[Bf-committers] development of conditional pass-thru node for compositor

pete larabell xgl.asyliax at gmail.com
Wed Apr 20 21:57:22 CEST 2011


Xavier,

Thanks. I was actually more concerned with the == and != operators. Do
you have a quick way of doing those with regular nodes too? That is
the main goal for me.

As far as full sample goes, yeah. :) I know all about it. :)

Had to deal with it whilst making another node...
http://www.youtube.com/watch?v=VcjEfoNIHZs

I ended up changing the way I do that now, so it no longer works like
in that vid, but you'll get the idea.

Thanks again!
Pete

On Wed, Apr 20, 2011 at 2:17 PM, Xavier Thomas
<xavier.thomas.1980 at gmail.com> wrote:
> Also, if you want anti-aliasing to work on setup like this, you should
> set "Full Sample Anti-Aliasing" ON.
>
> 2011/4/20 Xavier Thomas <xavier.thomas.1980 at gmail.com>:
>> hi Pete,
>>
>> I thought you might be interested in a fast and easy way to do this
>> with blender nodes:
>> http://www.pasteall.org/pic/11274
>>
>> 2011/4/20 pete larabell <xgl.asyliax at gmail.com>:
>>> indeed, certainly not saying it has to go into blender :)
>>>
>>> Just wanted an easy and fast way to get overlapping parts of alphas
>>> and such for an experiment. thought someone could find it interesting.
>>> Perhaps I'm not familiar enough with the other nodes, but i didn't see
>>> a really simple way to do all the tests i wanted to be able to do.
>>>
>>> On Wed, Apr 20, 2011 at 1:29 PM, Xavier Thomas
>>> <xavier.thomas.1980 at gmail.com> wrote:
>>>> Hi Pete,
>>>>
>>>> In fact the math node (in the convertor category) does something similar.
>>>> It does not have the "pass if true" and "pass if false" socket but
>>>> it's output can be used as mask to achieve the same result.
>>>>
>>>> I personally think your node is redundant, maybe more effective in
>>>> terms of performances but also less flexible.
>>>>
>>>> It is a pity that we cannot do nodes plugins in C so that one can
>>>> distribute some specialized node, because bloating the sets of
>>>> defaults nodes (provided with Blender by default) is not really an
>>>> option too, it will confuse user too much.
>>>>
>>>> Xavier
>>>>
>>>> 2011/4/20 pete larabell <xgl.asyliax at gmail.com>:
>>>>> not quite sure where this will end up, but it's working in it's
>>>>> current form, thought i'd share a quick render
>>>>>
>>>>> http://www.pasteall.org/pic/11269
>>>>>
>>>>> basic idea is:
>>>>>
>>>>> if(op1.pixel **TEST_HERE** op2.pixel){
>>>>>    out.pixel = pass_if_true.pixel;    //the 3rd input
>>>>> }else{
>>>>>    out.pixel = pass_if_false.pixel;   // the 4th input
>>>>> }
>>>>>
>>>>> obviously that's pseudo-code. :)
>>>>>
>>>>> current tests available are ==, !=, >, <, >=, <=
>>>>>
>>>>> plan to allow full RGBA sockets for at least the bottom 2 inputs and
>>>>> output socket.
>>>>>
>>>>> Cheers!
>>>>> _______________________________________________
>>>>> Bf-committers mailing list
>>>>> Bf-committers at blender.org
>>>>> http://lists.blender.org/mailman/listinfo/bf-committers
>>>>>
>>>> _______________________________________________
>>>> Bf-committers mailing list
>>>> Bf-committers at blender.org
>>>> http://lists.blender.org/mailman/listinfo/bf-committers
>>>>
>>> _______________________________________________
>>> Bf-committers mailing list
>>> Bf-committers at blender.org
>>> http://lists.blender.org/mailman/listinfo/bf-committers
>>>
>>
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at blender.org
> http://lists.blender.org/mailman/listinfo/bf-committers
>


More information about the Bf-committers mailing list