Trans Oddities Part Deux

Well, I tried Greg's suggestion of setting the limit weight to a very low number, but all that did is manage to scrounge up a few additional lux. This actually made things look worse, because my problems are with the really dense shades (which are already coming up with higher numbers than I'd expect).

I have a spreadsheet here with the results of all my tests so far. If anyone's interested in looking at it I'd be happy to share it. Perhaps there's a problem with my interpretation of this data. I dunno. But the tests indicate that whenever I try a 1% shade in any combination, I get very odd results from rtrace. And the whole reason I'm doing this test is because we were getting similar odd results on our real model; I was trying to simplify things so I could perhaps isolate something that was causing these oddities. So far, no joy. I shared my trans material definitions for all these tests earlier, but if anyone wants them to have a look I can send those as well.

My fear is that I'm simply trying to find too small a percentage of the available light, and relative comparisons between different shade materials will not be valid for this model. (?) I mean, this test model still has two pieces of diffuse glass in it, and the skylight is prety small, so already the building form/glazing is removing a ton of light from the space. Then when I start throwing shades in the skylight cavity, we're getting down to really tiny fractions of the exterior light. I dunno, I'd love to hear any thoughts.

···

----

      Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

Hi Rob,

Did you try Carsten's suggestion of setting -av 0 0 0 (i.e., render = -av 0 0 0 in the .rif file)? I should have realized that reducing -lw was going to increase your results rather than decreasing them. Setting -av to 0 will decrease them, at least slightly. If you say you're seeing similar anomolies in the actual model, though, perhaps there is something to it. Are you using mkillum to compute the output of your skylight? Have you tried that as well?

I think I'd have to see the actual model to be any real help at this point.
-Greg

···

From: Rob Guglielmetti <[email protected]>
Date: Tue May 13, 2003 9:22:50 AM US/Pacific
To: [email protected]
Subject: [Radiance-general] Trans Oddities Part Deux
Reply-To: [email protected]

Well, I tried Greg's suggestion of setting the limit weight to a very low number, but all that did is manage to scrounge up a few additional lux. This actually made things look worse, because my problems are with the really dense shades (which are already coming up with higher numbers than I'd expect).
...

Hi Rob,

before I engage myself in any further speculations: please mail me your
input-file (the one with the test-model you wrote about), then I'll do sme
testing here.

-cb

Hi Rob

My fear is that I'm simply trying to find too small a percentage of the
available light, and relative comparisons between different shade
materials will not be valid for this model. (?) I mean, this test model
still has two pieces of diffuse glass in it, and the skylight is prety
small, so already the building form/glazing is removing a ton of light
from the space. Then when I start throwing shades in the skylight
cavity, we're getting down to really tiny fractions of the exterior
light. I dunno, I'd love to hear any thoughts.

If there is some error with the indirect calculation arising from the
settings and very low light levels you may be able to pick this up by
radically scaling up the light coming from the sources, then pfilt down the
results by the same factor.

cheers
alex

···

*******************************************************
A. J. Summerfield [email protected]
Faculty of Architecture, University of Sydney
*******************************************************
Outside of a dog, a book is man's best friend.
Inside a dog, it's too dark to read. Groucho Marx
*******************************************************

alex summerfield wrote:

If there is some error with the indirect calculation arising from the
settings and very low light levels you may be able to pick this up by
radically scaling up the light coming from the sources, then pfilt down the
results by the same factor.

I'm not sure I follow you here. But your suggestion to use pfilt implies that I'm creating renderings, when in fact I'm not. I'm only using rtrace to computer some hard numbers on this project (and my test scene). Could you expand a little bit on your suggestion though? I'm always looking to learn new tricks ...

···

----

      Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

OK Rob,

I mean if you generate a CIE overcast sky with horizontal diffuse
illuminance of 10,000 lux (to use the example of John Mardaljevic in
Rendering with Radiance, p357) and since Radiance uses luminous efficacy of
179 lumens/watt:

!gensky -ang 45 0 -c -B 55.866

Run your test scene and repeat for for hdi of 100,000 lux

!gensky -ang 45 0 -c -B 558.66

Then it should produce results 10 times higher.
(You can do a similar thing for sunlight with gensky -R setting to set the
horizontal direct illuminance).

Anyway check that this test is running OK with the diffuser that is
'working' - then move onto the problem diffuser.

If it doesn't show the proportional relationship then perhaps some of your
settings at these low lighting levels mean that absolute rounding may be
having an undue effect. At low light levels small absolute errors will
produce large percentage errors in the quantitative results.

As you may be relying on a relatively few measurement points in very low
light levels - i reckon it would be worth doing some small renderings so you
can 'see' what is going on - for instance observing the distance between the
ambient calculation points causing light leakage problems (that appear as
splotches in images) or detecting modelling errors that are not significant
at the higher illuminance levels.

The renderings could be used like an overture calculation - ie reuse the
ambient file to recalculate your illuminance values and compare with
previous results. Again there should be consistency in the results.

hope this helps
cheers
alex

···

alex summerfield wrote:

If there is some error with the indirect calculation arising from the
settings and very low light levels you may be able to pick this up by
radically scaling up the light coming from the sources, then pfilt down the
results by the same factor.

I'm not sure I follow you here. But your suggestion to use pfilt
implies that I'm creating renderings, when in fact I'm not. I'm only
using rtrace to computer some hard numbers on this project (and my test
scene). Could you expand a little bit on your suggestion though? I'm
always looking to learn new tricks ...

----

    Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

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

*******************************************************
A. J. Summerfield [email protected]
Faculty of Architecture, University of Sydney
*******************************************************
Outside of a dog, a book is man's best friend.
Inside a dog, it's too dark to read. Groucho Marx
*******************************************************

Hey Rob,

How are you determining the reflectance of your trans materials, (measurement or cut sheet)?

-Jack

Rob Guglielmetti wrote:

···

Well, I tried Greg's suggestion of setting the limit weight to a very low number, but all that did is manage to scrounge up a few additional lux. This actually made things look worse, because my problems are with the really dense shades (which are already coming up with higher numbers than I'd expect).

I have a spreadsheet here with the results of all my tests so far. If anyone's interested in looking at it I'd be happy to share it. Perhaps there's a problem with my interpretation of this data. I dunno. But the tests indicate that whenever I try a 1% shade in any combination, I get very odd results from rtrace. And the whole reason I'm doing this test is because we were getting similar odd results on our real model; I was trying to simplify things so I could perhaps isolate something that was causing these oddities. So far, no joy. I shared my trans material definitions for all these tests earlier, but if anyone wants them to have a look I can send those as well.

My fear is that I'm simply trying to find too small a percentage of the available light, and relative comparisons between different shade materials will not be valid for this model. (?) I mean, this test model still has two pieces of diffuse glass in it, and the skylight is prety small, so already the building form/glazing is removing a ton of light from the space. Then when I start throwing shades in the skylight cavity, we're getting down to really tiny fractions of the exterior light. I dunno, I'd love to hear any thoughts.

----

     Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

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

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

Jack de Valpine wrote:

Hey Rob,

How are you determining the reflectance of your trans materials, (measurement or cut sheet)?

Hi Jack,

I measured the glass. For the shades, I'm estimating the reflectance and transmittance. We're trying to determine how dense the shades need to be to achieve a certain light level. The theory being that once we know what they need to be, we can specify that, and then measure actual samples to further validate the whole mess.

···

----

      Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

alex summerfield wrote:

OK Rob,

I mean if you generate a CIE overcast sky with horizontal diffuse
illuminance of 10,000 lux <snip>

Anyway check that this test is running OK with the diffuser that is
'working' - then move onto the problem diffuser.

Oh, now I get it. Thanks for explaining that to me. I will try that. I'm swamped with other things at the moment, so my investigation into this has stalled, but I will try your technique soon enough.

If it doesn't show the proportional relationship then perhaps some of your
settings at these low lighting levels mean that absolute rounding may be
having an undue effect. At low light levels small absolute errors will
produce large percentage errors in the quantitative results.

That was my fear.

As you may be relying on a relatively few measurement points in very low
light levels - i reckon it would be worth doing some small renderings so you
can 'see' what is going on - for instance observing the distance between the
ambient calculation points causing light leakage problems (that appear as
splotches in images) or detecting modelling errors that are not significant
at the higher illuminance levels.

Good point. They just take forever with all the bouncing and all. Byt maybe it's warranted.

The renderings could be used like an overture calculation - ie reuse the
ambient file to recalculate your illuminance values and compare with
previous results. Again there should be consistency in the results.

Good point (again), though I thought that rtrace didn't really benefit from an overture calculation. Perhaps that's YAM (yet another misunderstanding). Hmmm....

hope this helps

Yeah, I think it does, thanks! I appreciate your replies.

···

----

      Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

Hey Rob,

Do the shades all have the same reflectance or (Cr, Cg, Cb)?

I think you indicated that the glass is translucent, what is the makeup? How are you measuring it?

-Jack

Rob Guglielmetti wrote:

···

Jack de Valpine wrote:

Hey Rob,

How are you determining the reflectance of your trans materials, (measurement or cut sheet)?

Hi Jack,

I measured the glass. For the shades, I'm estimating the reflectance and transmittance. We're trying to determine how dense the shades need to be to achieve a certain light level. The theory being that once we know what they need to be, we can specify that, and then measure actual samples to further validate the whole mess.

----

     Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

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

.

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

Jack de Valpine wrote:

Hey Rob,

Do the shades all have the same reflectance or (Cr, Cg, Cb)?

No, and this is one of the things I have been wondering about. It was my understanding that when specifying trans, that transmittance was soley determined by the Td and Ts parameters (talking about the "Ehrlich assumptions" that Chas has set up in RwR). But perhaps that's it, the net transmission of the trans is calculated *after* the reflectance has been taken into account? I did a test of that while testing all these shades, where I had a 5% transmissive shade, and ran it once with the reflectance at 20%, and another where the shade was 70% reflective, but the results were very similar, so I thought that my original assumption was true, that transmission was unaffected by reflectance. But in looking at the data from the second pass of these tests, where Greg suggested the -lw bit, the values for the 70% reflective shade are much lower than the values for the 20% reflective shade. Hmmm...

However, the difference in reflectance of my shades is not that great. the 10% shade has a 30% grey reflectance, and the 5% and the 1% shades are 20% reflective. I don't see how a 10% change in reflectance could affect the calculated values by 100%, which is what's happenning. Hmmm, again.

Still, thinking about it right now, it seems like that might be it. It seems obvious. Now. I will test it out right now. If this is the culprit, I owe you a beer.

I think you indicated that the glass is translucent, what is the makeup? How are you measuring it?

The glass is another story, I won't bore you with the details, I'm already way over my bandwidth limit on this list this month, but it was extensively measured and validated against a 1/2 full scale physical mockup, and inspected by Greg. I'm confident in my glass material definitions, at least.

···

----

      Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

Hi Rob,

My two cents on the subject.

Have you tried raising -ab? I typically use -ab 5 when doing rtrace calcs - and this just comes from case studies I've done and read, it seems to get the calc within a 1% for most cases.

Also, the -av setting, for such a low light level calc seems like it could have the impact you are seeing, as has been discussed. Have you run this with a -av 0 0 0 setting yet?

Are the interior illuminances you are reporting an average using a grid of points or a single point?

Finally, the average workplane illuminance will only be linear (ie. half the Tvis = half the workplane illuminance) to the combined Transmittance of the skylight/shade system if the interior reflectance of the system is constant. I only briefly looked at how you defined the shades and noticed the reflectance is slightly different (although it doesn't seem different enough to give an 100% error). I also noticed the shades are mostly specularly transmissive. If the skylight is not completely diffuse transmittance, then you are getting some sort of sun patch on your workplane. If you are doing an average calc using a grid of points, a sun patch could cause the average workplane to seem off, although it may be off the same for each of your runs.

Also, perhaps a rendering of the room will reveal any oddities that are occuring. Whenever I get unintuitive results, a rendering will often show me whats happening.

Hope this helps.
Zack

···

--
Zack Rogers
Staff Engineer
Architectural Energy Corporation
2540 Frontier Avenue, Suite 201
Boulder, CO 80301 USA

tel (303)444-4149 ext.235
fax (303)444-4304

Hey Rob,

According to RwR (p. 325), actuallly the one place Rd has impact is on Argument 6 of the trans parameters. Based on some reverse calculations to figure out what you might have been doing, I also came up with 30% diffuse reflectance for the 10% shade and 20% for the 5%, 1%, 0.1% and 0.5% shades, thus my question. Based on my calculations here is what what you get:

10% shade at 30% reflectance:
Rd = 0.3
Rs = 0
Td = 0.005
Ts = 0.095
thus the following trans parameter line
7 0.4 0.4 0.4 0 0 .25 .95

10% shade at 20% reflectance:
Rd = 0.2
Rs = 0
Td = 0.005
Ts = 0.095
thus the following trans parameter line
7 0.3 0.3 0.3 0 0 .33 .95

FYI:
A6 = (Td+Ts)/(Rd+Td+Ts)
A7 = Ts/(Td+Ts)

I hope that I have done all my calcs correctly here. I expect others can point out if I am incorrect. I am not sure if this ends up pushing your study in the right direction. I will be interested to hear.

Also, I am quite interested in hearing about your methodology and experience measuring the glazing materials.

-Jack

Rob Guglielmetti wrote:

···

Jack de Valpine wrote:

Hey Rob,

Do the shades all have the same reflectance or (Cr, Cg, Cb)?

No, and this is one of the things I have been wondering about. It was my understanding that when specifying trans, that transmittance was soley determined by the Td and Ts parameters (talking about the "Ehrlich assumptions" that Chas has set up in RwR). But perhaps that's it, the net transmission of the trans is calculated *after* the reflectance has been taken into account? I did a test of that while testing all these shades, where I had a 5% transmissive shade, and ran it once with the reflectance at 20%, and another where the shade was 70% reflective, but the results were very similar, so I thought that my original assumption was true, that transmission was unaffected by reflectance. But in looking at the data from the second pass of these tests, where Greg suggested the -lw bit, the values for the 70% reflective shade are much lower than the values for the 20% reflective shade. Hmmm...

However, the difference in reflectance of my shades is not that great. the 10% shade has a 30% grey reflectance, and the 5% and the 1% shades are 20% reflective. I don't see how a 10% change in reflectance could affect the calculated values by 100%, which is what's happenning. Hmmm, again.

Still, thinking about it right now, it seems like that might be it. It seems obvious. Now. I will test it out right now. If this is the culprit, I owe you a beer.

I think you indicated that the glass is translucent, what is the makeup? How are you measuring it?

The glass is another story, I won't bore you with the details, I'm already way over my bandwidth limit on this list this month, but it was extensively measured and validated against a 1/2 full scale physical mockup, and inspected by Greg. I'm confident in my glass material definitions, at least.

----

     Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

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

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

Hi Zack!

My two cents on the subject.

Most appreciated. Thanks to you and all the others who have weighed in. We'll figure this out! First off, I haven't been able to test anything else today; right after I told Jack I would look at the relationship of reflectance to all this the phone rang, and a client changed my entire afternoon. You know how it goes.

Have you tried raising -ab? I typically use -ab 5 when doing rtrace calcs - and this just comes from case studies I've done and read, it seems to get the calc within a 1% for most cases.

OK, another thing to try. My tests on the gallery space showed that -ab 4 was sufficient, but the light levels were higher in those spaces. Maybe for these new areas I'm studying, I need to bump it up a notch. God, they're going to take forever to run, but it I get the "truth" as a result, it'll be worth it.

Also, the -av setting, for such a low light level calc seems like it could have the impact you are seeing, as has been discussed. Have you run this with a -av 0 0 0 setting yet?

No, I must try this.

Are the interior illuminances you are reporting an average using a grid of points or a single point?

Single point, again in the interest of time. Maybe I can try an average for this test model though.

Finally, the average workplane illuminance will only be linear (ie. half the Tvis = half the workplane illuminance) to the combined Transmittance of the skylight/shade system if the interior reflectance of the system is constant. I only briefly looked at how you defined the shades and noticed the reflectance is slightly different (although it doesn't seem different enough to give an 100% error). I also noticed the shades are mostly specularly transmissive. If the skylight is not completely diffuse transmittance, then you are getting some sort of sun patch on your workplane. If you are doing an average calc using a grid of points, a sun patch could cause the average workplane to seem off, although it may be off the same for each of your runs.

Yes, see my reply to Jack about that. It may be part of the cause, but as you say the differences in reflectance are not as great as the error. Yes, the shades have a large specular component, but they are sandwiched between two very diffuse pieces of glass, so no sun patch.

Also, perhaps a rendering of the room will reveal any oddities that are occuring. Whenever I get unintuitive results, a rendering will often show me whats happening.

Yeah, Alex suggests the same thing. Perhaps...

Hope this helps.

Yup, thanks. Hopefully tomorrow I can try all these new ideas you all have proffered. I can tell from these replies that you are really looking at my material definitions; I appreciate your time and thoughts, everyone!

P.S. Didn't I see an ArchEnergy rendering in Architectural Record a few months ago? I meant to post something to the list about it. I can't remember the issue or the context, but there was a rendering of that Phipps project you have on your website. Congrats!

···

----

      Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

Jack de Valpine wrote:

Hey Rob,

According to RwR (p. 325), actuallly the one place Rd has impact is on Argument 6 of the trans parameters. Based on some reverse calculations to figure out what you might have been doing, I also came up with 30% diffuse reflectance for the 10% shade and 20% for the 5%, 1%, 0.1% and 0.5% shades, thus my question. Based on my calculations here is what what you get:

10% shade at 30% reflectance:
Rd = 0.3
Rs = 0
Td = 0.005
Ts = 0.095
thus the following trans parameter line
7 0.4 0.4 0.4 0 0 .25 .95

10% shade at 20% reflectance:
Rd = 0.2
Rs = 0
Td = 0.005
Ts = 0.095
thus the following trans parameter line
7 0.3 0.3 0.3 0 0 .33 .95

You didn't have to reverse engineer my parameters, just ask! Here is a bit of the materials, as defined in my spreadsheet I created specifically to apply Chas' formulae from RwR. I make assumptions in the left column (Cr, Cg, Cb, Rs, Sr, Td, Ts), and the spreadsheet does the math to generate A1-A7.

Material Name: shade.10
Assumptions Trans Parameters
Diffuse reflectance, RED Cr 0.3000 A1 0.400000
Diffuse reflectance, GRN Cg 0.3000 A2 0.400000
Diffuse reflectance, BLU Cb 0.3000 A3 0.400000
Reflected Specularity Rs 0.0000 A4 0.000000
Surface Roughness Sr 0.0000 A5 0.000000
Diffuse Transmissivity Td 0.0050 A6 0.250000
Transmitted Specularity Ts 0.0950 A7 0.950000
Photopic Average refl. (calculated) Rd 0.3000
Total transmission 10.00%
Percent specular 95%

Material Name: shade.05
Assumptions Trans Parameters
Diffuse reflectance, RED Cr 0.2000 A1 0.250000
Diffuse reflectance, GRN Cg 0.2000 A2 0.250000
Diffuse reflectance, BLU Cb 0.2000 A3 0.250000
Reflected Specularity Rs 0.0000 A4 0.000000
Surface Roughness Sr 0.0000 A5 0.000000
Diffuse Transmissivity Td 0.0025 A6 0.200000
Transmitted Specularity Ts 0.0475 A7 0.950000
Photopic Average refl. (calculated) Rd 0.2000
Total transmission 5.00%
Percent specular 95%

> FYI:
> A6 = (Td+Ts)/(Rd+Td+Ts)
> A7 = Ts/(Td+Ts)

Well there it is, in black & white. Rd is in the equation for A6. It makes a difference. Shoot. Shoulda looked there in the first place.

I hope that I have done all my calcs correctly here. I expect others can point out if I am incorrect. I am not sure if this ends up pushing your study in the right direction. I will be interested to hear.

I will test it out tomorrow, for sure. Methinks that -ab 5, -av 0 0 0 and all the shades having the same reflectance will generate numbers that make a lot more sense. I hope so, anyway.

Also, I am quite interested in hearing about your methodology and experience measuring the glazing materials.

Not until you tell me all the parameters you used to make this (http://www.visarc.com/visarc/projects/images/bu_ab03.jpg) in fifteen minutes! =8-)

Seriously tho, I'm about to blast outta here for the evening, but tomorrow I will try and piece together a cohesive explanation of what we did to measure the glass. Something more detailed than "asked Greg for help", that is.

···

----

      Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

Rob Guglielmetti wrote:

> FYI:
> A6 = (Td+Ts)/(Rd+Td+Ts)
> A7 = Ts/(Td+Ts)

Well there it is, in black & white. Rd is in the equation for A6. It
makes a difference. Shoot. Shoulda looked there in the first place.

You have consulted this flowchart, right? (30% down the page)

  http://www.schorsch.com/rayfront/manual/transdef.html

That makes it visually obvious how all the input values
depend on each other. Unfortunately, you need to think like a
programmer for the sequence to actually make sense... :wink:

-schorsch

···

--
Georg Mischler -- simulations developer -- schorsch at schorsch com
+schorsch.com+ -- lighting design tools -- http://www.schorsch.com/

Wrong. Thanks for that, Georg. Very informative. Did you create
that? Sheesh, the people on this list are just great. Can't wait to
test all this stuff out tomorrow.

-R

···

On 14 May 2003, at 18:28, Georg Mischler wrote:

You have consulted this flowchart, right?

================
Rob Guglielmetti
[email protected]
www.rumblestrip.org

Hey Rob,

Nested below ...

Rob Guglielmetti wrote:

Jack de Valpine wrote:

Hey Rob,

[snip]

I will test it out tomorrow, for sure. Methinks that -ab 5, -av 0 0 0 and all the shades having the same reflectance will generate numbers that make a lot more sense. I hope so, anyway.

I will be interested in hearing how things work out.

Also, I am quite interested in hearing about your methodology and experience measuring the glazing materials.

Not until you tell me all the parameters you used to make this (http://www.visarc.com/visarc/projects/images/bu_ab03.jpg) in fifteen minutes! =8-)

Man oh man, you are really going to make me regret the statement ;-> That is certainly not a fifteen minute image! Although for that project it is definitely faster than some of the others, eg the atrium space images really crunch (~500 lights and illums). A 15 minute image is easy for exterior scenes under daylight conditions!

···

Seriously tho, I'm about to blast outta here for the evening, but tomorrow I will try and piece together a cohesive explanation of what we did to measure the glass. Something more detailed than "asked Greg for help", that is.

----

     Rob Guglielmetti

e. [email protected]
w. www.rumblestrip.org

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

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

Yes indeed, I should have mentioned this myself! Georg's flowchart is immensly helpful for understanding and determining trans parameters. Many thanks to Georg!

-Jack

Georg Mischler wrote:

···

Rob Guglielmetti wrote:

> FYI:
> A6 = (Td+Ts)/(Rd+Td+Ts)
> A7 = Ts/(Td+Ts)

Well there it is, in black & white. Rd is in the equation for A6. It
makes a difference. Shoot. Shoulda looked there in the first place.
   
You have consulted this flowchart, right? (30% down the page)

http://www.schorsch.com/rayfront/manual/transdef.html

That makes it visually obvious how all the input values
depend on each other. Unfortunately, you need to think like a
programmer for the sequence to actually make sense... :wink:

-schorsch

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