 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Ken wrote in message <369### [at] pacbell net>...
<snip>
> That concern is the grain structure that seems to be an inherent
>artifact of the process. While not as pronounced in the smoke of
>the above mentioned image, probably due to it's distance, it really
>shows up often in many other images. I wonder if this is going to
>be a fact of life with this feature or if there is a something that can
>be done to reduce the particle size while maintaining the density
>needed for realism.
<snip>
I've found increasing the intervals to around 30 smoothes the media quite
considerably, with little noticeable speed decrease (on my system anyway..).
Matt
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
About Atmospheric banding... I had problems with that even though I set
sample VERY high... then I realized that since I was running my Desktop
at 16bit color, I was viewing a 24bit image, undithered at
16bits...Changing my Desktop to 32bit cleared up the banding. Just a
comment (it doesn't fix the media grain artifact, but I must believe
that this graininess can be used somehow..)
Steve
Mike wrote:
>
> It's really just a function of the sampling. If you recall the
> particular look of atmosphere, it had a distinct banded look. Even
> though it was way off, it tended to look smoother because it was so in
> at least on direction. But that was just because it sampled at equal
> lengths along the viewing ray. Using jitter gave it a look similiar to
> media.
>
> The grainyness is objectionable though. Something I've been thinking
> about is that there must be some way to smooth media without taking
> extra samples. I figure it should involve comparing pixels and trying
> to balance them. I suppose a new keyword could let the user control
> this by allowing the user to give a maximum intensity difference that
> the pixels should be smoothed. The maximum for 24 bit images should be
> 1/255, but higher values would likely be specified. Perhaps the sampled
> pixels could be grouped into fours and cached, then they could be
> compared against each other. Then the intensities would be adjusted
> until neighboring pixels don't vary by more than the specified amount.
>
> I've heard that similiar techniques are used for stachastic renderers
> for the same reason. There is a certain loss of accuracy, but it
> doesn't require much extra time and results in more visually pleasing
> images. If you take that image and draw a freehand mask around it in a
> paint program, then apply a gaussian blur to the image, I bet the
> results would be very nice.
>
> comments/flames?
>
> -Mike
>
> Ken wrote:
>
> > I too will avoid comment on the specifics of this scene which is
> > note worthy in and of itself but one thing concerns me about this
> > image and other recent image postings using the new media feature.
> > That concern is the grain structure that seems to be an inherent
> > artifact of the process. While not as pronounced in the smoke of
> > the above mentioned image, probably due to it's distance, it really
> > shows up often in many other images. I wonder if this is going to
> > be a fact of life with this feature or if there is a something that can
> > be done to reduce the particle size while maintaining the density
> > needed for realism.
> > Smoke paticles are very fine particulate matter and the reason it
> > appears to us as it does is primarily due to shear volume. Might this
> > grain artifact be a problem of image resolutions and screen capabilities
> > or am I just jumping the gun on this still evolving feature ?
> >
> > I'm willing to entertain discussion on this but truthfully have as yet
> > not spent much time with it and am by no means a barometer of
> > it's potential.
> >
> > --
> > Ken Tyler
> >
> > tyl### [at] pacbellnet
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Ken wrote:
> Your system date is off a day.
Pardon my French but MotherF*^$er!!! I've had several people tell me it
was off and after several missed attempts I'm sure I got it right now.
I can't for the life of me understand why my message was listed as being
posted on Sunday. I'm looking at my system date and time now and it
says Monday, January 4th, 1999, 9:14...crap, it says AM! ARGGGGGGGGH!!!!
Um, I got it fixed now. Blood pressure dropping...sanity returning...
> I was working with fog in a recent image after trying unsuccessfully
> to get atmospheric media to work the way I wanted and thought
> to myself on more than one occassion how now it would be to
> be able to constrain the fog feature into a specific volume of space
> like you can media. The particle size used in the fog function is
> very nicely uniform in size and the density is controllable through
> filtering and transmission values. A few enhancements and more
> user control over that feature would make for great smoke,
> clouds, and various other similar items that have no real substance.
> The feasibibility of such an undertaking I no not but I can dream.
Actually, you can put fog in an object. It fun to do too cause it's
fast. The trick is to make an object and put some fog in it, then use
inverse. Now the inside is outside and the outside is inside. :)
-Mike
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Stephen Lavedas wrote:
>
> About Atmospheric banding... I had problems with that even though I set
> sample VERY high... then I realized that since I was running my Desktop
> at 16bit color, I was viewing a 24bit image, undithered at
> 16bits...Changing my Desktop to 32bit cleared up the banding. Just a
> comment (it doesn't fix the media grain artifact, but I must believe
> that this graininess can be used somehow..)
The problem with the image in question (FSaP) is clearly not color
discritization....
The problem is the simple Monte-Carlo technique used by POV. Each pixel
has the integral calculated independently from its neighboring
pixels. This is wasting information. More sophisticated Monte-Carlo
techniques can be used, with introduction of potentially considerable
complexity, which make use of the generally smoothly-varying nature
of the media fields.
The problem reminds me of the one in semiconductor device processing,
where Monte-Carlo is being used to predict the distribution of ions
which result from the bombardment of semiconductor device surfaces
with charged dopants. Some excellent results were demonstrated
at the latest International Electron Device Conference in San Francisco
of the use of some clever but relatively simple techniques to get
more out of each randomly sampled ion event. For example, one can
do "particle splitting" in which more than one particle shares part of a path,
but then part way through is split into multiple particles to generate
different random paths. But I digress....
The improvement of the interior methods would make an excellent
subject for a patch.
Dan
--
http://www.flash.net/~djconnel/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hey guys, I kind of liked the image :) at least until reading all these
techno-notes about it.
A great picture is bound to get more scrutiny than others though.
And I have a thought too. Looks like quite a bit of turbulence used,
depending on how it was applied, that can cause some level of graininess
too. The artist might have even preferred the look of it as is anyhow.
And if emission or absorption is used the crandy nature increases
drastically over pure scatter and density alone, from what I've seen. Be
willing to bet absortion, at least, was used.
The flaming parts were noticably ephemeral looking to me and caught my
attention more than the sooty smoke did. A tighter, torrent of flames
would have looked nice.
Want to say, btw, loved the pictures detail and "overall" look.
Dan Connelly wrote:
>
> Stephen Lavedas wrote:
> >
> > About Atmospheric banding... I had problems with that even though I set
> > sample VERY high... then I realized that since I was running my Desktop
> > at 16bit color, I was viewing a 24bit image, undithered at
> > 16bits...Changing my Desktop to 32bit cleared up the banding. Just a
> > comment (it doesn't fix the media grain artifact, but I must believe
> > that this graininess can be used somehow..)
>
> The problem with the image in question (FSaP) is clearly not color
> discritization....
>
> The problem is the simple Monte-Carlo technique used by POV. Each pixel
> has the integral calculated independently from its neighboring
> pixels. This is wasting information. More sophisticated Monte-Carlo
> techniques can be used, with introduction of potentially considerable
> complexity, which make use of the generally smoothly-varying nature
> of the media fields.
>
> The problem reminds me of the one in semiconductor device processing,
> where Monte-Carlo is being used to predict the distribution of ions
> which result from the bombardment of semiconductor device surfaces
> with charged dopants. Some excellent results were demonstrated
> at the latest International Electron Device Conference in San Francisco
> of the use of some clever but relatively simple techniques to get
> more out of each randomly sampled ion event. For example, one can
> do "particle splitting" in which more than one particle shares part of a path,
> but then part way through is split into multiple particles to generate
> different random paths. But I digress....
>
> The improvement of the interior methods would make an excellent
> subject for a patch.
>
> Dan
>
> --
> http://www.flash.net/~djconnel/
--
omniVERSE: beyond the universe
http://members.aol.com/inversez/POVring.htm
=Bob
Post a reply to this message
Attachments:
Download 'us-ascii' (1 KB)
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Bob Hughes wrote:
>
> Hey guys, I kind of liked the image :) at least until reading all these
> techno-notes about it.
>
Actually I believe everybody has done a great job of sticking
to the subject of the media grain problems and have avoided the
critiqueing of the image quite ethicaly. That is until ...
--
Ken Tyler
tyl### [at] pacbellnet
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Dan Connelly wrote:
> The problem reminds me of the one in semiconductor device processing,
> where Monte-Carlo is being used to predict the distribution of ions
> which result from the bombardment of semiconductor device surfaces
> with charged dopants. Some excellent results were demonstrated
> at the latest International Electron Device Conference in San Francisco
> of the use of some clever but relatively simple techniques to get
> more out of each randomly sampled ion event. For example, one can
> do "particle splitting" in which more than one particle shares part of a path,
> but then part way through is split into multiple particles to generate
> different random paths. But I digress....
I regress ...
What is the benefit of the above process. It sounds like
they are trying to increase electron flow while reducing
resistance at the junction. Mosfet applications ?
--
Ken Tyler
tyl### [at] pacbellnet
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Ken wrote:
>
> Dan Connelly wrote:
>
> > The problem reminds me of the one in semiconductor device processing,
> > where Monte-Carlo is being used to predict the distribution of ions
> > which result from the bombardment of semiconductor device surfaces
> > with charged dopants. Some excellent results were demonstrated
> > at the latest International Electron Device Conference in San Francisco
> > of the use of some clever but relatively simple techniques to get
> > more out of each randomly sampled ion event. For example, one can
> > do "particle splitting" in which more than one particle shares part of a path,
> > but then part way through is split into multiple particles to generate
> > different random paths. But I digress....
>
> I regress ...
>
> What is the benefit of the above process. It sounds like
> they are trying to increase electron flow while reducing
> resistance at the junction. Mosfet applications ?
Which process? The physical process? If so, yes -- the primary application
is the formation of MOSFETs. For example, As+ implanted into B-doped Si
results in "islands" of electron-rich (electrons coming from the As, which has
valence of 5, and thus has an extra electron to give to the lattice) Si connected
with electron-deficient regions (the B has valence of 3, and thus traps electrons).
An electrode above an insulating plate connecting the electron-rich regions
is used to control a sheet of electrons at the surface which, if formed,
closes the switch. If not, the electron-rich islands are isolated, and
the switch is "open".
The simulation process is used to predict the extent of the electron-rich
islands -- the goal is to keep them as dense and compact as possible... this
allows them to be placed in close proximity (separated by approximately
0.1 micrometers in cutting-edge technology), increasing the speed at which
the connection can be switched.
Dan
P.S. There are also devices with electron-deficient islands formed in an
electron-rich region.... in this case the conduction is via "holes" rather
than "electrons". Both devices are used in conjunction to form "CMOS".
--
http://www.flash.net/~djconnel/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Dan Connelly wrote:
> P.S. There are also devices with electron-deficient islands formed in an
> electron-rich region.... in this case the conduction is via "holes" rather
> than "electrons". Both devices are used in conjunction to form "CMOS".
Are these what they call depletion mode devices ?
P.S. Sorry for the off topic discussion folks and I promise to let it
go after this. It's just that I love this stuff.
--
Ken Tyler
tyl### [at] pacbellnet
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Ken wrote:
>
> Dan Connelly wrote:
>
> > P.S. There are also devices with electron-deficient islands formed in an
> > electron-rich region.... in this case the conduction is via "holes" rather
> > than "electrons". Both devices are used in conjunction to form "CMOS".
>
> Are these what they call depletion mode devices ?
>
> P.S. Sorry for the off topic discussion folks and I promise to let it
> go after this. It's just that I love this stuff.
>
Not necessarily -- they are PMOS ( as opposed to NMOS ).
Basically when one speaks of "electrons" in semiconductors one is
actually speaking of eigenstates of the macroscopic state of the
electron gas.... due to diffraction effects from the lattice,
quanta of electron charge manifest themselves as "quasiparticles" which
can have different energies at a given momentum. Some of these states
are effectively all occupied, others effectively so high in energy they
are never occupied. Two sets of states straddle the "Fermi level", which
is sort of a threshold energy required to achieve net charge neutrality
between electrons and protons. Those states above the Fermi level by a
few fractions of an electron volt are electron-like -- they tend to be light
and fast. Those just below the Fermi level are mostly occupied.... the states
which are NOT occupied act as positively charged quasiparticles called "holes".
It is the holes which act as the current carriers in PMOS devices.
They tend to be "heavier" and thus PMOS devices are slower than NMOS.
Basically, NMOS devices are used as pull-down switches (shorting things to ground)
while PMOS are pull-up devices (shorting things to the supply voltage). Since
switches are faster when they support a larger voltage drop, this guarantees at
least one of the two switches supporting a circuit node will have a large voltage drop
across it at any given time, allowing transitions to be executed quickly.
Anyway, enough of that.... time to go to work :).
Dan
--
http://www.flash.net/~djconnel/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
