[deadalvs]

.. I don't need the whole image at SL22 ..

Just remembered my wish .. from may 2007 :
http://maxwellrender.com/forum/viewtopi ... =2&t=22831
[ Material-ID-regions to drive the 'weigh' of the sample-distribution, though of course, that would mean a 'dynamic SL' per pixel. ]

[photomg1]

it was also suggested not so long ago , to say for example stop the main render at say sl16 then draw a render region over the area that needs more sl and let the render resume but only on that area.
Its either too hard to implement and idea like that (which I could well believe).My guess it's maybe viewed as some how cheating, or not being looked at as it wont work with animation.

Juan thanks so much for posting those links earlier a very interesting read , I didn't realise how fragmented the whole field is with many different techniques . I also thought it was a good proof that the
right technique (algorithm) should work best under every situation . Which obviously hasn't been found yet!

VCM looked interesting , did you have any thoughts on it?

[polynurb]

[ Material-ID-regions to drive the 'weigh' of the sample-distribution, though of course, that would mean a 'dynamic SL' per pixel. ]

i doubt it is that easy, even in the thread you posted there is an example of a glass sitting in front of the floor material.
but it could also be in a mirror, or the floor is first refracted and then visible in a mirror.. how would the software know where to start and where to stop?

theoretically, at a certain diffusion depth, isn't every pixel is somehow related to every other pixel?

i'd guess one would need to calculate some "pre pass"(bias.......) to determine how a certain material would behave in reflection/refraction in order to calculate the appropriate sl for a given pixel.

[deadalvs]

luckily it's a wishlist and not a 'wish and solve list'.

it's just a feature many people seem to wish, since mostly only 'some' materials or image regions have that 'issue'. how to solve this technically I do not know.

I just thought it's funny that some wishes remain wishes for a long time. but that comes with software development. I see that day by day in our own software too.

[polynurb]

luckily it's a wishlist and not a 'wish and solve list'.

hehe.. i do like the mat ID idea, i have had my own thoughts about it..just imagine you could add/remove objects without re-rendering the entire image!

[seghier]

is it possible if they create sl per objects
and in studio we choose for example sl=18 for render the scene and sl=22 for some objects.

[Fernando Tella]

is it possible if they create sl per objects
and in studio we choose for example sl=18 for render the scene and sl=22 for some objects.

That would be a nightmare to setup, no?

[seghier]

is it possible if they create sl per objects
and in studio we choose for example sl=18 for render the scene and sl=22 for some objects.

That would be a nightmare to setup, no?

why ? in v3 exist multi alpha for any object and you apply material for any object . so it's not nightmare to choose objects with sss material

[Fernando Tella]

Oops I was thinking about what deadalvs said about doing it per matID and quoted you by mistake. Per object seems more feasible.

[seghier]

Oops I was thinking about what deadalvs said about doing it per matID and quoted you by mistake. Per object seems more feasible.

no problem

[deadalvs]

That would be a nightmare to setup, no?

Personally, I'd make a 'per material' setting, which sets the 'sampling preference', 1 being the default for all materials.

The higher the value, the more likely, a pixel of that material in screen space is 'supersampled'. (again, this may speak completely against the nature of the 'linear SL' for the entire image.) If you have 1 material which needs a bit of focus, e.g. glass, when you have 30 materials, set it's preference to 5.

With more threads becoming available over time, you could do the same stuff via assigning more render threads to those same specific render regions > same logic.

If it's no problem to increase the SL (quality) of a rectangular selection in an image, it 'should be no problem' to engineer this a bit further to support any greyscale mask instead a rectangular mask.

[Asmithey]

Rendering speed is a non issue for me thanks to rendering farms. Even for test renders, if I need them fast. Pad your fee a bit and problem solved.

Lucky you, being able to pad any industrial design phase with $1000 for 10 high-quality test renderings. Unfortunately, in the academic and small studio environment, render farm usage by way of fee-padding is not an option.

It has nothing to do with luck really. It is all about communication with your client. First, I am speaking on a professional level here. People who are using Maxwell as a source of income. When I started telling my clients that I needed to add in rendering fees for my projects, they balked at first because they did not understand the rendering farm concept. Mainly because they did not understand the rendering process. But when they saw the fees were not that much and they got their rendering the same day, they were sold on the idea and ok with the fees.

[feynman]

People who are using Maxwell as a source of income.

Surely, that might be just so. However, in automotive, industrial, medical and packaging design, the source of income is the total service provided, not use and output of specific tools. Over the course of a project, certain tools available in the studio environment are constantly used to progress the project at hand in the best possible way. The pencil I grab to draw form variations to explain to my studio colleague cannot be billed, equal to the lump of modelling clay I may reheat to shape two alternative shapes that could be promising later on. Tools are consumed in the studio's operating expenses; the client usually does not care how many blocks of Clark foam have been sculpted on route to the next design solution. The same goes for the heavy churn of daily/weekly design process renderings, none of which usually reach the client's attention, as their use is for internal team, group or supplier communication.

[Polyxo]

It has nothing to do with luck really. It is all about communication with your client.

I think you completely missed the crucial part that feynmnan does not work in your Industry. Fulltime Visualizer vs Industrial Designer.
That makes a difference. Renderings are the main thing you deliver, in ID renderings likely don't even appear explicidly in the overall bill.

Imagine 3 people using Photoshop. The first does image retouching and compositing, the second uses the program for website wireframing,
the third paints Mangas all day. Do you think these guys will share the same opinion – say about the qualities of PS's brush engine?

[Mihai]

Regardless of who is missing who's point, or which field you work in - you should also try to make your own workflow as efficient as possible. Designers, as an example, I've seen a few working on a model and start FIRE on their second 30 inch monitor, in full screen.... Ok, yes it wasn't really intended for that, and it will take about 20x longer to see your test, compared to having it maybe 1/5th of the screen.

10 high quality test renders? I hope it's possible to explain to the customer that these are tests and the slight noise they see will not be in the final render. Or if you prefer, do some heavy noise reduction. You don't have to let each of those test renders be at 3000x3000 pixels and SL 18. There seems to be a contradiction here. You say - people won't notice if we add a sort of quality slider. If we lower bounces, if there is less shadow definition, less details in textures etc. they will still like it. On the other hand, you say it takes a really long time to show them several test renders. The difference between say SL 13 and SL 18 can be 30 min and 8 hours...noise reduction people.

So my point is, there are some simple quick things many users could do to greatly speed up their workflow. To me, many of them seem evident but from what I've seen, they don't seem to be to many users and it's almost like they want to sabotage themselves and work in the most painful way possible.