William F Pokorny <ano### [at] anonymousorg> wrote:

> My original surprise at the result in the lower right image over in pbi
> came from setting up a matrix I knew was out of bounds for typical use.
> I didn't expect things to work like two separate shear matrices; I
> didn't know what would happen, but the result surprised me due how
> dramatically it whacked the checker pattern.

Now that I've reviewed it all, your initial result was due to making x-hat and
y-hat the same vector.  <1, 1, 0> Essentially collapsing 3d space into a plane.
Attempting to apply the inverse and getting a parse error therefore makes sense.

> The isosurface animation was itself another way to verify behavior.
> There is the usual pattern space to function space inversion. I
> un-inverted the rotation so it looked like the checker rotation. I left
> the scaling as is because scaling up of the function's coordinate space
> shrinks the isosurface result here - keeping the results in a fixed
> view/frame, so to speak.

The isosurface looks like a strange, but interesting result to me, because
you're getting distortion of the square instead of the simple rotation like you
do in the checker.
It sort of looks like a camera iris - something to be pursued, I think.

Post a reply to this message

On 1/21/25 15:30, Bald Eagle wrote:
> The isosurface looks like a strange, but interesting result to me, because
> you're getting distortion of the square instead of the simple rotation like you
> do in the checker.
> It sort of looks like a camera iris - something to be pursued, I think.

Yes, I added a f_hypot(x,y) multiplier to the two 'rotation/shear terms' 
to create something more interesting than the square-frame rotating and 
shrinking.

Attaching two more mp4s. The one with the _BW.mp4 suffix removes that 
f_hypot() multiplier.

The one with the _SinCosHyp.mp4 suffix adds a sin() then cos() wrap of 
the f_hypot() multipliers. Done with the thought it might be 
interesting. :-)

Bill P.

Post a reply to this message

Attachments:
Download 'matrixplayasiso_bw.mp4.dat' (37 KB) Download 'matrixplayasiso_sincoshyp.mp4.dat' (30 KB)

William F Pokorny <ano### [at] anonymousorg> wrote:
>
> but I fully admit I'm just poking the matrix with a stick to
> see what happens. To some extent, I wonder about why it does what it
> does, but I care more about whether what it does might be useful.
>

That is also my usual method of playing with matrices, ha-- because I still have
a hard time understanding how they work and how to use them. I like your video
though; I need to experiment with your code example to see how I can break it!

------------
[In Windows 10]  I thought I should mention something: In the newsgroups'
standard web portal that I use, the Firefox 'previews' of your .mp4 videos do
not play -- with the message "Video cannot be played because the file is
corrupt." (which is not the case!) As a comparison: Josh's .mp4 animation in his
recent post  "Go home, RSOCP, yer drunk" plays fine in the Firefox preview.

So I downloaded your video-- and it plays fine in *most* of my Windows 10 media
players: VLC Media Player, SM Player, and even my old VirtualDub2. But it does
not play in Windows' own Media Player (a black screen)-- and actually breaks
Irfanview (the app hangs and has to be terminated via Windows Task Manager.)
Irfanview will usually play just about any video that I throw at it, so its
behavior in this case is very weird.

In VLC Media Player, I took a look at your first posted video's encoding
statistics:

Codec: H264-MPEG-4 AVC (part 10) (avc1)
Video Resolution: 450 X 450
Buffer Dimensions: 464 X 482
Decoded Format: Planar 4:4:4 YUV
Color Primaries: ITU-R BT.709

I did some research (yet again!) about video encoding and codecs, because it is
all so complicated to remember. Your video's stats are fairly standard stuff
AFAIU-- except for the 4:4:4 chroma subsampling.  4:2:0 is the typical scheme
for h.264; I use that myself. Take a look at Wikipedia's "h.264" page,
particularly the subsection "Feature support in particular profiles". It appears
that 4:4:4 is only supported in a particular high-end 'flavor' of h.264; perhaps
that presents a problem for Firefox, and/or for Windows 10's built-in video
codecs.

Or maybe you encoded the videos using 10-bit colors rather than the more
standard 8-bit?

I also did a search for possible Firefox problems regarding playback of
h.264 .mp4 files, but found nothing that would be helpful-- except for this
possibility:

"Yes, Firefox can play 4:4:4 videos, but only if your hardware and system
settings support it; this is because the capability to play 4:4:4 video depends
on your graphics card and the video codec used, not solely on the browser
itself."

(By the way: I had the same video-playback problems with your December 2024
post,
 "New f_dnoise(), modified f_dturbulence(). yuqk R17"
but forgot to mention it at the time.)

Post a reply to this message

For those of you that want to be able to do matrix calculations within POV-Ray,
I have made a library that can do some of that.

It can be found here:

https://github.com/t-o-k/POV-Ray-matrices
(NB: Some of the descriptions and examples are not finished yet.)

Below are some examples with this library that may be relevant to this thread.

--
Tor Olav
http://subcube.com
https://github.com/t-o-k


// ===== 1 ======= 2 ======= 3 ======= 4 ======= 5 ======= 6 ======= 7

#version 3.7;  // Should also work in version 3.8

global_settings { assumed_gamma 1.0 }

#include "matrices.inc"

// ===== 1 ======= 2 ======= 3 ======= 4 ======= 5 ======= 6 ======= 7

#declare No_Change_Transform = transform { }

#declare Rotate_Transform = transform { rotate 60*x }

#declare Shear_Transform =
    transform {
        matrix <
            1, 1, 0,
            0, 1, 0,
            0, 0, 1,
            0, 0, 0
        >
    }

#declare Translate_Transform = transform { translate <3, 1, 2> }

#declare Composite_Transform =
    transform {
        No_Change_Transform
        Rotate_Transform
        Shear_Transform
        Translate_Transform
    }

#debug "\nNo Change Transform:\n"
M_Print(M_FromTransform(No_Change_Transform))

#debug "\nRotate Transform:\n"
M_Print(M_FromTransform(Rotate_Transform))

#debug "\nShear Transform:\n"
M_Print(M_FromTransform(Shear_Transform))

#debug "\nTranslate Transform:\n"
M_Print(M_FromTransform(Translate_Transform))

#debug "\nComposite Transform:\n"
M_Print(M_FromTransform(Composite_Transform))

#debug "\n\n"

// ===== 1 ======= 2 ======= 3 ======= 4 ======= 5 ======= 6 ======= 7

#declare No_Change_Matrix = M_Identity(4);

#declare Rotate_Matrix = M_Rotate3D_AroundX(radians(60));

#declare Shear_Matrix =
    array[4][4] {
        {  1.0,  1.0,  0.0,  0.0 },
        {  0.0,  1.0,  0.0,  0.0 },
        {  0.0,  0.0,  1.0,  0.0 },
        {  0.0,  0.0,  0.0,  1.0 }
    }
;

#declare Translate_Matrix = M_Translate3D(<3, 1, 2>);

#declare Composite_Matrix =
    M_Mult(
        M_Mult(
            M_Mult(
                No_Change_Matrix,
                Rotate_Matrix
            ),
            Shear_Matrix
        ),
        Translate_Matrix
    )
;

#debug "\nNo Change Matrix:\n"
M_Print(No_Change_Matrix)

#debug "\nRotate Matrix:\n"
M_Print(Rotate_Matrix)

#debug "\nShear Matrix:\n"
M_Print(Shear_Matrix)

#debug "\nTranslate Matrix:\n"
M_Print(Translate_Matrix)

#debug "\nComposite Matrix:\n"
M_Print(Composite_Matrix)

// ===== 1 ======= 2 ======= 3 ======= 4 ======= 5 ======= 6 ======= 7

#debug "\n"
#error "No error, just finished"

// ===== 1 ======= 2 ======= 3 ======= 4 ======= 5 ======= 6 ======= 7

Post a reply to this message

"Kenneth" <kdw### [at] gmailcom> wrote:

> That is also my usual method of playing with matrices, ha-- because I still have
> a hard time understanding how they work and how to use them.

A matrix is nothing more than a statement of the end position of each basis
vector after the transformation.

For instance, when I do a shear along x, what I do is take "y" (<0, 1, 0>) and
shit it along the x-axis.  Maybe make it become <0.5, 1, 0>.

So matrix {
1, 0, 0,
0, 1, 0,
0, 0, 1,
0, 0, 0
}

becomes

matrix {
1, 0, 0,
0.5, 1, 0,
0, 0, 1,
0, 0, 0
}

Just think of the matrix operator as

matrix {
x,
y,
z,
translate <a, b, c>
}

and your matrix transform is

matrix {
new_x,
new_y,
new_z,
translate <a, b, c>
}

https://www.youtube.com/watch?v=kYB8IZa5AuE


https://www.f-lohmueller.de/pov_tut/trans/trans_000e.htm

Post a reply to this message

On 1/25/25 09:47, Kenneth wrote:
> I did some research (yet again!) about video encoding and codecs, because it is
> all so complicated to remember. Your video's stats are fairly standard stuff
> AFAIU-- except for the 4:4:4 chroma subsampling.  4:2:0 is the typical scheme
> for h.264; I use that myself. Take a look at Wikipedia's "h.264" page,
> particularly the subsection "Feature support in particular profiles". It appears
> that 4:4:4 is only supported in a particular high-end 'flavor' of h.264; perhaps
> that presents a problem for Firefox, and/or for Windows 10's built-in video
> codecs.
> 
> Or maybe you encoded the videos using 10-bit colors rather than the more
> standard 8-bit?

Hi Kenneth,

Excepting - maybe - the noise video where output grayscale and I 
remember trying the flag '-vf format=gray' to get a smaller mp4 file 
(made no difference), everything should be 8 bits because the png output 
is defaulting to 8 bits per channel.

Thank you for taking the time to investigate the problem you've seen 
with my ffmpeg encoded videos! I'm an amateur having encoded very few 
animations over the years - and I've posted fewer still.

I'd captured recommendations Dick Balaska made years ago for options 
(which included a '-pix_fmt yuv420p' setting), however, when I tried the 
command line late last year I kept getting errors I didn't understand.

Being lazy, I dropped back to a very simple command with a flag for 
frame rate, the input pattern for saved image frames and the output file 
name... Things seemed to work, so I went with ffmpeg's defaulting. :-) 
  
  
  
  
  
  


Next time I create an animation with ffmpeg, I'll try to get the 4:2:0 
chroma sub-sampling.

Bill P.

Post a reply to this message

"Bald Eagle" <cre### [at] netscapenet> wrote:
> "Kenneth" <kdw### [at] gmailcom> wrote:
>
> > That is also my usual method of playing with matrices, ha-- because I
> > still have a hard time understanding how they work and how to use them.
>
> A matrix is nothing more than a statement of the end position of each basis
> vector after the transformation.
> [clip]

Thanks for the useful info. Actually, the shearing ability of a matrix is one of
the (few) things that I do grasp. And William P's checkerboard rotation trick
has also been informative. But there are other more complex matrix transforms
that I have seen in newsgroup posts and include files that really baffle me, as
to how the 'magical' results are obtained. (Rune's old 'illusion.inc' file is a
prime example; I use it a lot, but its workings are a mystery to me.)

What I *need* to do is devote the time solely to learning the fundamentals-- and
without other POV-ray distractions. But that's difficult! Because there are
many *other* features that I only understand at a basic level...like
sophisticated functions! To truly learn all that I need to know, I could spend
25 hrs a day.

But every now and then, I have the urge to actually *render* something. ;-)

Post a reply to this message

On 1/25/25 19:22, Tor Olav Kristensen wrote:
> For those of you that want to be able to do matrix calculations within POV-Ray,
> I have made a library that can do some of that.
> 
> It can be found here:
> 
> https://github.com/t-o-k/POV-Ray-matrices

Thank you for the reminder. I'd looked at what you have there before - 
but my memory failed me in moment.

While I was "thinking aloud" about what was happening with the checker 
rotation/shear 'stick poking' play, your package would have saved me a 
chunk of time.

I've added a link to your page from my yuqk matrix documentation text 
file - which I hope will help me remember.

Bill P.

Post a reply to this message

"Kenneth" <kdw### [at] gmailcom> wrote:
(Rune's old 'illusion.inc' file is a
> prime example; I use it a lot, but its workings are a mystery to me.)

I have never used it, and the instructions puzzle me.
Perhaps if you could post a scene using it, and the intermediate render, then
maybe I can have some idea how to use it and see how it works.

Also, this seems like the sort of perspective correction that Francois LE COAT
would know all about.

- BW

Post a reply to this message

On 1/25/25 19:58, William F Pokorny wrote:
> Next time I create an animation with ffmpeg, I'll try to get the 4:2:0 
> chroma sub-sampling.

Alright. Found myself playing again today with a direct matrix transform 
  against yuqk's updated wrinkle pattern. The attached animation uses 
the sampling you suggested. Does it work in programs where failing before?

Bill P.

             matrix <
                 1.0, (frame_number-1)/60, 0.0,
                -(frame_number-1)/60, 1.0, 0.0,
                 0.0, 0.0, 1.0,
                 (frame_number-1)/57, (frame_number-1)/37, 0.0>

Post a reply to this message

Attachments:
Download 'wrinkles.mp4.dat' (922 KB)