Archived issue #0020740
ChoosePixelFormat fails to find the correct number of Depth bits
Description
On Intel 915 or 945 chipset the OCC view provided insufficient number of depth
bits (16) while there were at least 24 bits supported by the system.
It was discovered that the method ChoosePixelFormat was called with nDepthBits
== 32, and since there was no match it returned the first available format with
the required nColorBits == 32. This was the format "4" that has nDepthBits ==
16. It will be more relevant to choose the format "6" that has nDepthBits == 24
and nColorBits == 32.
This incident is bug because the customer was not satisfied by the presentation
of surfaces on the mentioned chipsets.
The correction may consist in calling ChoosePixelFormat in a loop that would
test couples of values (nDepthBits, nColorBits) from (32, 32) till (16, 24). The
first matching couple is deteced upon the next call of DescribePixelFormat and
used as the parameter of SetPixelFormat. See my attachment.
bits (16) while there were at least 24 bits supported by the system.
It was discovered that the method ChoosePixelFormat was called with nDepthBits
== 32, and since there was no match it returned the first available format with
the required nColorBits == 32. This was the format "4" that has nDepthBits ==
16. It will be more relevant to choose the format "6" that has nDepthBits == 24
and nColorBits == 32.
This incident is bug because the customer was not satisfied by the presentation
of surfaces on the mentioned chipsets.
The correction may consist in calling ChoosePixelFormat in a loop that would
test couples of values (nDepthBits, nColorBits) from (32, 32) till (16, 24). The
first matching couple is deteced upon the next call of DescribePixelFormat and
used as the parameter of SetPixelFormat. See my attachment.
Public activity
No public notes
Participants are labeled by their role within this record.
Related records