 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3e9ed72c$1@news.povray.org>, tom### [at] tomandlu couk says...
> "Jide" <jid### [at] kotisoonfi> wrote in message
> news:3e9ecb22@news.povray.org...
> > Gilles Tran wrote:
>
> > > I did find the interface horrible, a case example of how apparently smart
> > > and innovative interface design can turn out to be just dumb and clumsy
> > > (I'll rant about this some day as it's quite common, particularly in
> > > free/low-cost 3D modellers).
> >
> > I too have tried a free uncrippled version of Amapi over a year ago.
> > I didn't like the interface much either but luckily there's always Wings3d
> > ;)
> >
>
> Any worse the TrueSpace? (no, surely that's not possible).
>
Umm. If I remember rightly, yes it is worse than TrueSpace, worse some
object types (like NURBS) couldn't be used with some tools or converted
into a form that could be worked with using those tools. This quirk
resulted in my jumping to several other apps before finally deciding that
I didn't know any of them well enough to do what I wanted. However, it is
better than some others, so I have Amapi on the launchmate bar I have
used for years, along with 3D Canvas (which gasp!! actually loads DXF
objects as individual and 'visually unique' meshes, instead of lumping
them all into one huge mesh like some do), Hamapatch, Moray (really need
to either get the non-demo or maybe buy Rhino instead..), Poser and an
old copy of Breeze Designer.
In general.. Amapi's problem is that they supply very nice tools and
object types, but ignore the traditional multiple view style interface
and manipulation methods that most usable programs have in favor of a
single view and really hard to use one. TrueSpace does much the same, but
it at least lets you easily change viewing angles, etc. Amapi is
'supposed' to, but when rotating things or attempting to adjust view it
can be sluggish, unpredictable or just plain refuses to do what you want
it to. A major pain in the rear to work with.
> I can't help but wonder whether the whole principle of using a mouse to build 3d
> objects is basically flawed. Poser, for example, is a lot easier to use IMHO
> when you give up trying to drag limbs about with the mouse* and use the dials
> instead.
>
Hadn't thought of the dials, but Poser bugs the heck out of me anyway.
Why exactly for instance don't even the standard models contain
constraint information to prevent you from turning an ankle around 180
degrees from normal? I would rather see someone eventually come up with
something with Poser like features and some way to 'easily' model and
define morph data, so you can create something more detailed by altering
the model in a useful way (I.e. changing general distance between the
eyes, actual shape of a nose, etc.) and not have to hack the files to add
morphs for what should be built into the dang thing.
The first time I tried using it my reaction was, 'Gee this is annoying to
use. I wonder how much a real program like cosmetic surgeons use would
cost and if -it- could export to a useful format.' Poser was quite
disappointing and far more annoying than any 'real' modeller I have ever
seen.
However, you may be right about mice. The main issue as I see it is that
way back in the days of Autocad someone thought, 'Heh lets add a third
button to make it easier to access some options.' So, how many 3 button
mice did you see? Now practically every mouse has at least three (even if
one is 'under' the scroll wheel), but you are lucky if so called
professional modellers recognize that button for anything. They also
ignore the mouse wheels, never mind the fact that using it with the mouse
movement would give you 3 axis. Hmm.. Somehow that sounds familiar, but I
can't imagine what 'use' it would have in the 3D app. lol You shouldn't
blame stupidity and bad design for why it is hard to build things in
these programs. ;)
But one should not despair, some people have gotten fed up with this
silliness and have started building more accurate means to do 3D. They
only work with applications that support them and currently probably cost
$5,000-$10,000 dollar a piece for something that is little more than a
mouse that they gutted the scrolls and buttons out of and rearranged so
they work better, but won't be affordable for most people for years if
ever. Realizing the ridiculous price tag on these sorts of things, most
people designing new programs are trying to reinvent the reflective
sphere by coming up with new, innovative and often impractical or just
plain dysfunctional interfaces. Such is the world of copyright and 3D
technology. ;)
> * well, unless your aim is to create some kind of genetically deformed alien
> undergoing a cruel and unusual torture.
>
Exactly, but then again... At least it 'might' look unique, but I doubt
it. ;) lol
--
void main () {
call functional_code()
else
call crash_windows();
}
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
3e9ed72c$1@news.povray.org...
> I can't help but wonder whether the whole principle of using a mouse to
build 3d
> objects is basically flawed.
Well, my reference in 3D modelling is Rhino, which I have always found to be
one of the most intuitive piece of software ever (among complex ones of
course).
Often, it doesn't take much actually, and it's not even about the 3D aspect
itself, but just respecting some common conventions - or at least some
logic - about selecting, dragging, copying, pasting and deleting things, or
using some recognisable icons for usual tasks etc. The worst offender was
the very first version of Blender: some users (including me) couldn't figure
out how to quit the application and the only solution was to crash it.
G.
--
**********************
http://www.oyonale.com
**********************
- Graphic experiments
- POV-Ray and Poser computer images
- Posters
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Tom Melly wrote:
> So... what will Amapi allow me to do that I couldn't do with the SDL? (or
> at least not realistically).
I don't know Amapi (just tried it a view month ago) but generally a modeler
will allow you to make organic and / or realistic looking objects. You
can't do that with POV-Rays SDL. Most images posted at p.b.i are nice, some
are quite impressive, but none are really realistic. In my opinion the SDL
is not very well suited to build a scene. Not worth mentioning that this is
a very unnatural way to create something. So if you just want to create
some kind of 'interesting' images, SDL is okay. If your goal is realism,
you should forget it.
You mentioned to code realistic figures. There are lot's of other things you
can't really do with SDL.
--
http://www.render-zone.com
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
in news:3e9efa16@news.povray.org Andreas Kreisig wrote:
> I don't know Amapi (just tried it a view month ago) but generally a
> modeler will allow you to make organic and / or realistic looking
> objects. You can't do that with POV-Rays SDL.
I have to disagree! Anything can be done in SDL, if you have the right
macros and a lot of time. On the other hand the same can be said of
doing something simple in SDL that turns out to be something complex
with a modeller. There actually may be a reason for the fact that more
and more GUI modellers also incorporate a scripting language.
> Most images posted at
> p.b.i are nice, some are quite impressive, but none are really
> realistic.
Please define realistic.
> In my opinion the SDL is not very well suited to build a
> scene. Not worth mentioning that this is a very unnatural way to
> create something.
Yet again I have to disagree. SDL can be seen as a modelling language
or as an intermediate format between a GUI and the raytracer. If you
export your Maya models to POV-Ray, however "realistic" they are,
you're still using SDL.
> So if you just want to create
> some kind of 'interesting' images, SDL is okay.If your goal is
> realism, you should forget it.
Again, define realisim.
Ingo
ps. I'm not advocating the pure use of SDL here, POV-Ray IMO is 'just' a
tool to get a result. To get that result I may, or may not, use other
tools. And realism is IMO not the holy grail of graphics, ever seen a
negative light source in reality. No? Yet in POV-Ray you can use them.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Tom Melly" <tom### [at] tomandlucouk> wrote in message
news:3e9e625f$1@news.povray.org...
> This month's PC Pro has a free full version of Amapi (I can't remember
which
> version - sorry).
>
> As far as I can tell, it is the complete program without restrictions, and
has a
> full range of import/export plug-ins.
>
> Anyone know much about Amapi?
Get it if you can. Interface is extraordinary, but it's a
powerfull modeler. It was in version 3, also, when I first tried it. Beside
NURBS, it support subdividion surfaces, probably the best way for organic
shapes today. As I can see from their site, version 7 can have a clasic
interface.
Well, as i can see, you need some extra tools for another half of work, uv
mapping, suggested tool is UV mapper - this isn't good, really.
Blender have some nice things about subdividion surfaces, and modeling at
all (automatic creation of surfaces between objects), but don't have
undo-redo functions, and have a weak export.
Metasequia ( http://www21.ocn.ne.jp/~mizno/ ) is one of the best tools for
polygons at all, subdivdion surfaces, UV-mapping, but the only educational
version is available outside of Japan, with some export restrictions, for
formats like a 3ds, LWO, etc, but you can export UV-mapped models to
DirectX, and toPOV-Ray through 3d win. Complete, customizable interface is
available in English, but docs are not - however, it's enough intuitive. I
think there are some plugins for uv-mapped POV export, today.
HamaPatch is a powerfull bezier patch modeler, but UV mapping works only
with direct POV bezier patch export, which isn't smooth as exported meshes
in POV.
Some most important modeling things, probably, are: easy adding or removing
elements from your model, support for hardware accelaration, symetrical
modeling, and UV mapping editor. Only Metasequoia can do all of these
things, between mentioned modelers.
Some of free or cheap modelers are equaly powerfull as a high-ends 3d's,
esspecialy in cases of organic modeling. Subdividion surface models can be
exported and animated almost everywhere.
For organic models, POV-Ray SDL is, in my opinion, just a good toy for
spending time.
Anto
http://www.matkovic.com/anto
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Andreas Kreisig wrote:
> Tom Melly wrote:
>
>
>> So... what will Amapi allow me to do that I couldn't do with the
>> SDL? (or at least not realistically).
>
>
> I don't know Amapi (just tried it a view month ago) but generally a
> modeler will allow you to make organic and / or realistic looking
> objects. You can't do that with POV-Rays SDL. Most images posted at
> p.b.i are nice, some are quite impressive, but none are really
> realistic. In my opinion the SDL is not very well suited to build a
> scene. Not worth mentioning that this is a very unnatural way to
> create something. So if you just want to create some kind of
> 'interesting' images, SDL is okay. If your goal is realism, you
> should forget it. You mentioned to code realistic figures. There are
> lot's of other things you can't really do with SDL.
I'm much more mathematically inclined than artistically, so I really
like positioning object by co-ordinates. I cannot understand graphing
applications (even 2d ones) that don't allow that. If it doesn't, how
can you make something as simple as a bullseye, where you need several
concentric circles with a common center?
--
Tom A.
"I've got the whole world
In my hands" - John, 3 year old Megalomaniac
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
ingo wrote:
> in news:3e9efa16@news.povray.org Andreas Kreisig wrote:
>
>> I don't know Amapi (just tried it a view month ago) but generally a
>> modeler will allow you to make organic and / or realistic looking
>> objects. You can't do that with POV-Rays SDL.
>
> I have to disagree! Anything can be done in SDL, if you have the right
> macros and a lot of time. On the other hand the same can be said of
> doing something simple in SDL that turns out to be something complex
> with a modeller. There actually may be a reason for the fact that more
> and more GUI modellers also incorporate a scripting language.
In fact you really need a lot of time. But most images made with SDL are
somewhat geometrical. Scripting languages in GUI modelers have been used to
add modeling features or small tools. No one will use it to model something
with it. Okay, sometimes SDL may be faster when the object you want to make
is based on mathematical algorithms.
>> Most images posted at
>> p.b.i are nice, some are quite impressive, but none are really
>> realistic.
>
> Please define realistic.
A view examples: with a lot of patience you can make something like a car,
but you're not able to make a Porsche 911 or a VW Golf or whatever. You can
make something like a bottle, but it's very hard to make a Jack Daniels
bottle. You can try to make a human head but that's nearly impossible with
SDL. Most SDL based images I saw in the past are rather simplyfied (anyway
they're nice but that's not what I want to point out).
>> In my opinion the SDL is not very well suited to build a
>> scene. Not worth mentioning that this is a very unnatural way to
>> create something.
>
> Yet again I have to disagree. SDL can be seen as a modelling language
> or as an intermediate format between a GUI and the raytracer. If you
> export your Maya models to POV-Ray, however "realistic" they are,
> you're still using SDL.
Yes, the informations will be stored into mesh2 format. From this point of
view SDL is an interface like you mentioned it. But it's not very usefull
when you want to use it as a modelling language (depending on what you want
to do with it, of course).
> ps. I'm not advocating the pure use of SDL here, POV-Ray IMO is 'just' a
> tool to get a result. To get that result I may, or may not, use other
> tools. And realism is IMO not the holy grail of graphics, ever seen a
> negative light source in reality. No? Yet in POV-Ray you can use them.
You're right regarding realism. On the other side POV-Ray is an outstanding
Raytracer (I really love it) wich is able to create photorealistic images,
but most of the images made with it are ... chrome balls on checkered
floors in all variations. Realism is not the holy grail but it's one of the
most ambitious chalanges in CG. POV-Ray can do that but you can't reach it
when you only use the SDL. You definitive need other tools.
Well, there're some people who can do fantastic things with SDL but they've
years of experience and often third party tools. I'm too unpatient for that
and bevor I waste my time to learn the SDL en detail to make some balls or
cones I rather use Blender. :)
Regards,
Andreas
--
http://www.render-zone.com
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
in news:3e9f3175@news.povray.org Andreas Kreisig wrote:
> In fact you really need a lot of time. But most images made with SDL
> are somewhat geometrical.
Have a look at the results you can achieve with TomTree or Stricia,
nothing geometrical there.
>> Please define realistic.
>
> [...] You can try to make a human head but
> that's nearly impossible with SDL. Most SDL based images I saw in the
> past are rather simplyfied (anyway they're nice but that's not what I
> want to point out).
How many hours did it take DAZ to create Vicky? Does she look realistic?
Ingo
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Thu, 17 Apr 2003 20:15:16 +0200, "Gilles Tran" <tra### [at] inapginrafr>
wrote:
>Well, my reference in 3D modelling is Rhino, which I have always found to be
>one of the most intuitive piece of software ever (among complex ones of
>course).
Ditto that. The curious thing is, Rhino started out as an AutoCAD
plug-in and this is evident everywhere. It has kept all of the AutoCAD
good points in terms of interface (command line, object snapping &
extending etc.) but has gone to a completely new level of useability
by introducing new and logical elements which AutoCAD has never had
and will probably never have (it's too darn conservative).
AutoCAD is still better for drawing and sketching, though, but of
course that's what it was intended to do in the first place. That's
why I always model in Rhino and export to AutoCAD for the engineering
drawing when I make some design (like my current loudspeaker-to-be).
Peter Popov ICQ : 15002700
Personal e-mail : pet### [at] vipbg
TAG e-mail : pet### [at] tagpovrayorg
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Fri, 18 Apr 2003 01:04:12 +0200, Andreas Kreisig <and### [at] gmxde> wrote:
> A view examples: with a lot of patience you can make something like a car,
> but you're not able to make a Porsche 911 or a VW Golf or whatever.
http://news.povray.org/povray.binaries.images/19709/127888/MyCar.jpg
http://news.povray.org/povray.binaries.images/30145/214068/recar3dc5.jpg
http://news.povray.org/povray.binaries.images/22626/156738/susp.jpg
http://news.povray.org/povray.binaries.images/30945/220060/FORDTRCK.jpg
http://news.povray.org/povray.binaries.images/31112/221106/WALL.jpg
> You can
> make something like a bottle, but it's very hard to make a Jack Daniels
> bottle.
Have you seen kitchen images from JRG ?
http://news.povray.org/povray.binaries.images/21949/148872/spezie.jpg
http://news.povray.org/povray.binaries.images/22917/159099/forno.jpg
http://news.povray.org/povray.binaries.images/29029/205449/vaporeEfuoco.jpg
http://news.povray.org/povray.binaries.images/21145/141194/t_b.jpg
http://news.povray.org/povray.binaries.images/19254/124775/waxscatt.jpg
Have you seen images from Jaime ?
http://news.povray.org/povray.binaries.images/27109/191537/balanza-6.jpg
http://news.povray.org/povray.binaries.images/21968/149037/bolts.jpg
http://news.povray.org/povray.binaries.images/20868/138290/office.jpg
> Most SDL based images I saw in the past are rather simplyfied (anyway
> they're nice but that's not what I want to point out).
Oh sure, I can make a scene like
http://10best.raytrace.com/short2/ks8_800.jpg
But I can also make something like
http://10best.raytrace.com/short2/ks9_800.jpg
http://10best.raytrace.com/short2/ks7_800.jpg
http://10best.raytrace.com/short/ks3_800.JPG
http://10best.raytrace.com/short/ks2_800.JPG
How many clicks will you use in modellers to duplicate these scenes? All those
were created with less than 500 characters.
As Ingo pointed out POV is just one of tools used to connect idea with result.
That's why it is so common to hear something like: "The body is done in Rhino.
The texture is done in POVray. The wheels are also done intirly in POVray." POV
can be used in scenes in many ways: as modeller, as raytracer, as texturing
tool, to create image_maps for other scenes. And as all in one.
> I'm too unpatient for that
> and bevor I waste my time to learn the SDL en detail to make some balls or
> cones I rather use Blender. :)
You have to understand that everything what modeller do is just some interactive
action to get input for algorithm performed over primitive objects and textures
to make some complexity which looks soft, organic etc. And as such this action
can be recreated in scripting language with necessary programing capabilities
and set of 3D functions to create primitives. The difference is that in modeller
everything is 'on-screen' in windows while in SDL everything is 'in-manual' and
'in-brain' ;-)
The most important thing which limits possibilities of POV-Ray output is user:
his knowledge, willingness, free time. But these are reality reasons and
therefore some users use modellers and some use pure SDL. And sometimes they
even cooperate exchenging content in some portable format like SDL or OBJ. And
some postprocess their work either as 3D data (in SDL or in modeller) or 2D
image output (in SDL or in image manipulation format).
ABX
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
|
|
