Help please
Apache
in) in enough to work on a group of points what does one do? I have tried
cutting the piece to a seperate object and resizing it - No go. I tried all
the display options I thought might help. What am I missing - never ran
across this before.
TIA
Tesselator
"Apache" <Wh...@AmI.com> wrote in message news:Xns942D390AF6...@216.168.3.44...
Press [d], click on "Fixed Near Clip Distance", enter something small
like 1mm or 0.001mm or something.
--- or ---
Use the "[" and "]" keys to adjust the grid size on the fly.
--
I hope that helps.
Mark Dunakin
I sometimes (a lot of times, or most of the time) tend to forget all
about the Fixed Near Clip Distance.
I generally do the lazy method, with the [ and ] keys :)
Tesselator wrote:
--
MD arts
Mark Dunakin
m...@md-arts.com
http://www.md-arts.com
"Push the button, Frank!"
-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 100,000 Newsgroups - 19 Different Servers! =-----
Apache
news:bojg17$om2$1...@catv02.starcat.ne.jp:
The problem is with the object in modeler not in layout. Sorry I guess I
didn't state it very well. There is no near clip dist in modeler as far
as I can tell. Changing the grid doesn't work.
Anyway, When in modeler, I have hit the zoom limits. Even cutting and
pasteing the part on a seperate layer or in a seperate file doesn't fix
it. I can not get close enough to distingush seperate points. I don't
understand this limit. I mean we are not talking of a real object in the
real world. The other package I use has no zoom limits (Rhino). To
understand the size differences it's an inside corner of a curb on a
large shopping mall's sidewalk. So the area I need to fix is very small
compared to the model.
Jim Woodruff
two points that are .0005 mil (1.27 um) apart. You're building that sort of
resloution into a model the size of a shopping mall? How many points / polys
is this sucker going to have in it? Millions, maybe billions?
Jim
"Apache" <Wh...@AmI.com> wrote in message
news:Xns942DD1CF61...@216.168.3.44...
MkIII
to, resize it when finished.
LW handles big things bettter than small things (this has been a problem for
me also - since LW 6.0 stopped you zooming in so far)
"Apache" <Wh...@AmI.com> wrote in message
news:Xns942DD1CF61...@216.168.3.44...
---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.535 / Virus Database: 330 - Release Date: 01/11/2003
Apache
news:vqr9qol...@corp.supernews.com:
> Well, I just checked and at the max zoom in modeler I can see space
> between two points that are .0005 mil (1.27 um) apart. You're building
> that sort of resloution into a model the size of a shopping mall? How
> many points / polys is this sucker going to have in it? Millions,
> maybe billions?
>
> Jim
>
Not many points involved Jim. The curbs where they make a corner are
rounded The rounded curbs are to spec and have about 60 points. The
entire pad the buildings sit on has 725 points for side walks and what
not. The curb height is proportional to the buildings. I modeled a
indented space between the sidewalks and the curb. I wanted to narrow the
gap a bit by one of the corners and couldn't get close enough. It
surprised me. Well I have closed the gap and used a bump map to make the
gap. Since the camera will be moving at head height it should look ok.
What I want to know is why is their a zoom limit at all? Sounds like if
someone modeled something like an aircraft carrier they would never be
able to zoom in close to get just the bridge detailed. They would have to
make a model of the bridge seperately, I guess.
I was doing this from a autocad model.
Apache
news:3fadbe10$0$5411$cc9e...@news.dial.pipex.com:
> Make your entire model 10 times the size you need, then if you really
> need to, resize it when finished.
>
> LW handles big things bettter than small things (this has been a
> problem for me also - since LW 6.0 stopped you zooming in so far)
>
This is a small detail on a big model - Shopping mall.
I will try it on the model, but for now I just remodeled the part and
used a bump map to make the gap I was trying to make. If it looks funky I
will resize it.
Thanks,
BTW where you posting pixs of those great models you make?
Tesselator
Well I've read all the messages in this thread now
and I still have no clue what you're talking about.
I guess you've found a work around for whatever it is
but now you've got my curiosity up. Could you or
explain it differently or maybe put a screenshot
or two up if that's easier. You can put them here:
ftp://61.213.194.19/upload/Apache and link to it
there if you need space.
If it's a pain and you'ld rather just move on I'll
understand tho.
O :-)
"Apache" <Wh...@AmI.com> wrote in message news:Xns942EB2FF8...@216.168.3.44...
Jim Woodruff
between points in each viewport of around .0005 mils. That should be plenty
of resolution for what you are trying to do, no matter how "big" your
overall model is.
Okay... I just did a test and discovered what may be causing the issue
you're seeing. LW can only handle a certain degree of "resolution". This
resolution appears to be tied to how big the current "grid" cell dimension
you are viewing or working in and, more importantly, where this cell is
centered in Modeler space. If you have a very small object drawn within a
much larger object, but the small object is near the outer edges of the
larger object then you will have a resolution issue when trying to zoom-in
to the object. If you move the entire object (or at least the smaller one)
to the center of the Modeler editing "universe" then you can zoom all the
way the previously mentioned, very small editing cell resolution.
As an example: I just made a box that is 120,000,000 meters square centered
in the modeler editing space (probably a little larger than your shopping
mall ;-). If I attempt to zoom to the extreme corner of this huge box, the
largest grid cell I could zoom to was 10km. But, if I zoomed directly to the
center of the modeler editing space (x-0,y-0,z-0) then I could resolve a
grid cell dimension of .1um. And as I move farther out towards the edge of
the object the cell resolution accordingly drops off.
So the answer for your dilemma appears to be: temporarily move your overall
object so that the points you want to edit are in the very center of modeler
edit grid. You should be able to zoom-in with plenty of resolution to make
your edits, then reposition the model to its original location.
I hope this does the trick :-)
Jim
"Apache" <Wh...@AmI.com> wrote in message
news:Xns942EA37AF...@216.168.3.44...
Alessandro D. Petaccia
wrote:
>Hi Apache,
>
>Well I've read all the messages in this thread now
>and I still have no clue what you're talking about.
I think I have, as that happens to me quite often. Apache, if you're
working on a CAD-native model, there's a good chance the object will
be geo-referenced, thus when loaded in Modeler it'll be from meters to
kilometers away from axis origin: that will BADLY screw up Modeler's
zoom.
Try a rest-on-ground and see what happens...
ADP.
Apache
Your absolutely right!!! That did work. This is an important tip that
should be in those pdf files. I am so tired of 'undocumented features'
in every piece of software on the planet. QA is a term not in use much
today - to bad.
So where do I send the party girls and cash?
Apache
news:bokqnq$9tu$1...@catv02.starcat.ne.jp:
Mr. Jim ( said while lighting a candle and bowing toward the west)
Woodruff just figured it out and it's an important thing to know. Here
are the facts.
if you have a large model and want to work on a small group of points at
the outer edge of the model and find you can not zoom in close enough to
see the seperate points - move the entire model so that the area you need
to work on is at the center (0,0 doesn't have to be exact just somewhere
near the center.
It seems the amount one can zoom in depends on how far one is from the
center of the current grid.
Jim Woodruff
I've been using LW since version 1 (included with the original Amiga Video
Toaster) and I wasn't aware of this behavior in Modeler till this morning,
so we both learned something ;-)
Thanks for the offer, but I'd just blow the cash on more party girls. Then
I'd be broke, tired AND lonely ;-)
Jim
"Apache" <Wh...@AmI.com> wrote in message
news:Xns942E8B9D3B...@216.168.3.44...
Apache
news:780tqvovm4a97hc8c...@4ax.com:
Your right - I read Jim's response first so I'm out of party girls. I moved
the model and zoomed in until I could see the points I needed to work on.
Thanks.
Ernie Wright
> I've been using LW since version 1 (included with the original Amiga
> Video Toaster) and I wasn't aware of this behavior in Modeler till
> this morning, so we both learned something ;-)
Modeler started doing this in 6.0. It's an attempt to prevent the user
from exceeding the computer's limited floating-point precision. Prior
to the change, you could zoom in as far as you wanted, but Modeler would
start acting funny.
Someone here rediscovers these limits every couple of months. I've been
answering questions about them since 1995.
http://www.google.com/groups?selm=Pine.SUN.3.91.950830014510.20448A-100000%40access4.digex.net
http://www.google.com/groups?selm=Pine.SUN.3.91.950831025602.15324A-100000%40access4.digex.net
http://www.google.com/groups?selm=3B754A44.F689AFC0%40home.com
http://www.google.com/groups?selm=3B72F31B.A23BE6BD%40home.com
http://www.google.com/groups?selm=3ADC9DC6.90FB082%40home.com
http://www.google.com/groups?selm=3A4E647F.7B3FE755%40home.com
http://www.google.com/groups?selm=39DC1435.B21EDEF2%40home.com
http://www.google.com/groups?selm=391E31A8.3A3FE88F%40home.com
http://www.google.com/groups?selm=3704573E.AD3DAB36%40home.com
And when it's explained, people always say it should be in the manual.
I've also discussed that here several times. The jist of my position on
that is that if it were in the manual, nobody would find it. It's clear
people don't find it by Googling this group, or by reading it regularly,
even though it's been discussed here ad nauseum.
Jim Woodruff
Thanks for the detailed info, but keep in mind, I'm not the guy that
originated the question, I was addressing someone else's dilemma.
Like I mentioned, I have using Modeler for years and I have never run across
this "behavior", so it isn't an issue for me at all. I limit my modeling to
reasonably sized objects (about 100 feet, with most being a few feet) and I
have always modeled "to scale", so Modelers grid resolution has always been
more than enough for me.
One thing you did say was of interest to me though, "It's an attempt to
prevent the user from exceeding the computers limited floating-point
precision." Did you mean "exceeding the software's limited floating-point
precession" or the computers? I would think that the floating point
precession of a P4 (64-bit or 128-bit depending the application) is more
than enough.
Jim
"Ernie Wright" <ern...@comcast.net> wrote in message
news:FfqdnRVnK4p...@comcast.com...
Apache
Well Ernie that's the funniest suggestion I've heard in a while. Screw
the manual - search google. I was a test pilot in the military for many
years and I'm really glad your didn't write the manuals for the aircraft
I flew. It would seem to me perhaps when a Lightwave question is
answered here "ad nauseum" that Newtek should take a hint and think about
putting it in a manual. Apparently this has happened since 6.0 - seems
like Newtek has had plenty of time. I guess there might be some folks
that like to pay hundreds of dollars for software and play 'Where's
Waldo' with the operating instructions - I'm not one of them.
I don't read every line of every article in every newsgroup which might
contain a line of the Lightwave manual. But I did search 'earnie wright'
on Google and found several copies. So how much does a Pugmaster cost?
Ernie Wright
> Thanks for the detailed info, but keep in mind, I'm not the guy that
> originated the question, I was addressing someone else's dilemma.
I know. Your post was just a convenient place to hang my reply. Sorry
about that, it wasn't addressed to any single person.
> One thing you did say was of interest to me though, "It's an attempt
> to prevent the user from exceeding the computer's limited floating-
> point precision." Did you mean "exceeding the software's limited
> floating-point precession" or the computers?
The computer's!
> I would think that the floating point precession of a P4 (64-bit or
> 128-bit depending the application) is more than enough.
The choices are 32-bit or 64-bit (the same as they've been for every x86
FPU). LW, and many other programs, and all OpenGL hardware, use 32-bit,
which gives you 1 part in 8 million, not enough to tweak the bevel on
the curb of a shopping center.
There's no easy way around this, either. You can move the shopping
center in Modeler so that the curb is at the origin, and you can tweak
at atomic scales there, but when you move the shopping center back, your
tweaks are lost. Point coordinates 1 km from the origin are quantized
to 1/8 mm.
It doesn't do any good to make the shopping center larger or smaller,
either.
LW could use 64-bit, but that doesn't help you with the interface, since
that's still 32-bit OpenGL. 64-bit also doubles the memory cost for
every scene and object file.
Ernie Wright
> Well Ernie that's the funniest suggestion I've heard in a while. Screw
> the manual - search google.
I'm sorry if I embarrassed you by pointing out that this question is
frequently asked and answered. That wasn't my intention. I think it
*should* be in the manual. I'm just not optimistic that it would reduce
the number of times the question is asked, since it's hard for people to
connect the symptoms with the explanation. Numerous explanations have
been available in the archives of LW discussion forums for at least 8
years, and it hasn't seemed to help much.
It's a hardware limitation. It's not specific to LW. It affects every
piece of software on your computer that uses floating-point numbers.
Very few software manuals mention it.
Mark Dunakin
Ernie Wright wrote:
> Apache wrote:
>
>> Well Ernie that's the funniest suggestion I've heard in a while.
>> Screw the manual - search google.
>
>
> I'm sorry if I embarrassed you by pointing out that this question is
> frequently asked and answered. That wasn't my intention. I think it
> *should* be in the manual. I'm just not optimistic that it would reduce
> the number of times the question is asked, since it's hard for people to
> connect the symptoms with the explanation. Numerous explanations have
> been available in the archives of LW discussion forums for at least 8
> years, and it hasn't seemed to help much.
Yep. I think that this is the actual problem with people asking about
the same things over and over. It's because they don't know what or why
something is happening, so they can't vary well do any searches on
something they've never heard of or don't know what it's called. I've
never personally had any of these problems, as I'm like the other person
who mentioned that they never build models much bigger/smaller then a
certain size. But I too dought I would of known what would cause this
problem, much less what it might be called so that I could do a search
for it in the manual or otherwise.
................md :)
Apache
Ernie, I quit being embarrassed when I passed 50. You have to admit as
a computer code jockey that the weakest link in most software (other than
the user) is the documentation. It is the drudge work that few like to
do. I did a stint as a tech writer and it had to be the most boring work
on the planet. But it has to be done. The first place I went when the
problem happened was the 7.0 manual(pdf). Nothing there at all. As a guy
who thinks a floating point is something found in a poorly decorked wine
bottle, it would be nice to know a bit more on all the lightwave
functions. Unfortunately my blackboard is getting fairly full and I rely
on manuals to tweak the neurons now and then. I appriciate your effort to
inform the uninformed and the detail you go into with the explaination.
Thanks
Ernie Wright
> Ernie, I quit being embarrassed when I passed 50.
My kids immunized me.
> You have to admit as a computer code jockey that the weakest link in
> most software (other than the user) is the documentation.
I wrote the LW plug-in SDK documentation, so I know how hard it is to do
well. But documentation is the last mile. It doesn't matter how good
your technology is if users can't get it.
> I appriciate your effort to inform the uninformed and the detail you
> go into with the explaination.
Thanks. I really haven't figured out how to explain this concisely so
that the light bulb just goes on, but I keep trying.
Mark Dunakin
Ernie Wright wrote:
> Thanks. I really haven't figured out how to explain this concisely so
> that the light bulb just goes on, but I keep trying.
And we always appreciate it too :)
Thanx.....md :)
Ernie Wright
> And we always appreciate it too :)
> Thanx.....md :)
Thanks Mark.
CWCunningham
news:IdqdnXa7Ac0...@comcast.com...
> I really haven't figured out how to explain this concisely so
> that the light bulb just goes on, but I keep trying.
>
finally flips the switch. I tend to read manuals cover to cover, whether
they make sense or not, just because someday... I'll be doing something and
think, "I've read about this, halfway through *that* bookl, on one of the
left hand pages." Sure enough, I'll pop open the manual and flip through
looking on the left and boom, there it is, information. May not have meant
much the first time I saw it, but I know right where to look when the light
starts flickering.
Yes, thanks Ernie for your thoughtful contributions.
--
CWC
============================
Sometimes it's on a right hand page.
- Information found on a left hand page -
============================