Datasheet

PDF: 09005aef8202ec2e/Source: 09005aef8202ebf7 Micron Technology, Inc., reserves the right to change products or specifications without notice.
MT9D111__7_REV5.fm - Rev. B 2/06 EN
147 ©2004 Micron Technology, Inc. All rights reserved.
MT9D111 - 1/3.2-Inch 2-Megapixel SOC Digital Image Sensor
Firmware
Micron Confidential and Proprietary
• fd.search_f1/2_50 and fl.search_fl/2_60 are ranges of frequencies to look for:
fd.search_f1_50=0.2*fd.R9_step60-1;
fd.search_f2_50=0.2*fd.R9_step60+1;
fd.search_f1_60=0.2*fd.R9_step50-1;
fd.search_f2_60=0.2*fd.R9_step50+1
• fl.stat_min and fl.stat_max are detection thresholds; for flicker to be detected
fl.stat_max continuous frames must contain the searched frequency at least
fl.stat_min times
Other variables are as follows:
fd.windowPosH indicates current sampling window position. Variable fl.windowHeight
indicate window size. Do not change these variables.
Flicker can operate in automatic and manual modes. Automatic mode is enabled by
default, fd.mode[5] indicates current 50Hz/60Hz frequency selection. Make sure to set
flicker detection bits in sequencer states. To choose manual mode, set fd.mode[7]=1.
Bits fd.mode[6] is used for manual 50Hz/60Hz frequency selection. As frequencies are
selected, the auto exposure integration time step is assigned the appropriate
fd.R9_step50/60 variable value.
For the flicker avoidance driver to operate correctly, the sensor frame rate may not be
exactly 15 fps or 30 fps. Frame rate must be tuned off the exact frame rate. This is a nec-
essary condition to have the flicker bars rolling. For this reason the preview mode frame
rate is typically deliberately set not to exactly match 15 fps or 30 fps. However, in capture
mode, the frame rate is set exactly, assuming that the flicker has already been detected
and compensated during preview.
JPEG Driver
The JPEG driver programs Huffman table and quantization table memories, sets up the
MT9D111 to output JPEG compressed data, and handles error checking and handshak-
ing with the host processor.
Usage
JPEG is enabled using mode.mode_config. For example, to enable compression for cap-
ture set mode.mode_congif=16. jpeg.qscale1/2/3, and specify the quantization factor to
control compression ratio. Bigger number indicates more compression. mode.fifo* con-
figure spoof and non-spoof output and specify FIFO output clock divider. If necessary,
use jpeg.config for error handshaking with the host.
Table Programming
At power up initialization, the JPEG driver loads standard Huffman tables into Huffman
memory. Scaled versions of standard luma and chroma quantization tables and are
loaded into quantization memory. The calculation of the scaled quantization table is as
follows:
Scaled_Q = (scaling_factor * standard_Q + 16) >> 5
Scaling factor is bit 6:0 of JPEG driver variables qscale1, qscale2, and qscale3. The stan-
dard quantization and Huffman tables are Tables K.1–K.4 of the ISO/IEC 10918-1 Speci-
fication. Host processor can override Huffman and quantization memory with any
arbitrary Huffman and quantization tables through indirect register access.
If the host chooses to use the scaled standard quantization tables, bit 4 of JPEG driver
variable config must be set to “1.” Scaling factor can be changed at any time. Whenever it
is changed, bit 7 of the corresponding JPEG driver variable qscale must be set to “1,” and
the new value takes effect in the next frame JPEG encoding.