Thomas de Groot wrote:
> In povray.general, Stephen asked for a method to make a seamless patchwork 
> of height_fields (topic: Interlocking heightfields; May 11th). John 
> VanSickle was so kind to provide a couple of interesting macros. While 
> learning to use them, I made this simple test image consisting of 16 
> heightfields. I thought I could post it here, as info.

When I used these macros, I gave the height fields a high enough 
resolution so that there was no need for any tricks with the normals. 
There was no smoothing, and the texture applied to the height fields had 
no normal component.  To make the height fields rougher, I put more 
roughness in the pigment used to generate the height fields, so that the 
height fields really were rougher, instead of merely appearing to be.

Regards,
John

Post a reply to this message

"John VanSickle" <evi### [at] hotmailcom> schreef in bericht 
news:483166b3$1@news.povray.org...

> When I used these macros, I gave the height fields a high enough 
> resolution so that there was no need for any tricks with the normals. 
> There was no smoothing, and the texture applied to the height fields had 
> no normal component.  To make the height fields rougher, I put more 
> roughness in the pigment used to generate the height fields, so that the 
> height fields really were rougher, instead of merely appearing to be.
>

Yes, there are all kind of possibilities indeed to make the height_fields 
look interesting. Thanks John, for your further suggestions, which I shall 
certainly try out too.

Thomas

Post a reply to this message

...in particular, I overlooked that little parameter Test... :-)

Thomas

Post a reply to this message

"Thomas de Groot" <t.d### [at] internlDOTnet> schreef in bericht 
news:48318cbe$1@news.povray.org...
> ...in particular, I overlooked that little parameter Test... :-)
>
> Thomas
>
>

...which altogether, leads finally to this.

Thomas

Post a reply to this message

Attachments:
Download 'JVS_HFpatchwork.jpg' (89 KB)

Preview of image 'JVS_HFpatchwork.jpg'

Awesome!
;-)
Paolo
 >Thomas de Groot wrote:
> "Thomas de Groot" <t.d### [at] internlDOTnet> schreef in bericht 
> news:48318cbe$1@news.povray.org...
>> ...in particular, I overlooked that little parameter Test... :-)
>>
>> Thomas
>>
>>
> 
> ...which altogether, leads finally to this.
> 
> Thomas
> 
> 
>

Post a reply to this message

"Paolo Gibellini" <p.g### [at] gmailcom> schreef in bericht 
news:48342126$1@news.povray.org...
> Awesome!
> ;-)
> Paolo

This is just basic stuff as far as I am concerned. All the credit for the 
real work should go to John!!  :-)

Thomas

Post a reply to this message

Thomas de Groot wrote:
> "Paolo Gibellini" <p.g### [at] gmailcom> schreef in bericht 
> news:48342126$1@news.povray.org...
>> Awesome!
>> ;-)
>> Paolo
> 
> This is just basic stuff as far as I am concerned. All the credit for the 
> real work should go to John!!  :-)

The macro may need tweaking.  The rippling in the left-hand side of the 
image appears to be an artifact.

Regards,
John

Post a reply to this message

"John VanSickle" <evi### [at] hotmailcom> schreef in bericht 
news:4835644f$1@news.povray.org...

> The macro may need tweaking.  The rippling in the left-hand side of the 
> image appears to be an artifact.
>

I have been wondering about that. It appears indeed in the close up height 
field blocks.

Thomas

Post a reply to this message

Thomas de Groot wrote:
> "John VanSickle" <evi### [at] hotmailcom> schreef in bericht 
> news:4835644f$1@news.povray.org...
> 
>> The macro may need tweaking.  The rippling in the left-hand side of the 
>> image appears to be an artifact.
>>
> 
> I have been wondering about that. It appears indeed in the close up height 
> field blocks.

The macro (IIRC) calculates the number of pixels spanned in any 
direction by the heightfield, and sets the height field's resolution to 
that value.  It may be that when the line of view is parallel or 
near-parallel to the diagonal of the height field, that this is not 
quite enough.  Increasing the resolution of the HF (probably by no more 
than 40%) should fix this.

The ripple may also be the result of aliasing, in which case the same 
remedy--increasing the HF resolution--is called for, but there would 
also need to be aa + jittering in the render for the rippling to be 
eliminated, if this is the cause.

Regards,
John

Post a reply to this message

"John VanSickle" <evi### [at] hotmailcom> schreef in bericht 
news:483a9de9@news.povray.org...
>
> The macro (IIRC) calculates the number of pixels spanned in any direction 
> by the heightfield, and sets the height field's resolution to that value. 
> It may be that when the line of view is parallel or near-parallel to the 
> diagonal of the height field, that this is not quite enough.  Increasing 
> the resolution of the HF (probably by no more than 40%) should fix this.
>
That is correct. I have introduced a multiplicator to the Details parameter 
in the macro, which does a nice job, especially for larger images.

> The ripple may also be the result of aliasing, in which case the same 
> remedy--increasing the HF resolution--is called for, but there would also 
> need to be aa + jittering in the render for the rippling to be eliminated, 
> if this is the cause.

Tentatively, I introduced the smooth option in the macro, just to see what 
it does. The rippling is further generated I think by the pigment I used, 
where a downscaled wrinkle is present. The effect is somewhat lessened if  I 
scale it a bit up. However in larger renders, one clearly sees the mesh of 
rectangles so increasing the resolution works best.

All in all, using the macro is just to find the settings which will work 
best together, which is fun to do.

Thomas

Post a reply to this message