Renderworks 2011 versus Renderworks 2010 better or worse?

[gScott]

are the fun police 'disappearing' posts?

[VincentCuclair]

Apparently we have to keep it 'professional', one could demand the same of VWs in several cases :grin:

I think it was I who spoiled the 'fun'.......

Edited February 18, 2011 by Vincent C

[Dave Donley]

Yes threads will get locked and posts deleted if forum members go about insulting their peers. Not fun.

[gScott]

ooooh, they deleted another one!

and i haven't insulted anyone,

yet...

[Kaare Baekgaard]

I got lost in the previous thread with the same name, but here is my experience:

There is nothing wrong with the new rendering engine. It's the textures, that makes the difference. By default all 2011 textures are set to zero blurriness - and for a good reason: Even a slightly higher setting increases rendering times several times.

But to get a high degree of realism, some surfaces needs to be blurred to some extend - stainless steel and anodized aluminium being among them. In these cases, the old rendering engine is actually faster.

I hope, it is a glitch, that will be corrected eventually.

[Mattheng]

I read the previous post on Renderworks with the same name as this one and started having a go myself. Looks like I missed the boat with that thread but I have got some renders if anyone is still interested.

I liked billtheia's renders so they are pretty similar to his.

I also had a go with indirect lighting to see what happened.

The intensity of the environment lighting was set to 100% in 2010 and 200% in 2011, everything else was the same.

Like bill I think 2011 renders much more quickly with environment lighting. Even more quickly if you turn Indirect lighting on.

2010 renders noticeably faster than 2011 if there is no environment lighting and you use Ambient lighting with a light source for the sun.

I also found this somewhere else on this forum. Click the link and then at the bottom of the page click for a pdf about the change from 2010 to 2011. It actually makes a lot of sense.

http://www.vectorworks-training.co.uk/Buy_Vectorworks/Renderworks.aspx

It's by someone called Tamsin Slatter who seems to be pretty clued up.

Matt

[gScott]

matt

to return to the original contention of the thread,

your 2011 renders are a lot more washed out,

are there settings you might easily change to

get the same level of "vividness" in the 2010

render? obviously neither are "realistic" but

the 2011 renders are just boring.

could this be because you are mimicing a 2010

setup? would a default 2011 render be any different?

[billtheia]

It's almost as if the gamma is turned up in the VW2011 renders.

Compare the blue in the skies of the attached renders from VW2010 & VW2011. You'll see that the blue sky in VW2010 is quite a bit darker.

In fact, if you take the resulting VW2011 JPG file into photoshop and turn the gamma down to .45, you get a result much closer to the VW2010 render (see attachment.)

You can get a similar result by turning the intensity of the HDRI background down but that only helps the lighting, the sky remains washed out.

The problem is, I don't know which one is more realistic. I generally like the higher contrast of the VW2010 renders but I often found them to have too much contrast. The VW2011 renders do feel washed out but I wonder if it's only because I've compared them to the VW2010 renders.

I do know that I tried, several times, to get something close to photorealistic renders from VW2010 & earlier with mediocre results. Some of the renders I've seen from VW2011 are definitely more realistic.

For non-photorealistic renders (like those that I do most often,) I think that the contrast of VW2010 was a better fit.

[billtheia]

BTW, both of the examples above were modeled independently (not imported from previous version,) both use the same HDRI background set to 100%, and both have ambient light turned off.

For the record, I think that the default glass in VW2011 is more realistic (for use in windows) than the default glass in VW2010 (which always distorted the objects beyond too much) IMHO.

[billtheia]

I wonder if the "Linear Workflow" (part of the C4D engine) mentioned in the previous thread is the culprit here. Again, I'm not sure which engine gives more realistic (true-to-life) results.) It does seem, though, that, if it is the Linear Workflow, it might be a good idea to have an option to turn it off.

[billtheia]

For those who are interested, here's a comparison of HDRI lighting with 0, 1, & 2 bounces.

1 bounce changes the render only slightly and introduces some kind of glitch under the gable end. 2 bounces blows the whole thing out.

[Danielj1]

Bill,

It all depends on which HDRI background you use for lighting purposes, and its brightness. You can increase or reduce the HDRI's brightness to provide additional control. You can create a background that is brighter on one side than the other, and that will impact the image as well. You can also render in Custom RW and make adjustments to the Indirect Lighting, Soft Shadows and Environment Lighting settings to help with the scene's contrast. And, of course, you can adjust the Image Exposure (%) as well. So there are a number of settings that you can combine to avoid blowing the scene out (or to do it if that is your preference).

Dan J.

[billtheia]

Bill,

It all depends on which HDRI background you use for lighting purposes, and its brightness. You can increase or reduce the HDRI's brightness to provide additional control. You can create a background that is brighter on one side than the other, and that will impact the image as well. You can also render in Custom RW and make adjustments to the Indirect Lighting, Soft Shadows and Environment Lighting settings to help with the scene's contrast. And, of course, you can adjust the Image Exposure (%) as well. So there are a number of settings that you can combine to avoid blowing the scene out (or to do it if that is your preference).

Dan J.

All true, Dan.

I'm just trying to get my bearings with the new rendering engine. It behaves differently than the last one(s) so I'm trying to figure out where the differences are and how I need to (or if I need to) adjust.

I'm of two minds about the change to the C4D engine. I think that the new engine is, ultimately, a good thing and will result in better images in (usually) less time. I'm not thrilled, however, with the prospect of sinking more time into learning how to get what I need out of RW2011 - I couldn't possibly count the hours I have already spent learning previous versions of RW.

Thanks for the Getting Started with RW2011 booklet, BTW. I have only glanced at it thus far but it looks like it will be quite helpful.

[mike m oz]

It seems to me that the frustration that exists with the Renderworks 2011 is because Final Quality Renderworks no longer produces as good a quality render as it did in previous versions.

To get a good quality renders now users are having to use Custom Renderworks and make adjustments to settings that many don't understand. Its the uncertainty of this trial and error process which is the primary cause of the angst.

[Danielj1]

One key thing to keep in mind is that control of the rendering and lighting qualities are now two separate things, so even if you use FQRW, you must still go to the Lighting Options to make a quick couple of adjustments there, in order to get optimal results. As a reminder, the previous technology (i.e. Lightworks) went through a number of FQRW setting iterations throughout its life; FQRW in the 2010 version looked and acted differently than in the first version, years earlier; not that it makes much difference in this particular case.

I disagree that FQRW no longer produces as good a quality render as it did previously. If you have your Lighting Options set well, the resulting images are far superior to previous ones (in my opinion, anyway) with global illumination and zero artifacts to boot. And if you've worked with HDRI coming in through windows and bouncing off interior glass elements, it's not even just a quality difference any more--we can now produce images that were previously impossible to make.

I'm not promoting my book here. But do take a look at the settings for the different kinds of scenes shown there, for some initial hints on setting up files for rendering. You'll see a workflow that is quite different than previously. It may be inconvenient to figure out a new workflow when the previous one was useful to you. But I think you will ultimately be very pleased with the results and the rendering speed.

Dan J.

[Danielj1]

Here's an interior FQRW sample, with Lit Fog. First go-around with little or no tweaking done yet (sorry for the interior-only view; it's all I had on my home machine at the moment).

Dan J.

[img:left]http://www.danjansenson.com/FQRWlitfog.png[/img]

Edited February 21, 2011 by Danielj1

[Mattheng]

Ok, had another look and now have got some non environment lit renders as comparison.

These were a lot quicker than the environment lit ones.

Interestingly, as I mentioned above, 2010 is quite a bit quicker with ambient and a light source.

I also considered radiosity (in 2010 speak ) vs indirect lighting (in 2011 speak). This can be applied with and without environment lighting.

Even without environment lighting 2010 with radiosity runs extremely slowly, but does make a noticeable difference.

Indirect lighting without environment lighting in 2011 seems to make a minimal difference

I have actually got six files so I will have to split this over two posts.

First 2010:

Edited February 21, 2011 by Mattheng

[Mattheng]

Now 2011.

To be honest there isn't a lot of difference with the number of bounces, maybe around the window on the right of the house but it's not big.

It does seem like radiosity and environment lighting in 2010 make a big difference but take a long time, infeasibly so for radiosity.

Against this nothing 2011 makes a huge difference, radiosity (or indirect lighting as it's called) and environment lighting don't radically change things even the render times don't change by anything like the order of magnitude as they do in 2010.

I'll go away and have a deeper look but the final score so far seems to be.

Speed : without environment lighting or radiosity : 2010 wins by a healthy margin

with environment lighting or radiosity 2011 wins by a massive margin (potentially the 5-7 times that was mentioned in the marketing blurb, in fact if you use radiosity in 2010 I'd say this was conservative)

Look : obviously very subjective but, looking at billtheia's results as well as mine, 2011 does seem to wash out the details whereas 2010 gives more resolution.

Matt

Edited February 21, 2011 by Mattheng

[billtheia]

Funny, I was working on a similar comparison.

I turned off the HDRI to see how direct lighting worked in each VW2010 & VW2011.

The following renders have NO ambient lighting, just a single, directional light source set at 100%. For VW2010, I did FQRW & Final Quality Radiosity. For VW2011, I did FQRW with 0, 1 & 2 bounces.

I didn't time the renders but VW2010 radiosity was by far the slowest. The bounce-less FQRW was probably a bit slower in VW2011 than in VW2010. The bounced FQRW renders in VW2011 were slower than they were without bounces but were definetly faster than VW2010 radiosity.

As for quality, I'll let you all decide.

[billtheia]

And here are a couple of VW2011 renders with bounced light & soft shadows.

[billtheia]

It just occurred to me that we were always complaining that it was almost impossible to get RW2010 & earlier to render a white surface that looked white. Well, there's no doubt that my model is white in RW2011.

[Mattheng]

bill, I was actually colouring mine grey so that I could see them, here are some non environment lit ones in white.

It looks a bit like a blizzard has hit by the time you get to four bounces.

I will do another post for the environment lit equivalents.

[Mattheng]

Here are the white environment lit versions with a lightsource to add shadows.

It does seem that you have to be careful with the level of the environment lighting in 2010 when you add a light source otherwise it blows out a bit.

2011 just seems crazy even by two bounces, looks like you are better off without the indirect lighting.

The gamma correction point that you made, Bill, seems to be a good one. The Linear workflow C4D issue is all that I can think it might be. Can we turn it off?

Matt

[billtheia]

I'd turn ambient lighting off (in VW2011.) With indirect lighting on, you really don't need it and you'll get better contrast.

[Mattheng]

Bill,

I reckon I could reduce the intensity by about 50% as well but it was set to a level of 100 which I read in one of the many bits of information is the correct number for 2011.

I'm just trying to get my bearings with the new rendering engine. It behaves differently than the last one(s) so I'm trying to figure out where the differences are and how I need to (or if I need to) adjust.

I'm of two minds about the change to the C4D engine. I think that the new engine is, ultimately, a good thing and will result in better images in (usually) less time. I'm not thrilled, however, with the prospect of sinking more time into learning how to get what I need out of RW2011 - I couldn't possibly count the hours I have already spent learning previous versions of RW..

I have to say that I'm with you on this one, Bill. 2011 looks intriguing but it certainly doesn't work the same way as 2010.

I was getting into HDRI lighting in 2010 and I can't get the same effect in 2011. (yet, I'm still trying)

I would love to try turning off the gamma correction built into 2011 (Linear workflow?) to see what difference that makes.

The good thing about 2011 is that the render times might double but they won't rise exponentially as in 2010 with HDRI and, especially, radiosity.

(Have to say I haven't tested Kaare's idea about blurry textures, this might be its Achilles heel)

The bad thing is it looks different and, if you don't care about HDRI lighting, FQR is quite a bit slower.

Matt