|
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I say "sorta final" because, in actuality, there's quite a bit more that
could be done with this scene, both in terms of render quality and
additional modeling/texturing, but the pic here took over 42 hours to
render as it is, and until I get a much better computer, this is as much
work as I'm willing to put into something that takes far too long even
to test.
Comments welcome, of course.. for all the good it'll do.
-Xplo
Post a reply to this message
Attachments:
Download 'room.jpg' (40 KB)
Preview of image 'room.jpg'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Xplo Eristotle wrote:
>
> I say "sorta final" because, in actuality, there's quite a bit more that
> could be done with this scene, both in terms of render quality and
> additional modeling/texturing, but the pic here took over 42 hours to
> render as it is, and until I get a much better computer, this is as much
> work as I'm willing to put into something that takes far too long even
> to test.
>
> Comments welcome, of course.. for all the good it'll do.
>
> -Xplo
>
42 hours for without aa ? What's taking that long apart from the tree ?
Btw, it looks very dark on my screen.
Christoph
--
Christoph Hormann <chr### [at] gmxde>
Homepage: http://www.schunter.etc.tu-bs.de/~chris/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Christoph Hormann wrote:
>
> 42 hours for without aa ? What's taking that long apart from the tree ?
The tree was having some bounding problems, and manual bounding didn't
seem to having any effect (literally, there was no change). So the
middle third of the picture was really slow.
The table directly underneath it was even worse, because (I think) of
the reflection blur.
> Btw, it looks very dark on my screen.
I was going for something like realistic lighting, and with all three of
the lights on (you can't see them because they're overhead, but there's
three of them), making the lighting any brighter washed out the color of
the wooden walls. I suppose I could have increased the light intensity
to simulate pupil dilation (or longer exposure, whatever), but you know,
even with those measures, a real-life dimly-lit room still looks dimly-lit.
I wonder what "photographic" post-processing, as was discussed on here
some weeks back, would do to the image.
Try looking at it on a black background, and with the lights turned down
to match the illumination in the picture.. the difference is really dramatic.
-Xplo
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
For those of you who are on PCs and find the picture really dark, it was
created on a Mac with gamma 1.8, and most of the tones are in a range
where gamma makes a very noticible difference. I could post a gamma 2.2
version, if I can figure out how to do proper gamma correction in
Photoshop (I'm *certainly* not going to re-render at 2.2!).
-Xplo
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <39F92956.4D7B17BF@unforgettable.com>,
inq### [at] unforgettablecom wrote:
> Christoph Hormann wrote:
> >
> > 42 hours for without aa ? What's taking that long apart from the tree ?
>
> The tree was having some bounding problems, and manual bounding didn't
> seem to having any effect (literally, there was no change). So the
> middle third of the picture was really slow.
Official version? Make sure you have "Ignore any "bounded_by" in scene"
unchecked. It is in the "Render Settings..." dialog, under the
"Optimization" pane.
--
Christopher James Huff
Personal: chr### [at] maccom, http://homepage.mac.com/chrishuff/
TAG: chr### [at] tagpovrayorg, http://tag.povray.org/
<><
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Xplo Eristotle wrote:
>
> The tree was having some bounding problems, and manual bounding didn't
> seem to having any effect (literally, there was no change). So the
> middle third of the picture was really slow.
>
> The table directly underneath it was even worse, because (I think) of
> the reflection blur.
In such cases i would do test renders without the tree so i could adjust the
lighting situation and only add it for the final render
> I was going for something like realistic lighting, and with all three of
> the lights on (you can't see them because they're overhead, but there's
> three of them), making the lighting any brighter washed out the color of
> the wooden walls. I suppose I could have increased the light intensity
> to simulate pupil dilation (or longer exposure, whatever), but you know,
> even with those measures, a real-life dimly-lit room still looks dimly-lit.
I don't know the effect in this special case apart from making things even
slower, but it's often worth trying radiosity in indoor scenes ...
> I wonder what "photographic" post-processing, as was discussed on here
> some weeks back, would do to the image.
I had a look at the picture's histogram and saw that it only uses values below
103. When doing color correction after the render, there are always losses
because of the limited accuracy (8 bit per channel), especially with further
reduced color space like here, and 16 bit pictures are only supported by very
few programs right now. (That matter was discussed before as you mentioned)
> Try looking at it on a black background, and with the lights turned down
> to match the illumination in the picture.. the difference is really dramatic.
I also tried Photoshop's Auto-Correction (i don't know the exact name in
english) and the result was quite interesting.
Christoph
--
Christoph Hormann <chr### [at] gmxde>
Homepage: http://www.schunter.etc.tu-bs.de/~chris/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Christoph Hormann wrote:
>
> Xplo Eristotle wrote:
> >
> > The tree was having some bounding problems, and manual bounding didn't
> > seem to having any effect (literally, there was no change). So the
> > middle third of the picture was really slow.
> >
> > The table directly underneath it was even worse, because (I think) of
> > the reflection blur.
>
> In such cases i would do test renders without the tree so i could adjust the
> lighting situation and only add it for the final render
Which is, in fact, what I did.
> I don't know the effect in this special case apart from making things even
> slower, but it's often worth trying radiosity in indoor scenes ...
Um, it DOES have radiosity. That's not ambient picking up the shadows.
-Xplo
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Chris Huff wrote:
>
> In article <39F92956.4D7B17BF@unforgettable.com>,
> inq### [at] unforgettablecom wrote:
>
> > The tree was having some bounding problems, and manual bounding didn't
> > seem to having any effect (literally, there was no change). So the
> > middle third of the picture was really slow.
>
> Official version? Make sure you have "Ignore any "bounded_by" in scene"
> unchecked. It is in the "Render Settings..." dialog, under the
> "Optimization" pane.
MacMegaPOV, actually.. but that same setting exists, and unchecking it
didn't help.
I'm wondering if using clipped_by instead of making it a difference
would eliminate the problem. I think that leaving a hole in the tree
might be acceptable, since it gets covered over by the face anyway.
-Xplo
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Three lights like this would make the room painfully bright, but for
those of you who find my original image too dark, here's a much brighter
version, courtesy of Photoshop.
-Xplo
Post a reply to this message
Attachments:
Download 'room2.jpg' (71 KB)
Preview of image 'room2.jpg'
|
|
| |
| |
|
|
From: Ross Litscher
Subject: Re: Japanese room - much lighter version
Date: 27 Oct 2000 17:03:05
Message: <39f9ed89@news.povray.org>
|
|
|
| |
| |
|
|
Holy non-antialiased batman!
now though i can see the tree better, and the blurred reflection wasn't
visible at all before.
muchbetter.
ross
Xplo Eristotle <inq### [at] unforgettablecom> wrote in message
news:39F9E677.276481C4@unforgettable.com...
> Three lights like this would make the room painfully bright, but for
> those of you who find my original image too dark, here's a much brighter
> version, courtesy of Photoshop.
>
> -Xplo
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |