TheBeat
Posts: 165 |
Hi Skybase,
I've been watching your tutorials yesterday and today. Thanks a lot, very well done. The Loop pt1. is very interesting. Did you ever make pt2, as I could not find it? And I've seen the video on your Image Synth (done 6 years ago). Is it possible to get that filter? Looks very interesting. I could not find it in my Library program. Thanks for your work, Beat The Stone Age did not end because they ran out of stones. Niels Bohr |
|||||||
Posted: April 25, 2019 9:47 am | ||||||||
SpaceRay
![]() |
Here is part 1 of the Loops video
Regretably, there is NO part2 for the awesome and very useful and interesting Loop video tutorial that is very helpful and useful is a great pity that he could not make the second part or any other additional videos, as it would be really very helpful and useful fr om a great expert like him and he knows very well how to make tutorial videos
What Image Synth Video? do you have a link for it?
IF it is not available wh ere the video is, then probably is not available, although really I do not know, have to see the video first to know what is about |
|||||||
Posted: April 25, 2019 12:02 pm | ||||||||
TheBeat
Posts: 165 |
The video is called 'FilterForge … image synth?' and the link is https://vimeo.com/58611956 The post is 6 years old. I just want to see how an image synth is done in principle (probably more ways to cook the thing), so another Image synth would be fine as well. I did not find any in my FF library program though. The Stone Age did not end because they ran out of stones. Niels Bohr |
|||||||
Posted: April 26, 2019 11:49 am | ||||||||
Yuya
![]() |
I haven't posted the filter and I won't release the original.
A few reasons: 1. Everybody can end up making what I make. I made that filter as a side-branding project for myself. I made a website called RGBland which consequently became a self-published book. My main intent wasn't for profit, I just wanted a thing that generated a gajillion patterns with style. You can see my efforts below. https://www.skybase.xyz/tumblr-feed Overall, whether it'd be for learning purposes or not I just prefer not to have the same content generated elsewhere. 2. The library makes it free-to-sell the output on any website. Shutterstock, Adobe stock, Envato has a bunch of authors who seem to just smash that next-variant button to generate a bunch of variations to sell high-quality on markets. It's 100% legal, I just find it morally questionable as an artist. Yeah sure I may have helped a few pockets with extra cash but I'm the one who spent a few hundred hours on some of the things I made, all put out for free. What I can say though is that I think I can make a lite-version of sorts just for this community. But I'm also going on the "no-promises" thing since I really don't know when I'll get to it. Most of this is just based on feeling. |
|||||||
Posted: April 27, 2019 10:41 am | ||||||||
Yuya
![]() |
Overall the video gives you all you need to kinda get started. It's really basic in terms of its design. If you understand how video synthesizers work, you just have to apply the same idea here and it'll do its magic.
Try this: make 3 noise patterns with different settings, input those into the RG and B channels of the Assemble RGB component. You're basically 90% there in terms of principle. In my case I used 3 separate profile gradients which I just manipulated until it generated something cool. All the rest of the complexities was me just playing around with the components so I can randomize the output. Each time I add a new thing it makes the output more complex. ![]() |
|||||||
Posted: April 27, 2019 10:50 am | ||||||||
Yuya
![]() |
||||||||
Posted: April 27, 2019 10:53 am | ||||||||
Yuya
![]() |
[What the... duplicate post deleted this]
|
|||||||
Posted: April 27, 2019 11:06 am | ||||||||
Yuya
![]() |
||||||||
Posted: April 27, 2019 11:16 am | ||||||||
TheBeat
Posts: 165 |
Thanks for the link, Yuya. Looking very good and a nice idea that it's regenerating.
I agree. I look at other people's work to learn, but always cook my own. I feel that there ere too many taking the lazy way. I feel you have the right attitude.
I had no idea how an image synth was working, but tour suggestion makes it clear. This would be the main principle. I can find lots of refinements and variations for myself. Thanks a lot. I will post some results if I think they're worth while. The Stone Age did not end because they ran out of stones. Niels Bohr |
|||||||
Posted: April 27, 2019 4:38 pm | ||||||||
earthbound |
If I publish work that's compelling in any way I hope that others will exploit it and copy it, if my license allows it. I disagree that it's unethical if the license is open and the knowledge public: nobody and nothing in the license expressed any wish otherwise.
And besides just legal argument, imitation is the sincerest admiration; if I develop a useful algorithm or collection of components, ethically I never consider that the algorithm "belongs" to me; I found a way to do something and if it liberates others to create things more easily, I think that's great. Moreover, with virtually infinite possibilities if a filter is done well, what is there to lose? There is more than enough "pie" for everyone; you can point to the other slices and boast that you helped that exist; this can in turn advertise your own work. After seeing these filter previews I was hoping I could toy with it; I'm disappointed I can't. I'll probably want to look for filters others have created with these principles, and if I find none, contrive and publish my own. It's too cool to pigeonhole with a silly mirage of "mine." You refer to the prophecy of the pastry that will bring balance to the Force. And you believe it's this...donut? |
|||||||
Posted: April 29, 2019 9:20 am | ||||||||
TheBeat
Posts: 165 |
earthbound, I am not sure if you're post is an answer to what I wrote. Of course I am willing to share, not only results, but also filters that I hopefully will make in the future. At the moment I am just experimenting and learning, nothing worth while sharing yet, but I am confident that will come.
The Stone Age did not end because they ran out of stones. Niels Bohr |
|||||||
Posted: April 29, 2019 9:59 am | ||||||||
Yuya
![]() |
That was me about 10 years ago. That spirit of sharing info and getting ideas all over the place was my thing. Then I lost it when I began working in the industry. The reality sucks. I have to charge $50 to 60 an hour for my work. A lot of it goes to paying my life (medical bills, water, gas, electricity, loans, internet, phone line, a roof over my head), software, hardware etc. It's expensive to live for what I do. Even side projects need to be accounted for in my schedule since I'm constantly doing something in trying to keep my head out the water. I have to value work or I'm otherwise I'm going to miss payments. Reality sucks doesn't it?
"Infinite possibilities" based on parameters that have been carefully tweaked and tinkered to make sure it outputs something useful, meaningful, and beautiful. It's really not that "infinite". A great filter is carefully adjusted so you get what you want and the more complex the filter, the more time it takes to carefully adjust it. Yes you lose time. The thing I need to separate here is intent of its creation. I consider a lot of my personal filters to be "art" and that its output is a consequence of its creation. I never intended the public to have what I made. RGBLAND is the same, it's my artwork and I never thought about sharing the source generator with the rest of the universe. In its own right, RGBLAND is the artwork in itself. It's just not a thing where I can dump it on a public forum and hope that everybody will use it with fairness, education, and understanding in mind.
My friend, I have to be very stern here, this is basically the issue: You're saying that exposure will pay my bills or that my name will be spread all over the place. That having more people use a generator I made increases the chances of me being paid off at some point. That's absolutely abhorrent and wrong. In the 10+ years of me making free filters not one has returned me hard-cash that I can use to pay my bills. Please understand that I don't work for fame or popularity or LIKES, I work for a living. No professional works like that and it's totally insulting to designers and artists who work insane hours to deliver high-quality artistry to this world. I'm disappointed that I had to deal with that post. Absolutely insulting. |
|||||||
Posted: April 29, 2019 11:39 pm | ||||||||
Yuya
![]() |
Now shall we brighten the day with the original topic? I'm going to go 180-degree turn here and lighten up. I hate writing stuff like that and it sucks to talk about reality it's off topic anyway. So from here on down let's just not talk about that and focus on the "image synth".
I did mention I'll probably work on a lite-version of the image synth on a no-promises basis. After writing my initial response I was thinking "this is a simple filter" in comparison to everything else I've made for myself. In general the principle is the stuff discussed above and it doesn't take too many brains to accomplish it. At around 4 am I finished my job and decided "hell, I'm awake, why don't I just make RGBland or something" and decided to drill on making something. So here's RGBLAND - UltraLite edition that kinda outputs random but overall "ok" looking images. You can really take this principle everywhere and make some neat things with it. RGBLAND - UltraLite.ffxml |
|||||||
Posted: April 30, 2019 1:40 am | ||||||||
Yuya
![]() |
There are two motivations for making this filter. Firstly, I wanted something that just generated a gajillion patterns. Secondly, I wanted it to make a filter that goes making surprises but predictable and not look like s**t. The point here is that each of these outputs should look designed but somewhat still random enough.
The Ultra-Lite edition isn't the original RGBland stripped down but it's entirely a new filter just working on the same exact principles I've described here. In general while it's different it's easy to just add even more on this and make your own insane node-tree spaceship creation. You'll need to understand 4 key elements: 1. The source (A simple gradient that's manipulated by the profile gradient). 2. A randomizer - These output black and white random signals (HDR too) that's used to drive the green input parameters. This effectively removes the need for controls and keeps everything in-filter hence allowing you to just smash that Next-variant button instead of wondering what each of the darn sliders do. 3. Signal operators - Each channel R G and B signals have their own thing going. Each is randomized which results in a lot of variation but since all signals share the same exact string of operations, the result doesn't look like crap. 4. The assembly - Try swapping this with other assembly nodes for fun. I personally like RGB the most only because it gives you the most variance. The others such as HSL gives you rainbows and all sorts of other effects. Here are some useful things: - Make sure each signal uses the exact same flow. Reordering or throwing in something different per channel works but I tend to stick to the same tree to avoid one channel fr om looking too different from others. - The Randomizers are capable of producing HDR output. This is because some of the parameters such as the WAVE curve generator has a section wh ere I can use a value of 4 to drive the wave type parameter. - To avoid really noisy images avoid just adding noise on noise on noise. Use low values for parameters that can quickly go awry. i.e. Frequencies, Repeat H/V. ![]() |
|||||||
Posted: April 30, 2019 1:59 am | ||||||||
TheBeat
Posts: 165 |
I agree. Artists are hardly rewarded enough, and the exceptions being grossly overpaid (in my opinion people like Jeff Koons). I can understand and sympathize with what Yuya is writing. If someone makes an excellent filter, for me that is making art as well. A good filter can be used for the makers own work, or be shared and hopefully getting some recognition. BUT it is always the maker's choice and we should respect if someone wants his or her filter to be only for own work. The difference between traditional Art and the technical side (e.g. programming) is getting smaller and smaller. I applaud that. As well with Art as with programming, you need a good basic (new) idea. How it is worked out is secondary for me. I feel the willingness of Yuya to help others, beginners like me, and I really appreciate that Yuya has reached out to all of us, by sharing a simpler version of his RGBland Filter. Thanks a lot, Yuya ![]() It is a pity that our existences are so dominated by having to earn money. That way reality does suck. The Stone Age did not end because they ran out of stones. Niels Bohr |
|||||||
Posted: May 1, 2019 8:57 am | ||||||||
TheBeat
Posts: 165 |
I did not have time to look into Yuya's RGBland filter, but just before it fell out of heaven
![]() The Stone Age did not end because they ran out of stones. Niels Bohr |
|||||||
Posted: May 1, 2019 9:01 am | ||||||||
Yuya
![]() |
I say keep drilling with what you got so far (whatever that is). Generally these explorations should give you a more conceptual understanding of the program than something where the end-result drives your goals.
The point of RGBLand wasn't really to have a specific end result but was more like "what if I do this? and that?" By doing so it kinda gives you further understanding which can be used as fuel for other ideas that require more technical knowledge. ![]() |
|||||||
Posted: May 2, 2019 9:42 am | ||||||||
TheBeat
Posts: 165 |
Yuya, I totally agree. It's great fun just to explore. And I learn a lot Thanks, Beat The Stone Age did not end because they ran out of stones. Niels Bohr |
|||||||
Posted: May 3, 2019 3:49 pm | ||||||||
earthbound |
I never said anything would pay your bills. Where I wrote:
-- you can reasonably wonder that I might mean it could lead to income. But you have assumed that is my intent or meaning, and that is a false assumption. My intent in creating art is only to create beautiful or interesting things, etc. The extent of any goals I have in advertising is to that end. I'm not necessarily in it for money or followers or "likes." That being my pretext, I assume it could be yours. It sounds like in many respects it can't be or isn't, but in no way does that mean that I insult you. I simply didn't know that for you making art without profit is a very restrictive option. I did not know that I lobby contrary to your goals, and in retrospect of course a lot of people have similar goals to yours, so I suppose in retrospect I could show more understanding of that. I'd really like it if your unpleasant reply and this my necessary reply were deleted, though. You refer to the prophecy of the pastry that will bring balance to the Force. And you believe it's this...donut? |
|||||||
Posted: May 4, 2019 9:58 pm | ||||||||
earthbound |
Thanks for sharing this filter!
You refer to the prophecy of the pastry that will bring balance to the Force. And you believe it's this...donut? |
|||||||
Posted: May 4, 2019 9:59 pm | ||||||||
TheBeat
Posts: 165 |
Yuya, I have looked at your RGBland-UltraLite filter. I learned a thing or two
![]() Question : in the many groups (all the same) that do the randomizing, you use first two LookUp components after the Perlin noise (with different x and y coordinates) which fetch a different shade of grey. Then their outputs are fed into a new Lookup. Why is this different from just using one Lookup, as the the Variation of the Perlin will be changed every time with a new Variant? When I select a new variations (Variation slider of the Perlin Noise), I see changes in the first Lookup as often as in the final Lookup. I am sure you have a very good reason to do it your way, but I don't see it ![]() Beat The Stone Age did not end because they ran out of stones. Niels Bohr |
|||||||
Posted: May 6, 2019 9:47 am | ||||||||
Yuya
![]() |
I'd love an idealized world where I can talk about beauty and interesting things and just live off of that. And really, I held that in my philosophy for years. To be fair it's a fantastic philosophy to hold onto. You gotta keep that somewhere even as you're doing art for a living since often times resources are what kinda connects people around. Frankly though, the one thing that got me more upset was the fact that you wrote this:
I wished you tried what I suggested first and I wished you'd post some results off of your own experiments. Plus I kinda explained it in the original video as well as shown some pretty clear idea of how the filter works via screenshots. Albeit simple, it did show the core structure of the original RGBLAND. The point here is: the filter's really really simple at its core. Expert or beginner a few hints here and there and some bit of exploration would have lead you to differing but similar results. Then it'd be just about designing those results to your liking and I could have guided you on that to make it even more fun. I can only sincerely hope you'd try starting your own version of things here. Post some cool results. I'd love to see those than writing these lengthy responses. lol |
|||||||
Posted: May 6, 2019 10:06 am | ||||||||
earthbound |
Well, my intent in toying with it is to experiment with it hands-on to see how it works and then either make my own different filter(s) that incorporate some ideas (or derivatives of ideas) from it, or modify it enough that it's substantially different (unique). So I think my intent is in line with what you'd hope for, but that it was easy to lose or not express that intent in writing.
In other words, I learn hands-on, in addition to via clear descriptions (but those help!) and seeing clear videos. So experiments like you suggest are certainly on my list. (My list of creative things to try is long though.) You refer to the prophecy of the pastry that will bring balance to the Force. And you believe it's this...donut? |
|||||||
Posted: May 7, 2019 9:46 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!
28 unregistered users.