Hi Massimo,
On 18/06/2016 15:41, Massimo Callegari via Development wrote:
Hi everyone,
today, instead of being the usual "complaining guy" I'd like to explicit my
biggest congratulations and gratitude to all the people behind the Qt3D module.
In particular to Sean Harmer and Paul Lemire at KDAB, who have been doing a
monumental job working during weekends, fighting against CI breakage and taking
care of an infinite number of JIRA issues and reports.
I've been following the Qt3D developments for a couple of years and I have to
say Qt3D is going to be an awesome addition to the Qt features set.
I tried as much as I could to support the developments by testing patches and
reporting issues and I hope I've been somehow helpful for the project.
Thank you for your kind words. Yes, please do keep the bug reports
coming and prodding us when needed.
I'm also very excited by the Qt3D editor and I'd like to congratulate with all
the Qt people behind it as well.
Even if it's in a very early stage, you can already feel it is going to be an
awesome tool !
+1. It's already shaping up to be a nice addition. I hope it can
continue to be extended and hopefully one day become a great plugin for
Qt Creator
So, well done everybody !
Thank you and I'd like to echo my own thanks to everybody who has helped
with code, reporting bugs, asking questions that pushed us to make sure
Qt 3D is flexible enough to work with as many rendering algorithms as
possible. I already have plenty of ideas on how to continue extending
the frame graph in 5.8 and beyond plus a whole bunch of other additions
that we discussed with Pasi and his team in Oulu recently.
Now I've got a question for the Qt3D guys who knows all the techy bits of the
current status of the module.
It is my intention to implement the technique described in these AMD slides:
https://developer.amd.com/wordpress/media/2012/10/Mitchell_LightShafts.pdf
I am wondering if Qt3D can already cover such usage or if there are missing
bits that are still not there or not planned at all.
I don't see anything in there that Qt 3D can't do with a suitable
configuration.
A plus for my project is if that technique can work in a deferred rendering
pipeline.
That technique relies upon alpha blended, view-aligned quads. It would
need to be applied after the deferred portion of the rendering as alpha
blending doesn't play nicely with deferred. So something like this might
be something to try:
1) G-buffer pass
2) Lighting pass of opaque geometry
3) Light shafts pass for the brightest n lights, where n is tuned to
what your hardware can handle.
Depending upon the scene you want to render you may be able to hand
select which lights trigger light shafts. For example, if you have 10
bright spot lights, use those with light shafts and just rely upon some
other post-processing effect like bloom for the dimmer lights or point
lights which can be done in a single post-proc pass.
If KDAB/Qt is interested in such usage case, I am more than willing to
contribute to the developments or even support a developer more skilled than me
in the 3D area with a monetary contribution.
Sounds like an interesting algorithm to implement. If you'd like to do
it on a commercial basis then feel free to contact us.
Have a nice weekend and thank you again for your kind words and help!
Sean
--
Dr Sean Harmer | sean.har...@kdab.com | Managing Director UK
KDAB (UK) Ltd, a KDAB Group company
Tel. +44 (0)1625 809908; Sweden (HQ) +46-563-540090
Mobile: +44 (0)7545 140604
KDAB - Qt Experts
_______________________________________________
Development mailing list
Development@qt-project.org
http://lists.qt-project.org/mailman/listinfo/development
