Maybe you refuse to bother because of all these spheres, but they're my
recent attempts at realistic raytracing. What do you think; is there a photo
hiding? What is definitely not photo? Why?

Regards,
Hugo

Post a reply to this message

Attachments:
Download 'Real_experiments.JPG' (72 KB)

Preview of image 'Real_experiments.JPG'

Very very nice photorealistic work.
N°6  (or 1 perhaps) is a photo ?

Mathieu

Post a reply to this message

Not N6, scuse me >> N°5

Mathieu

Post a reply to this message

"Hugo" <hua### [at] post3teledk> wrote in message
news:3caf13cc@news.povray.org...
> Maybe you refuse to bother because of all these spheres, but they're my
> recent attempts at realistic raytracing.

1) A photo, only the lamps reflections bother me somehow. This might be real
ball bearings on a plastic carpet.

2) Nearly a photo, the largest spheres material looks unreal. The black one
is great and needs to be shared with us.

3) Not a photo, the spheres material looks unreal. Same as the big in no. 2.

4) Dunno, without the sphere might be a photo, it's the highlight.

5) A photo, good plastic on the white object.

6) Nearly a photo, the grey object should be lifted slightly off the surface
or the edge needs a bit roundess.

Ari-Matti

Post a reply to this message

I think they're all renders, though some are good.

1. Very good, probably the best. but the pattern on the floor looks a bit like a
standard noise function variant. The blur adds so much to the realism.

2. Would look rendered even if it was a photo! (though I don't think it is).
This style of lighting/radiosity/camera would look great for real world objects.

3. The ball looks like it's been composited onto the backdrop. I think maybe the
shadow needs to be darker.

4. The shadow of the ball on the sticking up object seems to have a double image
effect, as if you had an area light of too low resolution.

5. very nice. The only give away that it's traced is that there's nothing else
in the scene (which would be seen in radiosity and reflections). That in itself
is possible if you take the photo in a studio, but the light source is so far
away that it looks like sunlight... um, I guess what I'm saying is it looks
exactly like two objects photographed in an infinite flat white desert!

6. There's radiosity artefacts at the base of the shiny thing. Also it has a bit
of that disconnected feel that was in 3.

Very nice on the whole. Now do it with real objects!

--
Tek
http://www.evilsuperbrain.com

Post a reply to this message

Nr. 4 is definitively not a photo. It stands for waxlike objects. But the
texture is too transparent, so you can see edges through, which is not
possible with real wax.
I think everything can be made in Pov-Ray. If anything, no. 5 can be a
photo.


Norbert

Post a reply to this message

#1: highlights are to perfect, or I'm missing the lamp in
    the reflections... averything else looks real here...
#2: naa... white material is a bit unreal, and the reflecting
    environment of the black sphere doesn't fit here ...
#3: To sharp sphere in contrtast to the "blurry" floor...
#4: again, the highlight on the sphere is talking to
    me: "I'm unreal... I'm traced"
#5: dunno ... can't find a detail, where I would be 100% sure,
    that it is a traced one. The shadows are very slightly blurred,
    so it would be a very small arealight, and, it must be at a huge
    distance, as the shadows are nearly parallel...
    This would be very hard in a studio, (and I see no other
    reflections), but who knows...
    The highlights are somehow "real" to me ...
#6: The Ref-Map of the "thing" doesn't fit, looks somehow the
    same as in #2, and is missing in the floor.
    IMHO, If you have the objects mirroring in the floor, you
    also have the sky there ...
    2nd, there is something with the shadows, but I can't tell what

So, if there's any pic of these one a photo, I'd stick on #5 ...
but I'm sure, you could also trace a pic like this one ...

Post a reply to this message

"Hugo" <hua### [at] post3teledk> wrote in message
news:3caf13cc@news.povray.org...
> Maybe you refuse to bother because of all these spheres, but they're my
> recent attempts at realistic raytracing. What do you think; is there a
photo
> hiding? What is definitely not photo? Why?
You mean that they aren't all photos ;) ?
the sphere in 3 is obviously not a photo (the sphere looks wrong), but
that's all I can tell.


--
____________________
 RAY

Post a reply to this message

Hugo wrote:
> 
> Maybe you refuse to bother because of all these spheres, but they're my
> recent attempts at realistic raytracing. What do you think; is there a photo
> hiding? What is definitely not photo? Why?

1. Not a photo. Sphere material seems unrealistic.

2. Not a photo, but it might be with different material for center and
right sphere. Reflection in left sphere and ground convincing, as is shading.

3. Definitely not a photo. Shadow on white sphere darker than shadow on
gray ground, especially considering only one highlight on sphere!

4. Definitely not a photo. Good texture on ground, but left object seems
blurry somehow. "Halo" effect on inside edge of sphere seems to have no cause.

5. Could be a photo. Extremely convincing.

6. Not a photo. Unsure exactly why, but heavy shading under the blue
objects where they (maybe) should be better lit may contribute to failure.

I would like to say, though, that all in all these are very nice, and
could probably put a lot of other renderers to shame, or at least match
them. Thanks go to everyone who's gone to lengths to play with radiosity
and lighting and show the rest of us how it's done; Kari Kivisalo comes
to mind, but there were many other contributors, including yourself as
we can see here.

-Xplo

Post a reply to this message

"Hugo" <hua### [at] post3teledk> wrote in message
news:3caf13cc@news.povray.org...
> Maybe you refuse to bother because of all these spheres, but they're
my
> recent attempts at realistic raytracing. What do you think; is there a
photo
> hiding? What is definitely not photo? Why?

    Hmm.. Nice, very nice. The dark sphere in number two is very close
to a black pearl, not sure about the two spheres beside it though... The
black pearl needs some lustre and then bingo.

   As for the rest? I'll stick my neck out and say all .pov apart from #
6.

   Just hope I'm right...  ;)

      Nice.

    ~Steve~




>
> Regards,
> Hugo
>
>
>

Post a reply to this message

Interesting how others are saying #5 could be a photo, I'd have to choose
that one too. Maybe it's the little top central bumps; or lack of other
textures, which can be scrutinized more readily were they there. Regardless
of what someone else said about #4's semi-transparency it draws my attention
most as being realistic.

Post a reply to this message

Thank you Ari-Matti for your comments! I am encouraged by all the replies I
got from different people. Also your useful analyse.

You ask about the black sphere in #2. It mainly consists of blurred
reflection and depends very much on the enviroment, as you can imagine. The
enviroment is a HDR light probe (High Dynamic Range image). Radiosity and 3
dim area_lights is used to illuminate the scene. Rendertime: 5 minutes on my
1 ghz AMD. Rendered with ML-POV.

If you want to try this yourself, I attached the .pov file, but you'll also
need these links:

http://martial.rameaux.free.fr/mael/MLPov.html
http://www.debevec.org/%7Edebevec/Probes/

Good luck!
Regards,
Hugo

Post a reply to this message

Attachments:
Download 'HDR_test1.pov.txt' (3 KB)

Hi Tek, thanks for your comments! No, they are not all renders. There is one
photo.  :o)  I'll reviel which one a bit later.

> 3. The ball looks like it's been composited onto
> the backdrop. I think maybe the shadow needs
> to be darker.

You're right. I clearly see what's wrong now. Well, it's my first experiment
with 17 randomly placed area_lights (nearly random) with random colors.
Another strong area_light gives the most visible shadow. No radiosity. I was
inspired by John Bradshaw's post, and I think the method can be improved.

John's post:
http://news.povray.org/povray.general/22572/


> 6. There's radiosity artefacts at the base of the shiny
> thing. Also it has a bit of that disconnected feel
> that was in 3.

It's good you mention this. I wasn't aware of it.

Regards,
Hugo

Post a reply to this message

Hi Nobert, and thanks for your reply! I knew that you wouldn't buy #4 but I
didn't realize it was too transparant. It's my first attempt for subsurface
scattering. I've seen better results in other raytracers but I still have to
learn.. Unfortunately rendertime jumps up.. It took some hours to render
this image because of media.

> I think everything can be made in Pov-Ray.
> If anything, no. 5 can be a photo.

Well, you've done amazing things!
Thanks for your vote. Seems like a good bet.

Regards,
Hugo

Post a reply to this message

> Very very nice photorealistic work.
Thank you!

> N°5 (or 1 perhaps) is a photo ?
I'm glad you're not too sure.  ;o)

Regards,
Hugo

Post a reply to this message

Hello Jan and thanks for your comments!

> #2: naa... white material is a bit unreal, and the
> reflecting environment of the black sphere doesn't fit

I'm surprised a little you think the black sphere looks wrong, since it
merely reflects the enviroment. But the white material is much more "dull"
and might be causing the confusion, and the black sphere doesn't really look
metallic (well some people here thinks it's a pearl).

> #5: dunno ... can't find a detail, where I would be 100%
> sure, that it is a traced one. The shadows are very
> slightly blurred, so it would be a very small arealight,
> and, it must be at a huge distance, as the shadows are
> nearly parallel... This would be very hard in a studio,
> (and I see no other reflections), but who knows...

Who knows, but you're right: The lightsource is at a huge distance.
Interesting for me that you mention the parrellel shadows because I didn't
notice this.

> #6: The Ref-Map of the "thing" doesn't fit, looks somehow
> the same as in #2, and is missing in the floor.
> IMHO, If you have the objects mirroring in the floor,
> you also have the sky there ...

Hmm, seems like a good point. But the enviroment ought to give correct
reflections, as it's made the same way as picture #2... The ground is
reflective too, but maybe not enough..

The only difference between #6 and #2 is that I added a "vignette" effect -
like a camera lens - that makes the image brighter in the middle. That
should give a better impression of a photo, though it's not a real-life
effect.


Regards,
Hugo

Post a reply to this message

> You mean that they aren't all photos ;) ?

Hehe, thanks Ray, that's a positive comment.

Regards,
Hugo

Post a reply to this message

Hi Xplo!

Thank you for the kind and useful comments! I'm going to reviel where the
photo is now: It's NUMBER FIVE as
almost everyone suspected, but apparently it was not too obvious.. It shows
that many "small" things contribute to realism and POV-Ray is surely on the
right track.

> 1. Not a photo. Sphere material seems unrealistic.

You're right. The blurred reflection trick in POV3.5 has some limits. I
haven't been able to make a good bumpy looking normal PLUS blur.. So here I
tried to have a small granite pattern, with a large blur, by mixing
functions {}.. Rendertime was a BEAST! About 14 hours on my 1 ghz AMD, but
the major slowdown happened when reflection ping-pong's between the 2
spheres.. I reduced the max_trace_level to 3 which gave a good speed-up, but
still...

> 5. Could be a photo. Extremely convincing.

Ahh, yes.. I shot this photo today from my window.. A piece of paper and two
simple objects on top.. It seems like there is still a way to go in POV-Ray.

Regards,
Hugo

Post a reply to this message

Hi Steve and thanks for your comment!

> The dark sphere in number two is very close to a
> black pearl ....... needs some lustre and then bingo.

I'll think about the lustre (just looked up the word in my dictionary)..
hmm..

> As for the rest? I'll stick my neck out and say all
> .pov apart from #6.   Just hope I'm right...  ;)

eehhh.. The photo is number 5, but it's interesting that it was not too
obvious.. I conclude that although there is still headway to make, our
favourite raytracer is doing good.. Maybe it would be a good idea to make
some full setups for realistic light with a set of good textures.. Though
there is already a good include file for radiosity settings.

Regards,
Hugo

Post a reply to this message

Hi Bob,

> Regardless of what someone else said about #4's semi-transparency
> it draws my attention most as being realistic.

You're the only one who gives so much credit for #4 ... thanks! It's also
the only picture that features subsurface scattering and it's still not seen
so often in CG... There is still headway to make, however.. :o)

Your guess for #5 as the photo is correct!

Regards,
Hugo

Post a reply to this message

After reflection, N°5 is easy to dectet as a photo, because of the
compression artifacts of your numeric camera on the white cylinder. Can you
try to make the N°5 in CG to look for the differences ? It will be nice...

Mathieu (france)

Post a reply to this message

Attachments:
Download 'nouveau-2.jpg' (21 KB)

Preview of image 'nouveau-2.jpg'

> Can you try to make the N°5 in CG
> to look for the differences ? It will be nice...
>
> Mathieu (france)

I had the same idea today.. It's worth to try sometime this week.  ;o)

Regards,
Hugo

Post a reply to this message

I could write everything I know about lighting and texture on a match book
cover, but to my untrained eye, they all look pretty convincing except for
3. I think that the banding on the shadows in pic 4 gives away more than
does the material. 2 is the most realistic of the renders, but I think that
1 could surpass it with a slight focal blur to hide the regularity of the
patterns.

 -Shay

Hugo <hua### [at] post3teledk> wrote in message news:3caf13cc@news.povray.org...

Post a reply to this message

Hi Shay, thanks!

> I think that 1 could surpass it with a slight focal blur to
> hide the regularity of the patterns.

Huh, hmm.. Well all these renders actually have very similar focal blur, and
I don't see a regularity in the pattern of picture 1 ?

But it's interesting to note the big difference in render time for these
pictures.. The real beast was number 1 because of the blurred reflection
(ping pong between the spheres).. The rest renders very quickly, except
number 4 with subsurface scattering.. But this tells me I should probably
leave some things to the future, with faster computers, and concentrate on
radiosity now, since that gives great images, well unless you need
transparant or blurred materials up front.

Regards,
Hugo

Post a reply to this message

Hugo <hua### [at] post3teledk> wrote in message
news:3cb1fd0f$1@news.povray.org...
>
> Huh, hmm.. Well all these renders actually have very similar focal blur,
and
> I don't see a regularity in the pattern of picture 1 ?
>

Maybe it's just me, but the textures in #1 look mathematical. It might just
be the strength of the textures in #1 as compared to the others that gives
this appearance.

 -Shay

Post a reply to this message