Hi(gh)!

And once more digging in the history of POV-Ray... and perhaps finding 
something to adopt for own projects.

Heiko Rappich asked on 1998-10-21 how to create realistic flames with 
POV-Ray 3.1 - he previously used the "halo" feature of 3.02 and wondered 
why the media looks so transparent (first and second images attached here).

(and I wonder why the rendering in 3.1 looks so much darker and grainier 
than the one done with 3.7... and how abysmally slow my old laptop is 
compared to my hexacore main computer!)

The original thread can be found here: 
http://news.povray.org/povray.binaries.scene-files/thread/%3C362E409A.6AEB3235%40t-online.de%3E/?ttop=426795&toff=1431

Then three fellow POVers came upw with suggestions, first Ken Tyler: 
"Lower filter value of pigment" (third image attached)

Hendrik Knaepen (a relative of Zeger Knaepen?): "increase emission from 
0.68 to 5" (fourth image attached).

Finally, Bob Hughes (yes, he already was around back then!): "adapt 
emission color to flame color, replace spherical with onion media 
pattern and decrease turbulence" (fifth image attached).

I rendered all "suggestion versions" with 3.1 on my laptop (did not try 
them in 3.7 yet) - to me, Hendrik Knaepen's suggestion looks most 
promising, I would lower the emission value to something like 3... what 
do you think?

See you in Khyberspace!

Yadgar

Now playing: Action (Streetheart)

Post a reply to this message

Attachments:
Download '1998-10-21 fire problem with pov-ray 3.1 (heiko.rappich) [rendered on 2019-05-12 by yadgar using pov-ray 3.1].png' (512 KB) Download '1998-10-21 fire problem with pov-ray 3.1 (heiko.rappich) [rendered by yadgar on 2019-05-12 using pov-ray 3.7].png' (247 KB) Download '2019-05-12 heiko rappich's fire problem with pov-ray 3.1, using ken tyler's suggestion (yadgar).png' (439 KB) Download '2019-05-12 heiko rappich's fire problem with pov-ray 3.1, using hendrik knaepen's suggestion (yadgar).png' (407 KB) Download '2019-05-12 heiko rappich's fire problem with pov-ray 3.1, using bob hughes' suggestions (yadgar).png' (314 KB)

Preview of image '1998-10-21 fire problem with pov-ray 3.1 (heiko.rappich) [rendered on 2019-05-12 by yadgar using pov-ray 3.1].png'

Preview of image '1998-10-21 fire problem with pov-ray 3.1 (heiko.rappich) [rendered by yadgar on 2019-05-12 using pov-ray 3.7].png'

Preview of image '2019-05-12 heiko rappich's fire problem with pov-ray 3.1, using ken tyler's suggestion (yadgar).png'

Preview of image '2019-05-12 heiko rappich's fire problem with pov-ray 3.1, using hendrik knaepen's suggestion (yadgar).png'

Preview of image '2019-05-12 heiko rappich's fire problem with pov-ray 3.1, using bob hughes' suggestions (yadgar).png'

On 12-5-2019 2:23, Jörg "Yadgar" Bleimann wrote:
> Hi(gh)!
> 
> And once more digging in the history of POV-Ray... and perhaps finding 
> something to adopt for own projects.
> 
> Heiko Rappich asked on 1998-10-21 how to create realistic flames with 
> POV-Ray 3.1 - he previously used the "halo" feature of 3.02 and wondered 
> why the media looks so transparent (first and second images attached here).
> 
> (and I wonder why the rendering in 3.1 looks so much darker and grainier 
> than the one done with 3.7... and how abysmally slow my old laptop is 
> compared to my hexacore main computer!)
> 
> The original thread can be found here: 
>
http://news.povray.org/povray.binaries.scene-files/thread/%3C362E409A.6AEB3235%40t-online.de%3E/?ttop=426795&toff=1431

> 
> 
> Then three fellow POVers came upw with suggestions, first Ken Tyler: 
> "Lower filter value of pigment" (third image attached)
> 
> Hendrik Knaepen (a relative of Zeger Knaepen?): "increase emission from 
> 0.68 to 5" (fourth image attached).
> 
> Finally, Bob Hughes (yes, he already was around back then!): "adapt 
> emission color to flame color, replace spherical with onion media 
> pattern and decrease turbulence" (fifth image attached).
> 
> I rendered all "suggestion versions" with 3.1 on my laptop (did not try 
> them in 3.7 yet) - to me, Hendrik Knaepen's suggestion looks most 
> promising, I would lower the emission value to something like 3... what 
> do you think?
> 

I agree. To make the container less visible, I should also do something 
to the density. Concentrate the fire towards the centre and dim it 
towards the rim, using a spherical density map.


-- 
Thomas

Post a reply to this message

> Hendrik Knaepen (a relative of Zeger Knaepen?): "increase emission from 
> 0.68 to 5" (fourth image attached).

He's my brother, but I'm quite sure he never posted here. Must have been 
me using his computer :)

Post a reply to this message

On 19/05/12 09:40, Zeger Knaepen wrote:
>> Hendrik Knaepen (a relative of Zeger Knaepen?): "increase emission 
>> from 0.68 to 5" (fourth image attached).
> 
> He's my brother, but I'm quite sure he never posted here. Must have been 
> me using his computer :)

Also: don't use fully saturated colors for these very bright things, 
like fire. So no rgb <1,.5,0> but rather rgb <1,.5,.01>

Post a reply to this message

=?UTF-8?Q?J=c3=b6rg_=22Yadgar=22_Bleimann?= <yaz### [at] gmxde> wrote:
> Hi(gh)!
>
> And once more digging in the history of POV-Ray... and perhaps finding
> something to adopt for own projects.

> (and I wonder why the rendering in 3.1 looks so much darker and grainier
> than the one done with 3.7... and how abysmally slow my old laptop is
> compared to my hexacore main computer!)

> I rendered all "suggestion versions" with 3.1 on my laptop (did not try
> them in 3.7 yet) - to me, Hendrik Knaepen's suggestion looks most
> promising, I would lower the emission value to something like 3... what
> do you think?
>
> See you in Khyberspace!
>
> Yadgar


Hello,

obviously we've similar "projects", here is, what I did with this source.
First I assume, it's darker because of the old standard assumed_gamma 2.2.
The source doesn't make problems with 3.7 IIRC.

I changed a bit of the media code -

media {
        emission <1,0.8,0.3>*2.3
        intervals 1
        samples 50
        confidence 0.999
        variance 0.001
        density {
                spherical
                ramp_wave
                color_map {
                        [0      srgb <0,0,0>]
                        [0.2    srgb <1,0,0>]
                        [1      srgb <1,1,0>]
                }
                turbulence 1.9
                frequency 2
                scale 3 warp {turbulence <1,8,1>} scale 1/3
                scale 0.7
        }
}

Norbert

Post a reply to this message

Attachments:
Download 'fire problem with pov-ray 3.1.jpg' (649 KB)

Preview of image 'fire problem with pov-ray 3.1.jpg'

Hi(gh)!

On 12.05.19 09:43, Zeger Knaepen wrote:

> Also: don't use fully saturated colors for these very bright things, 
> like fire. So no rgb <1,.5,0> but rather rgb <1,.5,.01>

You probably meant something like rgb <0.75, 0.5, 0.25>, as rgb <1, 0.5, 
0.01> would not yield any significant reduction in saturation...

See you in Khyberspace!

Yadgar

Now playing: Synthesist (Harald Großkopf)

Post a reply to this message

> You probably meant something like rgb <0.75, 0.5, 0.25>, as rgb <1, 0.5, 
> 0.01> would not yield any significant reduction in saturation...

that's a bit too little saturation I think, but you're correct, 
<1,.5,.01> is still way too saturated. <1,.125,.06> looks a lot better imho.

I seem to remember I once made a kelvin-to-rgb converter, in javascript. 
Can't seem to find it though, but a quick google gave me this:
http://www.tannerhelland.com/4435/convert-temperature-rgb-algorithm-code/

Post a reply to this message

Attachments:
Download 'mediacolortest_yours.jpg' (37 KB) Download 'mediacolortest_mine.jpg' (42 KB) Download 'mediacolortest_better_imho.jpg' (45 KB)

Preview of image 'mediacolortest_yours.jpg'

Preview of image 'mediacolortest_mine.jpg'

Preview of image 'mediacolortest_better_imho.jpg'

Zeger Knaepen <zeg### [at] povplacecom> wrote:
> Yadgar wrote:
> > You probably meant something like rgb <0.75, 0.5, 0.25>, as rgb <1, 0.5,
> > 0.01> would not yield any significant reduction in saturation...
>
> that's a bit too little saturation I think, but you're correct,
> <1,.5,.01> is still way too saturated. <1,.125,.06> looks a lot better imho.
>
> I seem to remember I once made a kelvin-to-rgb converter, in javascript.
> Can't seem to find it though, but a quick google gave me this:
> http://www.tannerhelland.com/4435/convert-temperature-rgb-algorithm-code/

Lightsys IV has kelvin-to-RGB conversion using POV SDL.  The relevant include
file is CIE.inc.

  http://www.ignorancia.org/index.php/technical/lightsys/

Post a reply to this message

On 14-5-2019 16:40, Zeger Knaepen wrote:
>> You probably meant something like rgb <0.75, 0.5, 0.25>, as rgb <1, 
>> 0.5, 0.01> would not yield any significant reduction in saturation...
> 
> that's a bit too little saturation I think, but you're correct, 
> <1,.5,.01> is still way too saturated. <1,.125,.06> looks a lot better 
> imho.
> 
> I seem to remember I once made a kelvin-to-rgb converter, in javascript. 
> Can't seem to find it though, but a quick google gave me this:
> http://www.tannerhelland.com/4435/convert-temperature-rgb-algorithm-code/

I prefer the "original" (?) emission value of <1, 0.8, 0.3>*2.3. Note 
however, that I used POV-version 3.8 for my image.

I guess that the background influences the look. Yours are all with a 
black, featureless, background.

-- 
Thomas

Post a reply to this message

Attachments:
Download 'hr_fire.png' (267 KB)

Preview of image 'hr_fire.png'

Thomas de Groot <tho### [at] degrootorg> wrote:

> I prefer the "original" (?) emission value of <1, 0.8, 0.3>*2.3. Note
> however, that I used POV-version 3.8 for my image.

> Thomas

Hi Thomas,

the emission value is made of the suggestions in the old thread.
Original emission value is 0.62.

Norbert

Post a reply to this message