of a story I know and love. My experience here has already given me some ideas about my next piece. Next up will likely be some final shots and narrowing it down to one primary render. :poly009:
Here are the final shots of Balamb Garden! Thank you for the support and contribution. :poly136:
I'm happy with the result after 10.5 weeks and I learned a lot. I finally did it!