Digital Art Zone

 
   
3 of 10
3
3Delight Discussion
Posted: 26 September 2012 04:47 PM   [ Ignore ]   [ # 31 ]
Member
Rank
Total Posts:  49
Joined  2004-05-24

You can get a real wiz bang system for under $500, easily.
Cheaper still if you buy a bare bones box, the parts separately and put it together yourself.
I have three words for you about where to shop: tigerdirect, tigerdirect, and tigerdirect.

Profile
 
 
Posted: 26 September 2012 06:16 PM   [ Ignore ]   [ # 32 ]
Power Member
Avatar
RankRankRank
Total Posts:  1874
Joined  2006-02-17
vrba79 - 26 September 2012 04:47 PM

You can get a real wiz bang system for under $500, easily.
Cheaper still if you buy a bare bones box, the parts separately and put it together yourself.
I have three words for you about where to shop: tigerdirect, tigerdirect, and tigerdirect.

tigerdirect=screw-job.  You’ll spend more money sending back defective goods and mis-shipped parts than you’ll save.  With tigerdirect it is caveat emptor to the max.

Kendall

 Signature 

Any opinions expressed in this post are those of Kendall Sears and may, or may not, be more, or less, valid than any other opinion.

The contents of this post are intended for the DAZ forum only, do not re-post any portion to any other forum without his permission.

Profile
 
 
Posted: 26 September 2012 06:24 PM   [ Ignore ]   [ # 33 ]
Member
Rank
Total Posts:  49
Joined  2004-05-24
Kendall Sears - 26 September 2012 06:16 PM
vrba79 - 26 September 2012 04:47 PM

You can get a real wiz bang system for under $500, easily.
Cheaper still if you buy a bare bones box, the parts separately and put it together yourself.
I have three words for you about where to shop: tigerdirect, tigerdirect, and tigerdirect.

tigerdirect=screw-job.  You’ll spend more money sending back defective goods and mis-shipped parts than you’ll save.  With tigerdirect it is caveat emptor to the max.

Kendall

Really? I’ve yet to have a problem with them, and I’ve been dealing with them for years.

Profile
 
 
Posted: 26 September 2012 06:32 PM   [ Ignore ]   [ # 34 ]
Power Member
Avatar
RankRankRank
Total Posts:  1874
Joined  2006-02-17

I’ve had dealings with at least 2 dozen folks over the years that have had very bad dealings with TD.  I’ve personally had one order messed up royally by them (for well over two thousand dollars), but after some “persuasion” they made good.  I’ve had a couple messed up slightly, but the cost to “fix” the problem was more than the amount lost.

There are some times that I will chance a purchase from TD, but it is only with extreme caution.  I will not recommend them to any of my
clients, as TD’s failure will reflect on me.

EDIT:  But this thread isn’t about TD, but about 3DL.  grin

Kendall

vrba79 - 26 September 2012 06:24 PM
Kendall Sears - 26 September 2012 06:16 PM
vrba79 - 26 September 2012 04:47 PM

You can get a real wiz bang system for under $500, easily.
Cheaper still if you buy a bare bones box, the parts separately and put it together yourself.
I have three words for you about where to shop: tigerdirect, tigerdirect, and tigerdirect.

tigerdirect=screw-job.  You’ll spend more money sending back defective goods and mis-shipped parts than you’ll save.  With tigerdirect it is caveat emptor to the max.

Kendall

Really? I’ve yet to have a problem with them, and I’ve been dealing with them for years.

 Signature 

Any opinions expressed in this post are those of Kendall Sears and may, or may not, be more, or less, valid than any other opinion.

The contents of this post are intended for the DAZ forum only, do not re-post any portion to any other forum without his permission.

Profile
 
 
Posted: 26 September 2012 06:35 PM   [ Ignore ]   [ # 35 ]
Member
Rank
Total Posts:  49
Joined  2004-05-24

Hey man, I ain’t calling you a lair, I’ve just never had any problems with them personally.
But yeah, we’re getting OT. Back on topic, I’ve got nothing but good things to say about using the Standalone version. I’ve been playing with it ever since mid day yesterday. I’ve been trying light set ups that take an obscene amount of time when rendered in-studio, and to my delight they render at a very quick pace in standalone, like Lightdome Pro.

Profile
 
 
Posted: 26 September 2012 06:40 PM   [ Ignore ]   [ # 36 ]
Power Member
Avatar
RankRankRank
Total Posts:  1874
Joined  2006-02-17
vrba79 - 26 September 2012 06:35 PM

Hey man, I ain’t calling you a lair, I’ve just never had any problems with them personally.
But yeah, we’re getting OT.

I’m not insinuating that you are. grin  Folks will have a lot of different experiences with any vendor.  I happen to deal with a large number of people in IT on a regular basis, and am more likely to run into more people who have had problems than most.  I just wanted to pass along a precautionary heads-up for those who may find TD’s prices tempting, but who may also be ill equipped to deal with the extra costs sometimes associated with buying from TD.

Kendall

 Signature 

Any opinions expressed in this post are those of Kendall Sears and may, or may not, be more, or less, valid than any other opinion.

The contents of this post are intended for the DAZ forum only, do not re-post any portion to any other forum without his permission.

Profile
 
 
Posted: 26 September 2012 07:00 PM   [ Ignore ]   [ # 37 ]
Member
Rank
Total Posts:  123
Joined  2011-11-11
Kendall Sears - 24 September 2012 09:46 PM

[BINGO.  There are many applications where one would need to render non-physical (non-nature) light.  This would include pretty much all movies as there is significantly more light in a movie scene than exists in nature.  Which is why 3Delight has been used in all of the Harry Potter and X-Men movies.

Unbiased rendering, for the time being, is not useful in production.

Kendall

The key to doing that in Lux is to light your scene as if you were lighting it for a movie. Now that said, if you’re absolutely doing things that won’t work physically (like say, a vampire scene and need the vampire to not cast a reflection) then you really can’t use a physical unbiased renderer.

 Signature 

Andrew Dacey
Photographer
Geek

Profile
 
 
Posted: 26 September 2012 09:39 PM   [ Ignore ]   [ # 38 ]
Addict
Avatar
RankRankRankRank
Total Posts:  12578
Joined  2007-11-06

...well, if I could get 4.5 to behave while setting up a scene. I can use the latest 3Delight Standalone as it should be able to interpret the shaders correctly.  Tried it with a scene set up in 3Advanced only to get a tonne of TDLMake errors and a lovely render of the .obj file.

As my current system is only a DuoCore, not losing any “performance” with the free licence.

 Signature 

...it’s five minutes to midnight…

I’d rather have a blue sky above me than a blue screen in front of me.

Profile
 
 
Posted: 26 September 2012 11:09 PM   [ Ignore ]   [ # 39 ]
Power Member
Avatar
RankRankRank
Total Posts:  1874
Joined  2006-02-17
adacey - 26 September 2012 07:00 PM
Kendall Sears - 24 September 2012 09:46 PM

[BINGO.  There are many applications where one would need to render non-physical (non-nature) light.  This would include pretty much all movies as there is significantly more light in a movie scene than exists in nature.  Which is why 3Delight has been used in all of the Harry Potter and X-Men movies.

Unbiased rendering, for the time being, is not useful in production.

Kendall

The key to doing that in Lux is to light your scene as if you were lighting it for a movie. Now that said, if you’re absolutely doing things that won’t work physically (like say, a vampire scene and need the vampire to not cast a reflection) then you really can’t use a physical unbiased renderer.

Where unbiased renderers fall down is anywhere where full daylight/moonlight isn’t available or wanted.  Go outside in the countryside (where there’s no light pollution) on a night with a new moon (or before the moon has risen).  One will see exactly ... nothing.  (I live in such a place.)  Unbiased renderers will give one exactly that in a nighttime scene: nothing.  And they would be right.  However, try recording a scene that way… disembodied voices anyone?  Watch any movie scene set at night away from “neon”.  The train yard, the cemetery, the warehouse, the abandoned boiler room, the slum alleyway.  Where is all of that light coming from naturally?  Nowhere.  In real life, physically, there would be little to no light, but in the movie scene we can see clearly.  This is where the unbiased renderers fail; they’re supposed to.  Adding lights really doesn’t help, as they act like light would, which is exactly what one doesn’t want.

Similarly, many times lights are added even outdoors in full daylight.  The sun projects shadows that play h*ll with cameras, so flood lights will be used to soften the shadows, or even remove them.  Again, an unbiased renderer will give one physically correct light, but not the light needed for the purpose.

In this case, it’s using the right tool for the job.  Unbiased renderers are great when striving for photos, not so much for situations where the lighting, by its very nature, is contrived.

Don’t get me wrong.  I like Lux, I’ve liked it since its inception.  I’m interested in Octane, once it is a bit more capable.  I’ve written radiosity renderers in the past, as well as raytracers and scanline renderers, and other not so well known types.  They each have their place.  For the moment, unbiased rendering is limited to photo-realism.  However 90%+ of all production rendering is not “real light” and therefore is in the realm of biased renderers.

Kendall

 Signature 

Any opinions expressed in this post are those of Kendall Sears and may, or may not, be more, or less, valid than any other opinion.

The contents of this post are intended for the DAZ forum only, do not re-post any portion to any other forum without his permission.

Profile
 
 
Posted: 26 September 2012 11:23 PM   [ Ignore ]   [ # 40 ]
Addict
Avatar
RankRankRankRank
Total Posts:  12578
Joined  2007-11-06

...this is why I still like using LDP2, SLP as well as making up my own light sets.  Lighting in and of itself is a very powerful medium. and biased rendering brings out what I feel is a more “expressive” quality.

I’m not even totally won over by IDL except when using CG models with photo based backdrops/settings (which is one of it’s better strong points).

 Signature 

...it’s five minutes to midnight…

I’d rather have a blue sky above me than a blue screen in front of me.

Profile
 
 
Posted: 27 September 2012 10:04 AM   [ Ignore ]   [ # 41 ]
Member
Rank
Total Posts:  49
Joined  2004-05-24

This rendered in about a half a minute, using 3Delight standalone, using the lighting set that came with the scenery. The only post work done was a day-to-night shift in the color. I had D|S and Chrome running to boot.

Image Attachments
so_quiet.png
Profile
 
 
Posted: 27 September 2012 12:58 PM   [ Ignore ]   [ # 42 ]
Active Member
Avatar
RankRank
Total Posts:  676
Joined  2004-11-12

Kendall

Kendall Sears - 26 September 2012 11:09 PM
adacey - 26 September 2012 07:00 PM
Kendall Sears - 24 September 2012 09:46 PM

[BINGO.  There are many applications where one would need to render non-physical (non-nature) light.  This would include pretty much all movies as there is significantly more light in a movie scene than exists in nature.  Which is why 3Delight has been used in all of the Harry Potter and X-Men movies.

Unbiased rendering, for the time being, is not useful in production.

Kendall

The key to doing that in Lux is to light your scene as if you were lighting it for a movie. Now that said, if you’re absolutely doing things that won’t work physically (like say, a vampire scene and need the vampire to not cast a reflection) then you really can’t use a physical unbiased renderer.

Where unbiased renderers fall down is anywhere where full daylight/moonlight isn’t available or wanted.  Go outside in the countryside (where there’s no light pollution) on a night with a new moon (or before the moon has risen).  One will see exactly ... nothing.  (I live in such a place.)  Unbiased renderers will give one exactly that in a nighttime scene: nothing.  And they would be right.  However, try recording a scene that way… disembodied voices anyone?  Watch any movie scene set at night away from “neon”.  The train yard, the cemetery, the warehouse, the abandoned boiler room, the slum alleyway.  Where is all of that light coming from naturally?  Nowhere.  In real life, physically, there would be little to no light, but in the movie scene we can see clearly.  This is where the unbiased renderers fail; they’re supposed to.  Adding lights really doesn’t help, as they act like light would, which is exactly what one doesn’t want.

Similarly, many times lights are added even outdoors in full daylight.  The sun projects shadows that play h*ll with cameras, so flood lights will be used to soften the shadows, or even remove them.  Again, an unbiased renderer will give one physically correct light, but not the light needed for the purpose.

In this case, it’s using the right tool for the job.  Unbiased renderers are great when striving for photos, not so much for situations where the lighting, by its very nature, is contrived.

Don’t get me wrong.  I like Lux, I’ve liked it since its inception.  I’m interested in Octane, once it is a bit more capable.  I’ve written radiosity renderers in the past, as well as raytracers and scanline renderers, and other not so well known types.  They each have their place.  For the moment, unbiased rendering is limited to photo-realism.  However 90%+ of all production rendering is not “real light” and therefore is in the realm of biased renderers.

Kendall

Hmmmm .... OK, I’m not trying to pick a fight or derail the discussion (which is quite interesting), but I fail to see the logic here, or possibly my logic is completely flawed. Maybe you could help me understand your view point a bit better.

As I understand it, by definition, unbiased render engines use real light physical poperties (or real light physics) to “light” the scene. On a film/studio set, real lights are used to light the scene. So, by definition, if one had all the proper information regarding lighting and set/studio environment, would that not produce the exact same lighting as was found in the studio/set (assuming that the software had the proper tools to reproduce the lighting like softboxes, reflective umbrellas, etc.)?

A biased renderer uses light simulation “cheats” (for lack of an easier term), such as ambient occlusion, to provide rapid sampling/simulation of the effects of real light physics without encountering the overhead of actually performing the calculations for every mm of surface/light interaction required by an unbiased renderer. The beauty and utility of a biased render engine is the dramatic increases in rendering speed while using predictable approximations of lighting effects. The downside to biased render engines is that it can often be quite difficult to achieve equivalent lighting effects compared to unbiased render engines (caustics for example).

It’s my understanding that where unbiased render engines fall down in the production environment is not their inability to faithfully reproduce studio/set lighting. They “fail” in the production environment due to their inability to produce images quickly. The additional calculation times required for the ray-tracing of lighting based on real light physics would easily push production times from hours/days to months to render 5-10 minutes of film. At least that is my understanding and logic.

So, in your example above, where does the lighting come from for the filming of a night time scene? From lights, the moon, etc., unless they are using infrared cameras (and that would just look weird, unless that was the desired effect). Wouldn’t an unbiased renderer with the same lighting/environment, and camera/film setup achieve the same lighting as the nighttime scene of the set/studio? Take a visible spectrum camera outside when there is no moon, cloudy with no stars, and no ambient light from artificial lighting (like in/near a city), and you’ll get the exact same thing as your example above with any renderer or visible light spectrum camera (doesn’t matter if it’s a biased or unbiased renderer). Regardless of medium, no visible spectrum “light” in the scene means black render or black photo/film/image.

 Signature 

My Rendo Gallery
My RDNA Gallery
My DAZ 3D Gallery

Profile
 
 
Posted: 27 September 2012 01:36 PM   [ Ignore ]   [ # 43 ]
Power Member
Avatar
RankRankRank
Total Posts:  1874
Joined  2006-02-17
dustrider - 27 September 2012 12:58 PM

Hmmmm .... OK, I’m not trying to pick a fight or derail the discussion (which is quite interesting), but I fail to see the logic here, or possibly my logic is completely flawed. Maybe you could help me understand your view point a bit better.

Not picking a fight, nor derailing.  It has to do more with implementation than anything else.  In short, any renderer (with enough effort) can replicate the results of any other, barring deliberate design limitations.  We’re not going to get into those right now. grin

As I understand it, by definition, unbiased render engines use real light physical poperties (or real light physics) to “light” the scene. On a film/studio set, real lights are used to light the scene. So, by definition, if one had all the proper information regarding lighting and set/studio environment, would that not produce the exact same lighting as was found in the studio/set (assuming that the software had the proper tools to reproduce the lighting like softboxes, reflective umbrellas, etc.)?

You’ve answered the question right there.  Current unbiased renderers are predominately “direct” light renderers.  Most movie sets RARELY will use direct lighting, unless the director is looking for a specific effect.  Lighting will be filtered (providing limited wavelengths, or specularity), reflected or diffused (softening shadows, providing ambient), and any of a plethora of other modifications outside the scope of this discussion. 

A biased renderer uses light simulation “cheats” (for lack of an easier term), such as ambient occlusion, to provide rapid sampling/simulation of the effects of real light physics without encountering the overhead of actually performing the calculations for every mm of surface/light interaction required by an unbiased renderer. The beauty and utility of a biased render engine is the dramatic increases in rendering speed while using predictable approximations of lighting effects. The downside to biased render engines is that it can often be quite difficult to achieve equivalent lighting effects compared to biased render engines (caustics for example).

Speed is indeed one reason for the methodologies used, but the ability to simulate the myriad techniques used in “lighting” is the main reason for the “cheats” as you call them.  To the contrary, most biased renderers are fully capable of “physical light simulation.”  Those methodologies are rarely used because they have a deleterious effect on what is being attempted—the replication of “contrived” lighting.

It’s my understanding that where unbiased render engines fall down in the production environment is not their inability to faithfully reproduce studio/set lighting. They “fail” in the production environment due to their inability to produce images quickly. The additional calculation times required for the ray-tracing of lighting based on real light physics would easily push production times from hours/days to months to render 5-10 minutes of film. At least that is my understanding and logic.

So, in your example above, where does the lighting come from for the filming of a night time scene? From lights, the moon, etc., unless they are using infrared cameras (and that would just look weird, unless that was the desired effect). Wouldn’t an unbiased renderer with the same lighting/environment, and camera/film setup achieve the same lighting as the nighttime scene of the set/studio? Take a visible spectrum camera outside when there is no moon, cloudy with no stars, and no ambient light from artificial lighting (like in/near a city), and you’ll get the exact same thing as your example above with any renderer or visible light spectrum camera (doesn’t matter if it’s a biased or unbiased renderer). Regardless of medium, no visible spectrum “light” in the scene means black render or black photo/film/image.

OK, my question was rhetorical as I presumed the reader probably knew the answer.  In nighttime scenes, “blue” lighting is used to increase the recordable luminosity without increasing the ambience, and hence illuminating everything.  If you’ll look closely at those “nighttime” scenes you’ll see the “blue.”  It is not wholly dissimilar to “black” light which will only increase the luminosity of specific elements.  This is the reason one can see details in a “dark” area without getting the “nightvision” effect.  It is these types of effects that current unbiased renderers don’t handle.

Good questions.  As I stated above, it is not that unbiased rendering *cannot* do these effects, it is that the current unbiased renderers *don’t* handle these effects.  I’m sure that in due time this will change.  Also, your statements about product generation time… while previously quite true, are no longer germane.  With the right hardware, unbiased rendering can match, and in some cases, exceed the speed of biased rendering.

At the moment we have a distinction between biased and unbiased rendering only because the attention is on two products that only do only one part of the rendering equation.  Once, or if, Lux and Octane, can handle “contrived” lighting then the distinction will disappear.  It is the same with GPU rendering, it is special only because of it is lacking in a subset of the tools.  There is nothing inherently special about the tech that precludes its use in any one place.

Kendall

EDIT:  I think it may be appropriate at this point to recommend to the interested reader the 2nd edition of “[digital] Lighting & Rendering” by Jeremy Birn.  While it does not cover the absolute latest tech, it does give great insight into why specific rendering technologies are better than others for specific tasks.  Since it is only around 400 pages or so in length, it isn’t an overly hard read.

 Signature 

Any opinions expressed in this post are those of Kendall Sears and may, or may not, be more, or less, valid than any other opinion.

The contents of this post are intended for the DAZ forum only, do not re-post any portion to any other forum without his permission.

Profile
 
 
Posted: 27 September 2012 03:12 PM   [ Ignore ]   [ # 44 ]
Active Member
Avatar
RankRank
Total Posts:  676
Joined  2004-11-12
Kendall Sears - 27 September 2012 01:36 PM
dustrider - 27 September 2012 12:58 PM

Hmmmm .... OK, I’m not trying to pick a fight or derail the discussion (which is quite interesting), but I fail to see the logic here, or possibly my logic is completely flawed. Maybe you could help me understand your view point a bit better.

Not picking a fight, nor derailing.  It has to do more with implementation than anything else.  In short, any renderer (with enough effort) can replicate the results of any other, barring deliberate design limitations.  We’re not going to get into those right now. grin

As I understand it, by definition, unbiased render engines use real light physical poperties (or real light physics) to “light” the scene. On a film/studio set, real lights are used to light the scene. So, by definition, if one had all the proper information regarding lighting and set/studio environment, would that not produce the exact same lighting as was found in the studio/set (assuming that the software had the proper tools to reproduce the lighting like softboxes, reflective umbrellas, etc.)?

You’ve answered the question right there.  Current unbiased renderers are predominately “direct” light renderers.  Most movie sets RARELY will use direct lighting, unless the director is looking for a specific effect.  Lighting will be filtered (providing limited wavelengths, or specularity), reflected or diffused (softening shadows, providing ambient), and any of a plethora of other modifications outside the scope of this discussion. 

A biased renderer uses light simulation “cheats” (for lack of an easier term), such as ambient occlusion, to provide rapid sampling/simulation of the effects of real light physics without encountering the overhead of actually performing the calculations for every mm of surface/light interaction required by an unbiased renderer. The beauty and utility of a biased render engine is the dramatic increases in rendering speed while using predictable approximations of lighting effects. The downside to biased render engines is that it can often be quite difficult to achieve equivalent lighting effects compared to biased render engines (caustics for example).

Speed is indeed one reason for the methodologies used, but the ability to simulate the myriad techniques used in “lighting” is the main reason for the “cheats” as you call them.  To the contrary, most biased renderers are fully capable of “physical light simulation.”  Those methodologies are rarely used because they have a deleterious effect on what is being attempted—the replication of “contrived” lighting.

It’s my understanding that where unbiased render engines fall down in the production environment is not their inability to faithfully reproduce studio/set lighting. They “fail” in the production environment due to their inability to produce images quickly. The additional calculation times required for the ray-tracing of lighting based on real light physics would easily push production times from hours/days to months to render 5-10 minutes of film. At least that is my understanding and logic.

So, in your example above, where does the lighting come from for the filming of a night time scene? From lights, the moon, etc., unless they are using infrared cameras (and that would just look weird, unless that was the desired effect). Wouldn’t an unbiased renderer with the same lighting/environment, and camera/film setup achieve the same lighting as the nighttime scene of the set/studio? Take a visible spectrum camera outside when there is no moon, cloudy with no stars, and no ambient light from artificial lighting (like in/near a city), and you’ll get the exact same thing as your example above with any renderer or visible light spectrum camera (doesn’t matter if it’s a biased or unbiased renderer). Regardless of medium, no visible spectrum “light” in the scene means black render or black photo/film/image.

OK, my question was rhetorical as I presumed the reader probably knew the answer.  In nighttime scenes, “blue” lighting is used to increase the recordable luminosity without increasing the ambience, and hence illuminating everything.  If you’ll look closely at those “nighttime” scenes you’ll see the “blue.”  It is not wholly dissimilar to “black” light which will only increase the luminosity of specific elements.  This is the reason one can see details in a “dark” area without getting the “nightvision” effect.  It is these types of effects that current unbiased renderers don’t handle.

Good questions.  As I stated above, it is not that unbiased rendering *cannot* do these effects, it is that the current unbiased renderers *don’t* handle these effects.  I’m sure that in due time this will change.  Also, your statements about product generation time… while previously quite true, are no longer germane.  With the right hardware, unbiased rendering can match, and in some cases, exceed the speed of biased rendering.

At the moment we have a distinction between biased and unbiased rendering only because the attention is on two products that only do only one part of the rendering equation.  Once, or if, Lux and Octane, can handle “contrived” lighting then the distinction will disappear.  It is the same with GPU rendering, it is special only because of it is lacking in a subset of the tools.  There is nothing inherently special about the tech that precludes its use in any one place.

Kendall

EDIT:  I think it may be appropriate at this point to recommend to the interested reader the 2nd edition of “[digital] Lighting & Rendering” by Jeremy Birn.  While it does not cover the absolute latest tech, it does give great insight into why specific rendering technologies are better than others for specific tasks.  Since it is only around 400 pages or so in length, it isn’t an overly hard read.

Thanks for the clarifications! Now back into lurk and learn mode.

 Signature 

My Rendo Gallery
My RDNA Gallery
My DAZ 3D Gallery

Profile
 
 
Posted: 27 September 2012 03:42 PM   [ Ignore ]   [ # 45 ]
Power Member
Avatar
RankRankRank
Total Posts:  1426
Joined  2009-09-11

vrba79


When you say the standalone are you using the fee or paid version? I am also a Lux user after 7 months I am pretty well versed at using it and love the results it gives me, but this has me curious since i give 2 tier pricing for commissions using UE lighting… How does one send a render to it?

 Signature 

http://bobvan.deviantart.com/

http://www.flickr.com/photos/37909888@N05/sets/

Profile
 
 
   
3 of 10
3