Am 01.03.2017 um 12:36 schrieb Jim Holsenback:
> On 2/28/2017 4:17 PM, Alain wrote:

>>> Am 27.02.2017 um 16:21 schrieb Fudz4:
>>>
>>>> Thanks for the reply, yo said that photons are 'stored' at the
>>>> location they
>>>> land, is there a way to collect that data? I am tasked with using a
>>>> machine
>>>> learning algorithm to estimate the light density for example.
>>>
>>> There used to be a mechanism for saving photon data, for re-use in
>>> subsequent renders. Search for "save_file" and "load_file" photon
>>> settings.
>>>
>>> I'm afraid this mechanism is still dysfunctional in POV-Ray 3.7.x, so
>>> you may have to fall back to 3.6 instead.
>>>
>>
>> What do you mean by dysfunctional?
>> I routinely use that feature and did not encouter any obvious isue.
> 
> yep .... i was thinking the same thing

Sorry, my bad - in the process of verifying that the mechanism is alive
and kicking, I confused the parser code for radiosity
`save_file`/`load_file` with that for photon mapping
`save_file`/`load_file`.

Since the code in question no longer does anything except print a parse
warning saying that INI file options must be used nowadays (which is
indeed true for radiosity), but there are no such INI file options for
photon mapping, I erroneously concluded that photon save/load cannot
possibly be functional.

(Which begs the question why the bloops someone insisted that radiosity
had to use INI file options, when the SDL statements were apparently
good enough for photon mapping. I know I'm to blame for actually
/implementing/ the radiosity changes, but I'm pretty damn sure it wasn't
my call back then.)

Post a reply to this message