|
|
Le 17-09-02 à 03:03, Thomas de Groot a écrit :
> On 2-9-2017 6:31, omniverse wrote:
>> Alain <kua### [at] videotronca> wrote:
>>> Le 17-08-31 à 18:58, Loren a écrit :
>>>> media{ emission Red intervals 30 samples 100,100 }
>>> Yuck! That WAS ok in version 3.5 and older that used sampling method 2
>>> as the default. As of version 3.6, the sampling method is method 3 and
>>> it must use intervals 1 (the default value).
>>> Using method 3, more intervals only dramatically slow you down. It can
>>> also cause some artefacts.
>>> Also :
>>> 1) It only use a single value for samples. If a second value is used,
>>> it's silently ignored.
>>> 2) confidance and variance are also silently ignored as they are
>>> meaningless when you have only a single interval.
>>>
>>> Defaults for media in version 3.6+
>>> method 3
>>> samples 10
>>> confidance Not Applicable
>>> variance N/A
>>> intervals 1
>>> jitter 1
>>>
>>> Alain
>>
>> Well I'm learning something, again, because I had thought
>>
>> samples LesserInteger, GreaterInteger
>>
>> was valid for method 3. And I also thought samples 1,1 was the default.
>> Reading the 3.7 doc I don't find it saying the above you tell of, not
>> that I
>> don't believe you Alain, but I refer to the docs a lot and try to
>> believe what I
>> read there. :)
>>
>
> Alain is absolutely right. He is the one person warning us, again and
> again, for the method/intervals/samples misconception cropping up
> regularly in these ng's. Hail to the chief! ;-)
>
> As for the docs, samples LesserInteger, GreaterInteger, is only valid
> for method 1 and 2. I agree that there is an ambiguity where method 3 is
> concerned: paragraph 2.7.2.3 Sampling Parameters & Methods in the wiki (
> http://wiki.povray.org/content/Reference:Sampling_Parameters_%26_Methods
> ) does not state clearly that the second term is ignored when using
> method 3. This should be changed.
>
>
For the massive slowdown when using intervals >1, I've found out by
testings.
For the possibility of artefacts, it's writen in the docs.
Post a reply to this message
|
|