What do you think?

[omegaRED]

added #4

Upped the reflection value, so that the table is reflecting more on the glass ...

http://www.multimania.com/omega7/images/render/glass3.jpg

[Pneumatic]

So what IS in the glass? Is it a glass of water or what?

Wouldn't it be a lot easier to make it a glass of tar? That may save a few layers don't ya think?

[Indiana]

Originally posted by Jon P. Inghram
What'd you render them with? Reminds me of the good old days when I'd spend hours making all sorts of neat 3D renderings on an Amiga.

Speaking of it I still have two images floating around that I "created" (read: uninspiredly took together some objects and textures...) on the Amiga waaay back for a friend who wanted an raytraced pic with a logo:

EDIT: ***The following links contain nudity.***

http://www.indiana.claranet.de/trace1.jpg

http://www.indiana.claranet.de/trace2.jpg

Renderer for these versions was Cinema4D, although the scene originally was composed with Reflections. The original rendering with Reflections was done in 368x566 PAL HAM Interlaced on an Amiga500 and took about 24 hours with the good old 68000@7.14MHz....

P.S.: OmegaRed, the image is getting better and better, I really like the glass3 version the best. Just keep working on it, maybe adding some extra things would make the scene more "alive"? But it's very asthetic the way it's now.

[Pneumatic]

So, uh.......Indiana? I don't know if you noticed it, but there is a nekkid woman in that sphere! Or is that a self portrait like this?

[omegaRED]

Ok guys ... how about this one ...

http://www.multimania.com/omega7/images/render/glass4.jpg

The reason you couldnt see the bottom of the glass on the earlier pics is the lack of it. The glass didnt have a bottom ( on the inner side ). The reason for this is the way 3DSMAX treats materials. There is no way to create a "solid" object or material, all objects are hollow ... so the materials apply to the outside polygons, or if 2 sided, to both sides of a polygon. If i made the glass with bottom and a thicker side wall, there would be 2 reflections & refractions, because MAX would trace the ray as it hits the outside wall and then the inside wall ... which would produce another refraction and look silly ( trust me, i`ve tried ). So the solution is to use a thin ( read : 0 thickness ) wall, but 2 sided, so the inner walls would recieve reflections, but not produce another refraction.

John P. Ingram & Indiana : I remember the good old days on my Amiga 500 too ... i was fiddling around with Real 3D then ... those were the days.

KvH : Do you think the Ice Cubes should be clearer or what? I made then semi-transparent & "dirty" on purpose, to simulate the varied clarity, air pockets and amount of "freezing" in the ice ...

The liquid is supposed to be whiskey-like ... so the transparency is easy to percieve due to the Ice Cubes at the top, but it looks more opaque at the bottom because the color of the table is similar to the one of the liquid ... i hope you like the newest one better than the last couple.

[Maggi]

keep on fine tuning

not bad at all ...

but for http://www.multimania.com/omega7/ima...der/glass4.jpg :

- the table is way too glossy (looks like plastic)
- shadows of that glass are too sharp (looks artificial)
- that pattern on the wall in the background looks too aliased (displacement mapping ?) and seems to have a tiling problem
- you're anti-aliasing settings need some more fine-tuning



Don't get me wrong, I really don't wanna bash your attempts, but you asked for an honest opinion.

Could you post your MR rendering settings ?

Cheers,
Maggi

[omegaRED]

Hey sure Maggi ... thanks for your input too.

I just learned a couple of things to make shadows more blurry, the point is i`m using Ray Traced shadows and their settings arent the same as with the Shadow Map shadows.

As for the suggestions ...

Well, the table is NOT glossy ( 0 value ), but does seem to be too reflective now that you mention it.

Tiling of the wall, i`m aware of it ... just didnt pay much attention to it, i was focused on the Glass & the liquid. The aliasing of the wall is again due to a lower-res texture and low tiling value, thus stretching the texture & getting blurry effect ... i`ll fix it for the next render.

Aliasing, yeah sure it needs some work, because there isnt any . Do you mean using a different filter for the samples? IIRC i didnt encounter any setting for anti-aliasing apart from the filter.

As for the MR settings ... well nothing out of the ordinary, this is for the final renders ...

Min. Samples @ 0 ( occasionaly 1)
Max. Samples @ 2 ( when min=1, then this is 3)
Spatial Contrast = 10,10,10 ( RGB )
Filter = Box, still have to experiment w/ different filters, size 1-1

Shadow Mode = Segments ( still needs experiments w/ it )
Trace Algorithms, both Scanline & Ray Trace

Reflections = 4
Refractions = 4
Sum = 8
Acceleration = BSP
Size = 20
Depth = 40
Memory = 0

No Camera Effects or any Indirect Illumination effects ( slow machine )

Still needs work to be done, but i`m generally satisfied.

BTW Greebe told me you were into 3D modelling, would you mind showing me some of your work so far?

[Maggi]

Hi OmegaRED,

Min. Samples @ 0 ( occasionaly 1)
Max. Samples @ 2 ( when min=1, then this is 3)

try:

Min. Samples @ -3
Max. Samples @ 3

for final rendering



and for my own modelling skills, sadly I don't have enough time to finish anything, but I'll try and see if I can find something that's worth posting. (might take a while though, cos it's been a while since I've been using 3D)

For the shininess of that table, I recommend to scan a blank sheet of unbleached recycled paper (looks sort of grey and has plenty of tiny little irregularities) and apply it as reflectivity map (if there is such an option in 3DSM).

[omegaRED]

http://www.multimania.com/omega7/images/render/glass5.jpg

This is the latest one ... took 1hr to render. Maggi, i implemented your suggestions in this one. Should be better.

D`oh ... i dont like the liquid color ... oh no ... not again ...

The softer shadows lenghtened the render time by ~150% ... thats Area Lights. Samples is 5x5 for 2 Spotlights, resulting in virtually 50 light sources in this scene ... so the render time is accordingly long ... too bad ... i wanted to experiment more, but @1hr per render, i dont have the time nor patience to do that ... maybe i`ll experiment with other effects to achieve softer shadows yet decrease the rendering time.

Maggi, what package are you using for modelling?

[Maggi]

Your picture is really improving, well done !

Of course are those exploding rendering times a real PIA, but hey, we're talking about reproduction of natural elements, right ?



I started quite some time ago on my (g)olden A1230/882@50MHz, using stuff like Imagine and some peeks into Caligary and Real3D.
It became part of my profession about 5 years ago, when I joined a video postproduction's 3D department for commercials, starting as a trainee.
I had to learn Photoshop, Freehand, general digital picture manipulation techniques and acquired quite some knowledge of Softimage.
My career made a few interesting turns since then and on that way I crossed other interesting systems like Discreet Logic's Inferno/Flame/Flint/Edit/Combustion series, which are still one of my favourites and some learned some basics on Quantel's Henry & EditBox.
I had some interesting and neccessary encounters with video technique
(I prefer plugging the wires for myself and if you're alone on a weekend, you better know how to handle those machines)

Right now, I'm working in our VisualFX department, mainly on Digital Fusion and Inferno, but for morphing my trust is still in Elastic Reality. Sadly, it seems to be discontinued.

I also had a peek into Maya, but actually Real3D caught my interest (again), because I've seen very impressive stuff and a friend of mine never left Real3D since his Amiga days, so there's a good teacher in reach.

Dunno, future will tell ...

Back to your picture, you probably should start examining the irregularity of nature, ie. what it really looks like, how asymmetric and irregular natural surfaces and volumes are.
To give you a few hints, scratches and finger prints on glas & table, no repeating patterns anywhere (except you really can't see the pattern), little air bubbles enclosed in the icecubes and bubbling from bottom of glass, foaming a bit on surface, the glass itself need some variance of its refractive index etc. etc. etc.

...

man, you really got me drifting away there for a sec

LOL


Probably you need more refractions for all those cubes and glas surfaces to get rid of that dark center and btw, did you already check out MR's motion blur ?

Finally, it all boils down to how much time and effort you are willing to spend and how fast you can update your rig to keep pace with all those fancies ...

Keep 'em coming !

Cheers,
Maggi

[omegaRED]

http://www.multimania.com/omega7/images/render/glass6.jpg

Upping the Refractions really worked ... it really improved the realism of it.

Maggi, i added a Noise map for a Reflective map on the Table (of course, lower threshold and difference do there are only little irregularites in the reflection), plus a diffuse level on the wall & table so they dont look too "clean".

I didnt have the patience to wait another 2hrs for a render with soft shadows, so i turned them of ... the realism went down but not drastically i think.

IMO, it looks very good ... though there`s always room for improvement. I think i`ll add a couple of other objects in the scene ... an ashtray with a lit cigarette in it or a Magazine ... and a picture on the wall or on the table.

[KvHagedorn]

It does look good.. the bottom looks pretty realistic now. The ice cubes look better, but are still a tad unnatural looking.. maybe if you put a less wavy surface on them, and at angles that are not as sharp it would look right..

[omegaRED]

So i should make then clearer? How about the Specularity on them? Should i lower that?

[Pneumatic]

You made the background darker, and now it looks strange. For the glass to be throwing off that much of a shadow the wall behind it would be brighter.

I like the ice cubes.

[Jon P. Inghram]

Try moving that light source that's above and behind the camera a little to either side. The ice needs fewer bumps and sharper edges so that you can see into and though them better. If the software can do it make the table reflectivity blurry, otherwise just give it a very slight randomly dithered bump map to disperse the reflected image of the glass (not sure if that'll look right, worth giving a try). If you do the bump map thing be sure to turn the specular reflections off on the table or it might end up looking really odd if the lights catch in it. You also might consider reducing the reflectivity of the dark "cracks" in the table top, it might bring out the grain more. Perhaps put a colored light, blue or green perhaps. Just enough to make some faint contrasting colors show up on the edges of the glass and ice, but not enough to illuminate the scene otherwise.