Which image is the most photo-realistic - take the Rendering Challenge

For all of you PhotoWorks and non-PhotoWorks users (using other
rendering products), can you please take a moment to take the following
survey:

http://www.surveymonkey.com/s.asp?u=245782574173

At the end of the survey, you will be redirected to the Photo-Rendering
Challenge. Over the last few weeks the PhotoWorks team of developers
and management have been involved in a Photo-rendering challenge to see
who can create the most photo-realistic rendering. Please help us by
clicking on the link above, take the survey then cast your own vote for
the most photo-realistic image. When you are finished voting you will
be redirected to a page that reveals what rendering product was used
and how people voted.

This exercise had a duel purpose in that the PhotoWorks team just
wanted to do it for the challenge but more importantly we want to learn
from the challenge and the survey that is going out, where we can
improve.


Thank you for your participation.

Mark Biasotti
SolidWorks

 

Mark,

It would have been nice to see all of the renders with the same material
color/texture and from the same angle with the same lighting setup so that
all of the reflections were the same.
One can't help allowing orientation of the product (with all of the
reflections and highlights) to sway their opinions.
That said, it is obvious that SW is serious about a rendering package, and
for that I for one am glad. It is still a LOT easier for me to export out of
SW into Rhino and render using Flamingo. Maybe it's just familiarity but I
not a stupid person (OK some might argue) but I just don't get PW.

My 2 cents,
Muggs

 

Hi Mark,

It's great to see SolidWorks taking an larger interest in the
PhotoWorks product. For me it was easy spotting the PhotoWorks
renderings versus the other images in the survey. I'm excited to see
where all this information is going to lead us PhotoWorks users!

Rob

 

I had to read that a few times myself. Each can only have one # so as to
list them 1 thru 7.

But of course you knew that.

Muggs

 

LOL. Never even considered that.
I guess if you're a middle of the road kind of guy (or gal) then you're OK,
but if you're the type that everything is great or everything stinks then
you're hosed.

Muggs

 

I picked Maxwell, but I didn't like the lenses in any of them.

 

Paul, sorry for the confusion. This question is a ranking question
where you are suppose to enter numbers 1 thru 7 with a unique number
for each choice. I'll re-word the question to make this clear.

 

Yes, I agree that from a technical perspective, more variables could
have been eliminated, but we did lay down rules, and for the most part,
they were followed; yellow body, black hardware on a white tabletop.

Thanks for taking the survey and your feedback.

Mark

 

Mark...can you post the files? Perhaps Rob could set this up as the
next Photoworks challenge on his site. Let's see what some of these
Photoworks power users can come up with.

 

Mark,
First, great idea. Good way to promote the product and what it can do.


I have seen the new stuff in 2007, and it looks extremely promising.
For instance, the new materials libraries - at least as far as I have
dug into them - finally rock. They seem to satisfy that animated
discussion we had a few years ago with Gibson about what needs to be in
there.

Well done (so far as I have seen). Take everyone involved out for ice
cream or beers or whatever (not meant ironicly, in case someone
misreads that last line).

But I think you have fallen just-shy of an opportunity to make a really
compelling case for PWx that will satisfy even skeptics like me.

If you do a PWX challenge again, as a consumer evaluating renderers, I
would want to see more apples-to-apples samples.

-Same view and perspective for all images
-same lights (within the limitations of the comparison, of course - if
one package has an area light, in the other package put their
best-substitute light in the same location)
-same background/environment image (again, as is compatible with the
limitations of the renderers being evaluated) At least not if you used
comparable default environments when revealing the results.

If you did the above I would learn, for real, the assets and
liabilities of the packages.
Though the presentation is impressive, all I learned <for sure> was the
quality of the folks setting it up, which we have learned from Rob's
contests can vary wildly. I don't even know if you had Maxwell pros or
neophytes - nothing sours a comparison like the pssibility of cooking
the results (though I know you - I trust you gave it an honest effort)

Other things that would be worth noting in the renderings (maybe
revealed after voting):
-setup time
-rendering time (which could seriously ding maxwell as I understand it
from the newsgroup, having not tried it myself, ) PWx is pretty darn
fast compared to the field, and fast=money
-geometry translation time (again, a big ding to other renderers, and
the only reason I still use PWx. Excepting, perhaps large assemblies -
I wonder if other renderers have to spend a few mintues crunching
geometry each time you push the render button?)
-How much experience with each product? I know you guys target making
PWx as seamless and easy as possible - wouldn't it be a powerful pitch
if you could take ten novices, give them ten hours each, sick them
randomly on the different packages, and see what they could do with the
above constraints (views and lights already set?). Man, that would
make a case.

Again, Mark, good step in the right direction.
But as someone who is seriously considering my rendering options, I
want to see a compelling and realistic example of why I should not
stray. All this did was make me pause, and for me the rendering
quality of PWx was never a question - its all the other issues that
make me yearn for greener pastures.

You have a great opportunity to squash doubters if you can make the
business case for PWx by showing as-true-as-possible a comparison.

Ed

 

MAB - Thanks Ed, that means a lot coming from you.

MAB - And that would mean taking you out too since you've put in
hundreds of hours of critique and feedback on this product.

MAB - This started as a an-adhoc challenge between Joe and I 2 weeks
ago. All we did was made the results public and have them voted on. I
know that there could have been more variables eliminated but have you
ever tried to get 7 adults to play by the rules?

MAB - The renderings were done by myself, Joe Dunne, Mick Kellman, John
Cochrane, Ron Bates.

MAB - The Maxwell Rendering generally took over night while the PW
renderings took an hour or two. Lightwave took about 4 hours.

MAB - Yes, I would like to have this exercise done by Rob's Contest,
but I don't think that the customer will release the files, but I'm
checking into it.

MAB - You can see from the survey that you know what areas we are
suspecting we need to be working on, and the 350+ respondents so far
are confirming most of those with amazing precision.

 

-How much experience with each product? I know you guys target making

Hi guys. In response to the above comments. If the customer doesn't
release the files we could do another test using a different (publicly
releasable) file(s). I'm not sure how much time SW can or wants to spend on
something like this but, as you know my (or the PW user community's) contest
happens every month. All we really need to do is agree on an acceptable
model to post for the October contest. The next step; of course would be to
make sure the previously mentioned SW employees render the new model. We
could even change the rules a bit for the October contest to limit the
variables for a better comparison. I've been allowing people to post non PW
renderings for the contest (they just aren't eligible for voting) and
occasionally this does happen. Maybe the people who have posted in the past
using V-ray and Blender would give the October model a go? If we did this
for the October contest it would also give us time to get the word out for
maximum participation.

Rob