54 replies to this topic

[Funkdude]

First of all, Congratulations to the Genetica team! The new genetica is better then ever

I do have a small wish-list though:

-Being able in the results dialog, to assign an effect map to an alpha channel (Unity and Unreal 3 editor use this alot, so a specular map in the alpha of the colour map etc. would be great, saves doing it in another program. I know there is a workaround to do this in Genetica already, but in the results dialogue it would make more sense i guess.

-A dedicated scatter node for nodes, not just for environments, to make smooth heightmaps etc.

-A way to arrange, merge, and make folders inside the library of genetica, so we can delete and move files around, whilst being able to see what the file looks like.

Thats all for now, all minor stuff, but it might speed up my work-flow and that of others even more

[Lynda]

Vector path import.

[darrellp]

Modifiable parameters. One of the big reasons that I do work in both Genetica and Filter Forge is that FF has modifiable parameters. Huge advantage to be able to generate unlimited number of new textures without having to go dink around with/move nodes. Not only do you have to dink with them in Genetica, you have to understand what they do as opposed to having named parameters. Also, with named parameters I can pass one parameter to several nodes. There are all sorts of sample textures which have precisely tuned values so that, for instance, nails come out in the middle of boards. If you want to make the boards a bit wider, you have to figure out all the other values that depend on that and change those as well. Difficult both conceptually (figuring out where those values are crucial) and operationally (tediously changing all those places). Huge advantage to have one parameter called "board width" control all of that. I'm sure this would be a very difficult change to make but it would be by far the biggest improvement in my humble opinion.

[gordon]

Modifiable parameters. One of the big reasons that I do work in both Genetica and Filter Forge is that FF has modifiable parameters. Huge advantage to be able to generate unlimited number of new textures without having to go dink around with/move nodes. Not only do you have to dink with them in Genetica, you have to understand what they do as opposed to having named parameters. Also, with named parameters I can pass one parameter to several nodes. There are all sorts of sample textures which have precisely tuned values so that, for instance, nails come out in the middle of boards. If you want to make the boards a bit wider, you have to figure out all the other values that depend on that and change those as well. Difficult both conceptually (figuring out where those values are crucial) and operationally (tediously changing all those places). Huge advantage to have one parameter called "board width" control all of that. I'm sure this would be a very difficult change to make but it would be by far the biggest improvement in my humble opinion.

I have to agree here. This would be one of the single most important improvements that could be made.

I have 2 things that I would like to see added as well.

1) multiple render paths with a base name for output or the ability output normal maps with a texture.

2) a clip node that would allow for non-square renders or a way to clip final image by relative dimensions thus making it image size independent.

[Atlas]

Agreed on all accounts!

[Funkdude]

If non-square resolutions are supported this might be neat:

a "HDR" node, it would have inputs for multiple exposures (user specified number) and a curve tab where the tone-mapping can take place. That way, when for example i take multiple pics of a concrete wall, i can use the hdr node to combine them, and exclude any shadowing, making for an ideal diffuse map, and another hdr node with the same (referenced, for speed) pics to make a specular and height map. (With help of the synthesis node), (And then use procedural nodes for weathering etc).

I think this might be a nice addition and would also allow for super awesome photo editing.

Yes. I said super awesome.

[Edit: Super awesome is better then regular awesome.]

Edited by Funkdude, 11 July 2011 - 03:33 AM.

[MonsterMind]

............ and would also allow for super awesome photo editing.

This would expend a lot Genetica's use I think.

[Atlas]

a "HDR" node, it would have inputs for multiple exposures (user specified number) and a curve tab where the tone-mapping can take place. That way, when for example i take multiple pics of a concrete wall, i can use the hdr node to combine them, and exclude any shadowing, making for an ideal diffuse map, and another hdr node with the same (referenced, for speed) pics to make a specular and height map. (With help of the synthesis node), (And then use procedural nodes for weathering etc).

I think this might be a nice addition and would also allow for super awesome photo editing.

I recall putting a GTX together for that job a while back: http://www.spiralfor...a...ost&p=28647

That GTX assumes that each image has double the brightness of the previous one, ie, each exposure is one stop higher.

[gordon]

Just thought of a node that would be really useful.. well at least for me. It would be great if we could have a node for external reference. It could be part of the group nodes. The way I see this node working is that it would render the referenced gtx file and allow the resulting image to be referenced... This would be great when working on a set of images that all use the same base texture. So changing the set would just be a matter of changing the one gtx.

[MonsterMind]

Just thought of a node that would be really useful.. well at least for me. It would be great if we could have a node for external reference. It could be part of the group nodes. The way I see this node working is that it would render the referenced gtx file and allow the resulting image to be referenced... This would be great when working on a set of images that all use the same base texture. So changing the set would just be a matter of changing the one gtx.

How would this be different from using the "import image" node?

However, it could be nice to be able to reference an external gtx, sort of a "gtx subroutine". Of course one could create such gtxs today and copy and past them in the appropriate position but being able to reference them from a library is sort of intriguing.

[gordon]

How would this be different from using the "import image" node?

However, it could be nice to be able to reference an external gtx, sort of a "gtx subroutine". Of course one could create such gtxs today and copy and past them in the appropriate position but being able to reference them from a library is sort of intriguing.

Yes the image import is a workaround for this but it is not truly resolution independent like the source gtx would be. This concept is found in most modern computer languages and if working on a set of images would allow for change once and render many.

[Lynda]

Just thought of a node that would be really useful.. well at least for me. It would be great if we could have a node for external reference. It could be part of the group nodes. The way I see this node working is that it would render the referenced gtx file and allow the resulting image to be referenced... This would be great when working on a set of images that all use the same base texture. So changing the set would just be a matter of changing the one gtx.

Have you looked at the Batch Render feature in the Pro version? It lets you sub in a file full of GTX into a single node in your texture, and then render out the set of result images. These were made by swapping out the noise gtx node in my fudge striped cookie texture, using a swap file that contained a number of different custom noise gtx files.

Attached Images

• 

[gordon]

[quote name='Lynda' date='Jul 29 2011, 09:54 PM' post='32145']
Have you looked at the Batch Render feature in the Pro version? It lets you sub in a file full of GTX into a single node in your texture,

Yes I have. What your doing there is a base gtx with many input changes. What I was addressing was the revers situation where you have many base gtx files and need to change The same part in all of the files.

[Atlas]

Technically if you do a batch swap on the contents of a Group with a number of specially created GTXs (basically filters that have Reference nodes exposed at the top level) you can swap out filters for a single input. It's a pain to set up though.

Having said that, I agree with your overall sentiment. A way is needed for changes to automatically propagate through entire sets.

[Silkrooster]

Technically if you do a batch swap on the contents of a Group with a number of specially created GTXs (basically filters that have Reference nodes exposed at the top level) you can swap out filters for a single input. It's a pain to set up though.

Having said that, I agree with your overall sentiment. A way is needed for changes to automatically propagate through entire sets.

Maybe some kind of wizard where you could specify which node is to be swapped and let it run through how ever many files in the specified directory. Or something along those lines...