[x3d-public] rendering the X3D standard...is it extensive enough
Leonard Daly
Leonard.Daly at realism.com
Wed Oct 18 15:54:16 PDT 2017
On 10/18/2017 10:07 AM, John Carlson wrote:
> I guess there's no way to get the graphics card vendors into line,
> except to standardize the images output from the graphics cards. Why
> don't we do that? Voltages anyone?
John,
You appear to have forgotten about the primary input-to-user device --
the display. Most displays are digital now, so checking voltages is
probably not enough, you would need to check the digital values sent to
the device. But devices aren't calibrated and degrade over time.
The X3D Conformance testing is carried out against a reference set of
images/movies by visual inspection. When it was created it was
determined that it could not be a pixel comparison for the following
reasons:
* Different OS may apply different color transfer functions when
displaying results
* Different hardware (GPUs, displays, etc.) may use different
algorithms to render the output
* Display and pixel sizes may vary independently from small (think
phone) to large (think room-size cave).
* Human navigation is never exact or repeatable
An application that produces widely different color results for
different runs is not very deterministic.
X3D is neither mathematical, nor engineering -- it is computer "science"
(in the abstract) and computer engineering (in the implemented). It is
intended as Greg Couch mentioned to "archiv[e] the description of the
scene". It is operating at the boundary of what we can describe and what
we can implement.
Leonard Daly
>
> On Oct 18, 2017 1:00 PM, "John Carlson" <yottzumm at gmail.com
> <mailto:yottzumm at gmail.com>> wrote:
>
> I mean, are we mathematicians or engineers? We appear to be
> engineers. I'm a mathematician looking on in horror to what the
> engineers did to my wonderful computing device.
>
> On Oct 18, 2017 12:55 PM, "John Carlson" <yottzumm at gmail.com
> <mailto:yottzumm at gmail.com>> wrote:
>
> Here's the possible scenario, I have a vendor that says they
> meet the standard, but have wildly different renderings
> between runs. Do they get the X3D blessing because they
> implement the format, or is there an image standard that they
> much match, and who is doing the checking? Is it automated?
> It would seem not. What use is a standard which can't be
> verified?
>
> Yes, I am aware of the X3D resources thumbnails, I'm just
> wondering about any standard certification of browser renderings.
>
> I'm laughing too. X3D appears to be a joke.
>
> John
>
> On Oct 18, 2017 12:45 PM, "John Carlson" <yottzumm at gmail.com
> <mailto:yottzumm at gmail.com>> wrote:
>
> The question becomes, can I expect a single browser to
> generate identical pixels between runs, and if not, how is
> testing done?
>
> Thanks,
>
> John
>
> On Oct 18, 2017 12:31 PM, "Greg Couch" <gregc at cgl.ucsf.edu
> <mailto:gregc at cgl.ucsf.edu>> wrote:
>
> On 10/18/2017 08:53 AM, John Carlson wrote:
>
> I am wondering if the X3D standard is complete or
> extensive enough to provide for pixel perfect
> equality between browsers. This is important for
> archiving. Has this been a goal?
>
> Thanks,
>
> Johb
>
>
> Thanks for the laugh. The OpenGL/Vulkan/Direct3D/etc.
> specifications have never required pixel perfect
> equality between graphics cards, so it has not ever
> been a goal AFAIK. Are they close? Yes, very close,
> but not pixel perfect. The great thing about X3D is
> that you're archiving the description of the scene not
> the image of the scene.
>
> HTH,
>
> Greg
>
>
>
> _______________________________________________
> x3d-public mailing list
> x3d-public at web3d.org
> http://web3d.org/mailman/listinfo/x3d-public_web3d.org
--
*Leonard Daly*
3D Systems & Cloud Consultant
LA ACM SIGGRAPH Past Chair
President, Daly Realism - /Creating the Future/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://web3d.org/pipermail/x3d-public_web3d.org/attachments/20171018/fdbf86b9/attachment.html>
More information about the x3d-public
mailing list