sharing the ambient cache

radiance-general
1

Hi folks,

I should know the answer to this, but I don't. Let's say I have a script that performs some numeric analysis of a scene (rtrace) and then does some renderings as well. I typically crank up the settings quite a bit for rtrace, but relax them a bit for images (usually). My question is, can the rpict process(es) safely use the existing ambient cache left over from the high-accuracy rtrace run? Could running the rtrace at lower accuracy settings degrade any of the previously cached values? Just curious.

- Rob

2

Hi Rob,

Jack de Valpine or John Mardaljevic may know more about this than I do, as they have experience using different ambient settings in different runs, whereas I tend to stick with the same settings I start with. In general, it's safe to run with reduced ambient settings in subsequent renderings when you share an ambient file. Specifically, it's OK to relax the -ad and -as settings, and -aa and -ar as well. You can reduce -ab (though -ab 0 won't use the ambient file values at all), but I wouldn't recommend changing -av or -aw. (There is no speed advantage to doing so, anyway.)

-Greg

···

From: Rob Guglielmetti <[email protected]>
Date: December 21, 2005 3:14:25 PM PST

Hi folks,

I should know the answer to this, but I don't. Let's say I have a script that performs some numeric analysis of a scene (rtrace) and then does some renderings as well. I typically crank up the settings quite a bit for rtrace, but relax them a bit for images (usually). My question is, can the rpict process(es) safely use the existing ambient cache left over from the high-accuracy rtrace run? Could running the rtrace at lower accuracy settings degrade any of the previously cached values? Just curious.

- Rob

3

Greg Ward wrote:

Hi Rob,

Jack de Valpine or John Mardaljevic may know more about this than I do, as they have experience using different ambient settings in different runs, whereas I tend to stick with the same settings I start with. In general, it's safe to run with reduced ambient settings in subsequent renderings when you share an ambient file. Specifically, it's OK to relax the -ad and -as settings, and -aa and - ar as well. You can reduce -ab (though -ab 0 won't use the ambient file values at all), but I wouldn't recommend changing -av or -aw. (There is no speed advantage to doing so, anyway.)

Well, I'd certainly love to hear Jack and John chime in, but you have given me what I was looking for to proceed with confidence. Thanks!

- Rob

4

Hi Greg and Rob,

The strategy that we have always used is based on John's recomendations in RwR. For example:

parameter
  first pass (overture at 64 or 128 pixels)
  second pass (whatever your next resolution is)
-ad
  1024
  512 (1/2 of first pass)
-as
  512
  256 (1/2 of first pass)
-aa
  .15
  .25

In RwR, John does not recommend changing the other ambient parameters. However, I am intrigued to hear that Greg indicates that it would be ok to reduce -ar as well as -ab. It would be interesting to run some test and determine what good reduction parameters would be.

Now, we use the parameter reductions mainly for "pictures." I am interested to consider what kind of parameter reductions would work when moving from a tight grained high -ab quantitative simulation to something that could still use the ambient cache for producing "pictures," which is more clearly what Rob is thinking about.

-Jack

Greg Ward wrote:

···

Hi Rob,

Jack de Valpine or John Mardaljevic may know more about this than I do, as they have experience using different ambient settings in different runs, whereas I tend to stick with the same settings I start with. In general, it's safe to run with reduced ambient settings in subsequent renderings when you share an ambient file. Specifically, it's OK to relax the -ad and -as settings, and -aa and - ar as well. You can reduce -ab (though -ab 0 won't use the ambient file values at all), but I wouldn't recommend changing -av or -aw. (There is no speed advantage to doing so, anyway.)

-Greg

From: Rob Guglielmetti <[email protected]>
Date: December 21, 2005 3:14:25 PM PST

Hi folks,

I should know the answer to this, but I don't. Let's say I have a script that performs some numeric analysis of a scene (rtrace) and then does some renderings as well. I typically crank up the settings quite a bit for rtrace, but relax them a bit for images (usually). My question is, can the rpict process(es) safely use the existing ambient cache left over from the high-accuracy rtrace run? Could running the rtrace at lower accuracy settings degrade any of the previously cached values? Just curious.

- Rob

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

--
# Jack de Valpine
# president
#
# visarc incorporated
# http://www.visarc.com
#
# channeling technology for superior design and construction

5

Well, I should have made a caveat about reducing the -ab parameter. If your -av setting is not well-chosen, or if the lighting varies drastically over your scene, reducing -ab in subsequent runs could result in splotch artifacts. You would have to try it, but that's what I would expect.

-Greg

···

From: Jack de Valpine <[email protected]>
Date: December 22, 2005 8:20:14 AM PST

Hi Greg and Rob,

The strategy that we have always used is based on John's recomendations in RwR. For example:

parameter
  first pass (overture at 64 or 128 pixels)
  second pass (whatever your next resolution is)
-ad
  1024
  512 (1/2 of first pass)
-as
  512
  256 (1/2 of first pass)
-aa
  .15
  .25

In RwR, John does not recommend changing the other ambient parameters. However, I am intrigued to hear that Greg indicates that it would be ok to reduce -ar as well as -ab. It would be interesting to run some test and determine what good reduction parameters would be.

Now, we use the parameter reductions mainly for "pictures." I am interested to consider what kind of parameter reductions would work when moving from a tight grained high -ab quantitative simulation to something that could still use the ambient cache for producing "pictures," which is more clearly what Rob is thinking about.

-Jack

6

Jack de Valpine or John Mardaljevic may know more about this than I do...

Actually, I can't really expand much on what is in the book (p387):

Having created the ambient file with the “overture” calculation, you can, with
caution, relax some of the ambient parameters for the larger renderings. The
parameter revisions could be one or both of the following:
• Reduce -ad and -as by about 50%
• Slightly increase -aa (i.e., by 0.05 or 0.10)
The other ambient parameter settings should not be changed. If you do decide
to change any of the -ad, -as, or -aa settings after the “overture” calculation, you
should be aware that the modifications will not be reflected in the header of the
ambient file. Thus, you need to track both the picture and the ambient file headers
to obtain a complete record of the parameter settings for an image.

I seem to recall that Greg reckoned it was not a good idea to reduce the ambient resolution (ar) parameter when reusing the ambient cache. But I can't quite remember the reason why. I was going to end my contribution at this point. But then I did some tests. Load up:

http://www.iesd.dmu.ac.uk/~jm/pdfs/relax.pdf

What we have here is a test scene: a ridged box by a window and illuminated by a diffuse (i.e. glow) sky. The ar parameter limits the extent to which ambient sampling can convincingly shade the grooves. I populated the ambient file artest.af in creating the image at the top which only sees half the box. I then reused the ambient file with lower resolution settings to create wide views of the box -- I thought that seeing the "boundary" between the already populated hi-res ambient settings and the low-res might be interesting. It was. Note that for each of the three low ambient resolution images I reused the *original* artest.af (i.e. I kept a copy).

In the left and bottom images I relaxed the aa and ad parameters respectively. The results are what most of us would probably expect. For the image on the right I relaxed the ar parameter. Now, for this image, the "new" part on the left-hand-side is what would be expected from a fresh run without any pre-existing ambient file -- ar 64 doesn't allow close-enough sampling and the grooves get some "flat" ambient shading [1]. The interesting effect is of course on the right -- the pre-existing ambient samples on this side have somehow produced this lovely frog-spawn pattern. Why? [2]

-John

PS. Dontcha just love this "I'll never quite master it" side to Radiance?

[1] When av is zero, as it was here, how is the ar-limited shading value arrived at?

[2] I have this vague feeling that I sort of might know the reason why. But it is all too hazy for me to attempt to put it into words.

···

-----------------------------------------------
Dr. John Mardaljevic
Senior Research Fellow
Institute of Energy and Sustainable Development
De Montfort University
The Gateway
Leicester
LE1 9BH, UK
+44 (0) 116 257 7972
+44 (0) 116 257 7981 (fax)

[email protected]
http://www.iesd.dmu.ac.uk/~jm

7

Hi John,

Just goes to show, when it comes to the older algorithms, it's best to follow my old advice rather than my new advice... Clearly, I thought about this more carefully before, and there are good reasons for not changing the -ar parameter in subsequent runs. I think the artifacts you're seeing are due to the mix of new values and older ones, both of which get included in the interpolation step. The new values cover a wider area than the old ones (and wider than they should). The splotchy look is due to the way the interpolation algorithm works, and is even worse than I'd expect from reducing the -ab value.

-Greg

···

From: John Mardaljevic <[email protected]>
Date: December 22, 2005 11:18:29 AM PST

Jack de Valpine or John Mardaljevic may know more about this than I do...

Actually, I can't really expand much on what is in the book (p387):

Having created the ambient file with the “overture” calculation, you can, with
caution, relax some of the ambient parameters for the larger renderings. The
parameter revisions could be one or both of the following:
• Reduce -ad and -as by about 50%
• Slightly increase -aa (i.e., by 0.05 or 0.10)
The other ambient parameter settings should not be changed. If you do decide
to change any of the -ad, -as, or -aa settings after the “overture” calculation, you
should be aware that the modifications will not be reflected in the header of the
ambient file. Thus, you need to track both the picture and the ambient file headers
to obtain a complete record of the parameter settings for an image.

I seem to recall that Greg reckoned it was not a good idea to reduce the ambient resolution (ar) parameter when reusing the ambient cache. But I can't quite remember the reason why. I was going to end my contribution at this point. But then I did some tests. Load up:

http://www.iesd.dmu.ac.uk/~jm/pdfs/relax.pdf

What we have here is a test scene: a ridged box by a window and illuminated by a diffuse (i.e. glow) sky. The ar parameter limits the extent to which ambient sampling can convincingly shade the grooves. I populated the ambient file artest.af in creating the image at the top which only sees half the box. I then reused the ambient file with lower resolution settings to create wide views of the box -- I thought that seeing the "boundary" between the already populated hi-res ambient settings and the low-res might be interesting. It was. Note that for each of the three low ambient resolution images I reused the *original* artest.af (i.e. I kept a copy).

In the left and bottom images I relaxed the aa and ad parameters respectively. The results are what most of us would probably expect. For the image on the right I relaxed the ar parameter. Now, for this image, the "new" part on the left-hand-side is what would be expected from a fresh run without any pre-existing ambient file -- ar 64 doesn't allow close-enough sampling and the grooves get some "flat" ambient shading [1]. The interesting effect is of course on the right -- the pre-existing ambient samples on this side have somehow produced this lovely frog-spawn pattern. Why? [2]

-John

PS. Dontcha just love this "I'll never quite master it" side to Radiance?

[1] When av is zero, as it was here, how is the ar-limited shading value arrived at?

[2] I have this vague feeling that I sort of might know the reason why. But it is all too hazy for me to attempt to put it into words.

-----------------------------------------------
Dr. John Mardaljevic

8

Greg Ward wrote:

Hi John,

Just goes to show, when it comes to the older algorithms, it's best to follow my old advice rather than my new advice... Clearly, I thought about this more carefully before, and there are good reasons for not changing the -ar parameter in subsequent runs. I think the artifacts you're seeing are due to the mix of new values and older ones, both of which get included in the interpolation step. The new values cover a wider area than the old ones (and wider than they should). The splotchy look is due to the way the interpolation algorithm works, and is even worse than I'd expect from reducing the - ab value.

Hi gang,

Thanks to everyone who got involved in this thread. John, your analysis proves what I had been fearful of. Very interesting!

Radiance: You'll never quite master it (but you'll have fun along the way).

- Rob Guglielmetti