Streaming
Capture from an MCI Device
MCI devices
augment the capture operation in real-time capture and step-frame capture. You
can specify the MCI device, such as a videodisc or video-cassette recorder
(VCR), acting as the video source for your capture operation by using the WM_CAP_SET_MCI_DEVICE message (or the capSetMCIDeviceName macro) and specifying the
name of the device. You can also retrieve the device name currently set by
using the WM_CAP_GET_MCI_DEVICE
message (or the capGetMCIDeviceName macro).
In real-time
capture, the capture window synchronizes the capture operation and compensates
for delays associated with positioning the MCI video source and initializing
the resources (such as capture buffers) required for capturing data. The
capture window expects a valid MCI video device to be installed in the system
for capturing data this way.
Specifications
for controlling an MCI device are stored in the members of the CAPTUREPARMS structure. MCI-compatible
video sources include VCRs and laserdiscs. If the fMCIControl member of
this structure is set to TRUE, the capture window coordinates MCI operation.
The capture window uses the parameters specified in the dwMCIStartTime
and dwMCIStopTime members to obtain the starting and stopping positions,
in milliseconds, of the sequence. If the value of fMCIControl is FALSE,
the video source is not treated as an MCI device and the contents of dwMCIStartTime
and dwMCIStopTime are ignored.
You can use
Media Player to quickly verify that an MCI video device is properly connected
to the system. Playing a device with Media Player verifies that the MCI
configuration for the device is correct. If an image appears on the video
display, the video source is connected properly to the capture hardware.
In step-frame
capture, the capture window synchronizes the capture operation and compensates
for the delays associated with positioning the MCI video source and
initializing the resources required for capturing data. In addition, the
capture window ensures that no frames are dropped; it steps through the video
frames individually, ensuring that the frame is captured and stored before
capturing the next frame in the video stream.
Specifications
for controlling step-frame capture are stored in the members of the CAPTUREPARMS
structure. Step-frame capture uses the following members in addition to the
members used for real-time capture: fStepMCIDevice, fStepCaptureAt2x,
and wStepCaptureAverageFrames. If the fStepMCIDevice member is
set to TRUE, the capture window coordinates step-frame capture. The capture
window uses the parameters specified in the dwMCIStartTime and dwMCIStopTime
members for the starting and stopping positions, in milliseconds, of the sequence.
The capture window uses fStepCaptureAt2x to determine if the capture
hardware should capture video frames at twice the normal resolution and uses wStepCaptureAverageFrames
to specify the number of times each frame in the capture operation is sampled.
If fStepMCIDevice
is FALSE, real-time capture is used instead of step-frame capture and the
contents of fStepCaptureAt2x, and wStepCaptureAverageFrames are
ignored.
If a
step-frame capture is specified and fStepCaptureAt2x is set to TRUE, the
capture hardware captures at twice the specified resolution. (The resolutions
of both the height and width are doubled.) The software interpolates the pixels
in the higher resolution image to produce the image at the specified
resolution. This form of averaging can improve the edge definition of images in
a frame. You can enable this option if the hardware does not support
hardware-based decimation and you are capturing in the RGB format.
Note If your
hardware supports hardware-based decimation, it can capture samples at a higher
rate than specified and use these additional samples to obtain color
definitions that are more consistent with the original image. The additional
samples are discarded after they are used, and the hardware passes samples to
the capture driver at the specified rate.
If a
step-frame capture is specified, the wStepCaptureAverageFrames member
specifies the number of times a frame is sampled when creating a frame based on
the average sample. This averaging technique reduces the random digitization
noise appearing in a frame. A typical value for the number of averages is 5.
For more
information about MCI, see MCI.