Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - maus

Pages: 1 2 3 [4] 5
46
Yah I have noticed something similar, when scaling down only the height of a white rectangle and placing it using the FX-Map edges get blurred, the width seams to scale up on its own and the white becomes gray!?


47
Thanks for the tip.

But my Input Images have different sizes then the Output of the FX Map or Pixel Processer...

I am trying to make a compound image of sorts, and knowing the size of all the different Input Images is important to the placement logic.

It looks like we can Sample the Color of an Input Image at X,Y but there is no way to get the Size an Input Image...

^^

48
Thanks but the $size variable "returns the size of the current node" not the size of an Input Image...
https://support.allegorithmic.com/documentation/display/SD5/Variables

Thanks for your help though.

^^

49
Try a Histogram Range node.

^^

50
Cool thanks for your help!

I love what you all are doing.

^^

51
Try something more like this:
(I should have made the last node on the right the output node, derp)
The output node has to be able to reach all the nodes in your graph traveling from Right to Left for them to be calculated.

^^

52
Yah I was going back and forth from the root graph and into an FX node with a very large function graph that I had been working on. So I am pretty sure there where a few divisions in the logic.

The first node that went blank and zero time was the Normal Color node and then the Uniform Color node after that it seamed to affect any node I copied and pasted or duplicated.

After restarting SD and then the computer the copied/duplicated nodes that where saved worked fine.

I did check the task manager and SD was not on the list.

Thanks for all your help.

^^

54
Ok so not sure where to start...

First: Help > Report a Bug... just takes you to the documentation so not sure what that's about.

When you change between graphs (double click on them in the Explorer or right click "Display outputs in 3D view") any outputs that are not used in the current graph still show in the 3D view as the old outputs. So say your last graph had an AO output and the graph you just opened dus not, the old graphs AO output is still being displayed on the model in the 3D view...

After working for a while any copped or duplicated nodes just stop working. They say 0seconds and never update in the 2D view.

After getting frustrated I tried exiting Substance Designer and reopening it... well now looks like SD can no longer be reopened after you just closed the program. I had to restart my computer just so I could start SD again.

So yah lots of wired stuff...

I hope this helps you all to fix the bugs, let me know if there is something I can do to help.

^^

Maus

55
Is there a way to get the size in pixels of an Input Image?

Thanks for your help.

^^




56
Something a bit like the scene statistics in 3Ds Max would be cool.

Just some small text in the corner that could be turned on an off if needed that would display some helpful info like:

Total Nodes in the Graph
Total Milliseconds to calculate the Graph
Number of Nodes Selected
Processing time of Selected Nodes in Milliseconds

Just a little text overlay to display helpful information that will help us to develop better substances.

^^

Maus

57
I just think it would be cool to get to see a real time 3D preview of substances.

Plus linking back and forth from an art sharing site SketchFab would be good natural link generation.

^^

Maus

58
Hey Cyrille,

Sorry for not getting back to you sooner. My computer had a critical failure a few days after I made this post and I have been with out it for a few weeks now.

I totally agree with Waffenwars even a loop node with restrictions would be better then no loop node.

I know speed is important for real time rendering of substances, but speed is not always a top priority for some of us. Most if not all the substances in my project will only be calculated once during the first time the game is installed, so I am not to worried about speed.

In the case of sampling areas from input images. If you want more then just a few samples to be used to calculating the output pixel it blows up the size of your graph. Just sampling the 8 pixels adjacent to the current pixel and averaging them is a graphing nightmare. Now image you would like to sample a rectangle area (x1, y1, x2, y2) well aside from needing a loop to make all those samples in a variable size area even a fixed size area 4x4 pixels becomes a ridiculous task to graph.

Printing a pixel at a specified location, shouldn't add any more pixels to the proses. Say I have a 1024 x 1024 image, if I don't specify an XY position for the output then the Current XY position is used. So its still only 1,048,578 pixels... but now we have the option of saying where we want processed pixel to be output to.

Think of it as pushing a pixel instead of the current proses witch must pull all the information that is needed to the current pixel. In some cases just being abele to specify where the output pixel should be located could achieve the target effect with out say having to sample all of the surrounding pixels thus significantly speeding up the graph.

Storing some info in variables. This could actually be used to speed things up. Say I have to make a bunch of samples every single pixel right, well if I could store those samples in a variable or 2, I can reference that stored information on subsequent pixels and prevent resampling and recalculations of that data for every pixel.

More functionality really will help make the Pixel Processer faster! Because it will give the users more ways they can be creative when developing substances.

How about adding a Brake node so we can stop possessing pixels when where dun doing what ever it is where doing that could help save processing time. Or maybe we could specify the areas of the image to be processed with a mask for instance, thus saving more time.

More functionality really is less processing time, and if we fined some way to crash substance with a crazy graph, we wont blame you guys for taking away the training wheals, well thank you for having the freedom to get a few bumps and scratches while making something cool!

Thanks again for making such an awesome tool!

^^


59
I love the Pixel Processer node, but it seams every time that I set out to do something cool with it, I come up short...

Cant loop.
Cant sample an area from an input image.
Cant print to a pixel at a specified location.
Cant store information in a variable to use in subsequent pixels.
And so on...

After great frustration and days of trying to out clever the limitations of the Pixel Processer graphs, I fined my self getting stuck on little things that feel like they should be simple, only if we had just a few more nodes with witch to manipulate the data...

Graph editors aside, if there where a way for us to script custom nodes we could extend the functionality and make more cool nodes to share with the community.

Thanks for considering my requests, I really love SD even though nothing I have tried to make has really worked out... yet! I will keep trying and excitedly waiting for new features.

^^


60
Actually the Tri-Planer projection node projects images upside down, inside out and backwards... Just try projecting some text onto a model if you want to see what its really doing.

Pages: 1 2 3 [4] 5