Digital Art Zone

 
   
3 of 6
3
How do I get a reasonable light? Less intense shadows?
Posted: 28 September 2012 02:56 PM   [ Ignore ]   [ # 31 ]
Active Member
RankRank
Total Posts:  506
Joined  2004-06-12
_ PJF _ - 28 September 2012 02:23 PM

Try C - You haven’t noticed…

Oh yes I have noticed, and don’t think I haven’t noticed the sarcastic use of bold on the letter C, either.

I’m not saying other things aren’t at play or wrong, because they are, that’s what I’m saying; you can’t expect to get accuracy out of a ray-tracer unless you feed it accuracy.  Because he’s not feeding it accuracy, there are other things at play, and because of it, I used the example that setting your Refraction to “water” won’t make it accurate.

You know very well that in Bryce, Refraction is actually hard-wired to Reflection.
Consider yourself slapped with a kipper ;-)

Profile
 
 
Posted: 28 September 2012 03:09 PM   [ Ignore ]   [ # 32 ]
Member
Avatar
Rank
Total Posts:  139
Joined  2003-10-09

It’s C for comedy gold.

You need to go back and pay much more attention to the fifth element.

Profile
 
 
Posted: 28 September 2012 08:22 PM   [ Ignore ]   [ # 33 ]
Active Member
RankRank
Total Posts:  506
Joined  2004-06-12

Not sure what it is I’m supposed to have wrong:

- The Refraction is wrong.
- Ray-tracers do exactly what you tell them.
- Feed it right and it renders right.
- Details matter.

That’s effectively all I’ve said, I honestly don’t know why you’re having to fight it with wit.  The few times I tried it I didn’t have any problem making water look right in Bryce, in fact Bryce is one of the easiest renderers out there for water and glass, stuff like that.

I haven’t read all that stuff between you and Rashad yet (only skimmed them here and there), so you’ll have to forgive me if I’ve repeated something you already said, but If I said something really dumb then point it out because I’m buggered if I know what it is.  You’re acting as if you think you’re reading me like a book (IE: nodding).

Whatever it is you think you’re reading, it’s a new one on me because I haven’t a bloody clue what you’re getting at!
Anyway, already spent too much time again, busy.

Profile
 
 
Posted: 28 September 2012 08:37 PM   [ Ignore ]   [ # 34 ]
Active Member
RankRank
Total Posts:  506
Joined  2004-06-12

Just logged back in again, specially for you:

If you’re talking about caustics, yes of course that makes a difference, but that’s irrelevant to the point I was making which was simply that the Refraction is wrong.  It’s picking up too much detail.

Duh!

Profile
 
 
Posted: 28 September 2012 11:05 PM   [ Ignore ]   [ # 35 ]
Active Member
Avatar
RankRank
Total Posts:  716
Joined  2006-05-26

There are a few unique considerations with the water. For one, the entire scenario is intended to be viewed from any angle at any time. This presents a problem for the water because thought it might look okay from one angle move the camera around 180 degrees and it wont look right anymore. There is no master material setting that looks right from all angles under any type of light

I tried applying refraction 133 which Bryce says is water but alas it still wasnt right. The first version used refraction 133 the second uses 112. The water looked too much like solid crystal at 133, dependent of course on the viewing angle.

The main thing I see in those example shots from Len is the caustics of the water, which in my defense I did try to account for in the second version, you can even see the upward caustic on one of the rocks near the shore at the center of the image, but the problems are a little deeper than that. At this shallow depth caustic are essential and impossible to ignore. But we know that caustics arent right in Bryce. So though it is true that a raytracer will do what you tell it to, it doesnt always do it properly.

Because we don’t have displacement for infinite objects like water slabs I had no choice but to use terrains for the dynamic water surface. To avoid observable seams from terrain tiling due to the transparency of the water material I had to try and create all of the ocean that touches the shore with a single terrain object. This greatly limited the amount of detail I am able to use for the waves.

FYI the terrain is set to “solid” so the water should behave volumetrically, instead of just being a transparent wavy surface.

These MegaScenes work differently than more simplified scenes which is why I am studying them. Lots of theoretical ideals that work in less complex situations begin to fail or at least to not work as expected. Not everything scales up as it “should.” MegaScenes present a whole new set of problems.

Essentially, if I had my way I would be using a more detailed water surface and therefore more subtle waves that arent so large. Smaller waves would fit much better for sake of illusion, but as mentioned there are limits to terrain resolution and while 4096 seems like a lot when you start reaching for this level of detail 4096 at this grand of a scale it just isnt enough.

I will be away for a few days so if I dont respond please keep the convo going.

Mark,
I think you’re right. The OP is probably completely lost after our debate. But then again I’m a bit lost myself, so I can’t judge. Trial and error is always a good way to do things when the time permits.

 Signature 

Please view my Daz3d User Gallery
http://www.daz3d.com/gallery/#users/465/

Profile
 
 
Posted: 29 September 2012 05:49 AM   [ Ignore ]   [ # 36 ]
Active Member
RankRank
Total Posts:  506
Joined  2004-06-12

Rashad, just before I dash (and I hope no one thinks I’m being crude by posting it again):
http://www.flickr.com/photos/andybateman/2402808893/

Although that image was added to the end of my post for fun, looking at it from a technical angle it’s actually the best image of the three to teach stuff about water, and here’s why:

- We know the water is accurate because it’s a photo (there’s no arguing with it).
- Look at the shallow water over the sand, it’s clear, even though it looks different at depth (make your water clear).
- The clear water is covering sand (make sure the sand is the right colour).
- The sky is reflected and has the effect of colouring the water (make sure the sky is the right colour).
- The ripples are real (copy them).
- The photo is taken at one angle (copy that angle).

Judge what you have against the photo, and tweak as needed.  When you’re happy, move the bloody camera and see what happens, and because you put the time in copying a photo as close as possible at one angle, you’re well on the way for it looking good at other angles. 

So you take a note of the settings at that angle and then find a very similar environment at a radically different angle, and again, copy it as close as possible.  Note down the settings.  Final step is to take the difference between the settings you arrived at for each angle, and use those as a base.  If you do it right, you’ll be amazed to find it will look remarkably good at any angle you set it.

Plain, simple, logic.

To improve it further (and only when you have the base right), you need to start adding the right volume colour to the water, but when you do that, don’t go overboard and don’t colourise it, try to keep any volume in grey because the main colouring effect for water like that, is the environment, not the water content.  It’s really only stuff like a muddy river that would require a coloured volume (in that case, brown from the mud).

I just read some more of that stuff from you and Peter, and while I’m not debating it (know nothing about it), it was basically what I call “Boadroom Syndrome”, because just like in a boadroom full of marketing wise-asses, you’ll both sit there and witter about technicalities until you’re blue in the face, when actually, getting the job done effectively is much simpler than that.

Think logically, it’s that simple, works for me every time!  I automatically know I can render whatever I want in Bryce and make it look like a photo if I put the time in.  I can do that even if it’s something I’ve never rendered before.  It’s not magic or talent, it’s just logic and common sense - such as approaching things in the manner I outlined above.

Profile
 
 
Posted: 29 September 2012 02:21 PM   [ Ignore ]   [ # 37 ]
Power Member
Avatar
RankRankRank
Total Posts:  1590
Joined  2007-12-10
pumeco - 29 September 2012 05:49 AM

http://www.flickr.com/photos/andybateman/2402808893/

Although that image was added to the end of my post for fun, looking at it from a technical angle it’s actually the best image of the three to teach stuff about water, and here’s why:...

confused Is there water in that photo?... Can’t say I noticed.

 Signature 

I’ve got nothing of my own to promote, no models, no tutorials, no website, nothin’, nada, nill, nowt.

Profile
 
 
Posted: 30 September 2012 12:35 AM   [ Ignore ]   [ # 38 ]
Active Member
Avatar
RankRank
Total Posts:  927
Joined  2003-10-09

All we get washing up on the beaches here are horseshoe crabs and maybe an occasional whale. Sure would love to find one of those in the link washing up on my beach. tongue wink

Profile
 
 
Posted: 03 October 2012 08:28 AM   [ Ignore ]   [ # 39 ]
Active Member
RankRank
Total Posts:  506
Joined  2004-06-12

@TheSavage64
Water? :-D


@LordHardDriven
Apparently there’s a seemingly endless supply of these beautiful creatures lining the beaches (and streets) of Ibiza* :-P
*Van conversion to camper-van with double bed, and destination “Ibiza” are purely coincidental.

Profile
 
 
Posted: 05 October 2012 08:59 AM   [ Ignore ]   [ # 40 ]
Member
Avatar
Rank
Total Posts:  139
Joined  2003-10-09
pumeco - 29 September 2012 05:49 AM

- We know the water is accurate because it’s a photo (there’s no arguing with it).

Spot on, Len.


Well, here we are a week later. I really wanted Rashad to deliver the punch line since it was his post that Len was commenting on originally. But I think Rashad is very distracted at the moment and hasn’t noticed either…

Len, you know that fifth image you have an issue with, the one you’ve made lots of posts about and expended many, many, many words arguing with?

I’m afraid you’re going to have to take it up with a higher power than Rashad about the problematic refraction.

Because, you see, that fifth image, along with the other two of the last three where the water looks good…


.


.


.

 

- it’s a photo.

:mrgreen:


Think I’ll go and make a nice cup of tea.

 

Profile
 
 
Posted: 05 October 2012 03:25 PM   [ Ignore ]   [ # 41 ]
Active Member
Avatar
RankRank
Total Posts:  716
Joined  2006-05-26
_ PJF _ - 05 October 2012 08:59 AM
pumeco - 29 September 2012 05:49 AM

- We know the water is accurate because it’s a photo (there’s no arguing with it).

Spot on, Len.


Well, here we are a week later. I really wanted Rashad to deliver the punch line since it was his post that Len was commenting on originally. But I think Rashad is very distracted at the moment and hasn’t noticed either…

Len, you know that fifth image you have an issue with, the one you’ve made lots of posts about and expended many, many, many words arguing with?

I’m afraid you’re going to have to take it up with a higher power than Rashad about the problematic refraction.

Because, you see, that fifth image, along with the other two of the last three where the water looks good…


- it’s a photo.

:mrgreen:


Think I’ll go and make a nice cup of tea.

 

Ah, I too was confused by what he meant about the refraction being wrong. At first I too thought he was saying that the refraction was wrong in the final image, but I decided that he was saying the last image simply proves that the refraction is wrong in the two example renders I supplied.

Either way I am trying to see what more I can do with the water, but I might have hit a wall of sorts. I welcome all insight.

 Signature 

Please view my Daz3d User Gallery
http://www.daz3d.com/gallery/#users/465/

Profile
 
 
Posted: 05 October 2012 04:50 PM   [ Ignore ]   [ # 42 ]
Member
Avatar
Rank
Total Posts:  139
Joined  2003-10-09
Rashad Carter - 05 October 2012 03:25 PM

...but I decided that he was saying the last image simply proves that the refraction is wrong in the two example renders I supplied.

Just in case the Lenster tries to grasp at that straw, let me quote the start of his lessons:
“Comparing to render number five:”
:wink:

 

Rashad Carter - 05 October 2012 03:25 PM

Either way I am trying to see what more I can do with the water, but I might have hit a wall of sorts. I welcome all insight.

If you’ve hit a wall, Rashad, I’m sure it’s only temporary.

I thought Len was a bit harsh in his criticism (or he has a very smart dog); I thought your first was a good basis for progress in terms of overall realism.

In all three photos you posted as examples, the water has much less obvious surface character. Less reflection, more blurry reflection, and less specular reflection. You see more into the depth of the water in the photos, even though any details are very indistinct (maybe you could use blurry transmissions, too).

You might need to increase the height range of the sand terrain. At the moment the water looks like a thin sheet, and if you lessen its surface character you’ll need more underneath to be on show (even if very blurred).

The hard edge of the water is a 3D giveaway. That’s a toughie, particularly as you are choosing to render small areas of mega-scenes and thus suffer problems when using terrains for the ocean. Without the fractal abilities of MojoWorld, you really are making yourself a rough row to hoe with mega-scenes.

 

Profile
 
 
Posted: 07 October 2012 07:28 AM   [ Ignore ]   [ # 43 ]
Active Member
RankRank
Total Posts:  506
Joined  2004-06-12

WTF LOL

Well, in my defense, I did say the last one was the best one :mrgreen:

And anyway, have you ever heard of PL filters?  I reckon someone used a PL filter then, and because of the reflection being removed by the PL, it made the detail in the water look unnaturally more detailed because it isn’t being masked by reflection.

- Plus, the photo was only small so that was cheating.
- Plus, you are both amateurs compared to me because I am the most bestist.

Anyway, my ol’ bud Rash’ knows that I was exaggerating when I said my dog could render better water.  I’ll be honest and tell you right now, he cannot.  I caught him trying to render himself a Japanese Akita once ‘cause he’s good at fur and stuff.  But water, nope, he’s crap at water.  Now if you don’t mind, I have a love mobile to design, and funny enough, water is on my mind right now because today, I’m eyeballing those “Berkey” water cleaning systems, and until recently, I thought “Berkey” was just another name for you peasants :mrgreen:

Profile
 
 
Posted: 09 October 2012 11:42 AM   [ Ignore ]   [ # 44 ]
Active Member
Avatar
RankRank
Total Posts:  716
Joined  2006-05-26

I’ve taken the feedback and attempted to apply it to the scene. The vegetation has been rebuilt since the last upload so that the needles aren’t so regular as well as a few other changes.

The water in the first render is as you saw it in the previous series. The second render is my attempt at fixing up the refraction as Len pointed out. I also changed the caustic pattern to be a bit less frequent and I also removed the shoreline foam because the water isn’t disturbed enough to produce much of it. The example photos uploaded by myself and by Len show little if any foam so for now I’ve tried to skip it. The third render is just me throwing in some basic models for fun so I can test the upward caustic pattern as well. There are two caustic gels applied one facing down and one facing upward which we can see on the side of the ship. The caustic lights are True Parallel light with “Infinite Width” selected. I should also state that the Sun position has been changed for the third render, shining more directly instead of from behind as in the first two renders.

The water color has been an interesting study. When I removed all color from the Transparent and Volume Color Channels I found the water did not reflect the blue sky enough to gain any real color, so it looked unnaturally colorless. I had to add the color directly to the water. via the Transparent and Volume Color channels. I am using an altitude filter that progresses from white to deep blue at the base of the water terrain to create the look of increased particulate concentration at deeper depths. But alas it seems to create an odd effect where the waves that are tallest render as slight clearer water than the troughs of the waves. Not physically accurate but at least it is a start.

As always feedback is greatly appreciated. Thanks a ton guys!

EDIT: I had a terrible time uploading these images in the proper sequence. Rather than to present the images in the order they were uploaded and named it instead decided to order them based on file size. ARGHH!!!!

Image Attachments
003.jpg002.jpg001.jpg
 Signature 

Please view my Daz3d User Gallery
http://www.daz3d.com/gallery/#users/465/

Profile
 
 
Posted: 10 October 2012 06:34 AM   [ Ignore ]   [ # 45 ]
Active Member
RankRank
Total Posts:  506
Joined  2004-06-12

You’re just messing around, surely :-P

I still think the water is bad considering it’s a RashRender.
The edge of the water, the sand colour, what’s that all about?

One of the most obvious problems with it is the sudden change in colour between the water and sand.  It makes the water look like a shiny sheet that can be peeled off’ the sand!  Clear water doesn’t look like that (perfect excuse to look at that arse pic again for a perfect example). 

In order to get clear blending into a deeper colour as the water gets deeper, you need to apply volumetrics correctly.  Remember the trick Peasant Brinnen discovered for giving Bryce exactly the same feature as In-Scattering used in Carrara?

That’s what you need to be doing, and if you already are, the problem might be how you are doing it.  Let’s say you took a cube and made it into a volume of water just to test it.  Only the top of the cube has wave shapes on it.  Then you sink that cube into the sand so that the top of the water is where you want it.

That might sound sensible but it’ll never work.

And that’s where the problem lies with volumetric water, because although you got it looking right on the cube, the volume isn’t going to change depending on how far you sink it into the sand.  The way to deal with that is easy, you make the volume the same shape as you need it (or at least very close to it).  In other words, a volume of water that is covering the sand at the edges needs to be a thin wedge shape, not a cube, and it;s edge needs to lie just a fraction below the sand.  It matters because when Bryce calculates the soft edge of a volumetric, that soft edge is always defined by the actual shape of the volume shape (the object).

You might even be able to do a boolean on the water/sand to make the volume the right shape.

Profile
 
 
   
3 of 6
3