Duong, Ninh (Ninh.Duong++at++dsto.defence.gov.au)
Thu, 4 Nov 1999 11:31:30 +1030
I had a look at the Z Buffer resolution in Vega and SensorVision a while
back and this is what I concluded:
(1) The mapping of LOS (line of sight) range to a Z Buffer value is linear
between the near and far clipping planes in orthonormal projection mode.
(2) The mapping function is much more complicated in perspective projection
mode. It appears to have a hyperbolic form. This function seems to come
about because of the matrix equations they use to perform the
transformations.
I'm not sure if this answers your question but I hope it helps anyway.
Ninh
-------------------------------------------------------------
Dr. Ninh Duong
Research Scientist
Defence Science and Technology Organisation
Ph: +61 8 8259 6259
Fax: +61 8 8259 5616
Email: ninh.duong++at++dsto.defence.gov.au
Mail: PO Box 1500, Salisbury SA 5108, Australia
> -----Original Message-----
> From: Rudi Schmidt [SMTP:Rudolf++at++Schmidt.net]
> Sent: Thursday, November 04, 1999 7:20 AM
> To: info-performer++at++sgi.com; info-vega++at++spiffy.paradigmsim.com
> Subject: Z-Buffer Resolution ?
>
> Hi pfFolks!
>
> I'm running a gunnery simulation on an Onyx2 InfiniteReality, with a
> viewing range from 1m to 12km.
> In order to select the correct visual, I need to understand more about
> the actual implementation of the Performer's Z-buffer on an
> InfiniteReaity, and especially its resolution.
> The IR's visuals give me a choice between 15 bit, 16 bit or 23 bit
> resolution.
> QUESTION: Can somebody explain the mapping of these three versions of
> Z-buffer to my viewing range as stated above (near=1.0m, far=12000m),
> depending on the distance of an object in the scene (e.g. what is the
> Z-resolution in 6 kilometres distance from the observer ?
>
> Also, I did read -in a document called "Onyx2 Technical Report" about
> the fact that IR's z-buffer would be compressed, and that an actual 15
> or 16 bit on an IR are euqivalent to a traditional 24 bit z-buffer, as
> well as IR's 23 bit Z should be the same as conventional 32 bit.
> QUESTION: Can somebody explain this "Z-compression" ?
>
> pfThanks in advance.
>
> -- Rudi Schmidt, KMW
> --------------------
> Rudolf Schmidt
> Geltendorf (Munich)
> mailto:Rudolf++at++Schmidt.net (home)
> mailto:Rudolf.Schmidt++at++KMWeg.de (office)
> -----------------------------
>
This archive was generated by hypermail 2.0b2 on Wed Nov 03 1999 - 17:02:53 PST