[dd_]

i have to agree that discussing fry here isnt a good idea and should be brought up on a forum that isnt owned by a render engine company.
i think there has been enough heated words said and there isnt a need to produce more... is there?

[pluMmet]

i have to agree that discussing fry here isnt a good idea and should be brought up on a forum that isnt owned by a render engine company.
i think there has been enough heated words said and there isnt a need to produce more... is there?

I've always thought that bring up stuff like this belonged on this forum. Bringing it up on CGtalk in front of people who have no idea what is going on seems unfair to NL?

I must also say that the old adage applies "The easiest way to get over a girl is to go out with another girl."

You've found your new girl dd_. Your thoughts are coming from another standpoint now.

[dd_]

lol good job my wife aint reading this reply.
i still use maxwell and i still use fry oh and i still use max and i still use c4d. i just think that there is really no need to fuel a fire within a fire. trust me when i say if there was a software that offered me a c4d stroke 3ds max i would look towards it. its that easy really. so my thoughts are comming towards a lets not start another fire

and like everyone here im waiting for another update to improve maxwell, im also waiting for the vray port to c4d too

[glypticmax]

Motivating improvements by consumer action is a very tricky process.

Back to work.
cheers.

Yes but without action it has no effect. Please be respectful of those who are effecting action.

I'm probably woefully uninformed, but since joining the Forum last May, I haven't noticed anybody taking credit for effecting or expediting NL's development of Maxwell with posts on the Forum. I have noticed a lot of pointless aggression, rants and other catagories of venting about issues both legitimate and otherwise. And I admit that at times I get frustrated reading such posts, and shoot from the hip. Which does little but help lower my blood pressure after reading something that didn't improve my render skills which why I come here in the first place. My pet peave are posts aimed at individual NL employees that are only trying to do their job under sometimes difficult circumstances.

Also if you know a better way to effect NLs action please share your thoughts when you get around to responding.

I can't really address how to change NL's development of Maxwell. They are a privately held company that can do as they please. They will rise and fall based customer satisfaction with their product.
None of the ideas below should be construed as a suggestion on my part with regard to addressing any issues with NL.
However, there are some basic concepts about exerting influence by disaffected individuals that seem to have been totally ignored.
1) Create a group of like minded individuals that are willing to invest some time and maybe some money to rememdy their problem. The larger your group the greater the liklihood you have some success.
2) This group should establish a priority list of goals. One list could have Most Desireable Goals and another list of Most Obtainable Goals. Getting a little success with an Obtainable Goal (opening productive dialog, perhaps) could help validate the efforts fo the group, and draw additional supporters or members. You should assume the Most Desireable Goals are also the most difficult to achieve.
3) Develop strategies to communicate these Goals to those that can remedy them. Emails, Certified Mail, lawyers, etc are all proven techniques. Going public would be a last resort.
4) If your group is large enough and well financed consider legal action. If the transgression is *actionable* in court, this could be a valueable negotiating tactic.
A cohesive group with reasonable goals, participating in a productive dialog seems to offer a far higher probability of success than random public venting by a disparate group of individuals that share little more than a common level of discontent. Regardless of how valid some of the points may be by the individual communications, if the emotional level is too high, it is very easy for it to obscure the validity of the argument.

I have no clue how many license holders of Maxwell their are. I do know some companies and individuals own dozens of licenses. All license owners should be regarded as *stakeholders*. It really is up to those individuals to get themselves together if they believe such action would effect a positive change.

Personally, I don't see it happening. And because we are talking about software development that is notoriously umpredictable under the best of circumstances, there is nothing to suggest any group action would yield benefits.

Personally, I'm very, very happy with Maxwell Studio and look forward to developments as they are deemed ready to release by NL. If I was involved in creating animations, render farms or worked in a very time sensitive architectural field, I can easily imagine I would not feel this way. But for gem and jewelry renders, Maxwell Studio ROCKS. Nothing is even in second place with regard to ease of use and predictable, high quality renders.

Time to go work on some 12th Cent BC Egyptian knot design intagios.
cheers.

[Jozvex]

Perhaps I should have been more clear. Please do not link my fryrender images here in any way. They are intended only for registered members of the fryrender community.

Sorry Frances I missed where you said that. I've edited it out.

[Frances]

Thanks Jared.

[Jozvex]

This always puzzled me, but i have close to no knowledge about now a raytracer works technically, rather than theoretically.
Anyway, here my question on biased bucket renderers: Since oftentimes the whole scene contributes to the bucket being rendered, may it be in terms of a shadow, diffuse or specular lighting, how can the engine compute the result without having the whole scene cached?

It can do it because raytracing is 'backwards tracing' rather than 'forwards tracing'. In real life 'rays' get shot out of a light source, bounce around etc and then a small portion of them end up entering our eyes. We see stuff due to 'fowards tracing' (basically 'real tracing').

In raytraced bucket renderers like Mental Ray, they use 'backwards tracing', where the rays are shot out of the eyes (the camera) and work their way back to the light sources (which it knows the location of).

It shoots out an 'eye ray' for every pixel in the bucket. They hit the first surface they see. The surface goes "hey, I'm reflective" and the scene goes "there are raytraced shadows in this scene", so new rays are shot out from the points on the surface where the eyes rays hit. 'Light rays' head towards all the known light sources, and if they get blocked along the way by another object, then the surface point they came from must be in shadow. 'Reflection rays' (or sometimes just called 'secondary' rays) shoot out along the angle of reflectance looking for a new surface to appear in the reflection. The reason they use backwards tracing is because it's WAY faster than just "seeing which small amount of rays manage to enter the camera" like in real life.

Samsam is right about exploring volumetric subdivs along the ray path. Before the scene starts testing rays, the scene is chopped up into 'voxels' (volume of pixels). In Mental Ray this is called the 'BSP Tree' (binary space partition tree... eek!). I think in Renderman renderers it's called a 'KD Tree'? Anyway, you chop the scene up into voxels so that when you're testing rays, you can test just one voxel of polygons at a time rather than testing the whole scene with every ray.

In Mental Ray (and in VRay etc too I'm sure) you can tell the renderer how big or small the voxels will be. Having fewer, big voxels uses less memory (voxelizing the scene uses up lots of RAM) but takes longer to render because the rays have more polygons they could potentially run into in each voxel. Lots of tiny voxels uses more RAM but the rays only have to test like 15 polygons at a time along its path so it's much faster to raytrace.

Perhaps it's easier with a narrow bundle of ray traced rays to open up & explore a collection of volumetric subdivs along the vector of the reflection while with MntCrlo-ish rendering everything really does need to be there all the time as you are dealing with a more spherical & diverse set of samples.

You're right that a narrow bundle of rays is faster. When we use glossy/blurry reflections, or Final Gather, or Ambient Occlusion etc where the rays are really spread out, it's much slower as you know. However the whole scene doesn't need to be loaded at once. Don't forget that when we talk about rays being 'shot out' they don't all go bang and shoot out at once. The renderer can test each ray one at a time or in bundles of close rays, to still only need a small section of the scene loaded at a time. Then in moves onto the next set of rays. That's why it's slow.

Wow I hope some of what I said made sense!

[zoppo]

... a lot interesting stuff

hey, thanks! that _was_ interesting, didn't know that.

and to the "should fr be discussed in the mw forum" discussion: i like to read about fr here. i got to cover so many forums i'm glad if i haven't got to watch just another on. and it's interesting to compare the different approaches of these two so similar and in the same time so different products - and who could do this if not people who own both. so where should they go?
do we really need a special mw-fr forum? and if someone there brings in busyray - is he also told to shut up?

[-Adrian]

Thanks for taking the time to explain Jozvex, good stuff!

[x_site]

lets put this business to rest once and for all:

Is it true that:

1] Founders of FR were part of the NL team and aledgedly 'took' the stating code from MR?
2] NL are contesting this issue in court and the future of FR is on the line?

ok i was taking with someone in a well informed position and this is what was transmitted to me... sounds like a spanish drama to me and i hope does not affect too much the work of both parties///

[jeffg]

lets put this business to rest once and for all:

Is it true that:

1] Founders of FR were part of the NL team and aledgedly 'took' the stating code from MR?
2] NL are contesting this issue in court and the future of FR is on the line?

They say on their site that there is absolutely no relationship. They say they have never even met the NL people.

[VisualImpact]

ok i was taking with someone in a well informed position

Yeah, what were you taking

[Julian]

[-Adrian]

[Leonardo]