14:20 - you eluded to an interesting challenge. You should create a program that creates a font based off of this noise so that each time you write down the letter "o" for example you'll get one of these subtly off circles that is constantly different, and then extrapolate that to the entire alphabet so when you type in your keyboard you get the computer "handwriting" your sentence. Each letter would look unique, similar to handwriting, whereas fonts each letter is made exactly equal.
Well, this would be great in a video game or something like that, you wouldn't be able to use it as easily for messenger apps, of course you could have it to where it has each person see a different variation of the same text, and no one would likely notice.
Love your channel. At first it felt to me as "obnoxious youtuber tries to teach basic programming techniques to people who never tried programming". But later you did such an amazing job with some advanced algorithms like quadtree and some machine learning you've earned a lot of my respect. Keep up the good work, dude!
Unbelievable, i needed this kind of thing right now for a project i'm working on, i was going to watch back your asteroids video cause i remember from there something similar, all in all...perfect timing, you are the chosen one!
Love from France Dan! You are so inspiring and positive. I began coding thanks to you and it was the best thing that could happen to me. Merci beaucoup!!!
Dude, I am watching your videos for a while now, and i have never seen someone with this kind of sunshine-happyness while coding. Your videos are great! Keep it like this. Awesome work!
Never realized this was it's own concept. Used a very similar approach about 6 or 7 years ago to test attenuating light geometry. The renderer I was working on used calculated geometry to represent lights, everything from spheres, cones, toriods to exotic super shapes. After implementing improved perlin noise in HLSL, up to 4 dimensions I was itching to apply it all over the place in the renderer. One of the first things I did was sample noise space at each vertex of the geometry for the lights and passing that into my lighting equations as an attenuation of the radius. Made lights have the familiar imperfection we see in real life and when animated across time, made for very convincing incandescent light sources even though I kept the attenuation so small that the final image showed no more than a delta of 10 in RGB values. Didn't even occur to me while doing that, that I was playing with a rather interesting usage of perlin noise. Then again are there really any non-interesting uses of perlin noise?
This is fantastic. Cyclic noise feels like the kind of thing a lot of people have wanted and I didn’t realize it existed without summing some sort of time-shifted noise function or some other mumbo jumbo. This is a super elegant solution! I’m wondering if circles would be any better than, say, a square in this context. Surely any closed loop would have this property, but intuitively it feels like a circle is the “most random,” even though something like a square is less computationally taxing. Also, it would be interesting to see what would happen if you put multiple shapes in a perlin noise “field” that acted as a displacement map. If you varied each vertex in a shape by the value of the noise at that vertex, i’d imagine you’d get a really cool effect by moving that shape around or varying the field in 3d. Anyways, great video!
What an awesome video for inspiration! You could theoretically make a 3d random or self-avoiding walk through spherical coordinates (or a 2d polar random walk) and use that as the coordinates for your perlin noise to be able to just have so many random loops
@Manan Karnik No, it is *not*. IN processing and p5js at the very least it returns a fractal noise - that is NOT Perlin noise. Fractal noise is a form of noise generated by adding several different noises together. In case of processing/p5js that becomes apparent when you look at the "NoiseDetail" function: It set the number and falloff for the used noise-layers. Perlin Nosie itself is a very regular grid-based noise with uniform feature size and smooth transitions. Adding several layer of that with decreasing size and amplitude gives the nice rugged result that resembles a terrain.
According to the documentation, it's Perlin noise. processing.org/reference/noise_.html It is of course possible that the documentation is wrong. If it is, it's understandable where the confusion comes from.
Great video! You inspired me to make Snake (the game) in JavaScript. The twist is that i am not using any form of grid other than just the loose pixels... It's really close to being done. Thanks again for the inspiring video's
Was struggling with this problem, thanks for the solution. Tried this method for my music visualizations and it worked well. Now I think I'll try to improve it by adding a sound reacting component to make it interactive.
The segment at 16:34 looks like the continent of Antarctica is morphing with Earth's rotation. Trippy. Another cool way to implement 3+ dimensions is to make the extra dimensions color based or sound based. Then a rotation/translation/etc., would change colors and sounds.
This is a really fun one to add alpha clearing to, so you get blurred lines as the shape morphs. So instead of background(0) you would use color(0,0,0,25); rect(0,0,width,height); and this will draw a rectangle over the entire screen that is slightly faded out and because these are layered on top of each other each previous generation of line is more blurred out and so slowly transitions to black. Hopefully this makes sense, not sure how to format code in comments.
As well as changing the Z-Offset, could you also move the centre of the circle (that is sampling values in 2D perlin noise)? You could also play with that circle in other ways... Maybe use an oval and slowly change it's shape over time?
18:03 in he plays the cards and lures you in with that oneliner, your passion for code ignites "if......the creative possibilities expl...." o yeszz :)
First of all, you rock. Thank you for your videos. As a newbie, I feel overwhelmed, but inspired by your videos. I guess it's like someone learning to play the drums and watching Neal Pert thinking, 'I have a long way to go' and 'man, that's awesome' all at the same time. What is your advice to the noob? I am currently more than half way through the codecademy full stack web development course and plan to finish that, but was curious to hear from the community what "best practices" are recommend for new guys/gals.
I was thinking about whether you could make a seamless 2-dimensional texture (there was a video were you used Perlin noise to create a vector field and had particles flow through it, but it had artefacts at the edges where the vector directions would change abruptly). My first thought was to use the surface of a 3D torus to give a looping 2d texture in the same way you're using the edge of a 2D circle to generate cyclic 1D noise here. Then I twigged that the would cause distortion in the texture, since the outer edges of the Torus would sweep through a larger length of the noise pattern than the inner surface for the same angle. You could minimise it by turning the minor circle so it's tangent to the major circle rather than radial (so your torus would be flattened as if you sucked all the air out of it) but it'd still have a slight distortion to it unless you actually projected the circle onto the surface of a cylinder (sounds complicated, but probably just needs the coordinate calculations to come in the right order). Part of me wonders if it would be simpler just to write a Perlin noise class that interpolates smoothly between the edges, but that would limit the scale of the noise to integer fractions of the screen space, and might cause issues if you wanted to create images with odd aspect ratios.
I don't know what you found here, but I'm still studying it. I think you found something extremely important on accident. That phase thing is incredible!
As the phase changes, the orbit of (cos(A+phase), sin(A)) over A morphs as follows: Phase=0: counterclockwise circle Phase=pi/2: oscillating along the straight line between (1,-1) and (-1,1) Phase=pi: clockwise circle Phase=3pi/2: oscillating along the straight line between (-1, -1) and (1,1) And between those, it is various elliptical shapes.
What if, when moving through 3D Perlin space, instead of looking at circles in x,y with increasing z, you looked at circular cross-sections of a torus? Then it’d definitely loop :D
12:48 No, you couldn't just split up a into two different variables and increment them in another way. Then you wouldn't end up at the point where you started in the 2-dimensional perlin noise space. Or I just misunderstood what you said you could do.
It looks like the perfect technique for modeling a realistic planet. Just apply the exact same technique to spherical coordinates instead of polar coordinates. Maybe add a couple octaves of additional noise or an underlying sine wave pattern to help create larger scale areas like continents. At any rate, it certainly gives me a better idea of where to start than anything else I've found. Thank you.
Do you remember your Perlin noise example with moving particles? Please, try to make the same staff but let particles move like drawing contours of input image.
This is crazy. Last week i coded a wobbling blob of space jelly in a perlin noise loop for my space shooter and mine DOES have this visual artefact. Now I can fix it
During your explanation I had another idea. What if we take 1D perlin noise, then take the end points, draw a line between them, and then shear the entire graph, so the line stays perfectly flat. Essentially, when we take the noise value, we subtract some amount from it. That amount ranges from the start point value at t=0 and the end point value at t=1. Just simulating it in my brain, I don't think it would be noticeable
Is there any possible chance you can arrange to teach professors and teachers how to make animations like this? Your graphics really help to understand Mathematics and Physics. I think it would help quite a lot in the class room!
Just a thought. If the problem is creating some function(a) = b. Could there exist a Problem-Focused-Number-System? I mean since there is more than one number system. There has to be a reason. Like there are more than one problem. All a little different. So maybe there exist a way to adapt to each problem that could arise in the creation path. Why should the system be set from the beginning is what I guess. Then if a machine learning model() can classify a problem it can jump between problem solutions.
Continuously good videos. Can you do one with the polar noise mapping seamlessly to a sphere. it is tricky as you change the radius the polar mapping the noise texture distorts.
What if you moved where you were getting your perlin noise valises from, instead of moving around the circle? Say, having the circle trace around another circle in the perlin plain? Seems like this might be a cool way to add some more “randomness” to the space without having repeats before the end.
seems like the shape of these perlin noise values might be different than the 1 dimensional ones. and depending what angle it might be a bigger or smaller step. Not 100 percent sure.
So. You use one noise loop to make sure that the circle is perfectly closed. At the end you also introduced the possibility of a z-offset. Could you use 4 dimensional perlin noise, use dimensions 1 and 2 for the perfectly closed circle, and move around dimensions 3 and 4 in another circle with a z-offset and w-offset so that it creates a perfectly looping animation?
If you want a repeating pattern why not use a sum of sinewaves with period of 1, 1/2, 1/3 ... etc? You will have way more controls (amplitudes and phases for each wave), not just xyz for perlin. It will also have better resolution.
@@TheCodingTrain not even "better" resolution. Basically infinite resolution. At least as far as floating point comes if sin/cos is implemented correctly. Depends on compiler settings :)
I am new to Perlin Noise, does it have a probability to generate the same ?? lets say random order, even not immediately but after running some time ????
While moving through 3D perlin map with zoff, does it loop? Because now that we are moving in a straight line on the z axis, the values will not repeat their initial values
the question is why it has to be a 2D perlin noise instead of a simple 1D perlin noise which automatically starts and ends at the same point... you can just interpolinate the perlin noise array after replacing the last value with a copy of the first value. The first and last component will always stay the same that way. (for reference check out javidx9's video about perlin noise...)
He is proofing in the video that 1d perlin noise is not what he needs, he is simply circling around by the x offset in the 2D perlin noise and taking the y off set that it would match up
@@omereli1062 no hes using a predefined noise function. I was refering to a method which does the same while generating the noise. By interpolination. If you have a random Array and put the same Value at the start and beginning and interpolinate it you end up with a perlin noise that starts and ends at the same point.
I have said this on several of your other videos - but this is *NOT* perlin noise. Processing, p5js and several others might call it that, but this is a form of fractal noise - several layers of smooth noise of different frequencies added on top of each other. Also - Perlin noise is inherently closed. If you set the correct scale you automatically get your desired perfect closing loop. This also works for creating 2D perlin noise and making it tileable.
Thanks for the feedback! From what I understand, the implementation in Processing is "classic Perlin noise" from 1983. I agree that I could be better about being clear about this. Actually I'd love to make a video that shows the differences between this classic "VALUE" (?) method vs Simplex vs Perlin. It would be wonderful to have a Processing library with these different implementations. Maybe this can be a Coding Train project!
@@TheCodingTrain " From what I understand, the implementation in Processing is "classic Perlin noise" from 1983." the noise-function of processing is giving you a fractal noise. It might very well be using a form of Perlin-noise internally to generate the octaves, but what you get is certainly no longer Perlin.
This dude resonates positive energy.
radiates maybe
Truly earns his title of "Bob Ross of programming".
his new title you just assigned to him.
Which I like
random = happy accidents
"Blob Ross"
Sorry i had to
14:20 - you eluded to an interesting challenge. You should create a program that creates a font based off of this noise so that each time you write down the letter "o" for example you'll get one of these subtly off circles that is constantly different, and then extrapolate that to the entire alphabet so when you type in your keyboard you get the computer "handwriting" your sentence. Each letter would look unique, similar to handwriting, whereas fonts each letter is made exactly equal.
Love this idea! You should make it!
Well, this would be great in a video game or something like that, you wouldn't be able to use it as easily for messenger apps, of course you could have it to where it has each person see a different variation of the same text, and no one would likely notice.
Love your channel. At first it felt to me as "obnoxious youtuber tries to teach basic programming techniques to people who never tried programming". But later you did such an amazing job with some advanced algorithms like quadtree and some machine learning you've earned a lot of my respect. Keep up the good work, dude!
It's dark in the room, quiet. Daniel stirs slightly in his slumber.
Suddenly he is sat bolt upright!
"Perlin Noise!"
It's never dark in the room whenever Daniel's around.
lol this sounds like a creation myth
The edits in this video are sooooo good!
Unbelievable, i needed this kind of thing right now for a project i'm working on, i was going to watch back your asteroids video cause i remember from there something similar, all in all...perfect timing, you are the chosen one!
Love from France Dan! You are so inspiring and positive. I began coding thanks to you and it was the best thing that could happen to me. Merci beaucoup!!!
I just started making an Asteroids game today! Loved the video! Thanks, Dan!
Dude, I am watching your videos for a while now, and i have never seen someone with this kind of sunshine-happyness while coding. Your videos are great! Keep it like this. Awesome work!
Perlin noise is so awesome, I love learning more about it from this channel
Never realized this was it's own concept. Used a very similar approach about 6 or 7 years ago to test attenuating light geometry. The renderer I was working on used calculated geometry to represent lights, everything from spheres, cones, toriods to exotic super shapes. After implementing improved perlin noise in HLSL, up to 4 dimensions I was itching to apply it all over the place in the renderer. One of the first things I did was sample noise space at each vertex of the geometry for the lights and passing that into my lighting equations as an attenuation of the radius. Made lights have the familiar imperfection we see in real life and when animated across time, made for very convincing incandescent light sources even though I kept the attenuation so small that the final image showed no more than a delta of 10 in RGB values. Didn't even occur to me while doing that, that I was playing with a rather interesting usage of perlin noise. Then again are there really any non-interesting uses of perlin noise?
The edits were so fun!
Very inspiring. Thank you. Once again, I need to go and program my brains out.
Kudos to the subtitler! Great work!
You're such an amazing teacher. I finally understand what 2D perlin noise means.
Im 0 seconds in to the video and the like button is already pressed. Love your videos!
I am so glad you finally did this! I have been wondering about how to fix that blob for over a year now!
This is fantastic. Cyclic noise feels like the kind of thing a lot of people have wanted and I didn’t realize it existed without summing some sort of time-shifted noise function or some other mumbo jumbo. This is a super elegant solution!
I’m wondering if circles would be any better than, say, a square in this context. Surely any closed loop would have this property, but intuitively it feels like a circle is the “most random,” even though something like a square is less computationally taxing.
Also, it would be interesting to see what would happen if you put multiple shapes in a perlin noise “field” that acted as a displacement map. If you varied each vertex in a shape by the value of the noise at that vertex, i’d imagine you’d get a really cool effect by moving that shape around or varying the field in 3d.
Anyways, great video!
You channel is always so exciting! Thanks for all the great content over the years.
As always You made it very clear :).thanks.
I love how human and practical your code is. We don’t have to talk about the semi colons
What an awesome video for inspiration! You could theoretically make a 3d random or self-avoiding walk through spherical coordinates (or a 2d polar random walk) and use that as the coordinates for your perlin noise to be able to just have so many random loops
Man, I like emotion you put on this, you got a new subscriber.
Perlin noise is always positive
That made me really happy :)
I love how I found this video and it actually helped me solve a problem I was working on 😅
Best teacher on the internet
Polar Perlin Loops, part of a balanced developer’s diet.
edit: oops I did a mispelliginationating but its fixed now
But he didn't use Perlin noise.
@Manan Karnik No, it is *not*.
IN processing and p5js at the very least it returns a fractal noise - that is NOT Perlin noise.
Fractal noise is a form of noise generated by adding several different noises together. In case of processing/p5js that becomes apparent when you look at the "NoiseDetail" function: It set the number and falloff for the used noise-layers.
Perlin Nosie itself is a very regular grid-based noise with uniform feature size and smooth transitions. Adding several layer of that with decreasing size and amplitude gives the nice rugged result that resembles a terrain.
@@ABaumstumpf It was a joke k dude. Calm.
According to the documentation, it's Perlin noise. processing.org/reference/noise_.html
It is of course possible that the documentation is wrong. If it is, it's understandable where the confusion comes from.
Great video! You inspired me to make Snake (the game) in JavaScript. The twist is that i am not using any form of grid other than just the loose pixels... It's really close to being done. Thanks again for the inspiring video's
How does he have all of that energy
Keep up your work!
Was struggling with this problem, thanks for the solution. Tried this method for my music visualizations and it worked well. Now I think I'll try to improve it by adding a sound reacting component to make it interactive.
EXACTLY the information I was looking for! Thank you so much
you're a national treasure my friend 10/10
The segment at 16:34 looks like the continent of Antarctica is morphing with Earth's rotation. Trippy. Another cool way to implement 3+ dimensions is to make the extra dimensions color based or sound based. Then a rotation/translation/etc., would change colors and sounds.
Please try this!
I swear Mr Shiffman, you acknowledge every comment. Thanks for another intriguing video.
I do miss a few!
This is a really fun one to add alpha clearing to, so you get blurred lines as the shape morphs. So instead of background(0) you would use color(0,0,0,25); rect(0,0,width,height); and this will draw a rectangle over the entire screen that is slightly faded out and because these are layered on top of each other each previous generation of line is more blurred out and so slowly transitions to black. Hopefully this makes sense, not sure how to format code in comments.
You would be a great teacher! I can tell it on the way you leave the mouse in the middle of the images you want to show.
he is a *professor*
@@prakharlondhe3876 and this was a *joke*
This guy has crazy math/science teacher vibes
As well as changing the Z-Offset, could you also move the centre of the circle (that is sampling values in 2D perlin noise)?
You could also play with that circle in other ways... Maybe use an oval and slowly change it's shape over time?
18:03 in he plays the cards and lures you in with that oneliner, your passion for code ignites "if......the creative possibilities expl...." o yeszz :)
this channel is a real inspiration for stuff to do lol
You can get a 2d tileable perlin noise texture by cutting a thorus through 3d perlin noise
Love from Italy ❤️♥️🇮🇹, #LoveFromItaly5
I love this !
First of all, you rock. Thank you for your videos. As a newbie, I feel overwhelmed, but inspired by your videos. I guess it's like someone learning to play the drums and watching Neal Pert thinking, 'I have a long way to go' and 'man, that's awesome' all at the same time. What is your advice to the noob? I am currently more than half way through the codecademy full stack web development course and plan to finish that, but was curious to hear from the community what "best practices" are recommend for new guys/gals.
This dude is the best!
Your videos always inspires me ♥
I was thinking about whether you could make a seamless 2-dimensional texture (there was a video were you used Perlin noise to create a vector field and had particles flow through it, but it had artefacts at the edges where the vector directions would change abruptly). My first thought was to use the surface of a 3D torus to give a looping 2d texture in the same way you're using the edge of a 2D circle to generate cyclic 1D noise here.
Then I twigged that the would cause distortion in the texture, since the outer edges of the Torus would sweep through a larger length of the noise pattern than the inner surface for the same angle. You could minimise it by turning the minor circle so it's tangent to the major circle rather than radial (so your torus would be flattened as if you sucked all the air out of it) but it'd still have a slight distortion to it unless you actually projected the circle onto the surface of a cylinder (sounds complicated, but probably just needs the coordinate calculations to come in the right order).
Part of me wonders if it would be simpler just to write a Perlin noise class that interpolates smoothly between the edges, but that would limit the scale of the noise to integer fractions of the screen space, and might cause issues if you wanted to create images with odd aspect ratios.
good video Dan!
I don't know what you found here, but I'm still studying it. I think you found something extremely important on accident. That phase thing is incredible!
As the phase changes, the orbit of (cos(A+phase), sin(A)) over A morphs as follows:
Phase=0: counterclockwise circle
Phase=pi/2: oscillating along the straight line between (1,-1) and (-1,1)
Phase=pi: clockwise circle
Phase=3pi/2: oscillating along the straight line between (-1, -1) and (1,1)
And between those, it is various elliptical shapes.
I use 3d fractal noise to make mountains on the 2D surface of planets in my game I'm making. You're giving away my secrets!
What if, when moving through 3D Perlin space, instead of looking at circles in x,y with increasing z, you looked at circular cross-sections of a torus? Then it’d definitely loop :D
This guy is amazing!
12:48 No, you couldn't just split up a into two different variables and increment them in another way. Then you wouldn't end up at the point where you started in the 2-dimensional perlin noise space.
Or I just misunderstood what you said you could do.
Anyone else remember the debate about whether it was polar noise, perlin noise, polin noise or perlar noise?
It looks like the perfect technique for modeling a realistic planet. Just apply the exact same technique to spherical coordinates instead of polar coordinates. Maybe add a couple octaves of additional noise or an underlying sine wave pattern to help create larger scale areas like continents. At any rate, it certainly gives me a better idea of where to start than anything else I've found. Thank you.
Chaotic good coding.
Nice editing!
Omg, that's so amazing
Hey I just found your channel and I love it.
Do you remember your Perlin noise example with moving particles? Please, try to make the same staff but let particles move like drawing contours of input image.
This is crazy. Last week i coded a wobbling blob of space jelly in a perlin noise loop for my space shooter and mine DOES have this visual artefact. Now I can fix it
During your explanation I had another idea. What if we take 1D perlin noise, then take the end points, draw a line between them, and then shear the entire graph, so the line stays perfectly flat.
Essentially, when we take the noise value, we subtract some amount from it. That amount ranges from the start point value at t=0 and the end point value at t=1.
Just simulating it in my brain, I don't think it would be noticeable
Okay, the gag at 6:02 is hilarious. ¯\_(ツ)_/¯
Awesome! Finally fixed the broken blobby!
That is my favorite video of yours
wonderful wonderful wonderful wonderful wonderful wonderful wonderful wonderful wonderful wonderful wonderful wonderful
Actually, negative values for xoff & yoff are supported in the p5 Perkins noise function
I’m here for the hands gestures.
Nice video!! I challenge you to make a drawing predictor (simple neural network that calculates the next points locations from few starting points)
Great video!
Is there any possible chance you can arrange to teach professors and teachers how to make animations like this? Your graphics really help to understand Mathematics and Physics. I think it would help quite a lot in the class room!
Can you explain how to code the *map* and *noise* functions?
Just a thought. If the problem is creating some function(a) = b. Could there exist a Problem-Focused-Number-System? I mean since there is more than one number system. There has to be a reason. Like there are more than one problem. All a little different. So maybe there exist a way to adapt to each problem that could arise in the creation path. Why should the system be set from the beginning is what I guess. Then if a machine learning model() can classify a problem it can jump between problem solutions.
/** 적외선 센서에 물체가 감지되면, 시리얼 모니터에 "Detected"라는 문장을 출력*/int sensor = A0; // 센서핀은 A0번에 연결int val;void setup() { Serial.begin(9600); pinMode(sensor, INPUT); // 센서값을 입력으로 설정 Serial.println("arduino starts");}void loop() { val = digitalRead(sensor); // 센서값 읽어옴 if (val == LOW) { // IR센서는 LOW ACTIVE로 탐지 시 LOW값을 전송함 Serial.println("Detected"); delay(300); } else Serial.println("0"); delay(300);}
*give me the polar perlin noise LÖÖPS brother*
Continuously good videos. Can you do one with the polar noise mapping seamlessly to a sphere. it is tricky as you change the radius the polar mapping the noise texture distorts.
Oh I would love to try this!
What if you moved where you were getting your perlin noise valises from, instead of moving around the circle? Say, having the circle trace around another circle in the perlin plain? Seems like this might be a cool way to add some more “randomness” to the space without having repeats before the end.
how do you make the noise loop follow a certain shape like the heart you showed in the start
seems like the shape of these perlin noise values might be different than the 1 dimensional ones. and depending what angle it might be a bigger or smaller step. Not 100 percent sure.
So. You use one noise loop to make sure that the circle is perfectly closed.
At the end you also introduced the possibility of a z-offset. Could you use 4 dimensional perlin noise, use dimensions 1 and 2 for the perfectly closed circle, and move around dimensions 3 and 4 in another circle with a z-offset and w-offset so that it creates a perfectly looping animation?
Just amazing
If you want a repeating pattern why not use a sum of sinewaves with period of 1, 1/2, 1/3 ... etc? You will have way more controls (amplitudes and phases for each wave), not just xyz for perlin. It will also have better resolution.
Great suggestion!
@@TheCodingTrain not even "better" resolution. Basically infinite resolution. At least as far as floating point comes if sin/cos is implemented correctly. Depends on compiler settings :)
21:35 that shooting sound resonates in my nose
A finical note is that these noise additions are not exactly 1d Perlin noise. It was not stated though. Anyway, nice video.
Now I want to see rolling Perlin noise
Do a next challenge on Hilbert Curves and other space filling curves :)
Amazing. Thank you so much
Wouldn't 4-dimensional Perlin noise allow you to loop the changing blob by using the same process that you use on x and y?
Yes!
This would be cool to use to generate random terrain
See: ruclips.net/video/IKB1hWWedMk/видео.html
I am new to Perlin Noise, does it have a probability to generate the same ?? lets say random order, even not immediately but after running some time ????
While moving through 3D perlin map with zoff, does it loop? Because now that we are moving in a straight line on the z axis, the values will not repeat their initial values
Never though noise would be that useful
Welcome back to Polar Perlin Noise Loop Thing of a Blob
Perfect spin loader
thank you for your help:)
The mister rogers of computer science.
the question is why it has to be a 2D perlin noise instead of a simple 1D perlin noise which automatically starts and ends at the same point...
you can just interpolinate the perlin noise array after replacing the last value with a copy of the first value. The first and last component will always stay the same that way.
(for reference check out javidx9's video about perlin noise...)
He is proofing in the video that 1d perlin noise is not what he needs, he is simply circling around by the x offset in the 2D perlin noise and taking the y off set that it would match up
@@omereli1062 no hes using a predefined noise function. I was refering to a method which does the same while generating the noise. By interpolination. If you have a random Array and put the same Value at the start and beginning and interpolinate it you end up with a perlin noise that starts and ends at the same point.
@@TheHyperplayer he is using a build up perlin noise function I'm not sure if it can do that
@@omereli1062 Well the method I was refering to needs a tweak to the Perlin Noise generation so a predefined Method might not work with it.
Oh interesting technique, thanks for the reference!
Please make visualization of blackhole using its formula, like interstellar
I have said this on several of your other videos - but this is *NOT* perlin noise.
Processing, p5js and several others might call it that, but this is a form of fractal noise - several layers of smooth noise of different frequencies added on top of each other.
Also - Perlin noise is inherently closed. If you set the correct scale you automatically get your desired perfect closing loop. This also works for creating 2D perlin noise and making it tileable.
Thanks for the feedback! From what I understand, the implementation in Processing is "classic Perlin noise" from 1983. I agree that I could be better about being clear about this. Actually I'd love to make a video that shows the differences between this classic "VALUE" (?) method vs Simplex vs Perlin. It would be wonderful to have a Processing library with these different implementations. Maybe this can be a Coding Train project!
@@TheCodingTrain " From what I understand, the implementation in Processing is "classic Perlin noise" from 1983."
the noise-function of processing is giving you a fractal noise. It might very well be using a form of Perlin-noise internally to generate the octaves, but what you get is certainly no longer Perlin.