sn9c102.txt

trident tm5600的linux驱动

0x0A    [7:0]         Y sum outside Auto-Exposure area (low-byte)
0x0B    [7:0]         Y sum outside Auto-Exposure area (high-byte)
		      where Y sum = (R/4 + 5G/16 + B/8) / 128
0x0C    0xXX          Not used
0x0D    0xXX          Not used
0x0E    0xXX          Not used
0x0F    0xXX          Not used
0x10    0xXX          Not used
0x11    0xXX          Not used

The following table describes the frame header exported by the SN9C103:

Byte #  Value or bits Description
------  ------------- -----------
0x00    0xFF          Frame synchronisation pattern
0x01    0xFF          Frame synchronisation pattern
0x02    0x00          Frame synchronisation pattern
0x03    0xC4          Frame synchronisation pattern
0x04    0xC4          Frame synchronisation pattern
0x05    0x96          Frame synchronisation pattern
0x06    [6:0]         Read channel gain control = (1/2+R_GAIN/64)
0x07    [6:0]         Blue channel gain control = (1/2+B_GAIN/64)
	[7:4]
0x08    [ 0 ]         Compression mode. 0=No compression, 1=Compression enabled
	[2:1]         Maximum scale factor for compression
	[ 3 ]         1 = USB fifo(2K bytes) is full
	[ 4 ]         1 = Digital gain is finish
	[ 5 ]         1 = Exposure is finish
	[7:6]         Frame index
0x09    [7:0]         Y sum inside Auto-Exposure area (low-byte)
0x0A    [7:0]         Y sum inside Auto-Exposure area (high-byte)
		      where Y sum = (R/4 + 5G/16 + B/8) / 32
0x0B    [7:0]         Y sum outside Auto-Exposure area (low-byte)
0x0C    [7:0]         Y sum outside Auto-Exposure area (high-byte)
		      where Y sum = (R/4 + 5G/16 + B/8) / 128
0x0D    [1:0]         Audio frame number
	[ 2 ]         1 = Audio is recording
0x0E    [7:0]         Audio summation (low-byte)
0x0F    [7:0]         Audio summation (high-byte)
0x10    [7:0]         Audio sample count
0x11    [7:0]         Audio peak data in audio frame

The AE area (sx, sy, ex, ey) in the active window can be set by programming the
registers 0x1c, 0x1d, 0x1e and 0x1f of the SN9C1xx controllers, where one unit
corresponds to 32 pixels.

[1] The frame headers exported by the SN9C105 and SN9C120 are not described.


9. Supported devices
====================
None of the names of the companies as well as their products will be mentioned
here. They have never collaborated with the author, so no advertising.

From the point of view of a driver, what unambiguously identify a device are
its vendor and product USB identifiers. Below is a list of known identifiers of
devices assembling the SN9C1xx PC camera controllers:

Vendor ID  Product ID
---------  ----------
0x0458     0x7025
0x045e     0x00f5
0x045e     0x00f7
0x0471     0x0327
0x0471     0x0328
0x0c45     0x6001
0x0c45     0x6005
0x0c45     0x6007
0x0c45     0x6009
0x0c45     0x600d
0x0c45     0x6011
0x0c45     0x6019
0x0c45     0x6024
0x0c45     0x6025
0x0c45     0x6028
0x0c45     0x6029
0x0c45     0x602a
0x0c45     0x602b
0x0c45     0x602c
0x0c45     0x602d
0x0c45     0x602e
0x0c45     0x6030
0x0c45     0x603f
0x0c45     0x6080
0x0c45     0x6082
0x0c45     0x6083
0x0c45     0x6088
0x0c45     0x608a
0x0c45     0x608b
0x0c45     0x608c
0x0c45     0x608e
0x0c45     0x608f
0x0c45     0x60a0
0x0c45     0x60a2
0x0c45     0x60a3
0x0c45     0x60a8
0x0c45     0x60aa
0x0c45     0x60ab
0x0c45     0x60ac
0x0c45     0x60ae
0x0c45     0x60af
0x0c45     0x60b0
0x0c45     0x60b2
0x0c45     0x60b3
0x0c45     0x60b8
0x0c45     0x60ba
0x0c45     0x60bb
0x0c45     0x60bc
0x0c45     0x60be
0x0c45     0x60c0
0x0c45     0x60c2
0x0c45     0x60c8
0x0c45     0x60cc
0x0c45     0x60ea
0x0c45     0x60ec
0x0c45     0x60ef
0x0c45     0x60fa
0x0c45     0x60fb
0x0c45     0x60fc
0x0c45     0x60fe
0x0c45     0x6102
0x0c45     0x6108
0x0c45     0x610f
0x0c45     0x6130
0x0c45     0x6138
0x0c45     0x613a
0x0c45     0x613b
0x0c45     0x613c
0x0c45     0x613e

The list above does not imply that all those devices work with this driver: up
until now only the ones that assemble the following pairs of SN9C1xx bridges
and image sensors are supported; kernel messages will always tell you whether
this is the case (see "Module loading" paragraph):

Image sensor / SN9C1xx bridge      | SN9C10[12]  SN9C103  SN9C105  SN9C120
-------------------------------------------------------------------------------
HV7131D    Hynix Semiconductor     | Yes         No       No       No
HV7131R    Hynix Semiconductor     | No          Yes      Yes      Yes
MI-0343    Micron Technology       | Yes         No       No       No
MI-0360    Micron Technology       | No          Yes      Yes      Yes
OV7630     OmniVision Technologies | Yes         Yes      Yes      Yes
OV7660     OmniVision Technologies | No          No       Yes      Yes
PAS106B    PixArt Imaging          | Yes         No       No       No
PAS202B    PixArt Imaging          | Yes         Yes      No       No
TAS5110C1B Taiwan Advanced Sensor  | Yes         No       No       No
TAS5110D   Taiwan Advanced Sensor  | Yes         No       No       No
TAS5130D1B Taiwan Advanced Sensor  | Yes         No       No       No

"Yes" means that the pair is supported by the driver, while "No" means that the
pair does not exist or is not supported by the driver.

Only some of the available control settings of each image sensor are supported
through the V4L2 interface.

Donations of new models for further testing and support would be much
appreciated. Non-available hardware will not be supported by the author of this
driver.


10. Notes for V4L2 application developers
=========================================
This driver follows the V4L2 API specifications. In particular, it enforces two
rules:

- exactly one I/O method, either "mmap" or "read", is associated with each
file descriptor. Once it is selected, the application must close and reopen the
device to switch to the other I/O method;

- although it is not mandatory, previously mapped buffer memory should always
be unmapped before calling any "VIDIOC_S_CROP" or "VIDIOC_S_FMT" ioctl's.
The same number of buffers as before will be allocated again to match the size
of the new video frames, so you have to map the buffers again before any I/O
attempts on them.

Consistently with the hardware limits, this driver also supports image
downscaling with arbitrary scaling factors from 1, 2 and 4 in both directions.
However, the V4L2 API specifications don't correctly define how the scaling
factor can be chosen arbitrarily by the "negotiation" of the "source" and
"target" rectangles. To work around this flaw, we have added the convention
that, during the negotiation, whenever the "VIDIOC_S_CROP" ioctl is issued, the
scaling factor is restored to 1.

This driver supports two different video formats: the first one is the "8-bit
Sequential Bayer" format and can be used to obtain uncompressed video data
from the device through the current I/O method, while the second one provides
either "raw" compressed video data (without frame headers not related to the
compressed data) or standard JPEG (with frame headers). The compression quality
may vary from 0 to 1 and can be selected or queried thanks to the
VIDIOC_S_JPEGCOMP and VIDIOC_G_JPEGCOMP V4L2 ioctl's. For maximum flexibility,
both the default active video format and the default compression quality
depend on how the image sensor being used is initialized.


11. Video frame formats [1]
=======================
The SN9C1xx PC Camera Controllers can send images in two possible video
formats over the USB: either native "Sequential RGB Bayer" or compressed.
The compression is used to achieve high frame rates. With regard to the
SN9C101, SN9C102 and SN9C103, the compression is based on the Huffman encoding
algorithm described below, while with regard to the SN9C105 and SN9C120 the
compression is based on the JPEG standard.
The current video format may be selected or queried from the user application
by calling the VIDIOC_S_FMT or VIDIOC_G_FMT ioctl's, as described in the V4L2
API specifications.

The name "Sequential Bayer" indicates the organization of the red, green and
blue pixels in one video frame. Each pixel is associated with a 8-bit long
value and is disposed in memory according to the pattern shown below:

B[0]   G[1]    B[2]    G[3]    ...   B[m-2]         G[m-1]
G[m]   R[m+1]  G[m+2]  R[m+2]  ...   G[2m-2]        R[2m-1]
...
...                                  B[(n-1)(m-2)]  G[(n-1)(m-1)]
...                                  G[n(m-2)]      R[n(m-1)]

The above matrix also represents the sequential or progressive read-out mode of
the (n, m) Bayer color filter array used in many CCD or CMOS image sensors.

The Huffman compressed video frame consists of a bitstream that encodes for
every R, G, or B pixel the difference between the value of the pixel itself and
some reference pixel value. Pixels are organised in the Bayer pattern and the
Bayer sub-pixels are tracked individually and alternatingly. For example, in
the first line values for the B and G1 pixels are alternatingly encoded, while
in the second line values for the G2 and R pixels are alternatingly encoded.

The pixel reference value is calculated as follows:
- the 4 top left pixels are encoded in raw uncompressed 8-bit format;
- the value in the top two rows is the value of the pixel left of the current
  pixel;
- the value in the left column is the value of the pixel above the current
  pixel;
- for all other pixels, the reference value is the average of the value of the
  pixel on the left and the value of the pixel above the current pixel;
- there is one code in the bitstream that specifies the value of a pixel
  directly (in 4-bit resolution);
- pixel values need to be clamped inside the range [0..255] for proper
  decoding.

The algorithm purely describes the conversion from compressed Bayer code used
in the SN9C101, SN9C102 and SN9C103 chips to uncompressed Bayer. Additional
steps are required to convert this to a color image (i.e. a color interpolation
algorithm).

The following Huffman codes have been found:
0: +0 (relative to reference pixel value)
100: +4
101: -4?
1110xxxx: set absolute value to xxxx.0000
1101: +11
1111: -11
11001: +20
110000: -20
110001: ??? - these codes are apparently not used

[1] The Huffman compression algorithm has been reverse-engineered and
    documented by Bertrik Sikken.


12. Contact information
=======================
The author may be contacted by e-mail at <[e-mail]>.

GPG/PGP encrypted e-mail's are accepted. The GPG key ID of the author is
'FCE635A4'; the public 1024-bit key should be available at any keyserver;
the fingerprint is: '88E8 F32F 7244 68BA 3958  5D40 99DA 5D2A FCE6 35A4'.


13. Credits
===========
Many thanks to following persons for their contribute (listed in alphabetical
order):

- David Anderson for the donation of a webcam;
- Luca Capello for the donation of a webcam;
- Philippe Coval for having helped testing the PAS202BCA image sensor;
- Joao Rodrigo Fuzaro, Joao Limirio, Claudio Filho and Caio Begotti for the
  donation of a webcam;
- Dennis Heitmann for the donation of a webcam;
- Jon Hollstrom for the donation of a webcam;
- Nick McGill for the donation of a webcam;
- Carlos Eduardo Medaglia Dyonisio, who added the support for the PAS202BCB
  image sensor;
- Stefano Mozzi, who donated 45 EU;
- Andrew Pearce for the donation of a webcam;
- John Pullan for the donation of a webcam;
- Bertrik Sikken, who reverse-engineered and documented the Huffman compression
  algorithm used in the SN9C101, SN9C102 and SN9C103 controllers and
  implemented the first decoder;
- Ronny Standke for the donation of a webcam;
- Mizuno Takafumi for the donation of a webcam;
- an "anonymous" donator (who didn't want his name to be revealed) for the
  donation of a webcam.
- an anonymous donator for the donation of four webcams and two boards with ten
  image sensors.