Abstract: A computing system includes a codec device, a basic input-output system including both configuration data and feature data for the codec device, and a device driver for the codec device. The basic input-output system configures the codec device based on the configuration data. The device driver reads the feature data from the basic input-output system and enables one or more features of the codec device based on the feature data. In various embodiments, the device driver is WHQL certified and is included in an automated operating system upgrade of the computing system. Because the feature data is in the basic input-output system, the feature data is preserved during the operating system upgrade of the computing system. In some embodiments, the device driver enables one or more features of an application program based on the feature data.

Abstract: Wavetables for a wavetable synthesizer are divided into nonpaged and paged segments. The nonpaged segments are page locked in system memory, so that the wavetable may beginning playing when referenced. The paged segments are paged into memory as needed, and may be paged out of memory when no longer required. The segmentation of the wavetable is determined based on the maximum data rate for the wavetable and a maximum paging delay for the system. Wavetable segmentation is automatically tuned by monitoring actual paging delays and, taking into account a margin for error, updating the value of the maximum paging delay used to determine the required size for a nonpaged wavetable segment. An aggressive margin for error may be employed where an alternative mechanism is provided for handling overruns of the nonpaged wavetable segments.

Abstract: Wavetables for a wavetable synthesizer are divided into nonpaged and paged segments. The nonpaged segments are page locked in system memory, so that the wavetable may beginning playing when referenced. The paged segments are paged into memory as needed, and may be paged out of memory when no longer required. The segmentation of the wavetable is determined based on the maximum data rate for the wavetable and a maximum paging delay for the system. Wavetable segmentation is automatically tuned by monitoring actual paging delays and, taking into account a margin for error, updating the value of the maximum paging delay used to determine the required size for a nonpaged wavetable segment. An aggressive margin for error may be employed where an alternative mechanism is provided for handling overruns of the nonpaged wavetable segments.

Abstract: A MIDI playback system is described in which a sequencer is characterized by: logic for generating a sequence of events, the sequence comprising groups of events, each group comprising events to be executed within a first time interval, the groups being separated in the sequence by marker events for indicating the time intervals, the sequencer being arranged to send the sequence of events together to a synthesizer. A synthesizer is described comprising storage means to receive and store the sequences of events; logic for reading an event from the storage; and logic for determining if the event is a marker event, the synthesizer being arranged to wait, if the event is a marker event, a time equal to the first time interval before reading another event from storage. Using this arrangement, the sequencer is no longer required to schedule the MIDI commands precisely on time and therefore the timing services that a software implementation of such a sequencer might ask from a computer system are reduced.