What would you like to see in AM

[ludo_si]

Here are a few idea that I would like to see in AM. It's fine to talk about the evolution of his favorite software

 

modelisation:

Bevel

modeling on a polygon mesh and baking

 

UV:

real unwrap tools

 

Shader: volumétric shader as fire

 

Particle: objets (props) as particle

 

Modificator: explosion

 

texture: a texture to mix different shader

 

Hair: a cut option. wave and curl too.

 

rendering. CPU/ GPU enable

[Tore]

* Layers

* Displacement map sculpting

* Different options for bone's look (stick etc)

* Shadow only mesh (for compositing)

[ludo_si]

Layers, it's a fantastic idee too

 

UI: more easy to get 4 view and one view with a buton

[detbear]

Here are a few of my many wish list items:

 

1. FBX export with working rig in other(Wouldn't that be cool )

2. Camera Translate/ rotate Export for use in other Apps. The Vue export only works in Hash 14 and only with Vue 6. A script that would export cam data to several other Apps would be really helpful.

3. Animation Key POSE Picker window. (You can set up keys for any part of character and activate that key by clicking on a quick render picture of it in a window) For instance, You could have a hand pose library with fist, point, claw, etc. and small quick picks of each. Then you basically click the picture and the pose is keyed on the character and then can be adjusted as needed in the chor/ action. This could be set up/ customized for face, R/L Hand, Eyelids, Brows, etc.

 

4. THIS IS A HUGE ONE.....A CAD topography contour pluggin that allows cad landscapes to be translated into a mesh. Believe me, this is one of the most practical applications I can think of that would save massive amounts of time for those of us who often do projects where we have to place footings for buildings, install sidewalks at specific places, etc.. THE image based methods are not good enough. Seems like since the Cad arch data have specific height points(feet,inches,cent, etc.) This calculation could be done with some sort of plugin. THE REAL PROBLEM with using images(grayscale) is that it is not able to easily get a detailed, ground level surface. The only practical use of a greyscal map seems to be if you don't have to be specific and your camera is about 1,000 feet off of the ground.

[Xtaz]

GPU-based renderer ....

[detbear]

YEEEEESSSSSS!!!! GPU rendering.

[robcat2075]

3. Animation Key POSE Picker window. (You can set up keys for any part of character and activate that key by clicking on a quick render picture of it in a window) For instance, You could have a hand pose library with fist, point, claw, etc. and small quick picks of each. Then you basically click the picture and the pose is keyed on the character and then can be adjusted as needed in the chor/ action. This could be set up/ customized for face, R/L Hand, Eyelids, Brows, etc.

 

I know Kevin knows this, but for the onlookers... we have 95% of this feature already in "draggable poses". You can make a library with fist, point, claw, etc.... and drag them onto your character as needed to create the keys. No icon, however.

 

 

 

4. THIS IS A HUGE ONE.....A CAD topography contour pluggin that allows cad landscapes to be translated into a mesh. Believe me, this is one of the most practical applications I can think of that would save massive amounts of time for those of us who often do projects where we have to place footings for buildings, install sidewalks at specific places, etc.. THE image based methods are not good enough. Seems like since the Cad arch data have specific height points(feet,inches,cent, etc.) This calculation could be done with some sort of plugin. THE REAL PROBLEM with using images(grayscale) is that it is not able to easily get a detailed, ground level surface. The only practical use of a greyscal map seems to be if you don't have to be specific and your camera is about 1,000 feet off of the ground.

 

This sounds very do-able since it's just data conversion. But maybe not widely needed for most A:M users. I bet a programmer could be recruited to do this.

 

THE REAL PROBLEM with using images(grayscale) is that it is not able to easily get a detailed, ground level surface.

How are you creating that grayscale and what format are you using?

[detbear]

Grey_scale....

 

This often occurs when an architect or developer basically gives you a TIN file(at best) or even worse just a 2D CAD line drawing. Currently, in order to obtain a surface from this data, you have to somehow get a greyscale of it, and then convert that into a mesh. Now if you need a city block AND the project requires that an animation gets down to ground level with characters walking around and such, this can be very time consuming. The detail down to the inch...or even foot is difficult to obtain from an Greyscale...image map

 

My thought is.....Point data elevations(in a CAD drawing) must have a certain height number.........there should be some way to translate that height into a detailed Hash mesh. THIS is grueling work going the greyscale route and not very accurate. WELL.....not without going through other softwares/ pluggins and spending four times the hours that a pluggin would eleviate. I'm not a programmer, but it seems like this would be possible.

 

The goal is to obtain an editable mesh of ground topography based on Cad drawinging with elevation heights at specific places. That's what they usually send you(and it has no surface). During a development project, the client often wants to "DIG" out a sidewalk, or add a road that digs into the contours......Working with the splines is a bunch easier if you have the surfaces in Hash.

[robcat2075]

Grey_scale....

 

This often occurs when an architect or developer basically gives you a TIN file(at best) or even worse just a 2D CAD line drawing. Currently, in order to obtain a surface from this data, you have to somehow get a greyscale of it, and then convert that into a mesh. Now if you need a city block AND the project requires that an animation gets down to ground level with characters walking around and such, this can be very time consuming. The detail down to the inch...or even foot is difficult to obtain from an Greyscale...image map

 

My thought is.....Point data elevations(in a CAD drawing) must have a certain height number.........there should be some way to translate that height into a detailed Hash mesh. THIS is grueling work going the greyscale route and not very accurate. WELL.....not without going through other softwares/ pluggins and spending four times the hours that a pluggin would eleviate. I'm not a programmer, but it seems like this would be possible.

 

The goal is to obtain an editable mesh of ground topography based on Cad drawinging with elevation heights at specific places. That's what they usually send you(and it has no surface). During a development project, the client often wants to "DIG" out a sidewalk, or add a road that digs into the contours......Working with the splines is a bunch easier if you have the surfaces in Hash.

 

If you could get your grayscale in OpenEXR that would be much more accurate than an 8bit format.

 

But the density of such a thing sounds scary. Is it just a topographical grid that would suit you or do you need to have the various features like sidewalks economically splined out?

[detbear]

......It's a real headache to get a greyscale...nomatter what format. Resolution is a factor.

 

The ground contour is what you are really after. Many times(I've done sveral of these projects) ....The Cad drawings include points for structure "footings" and or roads etc. BUT just to have the ground layout is what is most important. Of course, the roads and footings are also height points. Maybe if a pluggin would have an option for mesh density...it would be helpful. That way you could add in or take away detail if needed to save data size.

 

Even among other Apps, this process is hard to get, so it would be a very marketable feature for hash. Even people who mostly use max or cad could benefit and use hash to iron out a contour and then export as obj.....Believe me, I've researched it in the past while under a pressing deadline. Currently, I only know of one App that does this and it's not even a mainstream one.

[zandoriastudios]

William,

Here is something I saw using Google Earth + Sketchup to get the terrain of a location. You could export a mesh from sketchup?

http://www.aecbytes.com/tipsandtricks/2006...0-sketchup.html

 

Here is a simpler tutorial I found:

1.Open Google Earth zoom in to the area you want to export

2.Open SketchUp go to Tools > Google Earth > Toggle Terrain then Tools > Google Earth > Get Current View

3.File > Export > 3D Model and under format choose .3ds

[detbear]

Hey,

 

That process works from Google Earth, But seems to import only large triangles. Maybe this process could be improved somehow.....Haven't done enough trial on it. But it doesn't get that "down on the ground" detail needed.

 

Looks promising though if more research provides an easy way to obtain better detail. Not every area has the same resolution in Google Earth.

[detbear]

If you want to see an example of getting a really good and flexible terrain generation for the detail I'm talking about, Watch what this App does with a CAD Contour drawing:

 

YOU need to run it forward to about 1:56....This is where the terrain lines are converted to mesh.

 

 

[detbear]

OK....I must apologize that this thread was probably not meant to expand on any one feature request so I will make this my last post on this...

 

I did find in Will's suggested process....In Sketchup....That you can use the "Sandbox" tools to increase the number of triangles in the mest and thus giving more detail to the terrain. This is a helpful thing.....However, I can't see having to work out all the triangles if it is not necessary. STILL...Hats off and thanks to Will as this is a workable import into hash at a higher detail than the Greyscale image process.

 

Thanks Will S.

[Tore]

Layers, it's a fantastic idee too

 

I have made a request for this (sort of) as an on off button on folders. Would work almost as layers, and would be very easy to implement, I guess. Hopefully we will see this in some coming A:M version.

 

http://www2.hash.com/reports/view.php?id=5846

[NancyGormezano]

Layers, it's a fantastic idee too

 

I have made a request for this (sort of) as an on off button on folders. Would work almost as layers, and would be very easy to implement, I guess. Hopefully we will see this in some coming A:M version.

 

Something is there already

 

create a folder in the chor. (right click on chor, new/folder).

 

Drag models/layers to newly created folder

 

right click on the folder (in the chor) - choose select children, then turn on/off visibility of one of the children, and then the visibility of all children in folder will turn on/off.

[robcat2075]

None of these ideas are insane wants or requests, but be aware... the development "budget" for A:M is extremely limited today.

 

We are fortunate that A:M continues to be maintained (Thanks, Steffen!) in an environment where other apps are going under or trying to sell themselves for next to nothing.

 

Many of these ideas could be plugins which would make them do-able by a 3rd party programmer, much as Jenpy did for the FakeAO plugin (If you haven't bought your "Pro" FakeAO plugin yet, do it!)

[robcat2075]

Layers, it's a fantastic idee too

 

I have made a request for this (sort of) as an on off button on folders. Would work almost as layers, and would be very easy to implement, I guess. Hopefully we will see this in some coming A:M version.

 

Something is there already

 

create a folder in the chor. (right click on chor, new/folder).

 

Drag models/layers to newly created folder

 

right click on the folder (in the chor) - choose select children, then turn on/off visibility of one of the children, and then the visibility of all children in folder will turn on/off.

 

I didn't even know folders could do that!

 

If your "layer" situation is such that folders won't do because some of your sets would have common members, you can use "Selection Filters"to group and quickly access any arbitrary set of items in the chor.

[detbear]

Hey...

 

The Jenpy FakeAO is awesome!!!!.....bought the pro version and it is well worth it.....I also recommend it.

 

If a pluggin converting Cad contour lines to a mesh was available, I would grab that one as well

[John Bigboote]

If a pluggin converting Cad contour lines to a mesh was available, I would grab that one as well

 

 

You mean like a re-topo tool?

[brainmuffin]

If a pluggin converting Cad contour lines to a mesh was available, I would grab that one as well

 

 

You mean like a re-topo tool?

 

 

I don't think that's what he's looking for, but retopo *might* do the job...

[Fuchur]

I already created a feature-request for the retopology-tool for A:M.

That would be one of the features that could make so many tasks much easier...

Including CAD-Imports, 3d-Imports from other programms, being able to use 3d-painter-tools with micro-modelling with A:M, etc.

 

I think any other micromodelling-stuff is undoable without using polygones.

 

See you

*Fuchur*

[robcat2075]

If a pluggin converting Cad contour lines to a mesh was available, I would grab that one as well

 

Does this particular format represent actual contour lines like this?

 

 

 

If you could convert those into splines somehow... hmmm... you'd still have to add spines to cross them, say from the peaks to the valleys, to create patches. But the result would be more economical splineage than a grid conformed to that shape as the current landscape wizard does.

 

A very serious topology guru would have to pursue something like this.

[rusty]

Hi,

 

I've seen and worked with all versions of A:M (NetRender Unlimited to be precise) since V8.5... except for Version 15. Up until 2009 I think (the release of version 15--I know time by AM versions better than years!), I purchased everything that was made for A:M. Plug-ins, 3rd party programs, training videos and books, a t-shirt and... my Hash coat! I'm a good customer.

 

Holding up a great big peace sign and waring 'I love AM' buttons pinned to my shirt, I would ask "What is the 'problem' back log?"

 

If you are caught up!!!!!!! Lets rock!

 

If not, catching up on that would be my only humble request.

 

Thanks!

Rusty

 

PS: I wrote to Martin three weeks ago (didn't know he was off in the bush) asking a question which is vital to me. I wouldn't think that this query would be against any forum rules (if it is I apologize! I'm not here to make waves) and, this info use to be available to me and others. The question is (drum roll):

What 'fixes' are being put off until the next version (version 16 I believe)? If there is any information I can get on this, I'd greatly appreciate it. I would change my current project to skip around any known problems.

[robcat2075]

"What is the 'problem' back log?"

 

There is one programmer working on A:M. That's Steffen. He's gotten lots done. The 2X speed up in V16 is a fantastic achievement that I would say outweighs any individual feature request. On top of that he's gotten lots of debugging done. And he's also added quite a few features.

 

He's aware of the whole AMReports log. I expect he has to decide "what will get the most improvement for the most reasonable amount of effort?"

 

I'd guess his priorities are like this

 

1 crashes

2 errant behavior

3 new features

 

 

Martin is mostly off the field now. If you have a question about A:M ask here, ask Jason@hash, or put it into AMReports.

[detbear]

Yes.....Robert...those are the type contour lines I'm speaking of.

[robcat2075]

Yes.....Robert...those are the type contour lines I'm speaking of.

 

What sort of app are those displayed in?

[rusty]

"What is the 'problem' back log?"

 

There is one programmer working on A:M. That's Steffen. He's gotten lots done. The 2X speed up in V16 is a fantastic achievement that I would say outweighs any individual feature request. On top of that he's gotten lots of debugging done. And he's also added quite a few features.

 

He's aware of the whole AMReports log. I expect he has to decide "what will get the most improvement for the most reasonable amount of effort?"

 

I'd guess his priorities are like this

 

1 crashes

2 errant behavior

3 new features

 

 

Martin is mostly off the field now. If you have a question about A:M ask here, ask Jason@hash, or put it into AMReports.

 

Been there, done that. Was a programmer back in the day.

 

Sometimes one programmer is by far the most efficient way to go. Less meetings anyway.

 

r

[Larry J B]

One feature if not already implemented is a KNIFE tool.

 

Imagine a cube. 2 cp on the top left to top right in front view.

 

*____*

| |

*------*

(* is CP, __ is spline)

 

Lets say you want to add control points and a spline down the middle. To do this you have to Add-lock and click everywhere you want to stitch a spline at an intersection of your choice. OR with the knife tool, click in the middle adn draw a line (or lasso type tool) and wherever the knife line mark interstects a CP, it will auto-stitch a spline.

 

*__~__*

| |

*--~---*

 

(~ is where knife tool intersects in a vertical line)

 

*__*__*

| } |

*--*---*

( '}' is newly autostitched line)

 

This would make my modeling dreams come true

 

-Larry B

[robcat2075]

-->

One feature if not already implemented is a KNIFE tool.

 

I believe the Cut Plane plugin does what you are describing?

[Rodney]

If a pluggin converting Cad contour lines to a mesh was available, I would grab that one as well

 

Don't 'cad contour lines' already convert very well via the DXF and AI importers?

 

 

The main issue with this image is how the numbers present in the image will interfere with the conversion of the contour lines.

Simply put, it's not an ideal candidate for conversion.

Otherwise it's a straightforward conversion via available raster to vector utilities.

Do you have such a converter in your toolbox already?

 

One thing you definitely don't want to do is convert vector lines into raster images and then go through the unnecessary process of converting back from raster to vector again. That will not only defeat the purpose but will degrade the quality and precision of the resulting splines and you'll lose image fidelity.

[thejobe]

i would love a "make random object button" but that's me just being silly hehe.

[robcat2075]

I just grabbed that example of "contour map" off the web, but I'll presume the data format doesn't encode the elevation as visible numbers.

 

But if you convert these contour lines to splines, you still don't have patches yet.

[zandoriastudios]

-->

One feature if not already implemented is a KNIFE tool.

 

Imagine a cube. 2 cp on the top left to top right in front view.

 

*____*

| |

*------*

(* is CP, __ is spline)

 

Lets say you want to add control points and a spline down the middle. To do this you have to Add-lock and click everywhere you want to stitch a spline at an intersection of your choice. OR with the knife tool, click in the middle adn draw a line (or lasso type tool) and wherever the knife line mark interstects a CP, it will auto-stitch a spline.

 

*__~__*

| |

*--~---*

 

(~ is where knife tool intersects in a vertical line)

 

*__*__*

| } |

*--*---*

( '}' is newly autostitched line)

 

This would make my modeling dreams come true

 

-Larry B

 

Just hitting the "A" key and drawing it in seems pretty straight-forward...I guess I don't understand. Have you read this "stiching" tutorial?

http://zandoria.wordpress.com/tutorials/te...ching-tutorial/

[thejobe]

I think what he was trying to say instead of manually placing new nodes to create a new "joint", when using the knife too on line, it will create nodes and connections on all corresponding line along that intersection. At least that's what I think he's saying.

[robcat2075]

Cut Plane will do that but all the cuts will be on one plane.

 

A knife tool that would be more freehand might not be too different from the Stitch process except that it would have to make 2 CPs at every click and be drawing a separate new CP ring for each of them so that both of the parts that are now cut apart would have a proper edge.

 

I think the elements are there to do it but it would take some programmer time. I can think of many "Spline continuity" circumstances that would be difficult to make sure that the computer got it right every time and didn't end up leaving confused splines strung between the two parts.

[dblhelix]

http://zandoria.wordpress.com/tutorials/te...ching-tutorial/

thank you!

[Malo]

For me it would be to update the logic common to all patches.

now:

 

This simplifies their creation, and would remove a button interface.

(and saves time to model).

 

I would push further reflection patches, by posing the question of the usefulness of any patches.

My own reflection, leads me to keep only those patches:

I may be wrong, but it is a reflection that seems useful to discuss for the future of AM.

[Xtaz]

What do you want to model with these splines ?????

I can't conceive a model that uses these features

the actual modeling system is simple and efficient.

[Larry J B]

Jobe, Got it right.

 

As for the knowing where the CP should connect in order should be easy enough to have the computer 'remember' what spline line was crossed first. then being on bird eye view of a cube as teh knife crosses one spline (intersects the point) to remember-mark it as first CP of teh cut. Then as the line intersects another spline, it will be remembered as 2nd point to connect to. Then as it intersects a third point, it can remember to add spline to That point next.

 

I guess the reason for this tool, and speeding things up is I know I can take a somewhat complex series of intersecting splines and pressing 'y' to get into the exact middle, between two splines THEN to a, click, click, click till all are connected I can lets say 'X' then draw a line-intersecting and auto inserting a spline segment instead of all the add, click, click, click, click.

[NancyGormezano]

Larry: Have you tried the plug-in wizards "cut plane"? and/or splitpatch? - I believe one or both probably do what you are looking for:

 

Split patch :

 

1) import a cube

2) select all cp's

3) right click - choose plugins/wizards/splitpatch

 

or

 

cut plane :

 

1) lathe something

2 with NO cps selected, right click in model window - choose plugins/wizards/cut plane

3) position plane where you want (can also rotate plane), choose options (cut, break, add spline ring) - hit apply

Edited April 20, 2011 by NancyGormezano

[tbenefi33]

Be able to draw pattern like boats, clothing, and etc, and print them out to scale.

[Rodney]

I guess the reason for this tool, and speeding things up is I know I can take a somewhat complex series of intersecting splines and pressing 'y' to get into the exact middle, between two splines THEN to a, click, click, click till all are connected I can lets say 'X' then draw a line-intersecting and auto inserting a spline segment instead of all the add, click, click, click, click.

 

There is also the Connect plugin.

It will adjoin/connect splines in 3D space within a user set tolerance from each other.

The plugin will recursively move through all splines and automatically create Control Points where the splines overlap each other.

[robcat2075]

-->

One feature if not already implemented is a KNIFE tool.

 

Imagine a cube. 2 cp on the top left to top right in front view.

 

*____*

| |

*------*

(* is CP, __ is spline)

 

Lets say you want to add control points and a spline down the middle. To do this you have to Add-lock and click everywhere you want to stitch a spline at an intersection of your choice. OR with the knife tool, click in the middle adn draw a line (or lasso type tool) and wherever the knife line mark interstects a CP, it will auto-stitch a spline.

 

*__~__*

| |

*--~---*

 

(~ is where knife tool intersects in a vertical line)

 

*__*__*

| } |

*--*---*

( '}' is newly autostitched line)

 

I'm not really sure what this describes. Could you show the progression with a simple model?

[Malo]

Hello Xtaz,

 

Precisely, that the current system.

 

And I agree, there are patches that are problematic for modeling. (Personally I would remove them, and keep only the three patches are in the last of the topic)

And The creation of 5-sided patches are not simplified. Currently, It is easier to create a patch 3 or 4 sides that create a 5-sided patches.

 

 

Sorry if I misspoke, I do not speak English.

[rusty]

Hi,

 

I have not read all of the posts here so if this is a duplication, please ignore it. If it already exists(!!) please let me know!

 

Copy group properties, paste group properties

 

This would copy any surface properties, any images, any materials and/or anything else (except the actual geometry and associated bones and decals) and then paste these onto another group.

 

Why when you can just take the target group and add it to the source group? It's more a case of having multiple groups needed for other for other reasons (mainly animation or decal stamping) and for saving time.

 

Rusty

[jakerupert]

>Larry: Have you tried the plug-in wizards "cut plane"? and/or splitpatch? - I believe one or both probably do what you are looking for:

 

Split patch :

 

1) import a cube

2) select all cp's

3) right click - choose plugins/wizards/splitpatch

 

or

 

cut plane :

 

1) lathe something

2 with NO cps selected, right click in model window - choose plugins/wizards/cut plane

3) position plane where you want (can also rotate plane), choose options (cut, break, add spline ring) - hit apply

 

This post has been edited by NancyGormezano: Yesterday, 06:02 PM

 

 

[robcat2075]

 

These are both modeling Wizards. Split patch requires that some CPs be selected.

[NancyGormezano]

If it already exists(!!) please let me know!...

Copy group properties, paste group properties

 

Select the group, right click, choose "create material from surface"

 

Not exactly the same - but works as well.

[jakerupert]

>

 

 

These are both modeling Wizards. Split patch requires that some CPs be selected.