SpaceRay
![]() |
||||||||
Posted: September 17, 2020 5:55 pm | ||||||||
SpaceRay
![]() |
This is very well done and I did not know that you have made it with small half triangular tiling as you have done other filters so they make a new tiled image together inside of each of the triangles of the Sierpinsky shape, is a good idea and I did not thought of it when I was making my own version, as I have done it with the plain full triangle, but you creatively have gone beyond and make it better and different than is usually available, and is good that you do it in your own way, so then is special.
Here is something I have done using your own Circular reasoning with the warhol filter ![]() ![]() |
|||||||
Posted: September 17, 2020 5:57 pm | ||||||||
SpaceRay
![]() |
WHAT I LIKE
Apart from previous post 1 - I like the way you can choose to make the sierpinsky in 5 layers and sel ect what parts you want to use, and additionally with the Max Depth that replaces the central triangle with another sierpinsky triangle inside that is a good idea. 2 - Like much the option to have the Tri-Parallel-Hex options so this way, not only you can do a triangle, you can also make a double triangle making a rhomboid and then making the Hexagon tiling 3 - As in other filters, it uses the well done and easy to use SEED view to create your own seed for the tiling of the triangles 4 - Like Preset 8 how is done with the missing central triangle,, and making three fractal frame style. WHAT I DO NOT LIKE 5 - Although I have some things that I do not like as much, Maybe I think that the main big background triangle (Layer 1 checkbox) has maybe a bigger size in comparison with the other triangles, and think that depending of the source image, it may sometimes look wrong. Will have to make more tests. 6 - From the tests I have done, I think that it would be really good if there could be an option to have the edges of the triangles visible and have a little width, so you could then clearly see optionally if wanted, wh ere each of the triangles are and the size of each one. 7 - From my personal point of view I think there may be missing perhaps and additional iteration and have also optionally the layer 6 with the smaller triangles filling the remaining space that is visible of the Layer 1. I think that maybe you did not add the layer 6, so it could be shown better the bakground of layer 1 and could not be needed, but as it is optional and is a Sierpinsky fractal it could be interesting to have it, or maybe I am wrong and it may not look good, do not know. I tried to make it myself but still do not understand how it works, and do not know how easy or difficult this layer 6 can be. |
|||||||
Posted: September 17, 2020 6:32 pm | ||||||||
SpaceRay
![]() |
||||||||
Posted: September 17, 2020 7:18 pm | ||||||||
SpaceRay
![]() |
Sorry about Point 6 that I thought it was easier, so for the edges of the triangles is not as easy as I thought because when you add the edges to the main triangle created first, it will be scaled down for the all the other smaller triangles and so it will be shown with a smaller edge and so it is wrong
And also that is needed to scale the triangle to be shown fully ![]() |
|||||||
Posted: September 17, 2020 8:08 pm | ||||||||
SpaceRay
![]() |
||||||||
Posted: September 17, 2020 9:50 pm | ||||||||
SpaceRay
![]() |
||||||||
Posted: September 17, 2020 11:43 pm | ||||||||
SpaceRay
![]() |
||||||||
Posted: September 17, 2020 11:45 pm | ||||||||
Rachel Duim
![]() |
||||||||
Posted: September 17, 2020 11:50 pm | ||||||||
SpaceRay
![]() |
||||||||
Posted: September 17, 2020 11:50 pm | ||||||||
SpaceRay
![]() |
||||||||
Posted: September 17, 2020 11:51 pm | ||||||||
Rachel Duim
![]() |
Here's the code. Busy, but will release it when I get the chance. FF9 and above.
Sierpinsky Crystals Too beta.ffxml Math meets art meets psychedelia. |
|||||||
Posted: September 17, 2020 11:52 pm | ||||||||
SpaceRay
![]() |
My own version uploaded above I think is not working correctly as the edges may not always work right and I think needs more work
The two images shown here above about the edges are saved as the last 2 presets in the filter uploaded
Hello, Glad that you agree and like my comments and is a good news that you may be doing and updated version with more things.
Thanks very much for showing your own version and will see how it works I still have to finish my own version of the Sierpinsky triangle that is different to yours and have 5 different colors and it has not the tiling of the images inside the triangles. |
|||||||
Posted: September 17, 2020 11:57 pm | ||||||||
SpaceRay
![]() |
Sorry to say that from my point of view and how I see it, this is wrong for 2 things 1 - The Biggest triangle edge of Layer 1 is under all of the other triangles near the edge and is not on the outside part, as you can see in my own version all the edges are NOT overlapping any other triangle 2 - I think that all the edges should around the same width, and as you have put it (and as I did put it also the first time) the edges are affected by the scale, so as you see the Layer 1 edge is thick, and then the Layer 2 is smaller, and then Layer 3 is smaller and so on. yours is similar to what I have done the first time as I have shown here ![]() because with this you have done is that the edges are ALSO scaled in the same way as the triangles so then the Line Thickness ONLY is true on the biggest triangle, and then on each reduction, the edge gets smaller and smaller, if this is what you want and what you like is good and not wrong. |
|||||||
Posted: September 18, 2020 12:36 am | ||||||||
SpaceRay
![]() |
I have seen that in the filter above it was wrong because it was badly positioned the edges over each triangle and so it was looking not good.
But after making it right, I made many tests and have not been able to make it possible to manage that just one slider with remapping can give different edge sizes, so I have decided to give a individual separate slider for each edge so you can decide how much of an edge you want for each layer and have it all the same equal or different I have put it in the first presets to see the difference Sierpinsky Crystals Individual Edges.ffxml |
|||||||
Posted: September 18, 2020 6:59 am | ||||||||
SpaceRay
![]() |
In the filter in previous post I have added a Polygon Contour with bevel width to each of the 5 layers.
And maybe I should add a threshold to each one to make them solid and not with gradient as it is now Here is an example of the above filter using different values of edges to make that visually all to be about the same width Of course, this is just a possible suggestion and you may have a different way to do it and you decide if you like it or not, as it is your filter, and do not have to be done as I like or want. ![]() |
|||||||
Posted: September 18, 2020 7:05 am | ||||||||
SpaceRay
![]() |
||||||||
Posted: September 18, 2020 7:27 am | ||||||||
SpaceRay
![]() |
||||||||
Posted: September 18, 2020 7:39 am | ||||||||
Rachel Duim
![]() |
Here's a sample of Sierpinsky Crystals with background color, edge color and 6th layer. In the queue. The first 4 layers have the same edge size now. Not practical to go below that level.
LATEST: the filter is updated. I found a couple of minor math errors, so another update... ![]() Math meets art meets psychedelia. |
|||||||
Posted: September 18, 2020 6:27 pm | ||||||||
SpaceRay
![]() |
First, thanks very much for considering and you did like to add the edges, background and the 6th layer
Great update and I like much how you have done the edges so well with only one slider, and how you made the 6th layer. As I am the one that made you update the filter and is slower, I think I has discovered why it happens, as I am making about 25 tests, and will give you a report later about what I have found that may be interesting to see the way FF works and is not your fault, as you do not know what will FF do when adding additional components to a filter |
|||||||
Posted: September 22, 2020 11:49 am | ||||||||
SpaceRay
![]() |
Before you have made the update with the 6th layer, I already have found the way to make additional layers and I now I have got the filter with 6 layers, 7 layers, 8 layers and 9 layers but the result makes not possible to update it in just one filter because it becomes much slower, so it should be in different versions
As you have done an awesome work making this sierpinski triangle in an amazing artistic way that nobody else have thought ( I have seen lots of different ones) and so I want to be able to have it all the possible iterations that can be visible As you have done a great work in the way that you can be able to select how much layers you want to use and is optional and see easily what looks best. Here is an example with 8 iterations that you can see at full size all the details with 8 layers, it may be less elegant and more tight and not as nice but is the full fractal. ![]() |
|||||||
Posted: September 23, 2020 2:16 pm | ||||||||
SpaceRay
![]() |
here below is a zoom version of the top part to see much better the details and the 8 layers that is done at 5000 x 5000
There is an error in the way the 7 and 8 layers positions that I need to fix and I can upload the filter here with the 8 layers. I am making a full report with the results of the tests, but meanwhile here is a short one made before tests with the first version and not with the new updated one Layer 5 600 x 600 = 6,19 sec ______5000 X 5000 = 3 minutes 41 sec. With Layer 6 600x600 = 22 sec _____________5000x5000 = 11 minutes 25 sec With Layer 7 600x600 = 1 min 2 sec ______________50000x5000. = 35 minutes 51 sec With Layer 8 600x600 = 3 minute 35 seconds ______________5000 x 5000 = 1 Hour 54 minutes With Layer 9 600x600 = 11 minutes 7 sec _____________5000 X 5000 = to be done ![]() |
|||||||
Posted: September 23, 2020 2:22 pm | ||||||||
Rachel Duim
![]() |
Very nice. The reason I did not go any deeper than 6 is illustrated in your tests. I normally won't release anything that is over 2 minutes at 600x600. But for a particular look, you can go deeper. I appreciate that the method I used is different than others you have seen... ironically I did not look at a lot of examples and just used my tiling method that has served me so well.
UPDATE: added a couple of cropping rectangles, now twice as fast or better. Coded it too quickly, hopefully this is the last update ![]() Math meets art meets psychedelia. |
|||||||
Posted: September 23, 2020 5:16 pm | ||||||||
SpaceRay
![]() |
You are right, after making the tests, I thought the same as you that increasing the layers would make the filter much slower, and should be done with care. And I agree with you that a filter that has over 2 minutes in 600x600 in not interesting to have unless it may be something really awesome and amazing that may be worth to wait for the result.
Your tiling method you have done works very well, and is very good and well done, and very easy to use and much customizable to position correctly the seed
I think this is related to the 27 Sept Update but I am sorry that I do not see any option in settings, or anything related to cropping rectangle inside the filter? where is this? thanks |
|||||||
Posted: September 28, 2020 2:42 am | ||||||||
SpaceRay
![]() |
The 27 sept update is really amazing, impressive and great how you have been able to make it much faster, is not twice as fast, is about triple as fast and sometimes even more!!!!! Congratulations! Please, would be possible to know what you have changed and optimized to be able to get it much faster? only if it is easy to explain 1- First Version original with 5 layers No edges 1.1 - Preset 2 and 8.................6.75 seconds 1.2 - Preset 6...........................5.29 1.3 - Preset 6 + Max Depth......7.84 1.4 - Preset 4 Hexagon............8.60 2 - Second Version - 6 Layers with Edges Preset 3 equals to Preset 2 Preset 7 equals to Preset 6 Preset 5 equals to Preset 4 2.1 - Preset 3...........................15.14 2.2 - Preset 7...........................24.10 2.3 - Preset 7 + Max Depth......49.01 2.4 - Preset 5 Hexagon...........50.43 3 - Third Version faster - 6 Layers with Edges 3.1 - Preset 3............................5.27 3.2 - Preset 7............................8.90 3.3 - Preset 7 + Max Depth.....12.41 3.4 - Preset 5 Hexagon...........17.03 3.5 - Preset 5 Hexagon 5 layers 8.98 seconds equals to 1.4 above that is 8.60 |
|||||||
Posted: September 28, 2020 2:56 am | ||||||||
SpaceRay
![]() |
You are the expert in making tiling textures as you have already shown with many filters and you are the creator of the
17 WALLPAPER GROUPS MADE BY RACHEL DUIM AVAILABLE IN FF LIBRARY So I suggest that you can also make this awesome sierpinski triangles make a seamless pattern too as it is easy to do because you already have made all the parts of it, just needs only to assemble them in the right way here is an example ![]() |
|||||||
Posted: September 28, 2020 6:39 am | ||||||||
SpaceRay
![]() |
and it does not slow down the filter as the 600x600 version takes 25.60 seconds
here is the filter for you to test it yourself with the 4th option in the slider, sel ect "Patern" I could not fit the double t I only added 6 components that I have separated below the filter with a note It is made from your latest 27 sept version with all your own presets, so you can test how it looks the pattern of any of your presets selecting the 4th option the above example is the added as a preset, although the above one is cropped I changed the selector from Parallel to Rhomboid that is the shape you get and so it fits the name Sierpinsky Crystals Pattern (1).ffxml |
|||||||
Posted: September 28, 2020 6:46 am | ||||||||
SpaceRay
![]() |
||||||||
Posted: September 28, 2020 11:37 am | ||||||||
SpaceRay
![]() |
deleted to make a new one
|
|||||||
Posted: September 28, 2020 11:43 am | ||||||||
SpaceRay
![]() |
Please, do not download the Pattern filter in the above post, is better the new one that is fixed available below in the next post.
NEW FIXED ERRORS VERSION AVAILABLE BELOW MADE BY ME TO HELP I really like and love much this filter and is really awesome that Rachel have done and I have been using it but always in low resolution, and now I made a 5000x5000 version and I have seen that there are some errors that I did not notice before, so I want to help to make it right and spent some few hours fixing it. 1 - The position of each triangle of the layers is not perfectly positioned, and you can not see it in low resolution, and I understand that Rachel could not be able to see it. There are some that are offset and some overlay the edges. 2 - The Edges are overlapping and cropping the source image of the seed more and more on layers from 1 to 4 and so the selected seed is not seen correctly in all the triangles. And also the Edge of Layer one is entering inside the design. Here you can see what I mean what is wrong and I think that of course is not Rachel fault, because I only notice this when making a high resolution. ![]() |
|||||||
Posted: October 19, 2020 3:58 am | ||||||||
SpaceRay
![]() |
So to help Rachel and make her awesome filter correctly and that she does not spend time fixing what is in the above post, I have spent some few hours to:
1 - very carefully position each one of the triangle correctly in each layer 2 - I have using the scale for each edge to be outside of the seed and so now all the triangles have exactly same seed with any cropping and Layer 1 does edge is not going inside the design I have seen that is nearly impossible to see the errors if you use the 600x600, you must use a very high resolution, so I have used a 5000x5000 Ladybug source image to do it, so this is why I know that Rachel could not do it at all. Here is the filter that is exactly the same than the one available on the Filter Library (with 5 new presets added) but is fixed the 2 errors told above, and is a pleasure to help you and congratulate again for making this filter. Sierpinsky Crystals 3rd version fixed errors.ffxml |
|||||||
Posted: October 19, 2020 4:12 am | ||||||||
SpaceRay
![]() |
I suggest that you update your filter on the library with this one here above if you want
and here is what it looks like the new fixed version using 6 layers with preset 17, that I think should be preset 2 or 3 Now all triangles are exactly the same with the same seed size and there is no error on positioning ![]() |
|||||||
Posted: October 20, 2020 1:50 am | ||||||||
SpaceRay
![]() |
after fixing your filter as shown here above, I have made also a new version of the pattern making so now I continue with what I have started above
So, here is the new the new filter for making seamless patterns using this awesome filter 1 - Now there is a 4th choice in the Tri-Rhomb-Hex-Patern (changed Parallel to Rhomboid as it about the same) 2 - There are some new presets to show what you can do with this new 4th option 3 - There are scale and offset settings to be able to zoom out or in in the pattern as wanted or to make it fit exactly to be seamless. Sierpinsky Crystals 3rd version fixed errors pattern.ffxml |
|||||||
Posted: October 20, 2020 2:05 am | ||||||||
SpaceRay
![]() |
Here is an example of what can be done with the new scale and offset settings from the post above here, to show a full triangle with the added additional sides and top
I forgot to tell that I have added many notes to each part of the filter to explain how it works and what it does in each step This is already available in the preset with the Note "Triangle Pattern" This image may be also seamless if you cut away the top part from the top black horizontal line upward ![]() |
|||||||
Posted: October 20, 2020 2:14 am | ||||||||
Kevin Worth
![]() |
I like this SpaceRay it looks great. Did you manage to do a circle version yet?
|
|||||||
Posted: October 20, 2020 4:21 am | ||||||||
SpaceRay
![]() |
Hello, thanks, but I think that you are answering in the wrong thread, as there is no circle version of sierpinski and I have just posted today on your own thread of your filter and the answer is for this |
|||||||
Posted: October 20, 2020 4:38 pm |
Filter Forge has a thriving, vibrant, knowledgeable user community. Feel free to join us and have fun!
33,711 Registered Users
+18 new in 30 days!
153,531 Posts
+36 new in 30 days!
15,347 Topics
+72 new in year!
16 unregistered users.