I'm not too keen on the 'composition' but I'm bound to a certain area (same POV at all times) to be visualized in different eras. This is supposed to become a spring post glacial (weichselien) river system (permafrost still present) , but I would like some input from guys who know tundras. I hope it's possible to get some more (color/texture) diversity in the rather even ground area, to make it less dull. Maybe 'islands' of certain vegetation are possible, I really don't know...
The crop is on final size but not the highest quality (book spread).
OK, it seems my last post got wiped off somehow, maybe too large. Here's another attempt at trying to simulate a 2x3km area of post-glacial tundra-steppe-like environment. Crowberry patches and flowers and grass in between, etc. There will be herds of animals in there, later. And some low POV detail renders, which will no doubt arouse more interest ;)
Any comment?
I'm really intrigued by the tributaries in some of the backgrounds. May I ask how you achieved that affect? (just a hint maybe?)
Excellent. I like you comment. It good to do that.
Just a painted mask in Photoshop, as easy as pie.
OK, next version, but I noticed I forgot to mask the procedural lilies out of the moving waters. Now I need to make believable cattails in the swampy lakes...
I also noticed some nasty bumps, even when flattening the snow areas, but that was because the terrain heights (derived from GISdata) was translated into an 8-bit greyscale. I now made that 16-bit, faded a little et voila, it's gone.
One cannot appreciate the magnitude and details of these images until you magnify and scroll through the compositions. Fine work Ulco.
Looks great Ulco.
Very nice tundra texturing overall Ulco. Lots of detail, all the areas and terrain types one would expect. I think the procedural lilies/water plants are fairly effective, but masking them by shore/depth might be good, and also I think there should be less of them (or only at the edges) in the moving water in the river network in the background. Otherwise it all looks great though.
- Oshyan
Thanks, guys.
@Oshyan: earlier post:
Quotebut I noticed I forgot to mask the procedural lilies out of the moving waters.
and; you may be right, but I have to wait what the scientists say. They are VERY picky and sent me comprehensive lists of species associations and environment specifications. But I hope not too picky, because this is largely procedural, and I wouldn't like to paint all sorts of masks for every species. The stagnant remains of what was a tributary as well, may consist of 'islands' of plants, or should be grown over totally, we'll see. One thing I can imagine is that the trees can be more, but clumped in depressions, and the willows in the tributaries als clumped more.
These rocks are max 3m and were later used by early settlers to build our famous 'hunebedden' (dolmens).
EDIT: and a part of the stagnant waters (with grass clumps)
Wow, great work, Ulco!
Interresting project Dune.
In many aspect it's remind me Svalbard in summer.
You put almost every ecosystem we could find in this archipelago between June and September.
I don't have a large experience about tundra and post-glacial environment but I've some comments according to what i saw maybe it can help you.
* threes are rare, tiny or non existants ! but there's an incredible low level life in tundra (mostly moss compose by different vegetals species)
This moss parts are really really thick and compact. It's like a 60 cm thick mattress full of water. glaciaal-test-3E-crop-2.jpg is very pertinent about this. I've saw many water hole like this (not sure about the sand around it ).
* the main problem to me is how you spread all those different ecosystem. You switch tOo quickly from one to another in a random way.
Every ecosystem are covering a much bigger area before switching to an another one but changes to an ecosystem to an another one are quick like you did. There's almost no in-between or blend.
The river in the second picture is exactly what i saw in Svalbard. This kind of ecosystem is present near sea shore for 1 or 2 km.
Further in land the river get finer and more "mono-path" and creates some kind of small canyon (10 meter depth) with less sand and more rocks. More you go in land more the vegetal part came closer from water.
*Water in sandy part is very yellow brown and non transparent. Water in tundra is crystal clear. River Water is clear further in land.
*you could add an new echosystem between tundra and sandy parts in some place : voronoi cracked clay. Like in deserts. But the clay is so "strong" water slip on it, it's almost like a non Newtonian fluid. A friend of mine loose his boots in this shit !!!! If you run it's solid if you stop or walk too slow you gonna be trapped. :P
*Rocks should be visible only on sandy rocky part. They are present in green parts but they're should be fully covered by moss. As i say the moss is very very thick. You can run and jump face on ground without any damage. It's "bouncy"and confortable.
Hope it help.
Anyway you render are already very interesting.
The close-up with the lake and snow (?) patches looks really excellent!
- Oshyan
Thanks very much Ndee! That's information I can work with, very interesting. I'll google Svalbard as well, see what comes up in terms of reference.
As a matter of fact I already had my reservations about the smallscale differences, and have wondered if it shouldn't be less diverse, but more equal, and depending on distance to the tributaries. This is an area where the highest parts (the sandy dune area in front) are only 18m above the bottom of the rivers. Very little height differences on the whole, except the front area.
Work to do! Thanks again!
Redone with larger biotope patches and an attempt to procedurally find depressions for the clumps of trees, that I was asked to put in. Still not sure about the 'ditches' leading to the tributary, they should probably be not sandy but grown over. The melting snow remains will be less bluish and widespread, more like snowheaps; it smoothes the heather too much here.
Next fase (crop): primeval forest and bogs. Still work to do...
Now that looks awesome!
Working on some open areas, where windfall has taken place. Most of the rest will be quite closed canopy, so some points interest there would be good.
May I ask what method are u using to color the populations?
A color powerfractal (or more) led through a transform shader with final position checked, fed into the color input of the object's default shader. The 'old method'.
Small windfall test.
Hmm...not sure I got that but will try. Your method looks better than what I use for sure :D
Awesome looking test btw.
Small sample, check inside object.
Quote from: Dune on October 17, 2014, 07:23:11 AM
Small sample, check inside object.
Thank u Sir! ;)
Will these images be inserts for a book? They are coming along nicely. Would like to see someone do a post-glacial for the Mississippi River area where eventually a diamond mine settled at the southern end.
Oo
Jolie coup.
They will be for a book (6 images), and I think some promotional posters for that book. 72x22cm images @300dpi.
Amazing, Ulco! The colours are fantastic. A typical "Glimmerveen"! ;)
Thanks. Still working on more gaps where trees have fallen and new growth is fighting for the new light, this one windfall is too simple (temporary SSS) and not very nice. I'll do it by PF.
I like the color variation among the trees. Is this the same setup you use for variation in the grass clumps?
The single dead trees are somewhat hard to read.
Maybe if some were a bit darker,it's wet enough there
for must and stuff.
Just a thought.
Glorious complexity and detail, love it!
@Bob: no here I added a masked (by PF+transform) surface shader with some orangy color, under the default shader.
There's 3x pinus sylvestris, 2x oak, 2x linden, 2x beech, some dead trees and shrub, and I used a masked PF+transform to take the leaves of some of the fallen pinus sylvestris. In the latest render, which I still have to finish, I have changed the rotation values so, some trees (from 1 pop) are still standing, other completely down.
I thought about mosses and rot and must on the dead trees, and indeed have a (again, PF+transform) on the dead trees with moss color added. Maybe I have to slide it more to the mossy side.
Update. I think this will be my concept for the scientists to shoot at. At the end there will be cloud shadows and mist perhaps. And birds.
Superb realism.
Looks nice Ulco.
It is realistic probably but still i would lower the contrast between the colors.Subjective of course.
nice realism Ulco!
I agree with Kadri, great image, perhaps a tad less contrast of colors (or saturation?). Just a little...
Also looking forward to seeing this at final render settings. Could you possibly post a crop render with a box around the area of the tree fall? Capture some of the living trees on all four sides. Render at AA16, default settings, Detail 0.75, and don't do any post work at all, no sharpening, etc. I'm just curious to see how it looks, if you don't mind. ;)
- Oshyan
Quote from: Dune on October 09, 2014, 02:17:06 AM.... This is supposed to become a spring post glacial (weichselien) river system (permafrost still present) , but I would like some input from guys who know tundras. ...
Ulco,
The long-term tundra has many occurrences of polygonal shapes caused by the ice forming into some startling shapes.
I took this photo(first) in 2010 from the air. I wish I knew how to manage a voronoi shader better to to this.
Also a shot of a tributary to the Yukon River showing the width of the channel is not constant. Old channels are dry and show up clearly from the air.
Thanks for your input guys. I'll try what you suggest, Oshyan. I know it will be a lot nicer than this (it's just a low-res concept). Regarding the patterns and old streams, I have to see what my 'employer' says, whether these (especially the patterns) did occur here. But it's good to have the knowledge and ideas, thanks!
The voronoi can be made, IMO. Try this sort of setup, maybe it should be warped on a small scale.
Some croptests. I (but that's me) don't see any difference between detail 0.75 AA 16 (1/16) and 0.65 AA6, or at least it's in some minute detail that you might find a difference. Anyway, the total render will be published at 300 dpi, so detail seen at screen resolution at 100% will be 1/4 that size in book (if I state that right). Even soft shadows don't do much in such wooded environment.
Great photorealistic works ! The only difference I would see between the two AA settings is at the little branch level of the dead trees, otherwise everthing else looks quite the same to me.
David.
Indeed, Soft Shadows makes only a very small difference. I definitely notice the improvement with higher AA and Detail, although I would guess most of that is coming from the AA and not the detail, so maybe better to try Detail 0.6, AA 12 or something. I think for heavy vegetation like this something higher than AA6 would be nice though. And no post-work sharpening.
Thanks for trying that. :)
- Oshyan
Lots of really nice stuff happening in there Ulco.
Always nice to see your (dedicated hard) work Ulco.
I'm far from a specialist when it comes to these topics, so I take everything for granted history wise :)
Oshyan has a good suggestion for rendering. AA12 is the max I'd go for.
At least, it's my new max. Did 16 for a short while, but you really have to look hard to see if it's better than 12.
AA12 definitely has that extra edge over AA8 though. So if you the time and resources I'd very much like to see that posted :)
Cheers,
Martin
I don't think it would really be very useful, as the image will be printed in a book at 300dpi and not viewed on screen at 100%. I mean, a lot of the really fine detail will be lost in the 'crude' dots of ink when printed anyway. I think. But it's worth a try, and if it doesn't take days to render, I'm fine with it.
Quote from: Dune on October 24, 2014, 05:07:50 AM
I don't think it would really be very useful, as the image will be printed in a book at 300dpi and not viewed on screen at 100%. I mean, a lot of the really fine detail will be lost in the 'crude' dots of ink when printed anyway. I think. But it's worth a try, and if it doesn't take days to render, I'm fine with it.
Yeah I see your arguments there, but if you read many photography websites you'll see that they value 1:1 pixel detail/sharpness for large scale prints (probably meaning anything over 50-75cm wide).
It remains to be seen though how much is true of that, it's the internets right.
So I think you may be right.
It would actually be nice to test that some time with a capable inktjet printer.
I did some printing of TG work and I must say that it usually looks better printed than on screen :) Your idea too?
Well, the group wasn't very thrilled with these birdseye views, but said it was their own fault. They wanted this, but never realized it's kind of boring. I knew it, and hope they will switch to a series of more detailed, lower POV images of the different eras. More interesting to make, though a bit of waste of time working on these wide vistas. We'll see.... to be continued anyway.
Quoteit usually looks better printed than on screen
I think everything looks better printed than on screen! assuming a competent printer was used, or one does a good job them selves.
Printing is the only way to make sure everyone who looks at it sees the same thing.
I like the bird eye views Ulco, I can't imagine how you would really feel it a waste of time? Maybe you just mean in a business sense? Time is money kind of thing? Sure lower more detail is better, but those are pretty nice.
Well, at least I learned some things of it, and I liked making them, but in terms of profitability it was a waste of time. Doesn't really bother me though. Going ahead full steam...
OK, here's a first iteration of a low POV for Post-Glacial. I'm adding more grasses, changing the colors (forgot to check world position somewhere), making some new arctic low stuff as well. It will be rendered at ~8500px wide, so quite a lot of detail will be visible up front.
The high stuff are juniper, but I have to make some smaller versions as well.
Just finished a bigger portion of the front. The only thing that bothers me (well, maybe more to come) is the sand color creeping up on the rocks here and there.
Looks good Ulco.
Fine work Ulco, especially like the close up of third image. Much detail here with the grasses and flowering plants.
Thanks. Another crop just finished. I'm going to make ephedra distachya, which will be fun.
Very nice image and detail.
That is a very nice plant Ulco; and, comes in a number of varieties. I'd like to see you try Raolia Lutescans which grows in NZ and in Greenland I think.
Truly excellent work!
I need to find out what vegetation is actually present in these sandy, dry slightly higher areas, as it has to be as scientifically correct (for the Netherlands) as possible. So, no NZ/Greenland plants, I'm afraid. By the way, That's a plant that is more easily done procedurally, I think. Too small and fine.
Stagnant water may in time produce bogs, but probably not in this depicted time.
You're right, it is best to stick with the authentic flora of the land. I do get taken up with vegetation that has the smoothness of moss yet displays minute textures and spreads and conforms to the shape of the terrain.
Quote from: Dune on November 16, 2014, 11:27:16 AM
Thanks. Another crop just finished. I'm going to make ephedra distachya, which will be fun.
XFROG makes one in their Red Sea collection. ephedra distachya is sea grape according to Mr Google so it may not be appropriate.
Mr Google gave me this
A number of plant species which grew in the Netherlands during the Weichsel are still found in Northern Europe and in the Alps; they moved their territory behind the receding glaciers. Dryas octopetala (white dryas), Saxifraga oppositifolia (purple saxifrage) and Salix retusa (creeping willow) are some examples.
http://www.zeeinzicht.nl/vleet/index.php?id=3207&template=template-vleeteng&language=2&item=Weichsel-glacial
Love the lichen/mossy rocks, beautiful detailing...
:)
Thanks Bobby, hadn't found that site yet. Luckily, I also have a bunch of scientists providing feedback, but they're not very fast in responding, and I want to get on with it. Just made a bison/wisent, but I still need to give it some postures and hopefully the texture will stick.
The Saxifraga would be nice to put in, my query is out.
Quote from: Dune on November 18, 2014, 03:26:39 AM
Thanks Bobby, hadn't found that site yet. Luckily, I also have a bunch of scientists providing feedback, but they're not very fast in responding, and I want to get on with it. Just made a bison/wisent, but I still need to give it some postures and hopefully the texture will stick.
The Saxifraga would be nice to put in, my query is out.
Here y go, can y tell I'm bored? Haven't yet adjusted to the white shite all about..no bike I stay home way more.
http://www.turbosquid.com/3d-models/3d-plant-saxifraga/759113
Thanks Bobby, but I'll make it myself (if I may add it). Bored? Why not play with the swamp you were planning, nice warm environment to forget about the white stuff ;)
Quote from: Dune on November 18, 2014, 10:12:15 AM
Thanks Bobby, but I'll make it myself (if I may add it). Bored? Why not play with the swamp you were planning, nice warm environment to forget about the white stuff ;)
Coincidentally the swamp is in TG2 at the moment(test render) as I wanted to do a huge render with lots of variety in the foliage, but I've been playing with TG3 so long I find it almost intolerable to not have textured view. Went into TG2 as I have the full version and want lush foliage but I may make it in TG3 after all...just merge some .obj veg in Poseray or Deep Exploration to get more variety in the pops with mixed groups for pops.
Quote from: Dune on November 18, 2014, 10:12:15 AM
Thanks Bobby, but I'll make it myself (if I may add it). Bored? Why not play with the swamp you were planning, nice warm environment to forget about the white stuff ;)
Here y go Ulco...light test WIP...a ways t go yet.
Quote from: Tangled-Universe on October 24, 2014, 06:14:04 AM
Yeah I see your arguments there, but if you read many photography websites you'll see that they value 1:1 pixel detail/sharpness for large scale prints (probably meaning anything over 50-75cm wide).
It remains to be seen though how much is true of that, it's the internets right.
So I think you may be right.
It would actually be nice to test that some time with a capable inktjet printer.
I did some printing of TG work and I must say that it usually looks better printed than on screen :) Your idea too?
Some go even further! :)
http://blog.mingthein.com/2014/02/27/introducing-the-ultraprint/
This fellow actually goes for a 720ppi value when printing and prints 8x12s - 10x15s from high resolution 36-39MP files with amazing results.
He calls them "ultraprints" and the detail is pretty amazing. Reminds me of large format contact prints. Basically, his findings show that at the resolutions his images are being printed, you can hold the print as close as possible to your eye and there will always be the impression of more image detail. The detail doesn't run out/you won't see artifacts from the printing process without a 3-5x loupe.
I have a 44" wide format gallery quality printer here in the studio and have been itching to work on things like this, but haven't had the time.
~Micheal
@ Bobby, the light is great. I would probably get more distance from the trees and show more water.
@ Micheal: that's very interesting, but challenging as well. So we need even more pixels. I recognize the fact that there's a certain microcontrast to take into account. Same as the difference between photo's taken with 'normal' camera's and bigger film sizes, it's often very obvious. I'll be making the renders at 8640px, for a 72cm wide print (I guess the printer prints at 300dpi standard), so that should yield at least some microcontrast in print.
Update. I will try to get the species into a more clustered configuration though. Some blades growing through stone, but that's for post. I'd also like some lichen on the bare sand and some Silene acaulis. But the reindeermoss will be hard, as it's very 'cloudy' and should be really highpoly, and then hundreds of instances.... don't know....
The detail and realism here is absolutely fabulous Ulco. My only small criticism/question is about the "ripples" in the sandy-looking texture. If it's meant to be sand ripples, I think it's a bit too rough-looking and large-grained. The sand should be smoother than the surrounding dirt, I think. Or at least have a different texture of some kind. If it's more meant to be something like rippled mud/dried earth due to varying water levels or something, then it might look more like this, but I still think some more distinction in simple roughness in those areas would be good. It reminds me of how things were in that stratified canyon scene I looked at, there was a lot of uniform roughness that just needed to be broken up some and it improved realism a good deal. It's more subtle here, but something to consider still.
- Oshyan
Yes, you may be right. I added the fake stones at 0.005m lately, but may reduce that to half the size and mask them by some pf. I think it needs more dirt and stuff too, but it's getting a rather busy file as it is already, so I have to choose carefully what to add (and how much time to continue spending on it).
You probably mean that the darker 'sand' is dirt. Well, I have to think about that too, should it be a darker (wet) sand or really dirt/soil plus the lighter sand. Don't know really, but maybe it's good to break the darker bits up to form more soily ground.
Mainly I meant that if you have wavy patterns like that, it's probably going to be assumed to be sand, in which case the texture doesn't look quite right (or at the least it should differ from the non-sand stuff around it more). If you removed that wavy pattern (which I generally like), then I probably wouldn't have questioned it, although the surface roughness *is* a bit uniform...
- Oshyna
Quote from: Dune on November 19, 2014, 03:26:19 AM
@ Bobby, the light is great. I would probably get more distance from the trees and show more water.
O K Dune, enough from this lil guy for now. I'm happy with the pic. Hope y all like it. 8)
I may take another run it this featuring the Swamp House as it's kinda cool itself.
re: glacial....well I'd say damn near perfect but agree with Oshyan's critique. Also encourage you to 'go for it' with the reindeer moss...never seen that stuff so I guess it's a selfish request. 8)
Looking good, Bobby, the lights work very well.
I don't think the moss is needed, as I just got some positive replies from my specialists. They weren't bothered by the sand, so that's easy. Changed a few things though, like more intrazonation (subzoning) of the displacement intersectioned dune valleys. I might have to loose the mammoths and add some wild pig/boar, which I still would have to make (mmm).
By the way; less flowers on the Silene acaulis in the final, as I made a separate distribution of them (green and flowers).
I guess it's more or less up to the Forest!
Quote from: Dune on November 20, 2014, 09:59:07 AM
Looking good, Bobby, the lights work very well.
I don't think the moss is needed, as I just got some positive replies from my specialists. They weren't bothered by the sand, so that's easy. Changed a few things though, like more intrazonation (subzoning) of the displacement intersectioned dune valleys. I might have to loose the mammoths and add some wild pig/boar, which I still would have to make (mmm).
By the way; less flowers on the Silene acaulis in the final, as I made a separate distribution of them (green and flowers).
I guess it's more or less up to the Forest!
Thanks Ulco..was fun to do....stay tuned for the House render as well.
Your renders are for me a constant inspiration......thanks. :)