V-Ray wizzard - request for contribution

[tricky]

At risk of being flamed, can I differ from all the other responses on this thread?

I think this is a terrible idea - it will basically allow any Tom, Dick or Harry to produce good results. We have invested a lot of time and money (and stress) in learning how Vray works from the ground up - its working for us, and gives us a competative edge. If a 'wizzard' was there, any old Cad-monkey could produce a good image.

Yes, it initially sounds nice, but from a purely commercial point of view, "No way! Don't do it!!!"

[tom schuelke]

HE HO .... !!!!!

- it will basically allow any Tom, Dick or Harry to produce good results. way! "

i allready do get very good results....


Well nice idea Vlado..

another thing to think of... for animations, best settings can very much differ..

is there a way, to optimise settings over an animation, meaning, to change parameters smothely ?

of course, the wizard would have to analyse the animation at varying frames, but maybe ... ?

nice idea. imo


Tom

[RErender]

I would like an addition for rendering animations.It would work like this:

Prior to calculating imaps for a ani path it would run along the animation path and determine at which frames new geometry is seen and then calculating the least amount of frames needed to expose all geometry. This could optimize the practice of rendering every nth frame for imaps.

[helpdesk]

Tricky....I don't quite understand your thinking with that response.
I believe anything that VLADO decides could make workflow easier and more efficient out of the box helps everyone. Sorry if you had to take so long learning the ins and outs of what works and how it works. Really that is a good thing that you do understand all that because it allows you to tweak "presets" that way you like. I don't think you will loose any business dude. I think it will definitely make VRAY much more marketable against Plugins such as "Maxwell"...which in my opinion is no match....sure it's great looking and all, but you have to tweak the h3!! outta it to get it even close to render times like VRAY.

[joconnell]

Tricky - One thing about it though is if the settings are really easy to get for a nice render, you'll have far more time to spend on the things that matter like material settings, light placement and colours. Maxwell is nice in the sense that the materials are more of a formality so you can concentrate more on your composition and photography.

Another way of thinking about it is theres a well known post company in soho where the head of 3d apparently went through a period of not hiring anyone that was better than him in an effort to artificially keep himself at the top. Not sure if this was true or not but if it was, he was keeping himself from learning what the better people knew and improving himself. Competition is good because it'll keep pushing you to improve things and be a better 3d op overall.

When it gets to the point where it's really easy to get a photorealistic render then what will be the things that make the difference? It'll be traditional skills like composition, editing, animation, direction and concept - You're probably a more artistic person than "any old cad monkey" so these things will rpboably come far more naturally to you than them and that will become the new difference putting your work above theirs

[joconnell]

Prior to calculating imaps for a ani path it would run along the animation path and determine at which frames new geometry is seen and then calculating the least amount of frames needed to expose all geometry. This could optimize the practice of rendering every nth frame for imaps.

This could cause a lot of problems, much like hiding and unhiding objects or doing polygon reduction on objects that fall outside a cameras field of view. Seeing as an object will have an effect on the light bouncing around a scene, it'll make a difference to the overall solution of the space you're in so you might get weird light solutions if you need to rotate back to look at an area, there might be a hard edged shadow from an object whos samples may have only been taken from a certain view but since it's already appeared in shot by your logic, it mightnt get sampled again for a different view - I reckon you're still better doing this by hand, you might be able to more accurately guess where you need your samples than vray since you know exactly the details that you need to bring out.

[RErender]

Prior to calculating imaps for a ani path it would run along the animation path and determine at which frames new geometry is seen and then calculating the least amount of frames needed to expose all geometry. This could optimize the practice of rendering every nth frame for imaps.

This could cause a lot of problems, much like hiding and unhiding objects or doing polygon reduction on objects that fall outside a cameras field of view. Seeing as an object will have an effect on the light bouncing around a scene, it'll make a difference to the overall solution of the space you're in so you might get weird light solutions if you need to rotate back to look at an area, there might be a hard edged shadow from an object whos samples may have only been taken from a certain view but since it's already appeared in shot by your logic, it mightnt get sampled again for a different view - I reckon you're still better doing this by hand, you might be able to more accurately guess where you need your samples than vray since you know exactly the details that you need to bring out.

Perhaps I wasn't clear enough.

The intent is not to cull objects from the scene, only to optimize the "every nth" rendering of imaps in an animation. For instance, as it stands the basic workflow is to render imaps every 50th frame or every 20th frame etc., this is rather ineficient as perhaps frames 1-120 might be a zoom in 120+ moving into a pan, where the calc for the first frame would suffice for frames 1 -120 you would need more frequent calc's for the pan.

The method I see working would be shooting a rough grid of rays from which one could ascertain whether they are hitting new geometry or not.

make sense?

[joconnell]

Sure yeah - What I meant was it would have to take into account objects that leave the frame and come back in so there aren't any holes left in the sampling - it'd have to almost do a pre pass for each object and find out when its visible in frame and not - it'd be almost like making a bsp tree for the animation - You may end up with inaccuracies if you dont shoot enough rays. I reckon this might still be more sensibly done by the person that made the animation in the first place. You could quite happily use the render dialog to specify the frames you want to render (1,3,100,150,160,300 or whatever) rather than having to go through a fairly large pre process to get correct results since vray doenst know in the grand scheme of things what frames actually matter in the end or not.

[RErender]

I'm not convinced that it would not work

The grid would not need to be that dense that it would take long to sample, maybe one ray per 10 pixels or so and could also be set to sample every nth frame. Each ray would return what object it intersects. Then the values for each frame could be looked at for overlap.

That an object leaves the view and comes back is usually not an issue since it will already be considered in a previous imap. Of course it will fall apart on some crazy camera moves but for 90% it should work.


Since we are looking at automation and convenience I do not think it is too far fetched.

[TRGraphics]

I agree with Tricky. I too have spent a great deal of time and effort learning the program, and still am, by the way. The "magic button" idea is not one I agree with. And if you think something like that won't cost you work, your very, very wrong. The very best quality is not what everyone in the architecture business is looking for. And if your clients know there is a "magic button" for descent quality renderings, your overall value to them will drop as well as more people start using vray, and they will. Which is obviously what chaos wants, I would if I were them.

Competition is a very real thing, and mastering your programs is what gives you the advantage, because you were willing to invest the time and hard work. I freelance, so this may effect freelancers more than established firms.

And to compare vray with Maxwell is obserd. Maxwell is finished as a viable product. Just read their own forum. And they had the so called "magic button". Of course, thats all they really had.

Just my 2 cents worth and there is every chance I could be wrong.

Tony

[flipside]

This stuff about losing business is the same as a few years back, when traditional lighting artists complained about GI renders. Any dick tom and harry would be able to create photoreal images without placing fill lights at every wall and under every table.

A wizard is a good thin imo, especially for learning from it. But nothing beats the actual knowledge of the software. So not just anyone will be able to steal your job. What will they say to to a client who wants a change, but they can't do it because they can only use the wizard and there is no 'make my walls brighter' or 'decrease the color bounce only on this wall' checkbox?

In a few years time, all you know now will also be obsolete because then CPU's will be good enough to render every still with ppt, which only needs one setting to control quality. But by then, a lot of other stuff will be implemented to make images even more real which requires even more cpu power, raising the standards for photoreal again. Have you ever thought about the fact that 99% of all vray settings are only there to fake and to speed up things? Take the speed issue away and you will never have to tweak ir map tresholds and glossy subdivs etc...

So bring on the wizard, and bring on the new load of artists who will finally be able to use vray without a degree in physics and math!

[rpc212]

I am not quite sure what to think about the wizard idea. Some clients appreciate shite so what do you do with them? They also can't think outta the box so they are clueless how to use good 3D images. For example one potential client I have uses absolute crap images to sell their developments and they actually think the images their architect produces are good. They have a show room with all these horrible 3D images everywhere . . . ugh! The shite comment was confirmed by a third party potential client I talked to who made a harsher comment about the other potential clients showroom.

I guess my fear would be the same as Tricky and TRGraphics as if this architect got a hold of a wizard his output would take a big, big leap forward not because of his skills in producing the image but because he bought Vray and can use a wizard.

At the same time I have confidence in myself that I would be able to compete against wizard weilding, non-knowledgable, psuedo 3D artists as, like Tricky and TRGrphics, I have invested a great deal of time learning and still am learning Vray. These other people will not have spent that time with Vray and there in lies the competative advantage. I know, at least to some degree, what I am doing whereas these aforementioned people won't. Anyway I guess I am a fence sitter.

[vlado]

Well, of course knowing what the settings do is always an advantage... It is like programming in assembly language - people can do it for small portions of the code where optimization is critical, and for the other parts of the code they leave the optimization to the compiler, which may not do the best job ever, but for the invested human effort (e.g. none ) it performs very well. Having a good optimizing compiler doesn't make you a good programmer though - it just takes a large technical burden off of your shoulders.

I should note that such a wizard is not at all easy to create as there are many many factors that must be considered, so it won't come in a week or two for sure. For now, it is just an idea...

Best regards,
Vlado

[djembe]

I'm a frequent lurker and infrequent poster, but this topic is very interesting to me.
Personally, I think the idea for a V-Ray wizard is a very good one and I've been hoping for something of the sort.
The argument that it will hurt professional profitability strikes me as a little absurd. (No offense intended to those who hold this opinion.)

My productivity would be improved by any features that allow me "dial in" good render settings more quickly and consistently. I am much more an artist than a scientist. I want to focus my skills on the subject/modeling, composition, lighting and mood (among other, more subjective considerations), not render theory and multiple test renders. To me, getting a job done well and quickly will always help my bottom line.

I spent the first decade of my professional career doing all of my work on a drafting table with technical pens, markers, gouache, airbrush, etc. I still remember quite clearly the attitudes of my colleagues when computer software began taking over their various fields of arch viz, graphic design and illustration. Many were certain that with the advances in software and hardware that any idiot could do their work and we'd all be competing for scraps. Of course, that didn't happen. Oh sure, we all had to redefine how we competed in the marketplace and how to leverage the new tools to our advantage, but what else is new? If a potential client is satisfied with renders that were created without a fine eye for detail, accuracy and good composition, then they were going to undervalue my efforts anyway.

I for one, hope that this idea is pursued further.
Thanks!

[Sawyer]

I think that no doubt this would make the competition greater. On one hand it's frustrating to think of loosing skill. We have taken the time and learned a skill set. Some one then says "Hey you no longer need to know that. You can just push these buttons". Pretty common in this computer world though. Trhuthfully I have my own wizard. Well I do the exact same things everytime to set up a scene. Sometimes I copy directly from a simular file. I don't tweak as much as I used to and I want to tweak less. If Chaos does not give me a wizard I will just refine my own.

Seems that something we all like about this is it's hard to do. We know stuff and it's not easy and we charge because we know stuff. I am better than a cad drafter because I have taught myself how to use these programs and I get better results. Now to many of use there is this real fear that soon the programs will be "easy". That cad drafter doesn't need to know that much and he is getting pretty picts. 5 years from now will Acad have v-ray and all you need to do is press the radio button that says int or ext and in 30 seconds it's rendered? These are my fears. They are not keeping me up at night mostly because few architectural firms I know care that much. Most could be using BIM but are in 2d and the thought of a full time 3d guy is not a reality as most places are understaffed. They really "could" do what I do but they don't and not because archi firms don't have creative people but because they can't afford the man hours off a set of drawings.

Anyway I rant.