Autochrome and Perlin Noise

Mustafa Umut Sarac · Jun 16, 2010

If you download and enlarge this image , you could see the starch grains.

A lots of people uses gaussian noise of the photoshop for to create a look like grainy autochrome but I think true color noise can be adapted by Perlin Noise.

Perlin Noise is what you see at shoot kill multi player games walls , ground everywhere. Realistic noise

It is used for all games and renderings and Perlin got academy award.

I dont know how you will use this information but it is very very important.

May be ALT printers will prepare their negatives by this way.

Or you can write a plug in to mimic the look of autochrome.

ALT people will enjoy and the digital ALT inkjet people also

Best ,

Mustafa Umut Sarac

Istanbul

Mustafa Umut Sarac · Jun 16, 2010

Thank you Charles for the image.

Umut

holmburgers · Jun 17, 2010

very interesting sir!

Mustafa Umut Sarac · Jun 17, 2010

If you want to print transparent filter look alike autochrome , stohastic distribution of noise will be rewarded also. But there is no such huge information at wiki and the wiki examples are not open to got the idea.
Game developer forum sites sometimes put examples.
I read dust , mud are used to make generation architecture look more realistic. I am remembering Tarkovsky , Stalker Bar and bedroom scenes and perlin noise creates this without artistic genius need.
Starch noise is so similar to this.

Mustafa Umut Sarac · Jun 17, 2010

http://www.noisemachine.com/talk1/index.html

Ken Perlin explains noise , I found that 3D modelling softwares gives a selection of textures to render on your object. May be not photoshop but MAYA or others would help to implement this.
Modelling softwares tutorials and forums are so rich and you can code everything.
May be you can find a algorithm and voila.
Well We are going in to 3D photography and new softwares will be definetely put this on the table.
I think APUG will switch to hologram printing from stereo pictures.
Well future will be very interesting really.

Kirk Keyes · Jun 17, 2010

Mustafa Umut Sarac said:
I think APUG will switch to hologram printing from stereo pictures.

You've have to go to the hybrid photo site to discuss that... ;^)

holmburgers · Sep 7, 2010

How could one make a screen-plate with a random RGB pattern (such as Perlin noise) that would mimic the irregular screen methods (like autochromes)? I personally have no idea and no software capable of doing such, but what would it take? I know that some examples are mentioned above, but I'm just re-asking the question to see if something can be done in practice.

Surely there's some CPU wiz that could pop out a screen in no time, right?!

Photo Engineer · Sep 7, 2010

Just like the printing method described in the other thread, but use a random number generator for the color. And, make the "dots" somewhat larger. On the Autochromes that I have compared with Dufaycolor, it seems that the Autochrome "dots" are larger but less noticeable than the regular Dufay "dots". OTOH, at least the Autochromes that I scanned look awful.

A large (4x5 or 5x7) Autochrome or Dufay plate looks better than a 2x3 plate at a normal viewing distance.

PE

holmburgers · Sep 8, 2010

Any ideas what programs are capable?

Photo Engineer · Sep 8, 2010

holmburgers said:
Any ideas what programs are capable?

I've mentioned before that Visual BASIC has this capability as does C++, both from Microsoft. VB is easier to program. I believe it is a free download from Microsoft now.

As for the program, you would draw a dot sequence on a "form" using the color and point commands and then you could print the form to the printer using transparency materials.

I have done this many times, and have seen the Dufay resau printed, but have not written such a program. I have written graphing programs though that do much the same.

PE

holmburgers · Sep 8, 2010

Thanks PE, sorry if I missed that this has been covered before. I'll see if I can't recruit my computer-savvy (nice way of saying geek! ) brother on this....

alexhill · Sep 8, 2010

Kirk Keyes said:
You've have to go to the hybrid photo site to discuss that... ;^)

You can do holography totally analog. I've never done it myself, but i've drooled over diy kits and Dead Link Removedbefore.

holmburgers said:
How could one make a screen-plate with a random RGB pattern (such as Perlin noise) that would mimic the irregular screen methods (like autochromes)? I personally have no idea and no software capable of doing such, but what would it take? I know that some examples are mentioned above, but I'm just re-asking the question to see if something can be done in practice.

Surely there's some CPU wiz that could pop out a screen in no time, right?!

If you are doing photogravures, then its very simple to add a random grain using an aquatint and quick etch. Its the rest of the process that will drive you nuts.

holmburgers · Sep 9, 2010

Yeah, that guy doesn't know what he's talking about (with all due respect )... unless lasers are somehow "digital", holography is no different than black and white photography.

As for the aquatint, that's interesting but I can't think of how that this could lead to three separate colors randomly distributed on a screen; nor do I need another process to drive me nuts! I'm very much intrigued by these types of printing methods though.

ntenny · Sep 9, 2010

What exactly characterises Perlin noise? I've looked around online a little and found mainly explanations of its uses and a few bits of code for generating it (mainly through spectral synthesis), but from reading the code it's not obvious to me what the underlying principle is.

I know a little bit about the uses of noise in the musical domain, where there's a sweet spot in the parameters for noise generation that makes stochastically generated music sound "just random enough". It wouldn't surprise me if there's a similar phenomenon in the visual domain, where some set of parameters can be used to make a "faux-autochrome" mask that looks neither excessively random nor excessively orderly.

-NT

Mustafa Umut Sarac · Sep 9, 2010

ntenny ,

I think , there is a easier way to decide your degree of noise. Otherwise you dive in to brain research on human perception of visual noise problem.
Easy way is to find the fractal degree of autochrome. Enlarge my book cover scan with a photocopier , divide it to squares and count the noise per square. When you find the ratio of noise count to square , you find the fractal degree. This is the most basic way to tell your fractal software to generate noise. May be you need to color seperate the book cover with a enlarger and count fractal count per base color.
Ultra fractal software is one of the best software to generate your noise and before you could ask help from their list. May be you can create noise on pictures with a code.

Best ,

Mustafa Umut Sarac

Istanbul

ntenny · Sep 9, 2010

Mustafa Umut Sarac said:
Easy way is to find the fractal degree of autochrome. Enlarge my book cover scan with a photocopier , divide it to squares and count the noise per square. When you find the ratio of noise count to square , you find the fractal degree.

This is not quite right---that's just counting the density of the noise but says nothing about distribution, which is critical to fractal dimension. But there are well-known ways to get an approximate fractal dimension from an image. (I say "approximate" because of course the dots on an autochrome aren't really fractal; they're a finite number of discrete points scattered on the area of the image. You're not trying to find the fractal dimension of the actual object, but the fractal dimension of a theoretical object of which the real one is a good approximation.)

It's easy to cook up noise of a specified fractal dimension using fractional Brownian motion (Mandelbrot sketches the algorithm in _The Fractal Geometry of Nature_, I believe, and one of the coffee-table fractal books from the 1980s has actual pseudocode).

What I'm trying to understand is, is noise with the right fractal dimension "Perlin noise" by definition, or is there some more specific characterisation?

-NT

Mustafa Umut Sarac · Sep 9, 2010

Nathan ,

Well , may be You are right . I got this idea from achitecture doctoral thesis which tries to model same architecture of some city street with based on square counting and different parameters like topography.
I gave same reaction with yours when I read this thesis.
Perlin noise awarded by academy to mimic same natural noise on natural things like mud , dirt etc.
I think Ken Perlin did a similar dimension research and also distribution research on nature.
But as I said before , Perlin noise can fall in to Brain research also. So it makes everything more complex.
I am remembering a photoshop plug in , extracts full image from extremelly dark image. I found the real power of digital there. It was about the eye sensor cells and their working mechanism.

May be this helps. You can contact with Ken Perlin .

http://cs.nyu.edu/~perlin/

Please share whatever you learned.

Umut

Mustafa Umut Sarac · Sep 9, 2010

I am not expert but another idea is to train a nonlinear neural networks code with autochrome prints and prepare a catalog and apply to your prints. Whatever you like is yours.

Another research is from MIT and preparing a phase space of different prints and extract a function. This can be powered with many variables.

Max MSP may be works if it does not only for sound.

Matlab is another tool.

Umut

htmlguru4242 · Sep 10, 2010

holmburgers - In relation to your computer-generated reseau program - If I have time this coming week, I'll throw together a random-color image generator in Python. It's probably easier to make something that spits out an image file of some kind that can then be taken into Photoshop and printed.

holmburgers · Sep 10, 2010

htmlguru4242 said:
holmburgers - In relation to your computer-generated reseau program - If I have time this coming week, I'll throw together a random-color image generator in Python. It's probably easier to make something that spits out an image file of some kind that can then be taken into Photoshop and printed.

Woohoo!

That's great htmlguru4242, I wondered if you were lurking in the shadows somehwere. I remembered from old discussions that you had made a reseau in the past.

Anyways, if you do make one that will be great. I think there'll be a few people (or more) that would put it to use; myself included.

alexhill · Sep 10, 2010

holmburgers said:
Yeah, that guy doesn't know what he's talking about (with all due respect )... unless lasers are somehow "digital", holography is no different than black and white photography.

As for the aquatint, that's interesting but I can't think of how that this could lead to three separate colors randomly distributed on a screen; nor do I need another process to drive me nuts! I'm very much intrigued by these types of printing methods though.

The aquatint idea would be for a cmyk 4 plate photogravure, and is actually part of the process. I am no expert on photogravure, and my attempts have failed miserably and at quite a cost

holmburgers · Sep 10, 2010

alexhill said:
...and my attempts have failed miserably and at quite a cost

As they would say in the sitcoms.... "aaawwwww"

alexhill · Sep 10, 2010

Haha

holmburgers · Sep 21, 2010

Ok,

My bro whipped this together. Not sure in what program... maybe Perl, maybe just PS, maybe both, IDK! It's gaussian noise with a stained glass filter applied.

Now, obviously it's not right because it consists of too many colors, but if you could limit the elements to just RGB, I think this noise pattern and effect would work well. My brother said "well it'd be interesting to see what it would do to use all the colors." HAH! What an amateur!

There are many historical precedents for people who thought that a rainbow screen-plate would somehow produce better color than just RGB. Unfortunately, these folks have no understanding of color theory. As for my brother, he's family so I tolerate it.

Mustafa Umut Sarac · Sep 22, 2010

Holmes ,

I found below code at internet. Its 2 dimensional perlin noise. POVRAY is free and render such codes but it is very slow.

Umut

function Noise1(integer x, integer y)
n = x + y * 57
n = (n<<13) ^ n;
return ( 1.0 - ( (n * (n * n * 15731 + 789221) + 1376312589) & 7fffffff) / 1073741824.0);
end function

function SmoothNoise_1(float x, float y)
corners = ( Noise(x-1, y-1)+Noise(x+1, y-1)+Noise(x-1, y+1)+Noise(x+1, y+1) ) / 16
sides = ( Noise(x-1, y) +Noise(x+1, y) +Noise(x, y-1) +Noise(x, y+1) ) / 8
center = Noise(x, y) / 4
return corners + sides + center
end function

function InterpolatedNoise_1(float x, float y)

integer_X = int(x)
fractional_X = x - integer_X

integer_Y = int
fractional_Y = y - integer_Y

v1 = SmoothedNoise1(integer_X, integer_Y)
v2 = SmoothedNoise1(integer_X + 1, integer_Y)
v3 = SmoothedNoise1(integer_X, integer_Y + 1)
v4 = SmoothedNoise1(integer_X + 1, integer_Y + 1)

i1 = Interpolate(v1 , v2 , fractional_X)
i2 = Interpolate(v3 , v4 , fractional_X)

return Interpolate(i1 , i2 , fractional_Y)

end function

function PerlinNoise_2D(float x, float y)

total = 0
p = persistence
n = Number_Of_Octaves - 1

loop i from 0 to n

frequency = 2i
amplitude = pi

total = total + InterpolatedNoisei(x * frequency, y * frequency) * amplitude

end of i loop

return total

end function

Autochrome and Perlin Noise

Attachments

Privacy & Transparency

Privacy & Transparency