POV-Ray : Newsgroups : povray.documentation.inbuilt : Corrections to documented media defaults. Server Time
24 Jun 2021 13:39:50 EDT (-0400)
  Corrections to documented media defaults. (Message 4 to 13 of 13)  
<<< Previous 3 Messages Goto Initial 10 Messages
From: Alain
Subject: Re: Corrections to documented media defaults.
Date: 15 Feb 2018 19:36:03
Message: <5a862773$1@news.povray.org>
Le 18-02-15 à 13:30, William F Pokorny a écrit :
> On the page:
> 
> http://wiki.povray.org/content/Reference:Media#Index_Entry_media_types
> 
> samples      : Min 1, Max 1
> 
> should be:
> 
> samples      : Min 10, Max 10
> 
> given we otherwise have the >=3.5 method 3 defaults documented.
> 
> Bill P.

It should be "samples 10"
min and max are only used with method 1 and 2.


Post a reply to this message

From: Alain
Subject: Re: Corrections to documented media defaults.
Date: 15 Feb 2018 19:38:25
Message: <5a862801$1@news.povray.org>
Le 18-02-15 à 16:03, Mike Horvath a écrit :
> On 2/15/2018 1:30 PM, William F Pokorny wrote:
>> On the page:
>>
>> http://wiki.povray.org/content/Reference:Media#Index_Entry_media_types
>>
>> samples      : Min 1, Max 1
>>
>> should be:
>>
>> samples      : Min 10, Max 10
>>
>> given we otherwise have the >=3.5 method 3 defaults documented.
>>
>> Bill P.
> 
> Off-topic, but your posts always show up as belonging to clipka in 
> Thunderbird. I think it is because you both use the same anonymous email 
> address.
> 
> 
> Mike

Show correctly here. Thunderbird under Windows 7.


Post a reply to this message

From: Kenneth
Subject: Re: Corrections to documented media defaults.
Date: 15 Feb 2018 23:15:01
Message: <web.5a8659e248f091b8a47873e10@news.povray.org>
William F Pokorny <ano### [at] anonymousorg> wrote:
> On the page:
>
> http://wiki.povray.org/content/Reference:Media#Index_Entry_media_types
>
> samples      : Min 1, Max 1
>
> should be:
>
> samples      : Min 10, Max 10
>
> given we otherwise have the >=3.5 method 3 defaults documented.
>

What's also missing from all of the various media wiki references I've looked at
is that, with method 3, a *single* samples value can be used instead...

samples  35

....but even then, it isn't clear if method 3 is the *only* method that can do
that. Do methods 1 and 2 *require* 2 values?

The docs do need updating about this-- but I'm wondering if, sometime within the
v3.5-to-3.7 interval, the default samples value changed more than once(?) In
which case, changing the docs' default to  samples 10 (or 10,10?)  might raise
questions for users of those older versions-- if samples 1,1  as the default
'came and went' with various versions.  (Of course, that predicament cold be
easily solved with an extra note in the docs, about old vs. new behavior.)


Post a reply to this message

From: William F Pokorny
Subject: Re: Corrections to documented media defaults.
Date: 16 Feb 2018 07:56:02
Message: <5a86d4e2$1@news.povray.org>
On 02/15/2018 07:36 PM, Alain wrote:
> Le 18-02-15 à 13:30, William F Pokorny a écrit :
>> On the page:
>>
>> http://wiki.povray.org/content/Reference:Media#Index_Entry_media_types
>>
>> samples      : Min 1, Max 1
>>
>> should be:
>>
>> samples      : Min 10, Max 10
>>
>> given we otherwise have the >=3.5 method 3 defaults documented.
>>
>> Bill P.
> 
> It should be "samples 10"
> min and max are only used with method 1 and 2.

Understand what you are saying, but here we are documenting what POV-Ray 
is setting as the default if the language level seen by the parser is >= 
3.5. This override happens on seeing the media keyword. A user can, for 
example, still specify method 2 in that media and then knowing that 
default max value is 10 matters.

Aside: I'm with you in not understanding why the parser sets the max 
value for language level >= 3.5 for method 3, but it does. If it didn't 
we'd need to leave Max 1 for the same reason - it helps users understand 
what defaults are when those defaults apply to the method they are using.

Bill P.


Post a reply to this message

From: William F Pokorny
Subject: Re: Corrections to documented media defaults.
Date: 16 Feb 2018 08:24:51
Message: <5a86dba3$1@news.povray.org>
On 02/15/2018 11:11 PM, Kenneth wrote:
> William F Pokorny <ano### [at] anonymousorg> wrote:
...
> 
> What's also missing from all of the various media wiki references I've looked at
> is that, with method 3, a *single* samples value can be used instead...
> 
> samples  35
> 
> ....but even then, it isn't clear if method 3 is the *only* method that can do
> that. Do methods 1 and 2 *require* 2 values?
> 
> The docs do need updating about this-- but I'm wondering if, sometime within the
> v3.5-to-3.7 interval, the default samples value changed more than once(?) In
> which case, changing the docs' default to  samples 10 (or 10,10?)  might raise
> questions for users of those older versions-- if samples 1,1  as the default
> 'came and went' with various versions.  (Of course, that predicament cold be
> easily solved with an extra note in the docs, about old vs. new behavior.)
> 

Agree further doc updates are needed. As I dig deeper into media I'm 
gathering a list of other documentation updates that would be helpful I 
think including some of what you discuss. Trying not to go after too 
much at once - and some behavior suppose we might change sooner rather 
than later.

For example, while I've never made use of photons beyond playing with 
the distribution sample scenes, I now know it's "method" for media is 
it's own thing. 'Method photons' - suppose I see it as most like method 
2 at this point - is not using at all the sample or jitter settings for 
example. ...and I think photon jitter put into the photon media 
depositing while debugging and got left in accidentally. I can see no 
reason to introduce the banding artifact - see our optics.pov scene 
(more visible if you reduce the media gather radius but it's there in 
the scene we ship). The default jitter should be the full 1.0 always as 
far as I can see.

We have this page:

http://wiki.povray.org/content/Reference:Sampling_Parameters_%26_Methods#Index_Entry_method_media

where it might be good to mention the effective method 'photons', how it 
is different, how it can and cannot be controlled by media block 
settings.  There is limited documentation where the photons feature is 
documented.

(Yes, that page lists the default method 3 aa_level as 4 when it's 3 as 
indicated on the initial wiki page of this thread.)

Anyway... Also so much of the whole of media behavior I still don't 
understand.

Bill P.


Post a reply to this message

From: Kenneth
Subject: Re: Corrections to documented media defaults.
Date: 16 Feb 2018 13:15:01
Message: <web.5a871e0048f091b8a47873e10@news.povray.org>
William F Pokorny <ano### [at] anonymousorg> wrote:
.....and I think photon jitter put into the photon media
> depositing while debugging and got left in accidentally. I can see no
> reason to introduce the banding artifact - see our optics.pov scene
> (more visible if you reduce the media gather radius but it's there in
> the scene we ship). The default jitter should be the full 1.0 always as
> far as I can see.
>

I agree. Although I haven't used photons in awhile, I noticed a long time ago
that jitter 0.5 wasn't enough to eliminate visible 'rings', especially seen
close-up. I never mentioned it, though :-(  And I no longer have the test
scene that I used. If 0.5 is still the default, it does need reworking for a
smoother more-random appearance.


Post a reply to this message

From: Alain
Subject: Re: Corrections to documented media defaults.
Date: 16 Feb 2018 13:43:48
Message: <5a872664$1@news.povray.org>
Le 18-02-16 à 07:56, William F Pokorny a écrit :
> On 02/15/2018 07:36 PM, Alain wrote:
>> Le 18-02-15 à 13:30, William F Pokorny a écrit :
>>> On the page:
>>>
>>> http://wiki.povray.org/content/Reference:Media#Index_Entry_media_types
>>>
>>> samples      : Min 1, Max 1
>>>
>>> should be:
>>>
>>> samples      : Min 10, Max 10
>>>
>>> given we otherwise have the >=3.5 method 3 defaults documented.
>>>
>>> Bill P.
>>
>> It should be "samples 10"
>> min and max are only used with method 1 and 2.
> 
> Understand what you are saying, but here we are documenting what POV-Ray 
> is setting as the default if the language level seen by the parser is >= 
> 3.5. This override happens on seeing the media keyword. A user can, for 
> example, still specify method 2 in that media and then knowing that 
> default max value is 10 matters.
> 
> Aside: I'm with you in not understanding why the parser sets the max 
> value for language level >= 3.5 for method 3, but it does. If it didn't 
> we'd need to leave Max 1 for the same reason - it helps users understand 
> what defaults are when those defaults apply to the method they are using.
> 
> Bill P.

Then, the various defaults should be given for each sampling methods.

Default :
Sampling method = 3
samples 10 (if you use two values, the second will be silently ignored)
intervals 1 (Don't change as it drastically increase rendering time)
confidance, variance are not used.

Optional methods are 1 and 2. For those the defaults are :
samples min 1 max 1
intervals 10
...


Post a reply to this message

From: Bald Eagle
Subject: Re: Corrections to documented media defaults.
Date: 16 Feb 2018 16:45:01
Message: <web.5a874fd448f091b8c437ac910@news.povray.org>
This is just something that I always wish there was something in POV-Ray where
you could just do something like:

#displaydefault {media}

#man media

or something similar.

I usually [try to] embed some sort of debugging output flag into my code so that
when it's enabled, the code spits out any (ok, every) relevant value.

Perhaps something to think about for 4.0 so that the software itself can answer
questions instead of / in addition to a separate documentation page.

Think "measurement", not "label".


Post a reply to this message

From: William F Pokorny
Subject: Re: Corrections to documented media defaults.
Date: 16 Feb 2018 20:04:06
Message: <5a877f86$1@news.povray.org>
On 02/16/2018 01:44 PM, Alain wrote:
> Le 18-02-16 à 07:56, William F Pokorny a écrit :
>> On 02/15/2018 07:36 PM, Alain wrote:
> 
> Then, the various defaults should be given for each sampling methods.
> 
> Default :
> Sampling method = 3
> samples 10 (if you use two values, the second will be silently ignored)
> intervals 1 (Don't change as it drastically increase rendering time)
> confidance, variance are not used.
> 
> Optional methods are 1 and 2. For those the defaults are :
> samples min 1 max 1
> intervals 10
> ...
> 

Better documentation on what settings are used by each media method 
would be helpful for sure, but today the defaults set are the same no 
matter the method defaulted or set. In other words, there are not 
different defaults set for methods 1 & 2 should the user specify one of 
those in the media.

Aside: Not universally true, but often keywords and their settings can 
be given in any order in a block {}. Further you can often later in a 
parse refer to a defined 'item' and override some part of it. Useful, 
but this freedom makes having different default sets or sanity checking 
a group of settings for consistency during the parsing phase more 
difficult. The only place where I know we've today got some pseudo post 
parse checking is with patterns and in that case the extra verification 
happens only with the pavement and tiling patterns.

Bill P.


Post a reply to this message

From: Mike Horvath
Subject: Re: Corrections to documented media defaults.
Date: 16 Feb 2018 21:52:16
Message: <5a8798e0$1@news.povray.org>
On 2/15/2018 4:17 PM, dick balaska wrote:
> On 02/15/2018 04:03 PM, Mike Horvath wrote:
>> Off-topic, but your posts always show up as belonging to clipka in 
>> Thunderbird. I think it is because you both use the same anonymous 
>> email address.
>>
>>
>> Mike
> 
> I don't see that. (Thunderbird/Ubuntu).
> You likely have anonymous in your address book and it substitutes.
> 


Yes. I since removed clipka from my address book (sorry, Christoph!), 
and Bill's name is showing up correctly now.


Mike


Post a reply to this message

<<< Previous 3 Messages Goto Initial 10 Messages

Copyright 2003-2021 Persistence of Vision Raytracer Pty. Ltd.