tjh.dev
Linux kernel mirror (for testing)
git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel
os
linux
1 <title>Image Formats</title>
2
3 <para>The V4L2 API was primarily designed for devices exchanging
4image data with applications. The
5<structname>v4l2_pix_format</structname> structure defines the format
6and layout of an image in memory. Image formats are negotiated with
7the &VIDIOC-S-FMT; ioctl. (The explanations here focus on video
8capturing and output, for overlay frame buffer formats see also
9&VIDIOC-G-FBUF;.)</para>
10
11 <table pgwide="1" frame="none" id="v4l2-pix-format">
12 <title>struct <structname>v4l2_pix_format</structname></title>
13 <tgroup cols="3">
14 &cs-str;
15 <tbody valign="top">
16 <row>
17 <entry>__u32</entry>
18 <entry><structfield>width</structfield></entry>
19 <entry>Image width in pixels.</entry>
20 </row>
21 <row>
22 <entry>__u32</entry>
23 <entry><structfield>height</structfield></entry>
24 <entry>Image height in pixels.</entry>
25 </row>
26 <row>
27 <entry spanname="hspan">Applications set these fields to
28request an image size, drivers return the closest possible values. In
29case of planar formats the <structfield>width</structfield> and
30<structfield>height</structfield> applies to the largest plane. To
31avoid ambiguities drivers must return values rounded up to a multiple
32of the scale factor of any smaller planes. For example when the image
33format is YUV 4:2:0, <structfield>width</structfield> and
34<structfield>height</structfield> must be multiples of two.</entry>
35 </row>
36 <row>
37 <entry>__u32</entry>
38 <entry><structfield>pixelformat</structfield></entry>
39 <entry>The pixel format or type of compression, set by the
40application. This is a little endian <link
41linkend="v4l2-fourcc">four character code</link>. V4L2 defines
42standard RGB formats in <xref linkend="rgb-formats" />, YUV formats in <xref
43linkend="yuv-formats" />, and reserved codes in <xref
44linkend="reserved-formats" /></entry>
45 </row>
46 <row>
47 <entry>&v4l2-field;</entry>
48 <entry><structfield>field</structfield></entry>
49 <entry>Video images are typically interlaced. Applications
50can request to capture or output only the top or bottom field, or both
51fields interlaced or sequentially stored in one buffer or alternating
52in separate buffers. Drivers return the actual field order selected.
53For details see <xref linkend="field-order" />.</entry>
54 </row>
55 <row>
56 <entry>__u32</entry>
57 <entry><structfield>bytesperline</structfield></entry>
58 <entry>Distance in bytes between the leftmost pixels in two
59adjacent lines.</entry>
60 </row>
61 <row>
62 <entry spanname="hspan"><para>Both applications and drivers
63can set this field to request padding bytes at the end of each line.
64Drivers however may ignore the value requested by the application,
65returning <structfield>width</structfield> times bytes per pixel or a
66larger value required by the hardware. That implies applications can
67just set this field to zero to get a reasonable
68default.</para><para>Video hardware may access padding bytes,
69therefore they must reside in accessible memory. Consider cases where
70padding bytes after the last line of an image cross a system page
71boundary. Input devices may write padding bytes, the value is
72undefined. Output devices ignore the contents of padding
73bytes.</para><para>When the image format is planar the
74<structfield>bytesperline</structfield> value applies to the largest
75plane and is divided by the same factor as the
76<structfield>width</structfield> field for any smaller planes. For
77example the Cb and Cr planes of a YUV 4:2:0 image have half as many
78padding bytes following each line as the Y plane. To avoid ambiguities
79drivers must return a <structfield>bytesperline</structfield> value
80rounded up to a multiple of the scale factor.</para></entry>
81 </row>
82 <row>
83 <entry>__u32</entry>
84 <entry><structfield>sizeimage</structfield></entry>
85 <entry>Size in bytes of the buffer to hold a complete image,
86set by the driver. Usually this is
87<structfield>bytesperline</structfield> times
88<structfield>height</structfield>. When the image consists of variable
89length compressed data this is the maximum number of bytes required to
90hold an image.</entry>
91 </row>
92 <row>
93 <entry>&v4l2-colorspace;</entry>
94 <entry><structfield>colorspace</structfield></entry>
95 <entry>This information supplements the
96<structfield>pixelformat</structfield> and must be set by the driver,
97see <xref linkend="colorspaces" />.</entry>
98 </row>
99 <row>
100 <entry>__u32</entry>
101 <entry><structfield>priv</structfield></entry>
102 <entry>Reserved for custom (driver defined) additional
103information about formats. When not used drivers and applications must
104set this field to zero.</entry>
105 </row>
106 </tbody>
107 </tgroup>
108 </table>
109
110 <section>
111 <title>Standard Image Formats</title>
112
113 <para>In order to exchange images between drivers and
114applications, it is necessary to have standard image data formats
115which both sides will interpret the same way. V4L2 includes several
116such formats, and this section is intended to be an unambiguous
117specification of the standard image data formats in V4L2.</para>
118
119 <para>V4L2 drivers are not limited to these formats, however.
120Driver-specific formats are possible. In that case the application may
121depend on a codec to convert images to one of the standard formats
122when needed. But the data can still be stored and retrieved in the
123proprietary format. For example, a device may support a proprietary
124compressed format. Applications can still capture and save the data in
125the compressed format, saving much disk space, and later use a codec
126to convert the images to the X Windows screen format when the video is
127to be displayed.</para>
128
129 <para>Even so, ultimately, some standard formats are needed, so
130the V4L2 specification would not be complete without well-defined
131standard formats.</para>
132
133 <para>The V4L2 standard formats are mainly uncompressed formats. The
134pixels are always arranged in memory from left to right, and from top
135to bottom. The first byte of data in the image buffer is always for
136the leftmost pixel of the topmost row. Following that is the pixel
137immediately to its right, and so on until the end of the top row of
138pixels. Following the rightmost pixel of the row there may be zero or
139more bytes of padding to guarantee that each row of pixel data has a
140certain alignment. Following the pad bytes, if any, is data for the
141leftmost pixel of the second row from the top, and so on. The last row
142has just as many pad bytes after it as the other rows.</para>
143
144 <para>In V4L2 each format has an identifier which looks like
145<constant>PIX_FMT_XXX</constant>, defined in the <link
146linkend="videodev">videodev.h</link> header file. These identifiers
147represent <link linkend="v4l2-fourcc">four character codes</link>
148which are also listed below, however they are not the same as those
149used in the Windows world.</para>
150 </section>
151
152 <section id="colorspaces">
153 <title>Colorspaces</title>
154
155 <para>[intro]</para>
156
157 <!-- See proposal by Billy Biggs, video4linux-list@redhat.com
158on 11 Oct 2002, subject: "Re: [V4L] Re: v4l2 api", and
159http://vektor.theorem.ca/graphics/ycbcr/ and
160http://www.poynton.com/notes/colour_and_gamma/ColorFAQ.html -->
161
162 <para>
163 <variablelist>
164 <varlistentry>
165 <term>Gamma Correction</term>
166 <listitem>
167 <para>[to do]</para>
168 <para>E'<subscript>R</subscript> = f(R)</para>
169 <para>E'<subscript>G</subscript> = f(G)</para>
170 <para>E'<subscript>B</subscript> = f(B)</para>
171 </listitem>
172 </varlistentry>
173 <varlistentry>
174 <term>Construction of luminance and color-difference
175signals</term>
176 <listitem>
177 <para>[to do]</para>
178 <para>E'<subscript>Y</subscript> =
179Coeff<subscript>R</subscript> E'<subscript>R</subscript>
180+ Coeff<subscript>G</subscript> E'<subscript>G</subscript>
181+ Coeff<subscript>B</subscript> E'<subscript>B</subscript></para>
182 <para>(E'<subscript>R</subscript> - E'<subscript>Y</subscript>) = E'<subscript>R</subscript>
183- Coeff<subscript>R</subscript> E'<subscript>R</subscript>
184- Coeff<subscript>G</subscript> E'<subscript>G</subscript>
185- Coeff<subscript>B</subscript> E'<subscript>B</subscript></para>
186 <para>(E'<subscript>B</subscript> - E'<subscript>Y</subscript>) = E'<subscript>B</subscript>
187- Coeff<subscript>R</subscript> E'<subscript>R</subscript>
188- Coeff<subscript>G</subscript> E'<subscript>G</subscript>
189- Coeff<subscript>B</subscript> E'<subscript>B</subscript></para>
190 </listitem>
191 </varlistentry>
192 <varlistentry>
193 <term>Re-normalized color-difference signals</term>
194 <listitem>
195 <para>The color-difference signals are scaled back to unity
196range [-0.5;+0.5]:</para>
197 <para>K<subscript>B</subscript> = 0.5 / (1 - Coeff<subscript>B</subscript>)</para>
198 <para>K<subscript>R</subscript> = 0.5 / (1 - Coeff<subscript>R</subscript>)</para>
199 <para>P<subscript>B</subscript> =
200K<subscript>B</subscript> (E'<subscript>B</subscript> - E'<subscript>Y</subscript>) =
201 0.5 (Coeff<subscript>R</subscript> / Coeff<subscript>B</subscript>) E'<subscript>R</subscript>
202+ 0.5 (Coeff<subscript>G</subscript> / Coeff<subscript>B</subscript>) E'<subscript>G</subscript>
203+ 0.5 E'<subscript>B</subscript></para>
204 <para>P<subscript>R</subscript> =
205K<subscript>R</subscript> (E'<subscript>R</subscript> - E'<subscript>Y</subscript>) =
206 0.5 E'<subscript>R</subscript>
207+ 0.5 (Coeff<subscript>G</subscript> / Coeff<subscript>R</subscript>) E'<subscript>G</subscript>
208+ 0.5 (Coeff<subscript>B</subscript> / Coeff<subscript>R</subscript>) E'<subscript>B</subscript></para>
209 </listitem>
210 </varlistentry>
211 <varlistentry>
212 <term>Quantization</term>
213 <listitem>
214 <para>[to do]</para>
215 <para>Y' = (Lum. Levels - 1) · E'<subscript>Y</subscript> + Lum. Offset</para>
216 <para>C<subscript>B</subscript> = (Chrom. Levels - 1)
217· P<subscript>B</subscript> + Chrom. Offset</para>
218 <para>C<subscript>R</subscript> = (Chrom. Levels - 1)
219· P<subscript>R</subscript> + Chrom. Offset</para>
220 <para>Rounding to the nearest integer and clamping to the range
221[0;255] finally yields the digital color components Y'CbCr
222stored in YUV images.</para>
223 </listitem>
224 </varlistentry>
225 </variablelist>
226 </para>
227
228 <example>
229 <title>ITU-R Rec. BT.601 color conversion</title>
230
231 <para>Forward Transformation</para>
232
233 <programlisting>
234int ER, EG, EB; /* gamma corrected RGB input [0;255] */
235int Y1, Cb, Cr; /* output [0;255] */
236
237double r, g, b; /* temporaries */
238double y1, pb, pr;
239
240int
241clamp (double x)
242{
243 int r = x; /* round to nearest */
244
245 if (r < 0) return 0;
246 else if (r > 255) return 255;
247 else return r;
248}
249
250r = ER / 255.0;
251g = EG / 255.0;
252b = EB / 255.0;
253
254y1 = 0.299 * r + 0.587 * g + 0.114 * b;
255pb = -0.169 * r - 0.331 * g + 0.5 * b;
256pr = 0.5 * r - 0.419 * g - 0.081 * b;
257
258Y1 = clamp (219 * y1 + 16);
259Cb = clamp (224 * pb + 128);
260Cr = clamp (224 * pr + 128);
261
262/* or shorter */
263
264y1 = 0.299 * ER + 0.587 * EG + 0.114 * EB;
265
266Y1 = clamp ( (219 / 255.0) * y1 + 16);
267Cb = clamp (((224 / 255.0) / (2 - 2 * 0.114)) * (EB - y1) + 128);
268Cr = clamp (((224 / 255.0) / (2 - 2 * 0.299)) * (ER - y1) + 128);
269 </programlisting>
270
271 <para>Inverse Transformation</para>
272
273 <programlisting>
274int Y1, Cb, Cr; /* gamma pre-corrected input [0;255] */
275int ER, EG, EB; /* output [0;255] */
276
277double r, g, b; /* temporaries */
278double y1, pb, pr;
279
280int
281clamp (double x)
282{
283 int r = x; /* round to nearest */
284
285 if (r < 0) return 0;
286 else if (r > 255) return 255;
287 else return r;
288}
289
290y1 = (255 / 219.0) * (Y1 - 16);
291pb = (255 / 224.0) * (Cb - 128);
292pr = (255 / 224.0) * (Cr - 128);
293
294r = 1.0 * y1 + 0 * pb + 1.402 * pr;
295g = 1.0 * y1 - 0.344 * pb - 0.714 * pr;
296b = 1.0 * y1 + 1.772 * pb + 0 * pr;
297
298ER = clamp (r * 255); /* [ok? one should prob. limit y1,pb,pr] */
299EG = clamp (g * 255);
300EB = clamp (b * 255);
301 </programlisting>
302 </example>
303
304 <table pgwide="1" id="v4l2-colorspace" orient="land">
305 <title>enum v4l2_colorspace</title>
306 <tgroup cols="11" align="center">
307 <colspec align="left" />
308 <colspec align="center" />
309 <colspec align="left" />
310 <colspec colname="cr" />
311 <colspec colname="cg" />
312 <colspec colname="cb" />
313 <colspec colname="wp" />
314 <colspec colname="gc" />
315 <colspec colname="lum" />
316 <colspec colname="qy" />
317 <colspec colname="qc" />
318 <spanspec namest="cr" nameend="cb" spanname="chrom" />
319 <spanspec namest="qy" nameend="qc" spanname="quant" />
320 <spanspec namest="lum" nameend="qc" spanname="spam" />
321 <thead>
322 <row>
323 <entry morerows="1">Identifier</entry>
324 <entry morerows="1">Value</entry>
325 <entry morerows="1">Description</entry>
326 <entry spanname="chrom">Chromaticities<footnote>
327 <para>The coordinates of the color primaries are
328given in the CIE system (1931)</para>
329 </footnote></entry>
330 <entry morerows="1">White Point</entry>
331 <entry morerows="1">Gamma Correction</entry>
332 <entry morerows="1">Luminance E'<subscript>Y</subscript></entry>
333 <entry spanname="quant">Quantization</entry>
334 </row>
335 <row>
336 <entry>Red</entry>
337 <entry>Green</entry>
338 <entry>Blue</entry>
339 <entry>Y'</entry>
340 <entry>Cb, Cr</entry>
341 </row>
342 </thead>
343 <tbody valign="top">
344 <row>
345 <entry><constant>V4L2_COLORSPACE_SMPTE170M</constant></entry>
346 <entry>1</entry>
347 <entry>NTSC/PAL according to <xref linkend="smpte170m" />,
348<xref linkend="itu601" /></entry>
349 <entry>x = 0.630, y = 0.340</entry>
350 <entry>x = 0.310, y = 0.595</entry>
351 <entry>x = 0.155, y = 0.070</entry>
352 <entry>x = 0.3127, y = 0.3290,
353 Illuminant D<subscript>65</subscript></entry>
354 <entry>E' = 4.5 I for I ≤0.018,
3551.099 I<superscript>0.45</superscript> - 0.099 for 0.018 < I</entry>
356 <entry>0.299 E'<subscript>R</subscript>
357+ 0.587 E'<subscript>G</subscript>
358+ 0.114 E'<subscript>B</subscript></entry>
359 <entry>219 E'<subscript>Y</subscript> + 16</entry>
360 <entry>224 P<subscript>B,R</subscript> + 128</entry>
361 </row>
362 <row>
363 <entry><constant>V4L2_COLORSPACE_SMPTE240M</constant></entry>
364 <entry>2</entry>
365 <entry>1125-Line (US) HDTV, see <xref
366linkend="smpte240m" /></entry>
367 <entry>x = 0.630, y = 0.340</entry>
368 <entry>x = 0.310, y = 0.595</entry>
369 <entry>x = 0.155, y = 0.070</entry>
370 <entry>x = 0.3127, y = 0.3290,
371 Illuminant D<subscript>65</subscript></entry>
372 <entry>E' = 4 I for I ≤0.0228,
3731.1115 I<superscript>0.45</superscript> - 0.1115 for 0.0228 < I</entry>
374 <entry>0.212 E'<subscript>R</subscript>
375+ 0.701 E'<subscript>G</subscript>
376+ 0.087 E'<subscript>B</subscript></entry>
377 <entry>219 E'<subscript>Y</subscript> + 16</entry>
378 <entry>224 P<subscript>B,R</subscript> + 128</entry>
379 </row>
380 <row>
381 <entry><constant>V4L2_COLORSPACE_REC709</constant></entry>
382 <entry>3</entry>
383 <entry>HDTV and modern devices, see <xref
384linkend="itu709" /></entry>
385 <entry>x = 0.640, y = 0.330</entry>
386 <entry>x = 0.300, y = 0.600</entry>
387 <entry>x = 0.150, y = 0.060</entry>
388 <entry>x = 0.3127, y = 0.3290,
389 Illuminant D<subscript>65</subscript></entry>
390 <entry>E' = 4.5 I for I ≤0.018,
3911.099 I<superscript>0.45</superscript> - 0.099 for 0.018 < I</entry>
392 <entry>0.2125 E'<subscript>R</subscript>
393+ 0.7154 E'<subscript>G</subscript>
394+ 0.0721 E'<subscript>B</subscript></entry>
395 <entry>219 E'<subscript>Y</subscript> + 16</entry>
396 <entry>224 P<subscript>B,R</subscript> + 128</entry>
397 </row>
398 <row>
399 <entry><constant>V4L2_COLORSPACE_BT878</constant></entry>
400 <entry>4</entry>
401 <entry>Broken Bt878 extents<footnote>
402 <para>The ubiquitous Bt878 video capture chip
403quantizes E'<subscript>Y</subscript> to 238 levels, yielding a range
404of Y' = 16 … 253, unlike Rec. 601 Y' = 16 …
405235. This is not a typo in the Bt878 documentation, it has been
406implemented in silicon. The chroma extents are unclear.</para>
407 </footnote>, <xref linkend="itu601" /></entry>
408 <entry>?</entry>
409 <entry>?</entry>
410 <entry>?</entry>
411 <entry>?</entry>
412 <entry>?</entry>
413 <entry>0.299 E'<subscript>R</subscript>
414+ 0.587 E'<subscript>G</subscript>
415+ 0.114 E'<subscript>B</subscript></entry>
416 <entry><emphasis>237</emphasis> E'<subscript>Y</subscript> + 16</entry>
417 <entry>224 P<subscript>B,R</subscript> + 128 (probably)</entry>
418 </row>
419 <row>
420 <entry><constant>V4L2_COLORSPACE_470_SYSTEM_M</constant></entry>
421 <entry>5</entry>
422 <entry>M/NTSC<footnote>
423 <para>No identifier exists for M/PAL which uses
424the chromaticities of M/NTSC, the remaining parameters are equal to B and
425G/PAL.</para>
426 </footnote> according to <xref linkend="itu470" />, <xref
427 linkend="itu601" /></entry>
428 <entry>x = 0.67, y = 0.33</entry>
429 <entry>x = 0.21, y = 0.71</entry>
430 <entry>x = 0.14, y = 0.08</entry>
431 <entry>x = 0.310, y = 0.316, Illuminant C</entry>
432 <entry>?</entry>
433 <entry>0.299 E'<subscript>R</subscript>
434+ 0.587 E'<subscript>G</subscript>
435+ 0.114 E'<subscript>B</subscript></entry>
436 <entry>219 E'<subscript>Y</subscript> + 16</entry>
437 <entry>224 P<subscript>B,R</subscript> + 128</entry>
438 </row>
439 <row>
440 <entry><constant>V4L2_COLORSPACE_470_SYSTEM_BG</constant></entry>
441 <entry>6</entry>
442 <entry>625-line PAL and SECAM systems according to <xref
443linkend="itu470" />, <xref linkend="itu601" /></entry>
444 <entry>x = 0.64, y = 0.33</entry>
445 <entry>x = 0.29, y = 0.60</entry>
446 <entry>x = 0.15, y = 0.06</entry>
447 <entry>x = 0.313, y = 0.329,
448Illuminant D<subscript>65</subscript></entry>
449 <entry>?</entry>
450 <entry>0.299 E'<subscript>R</subscript>
451+ 0.587 E'<subscript>G</subscript>
452+ 0.114 E'<subscript>B</subscript></entry>
453 <entry>219 E'<subscript>Y</subscript> + 16</entry>
454 <entry>224 P<subscript>B,R</subscript> + 128</entry>
455 </row>
456 <row>
457 <entry><constant>V4L2_COLORSPACE_JPEG</constant></entry>
458 <entry>7</entry>
459 <entry>JPEG Y'CbCr, see <xref linkend="jfif" />, <xref linkend="itu601" /></entry>
460 <entry>?</entry>
461 <entry>?</entry>
462 <entry>?</entry>
463 <entry>?</entry>
464 <entry>?</entry>
465 <entry>0.299 E'<subscript>R</subscript>
466+ 0.587 E'<subscript>G</subscript>
467+ 0.114 E'<subscript>B</subscript></entry>
468 <entry>256 E'<subscript>Y</subscript> + 16<footnote>
469 <para>Note JFIF quantizes
470Y'P<subscript>B</subscript>P<subscript>R</subscript> in range [0;+1] and
471[-0.5;+0.5] to <emphasis>257</emphasis> levels, however Y'CbCr signals
472are still clamped to [0;255].</para>
473 </footnote></entry>
474 <entry>256 P<subscript>B,R</subscript> + 128</entry>
475 </row>
476 <row>
477 <entry><constant>V4L2_COLORSPACE_SRGB</constant></entry>
478 <entry>8</entry>
479 <entry>[?]</entry>
480 <entry>x = 0.640, y = 0.330</entry>
481 <entry>x = 0.300, y = 0.600</entry>
482 <entry>x = 0.150, y = 0.060</entry>
483 <entry>x = 0.3127, y = 0.3290,
484 Illuminant D<subscript>65</subscript></entry>
485 <entry>E' = 4.5 I for I ≤0.018,
4861.099 I<superscript>0.45</superscript> - 0.099 for 0.018 < I</entry>
487 <entry spanname="spam">n/a</entry>
488 </row>
489 </tbody>
490 </tgroup>
491 </table>
492 </section>
493
494 <section id="pixfmt-indexed">
495 <title>Indexed Format</title>
496
497 <para>In this format each pixel is represented by an 8 bit index
498into a 256 entry ARGB palette. It is intended for <link
499linkend="osd">Video Output Overlays</link> only. There are no ioctls to
500access the palette, this must be done with ioctls of the Linux framebuffer API.</para>
501
502 <table pgwide="0" frame="none">
503 <title>Indexed Image Format</title>
504 <tgroup cols="37" align="center">
505 <colspec colname="id" align="left" />
506 <colspec colname="fourcc" />
507 <colspec colname="bit" />
508
509 <colspec colnum="4" colname="b07" align="center" />
510 <colspec colnum="5" colname="b06" align="center" />
511 <colspec colnum="6" colname="b05" align="center" />
512 <colspec colnum="7" colname="b04" align="center" />
513 <colspec colnum="8" colname="b03" align="center" />
514 <colspec colnum="9" colname="b02" align="center" />
515 <colspec colnum="10" colname="b01" align="center" />
516 <colspec colnum="11" colname="b00" align="center" />
517
518 <spanspec namest="b07" nameend="b00" spanname="b0" />
519 <spanspec namest="b17" nameend="b10" spanname="b1" />
520 <spanspec namest="b27" nameend="b20" spanname="b2" />
521 <spanspec namest="b37" nameend="b30" spanname="b3" />
522 <thead>
523 <row>
524 <entry>Identifier</entry>
525 <entry>Code</entry>
526 <entry> </entry>
527 <entry spanname="b0">Byte 0</entry>
528 </row>
529 <row>
530 <entry> </entry>
531 <entry> </entry>
532 <entry>Bit</entry>
533 <entry>7</entry>
534 <entry>6</entry>
535 <entry>5</entry>
536 <entry>4</entry>
537 <entry>3</entry>
538 <entry>2</entry>
539 <entry>1</entry>
540 <entry>0</entry>
541 </row>
542 </thead>
543 <tbody valign="top">
544 <row id="V4L2-PIX-FMT-PAL8">
545 <entry><constant>V4L2_PIX_FMT_PAL8</constant></entry>
546 <entry>'PAL8'</entry>
547 <entry></entry>
548 <entry>i<subscript>7</subscript></entry>
549 <entry>i<subscript>6</subscript></entry>
550 <entry>i<subscript>5</subscript></entry>
551 <entry>i<subscript>4</subscript></entry>
552 <entry>i<subscript>3</subscript></entry>
553 <entry>i<subscript>2</subscript></entry>
554 <entry>i<subscript>1</subscript></entry>
555 <entry>i<subscript>0</subscript></entry>
556 </row>
557 </tbody>
558 </tgroup>
559 </table>
560 </section>
561
562 <section id="pixfmt-rgb">
563 <title>RGB Formats</title>
564
565 &sub-packed-rgb;
566 &sub-sbggr8;
567 &sub-sgbrg8;
568 &sub-sgrbg8;
569 &sub-sbggr16;
570 </section>
571
572 <section id="yuv-formats">
573 <title>YUV Formats</title>
574
575 <para>YUV is the format native to TV broadcast and composite video
576signals. It separates the brightness information (Y) from the color
577information (U and V or Cb and Cr). The color information consists of
578red and blue <emphasis>color difference</emphasis> signals, this way
579the green component can be reconstructed by subtracting from the
580brightness component. See <xref linkend="colorspaces" /> for conversion
581examples. YUV was chosen because early television would only transmit
582brightness information. To add color in a way compatible with existing
583receivers a new signal carrier was added to transmit the color
584difference signals. Secondary in the YUV format the U and V components
585usually have lower resolution than the Y component. This is an analog
586video compression technique taking advantage of a property of the
587human visual system, being more sensitive to brightness
588information.</para>
589
590 &sub-packed-yuv;
591 &sub-grey;
592 &sub-y16;
593 &sub-yuyv;
594 &sub-uyvy;
595 &sub-yvyu;
596 &sub-vyuy;
597 &sub-y41p;
598 &sub-yuv420;
599 &sub-yuv410;
600 &sub-yuv422p;
601 &sub-yuv411p;
602 &sub-nv12;
603 &sub-nv16;
604 </section>
605
606 <section>
607 <title>Compressed Formats</title>
608
609 <table pgwide="1" frame="none" id="compressed-formats">
610 <title>Compressed Image Formats</title>
611 <tgroup cols="3" align="left">
612 &cs-def;
613 <thead>
614 <row>
615 <entry>Identifier</entry>
616 <entry>Code</entry>
617 <entry>Details</entry>
618 </row>
619 </thead>
620 <tbody valign="top">
621 <row id="V4L2-PIX-FMT-JPEG">
622 <entry><constant>V4L2_PIX_FMT_JPEG</constant></entry>
623 <entry>'JPEG'</entry>
624 <entry>TBD. See also &VIDIOC-G-JPEGCOMP;,
625 &VIDIOC-S-JPEGCOMP;.</entry>
626 </row>
627 <row id="V4L2-PIX-FMT-MPEG">
628 <entry><constant>V4L2_PIX_FMT_MPEG</constant></entry>
629 <entry>'MPEG'</entry>
630 <entry>MPEG stream. The actual format is determined by
631extended control <constant>V4L2_CID_MPEG_STREAM_TYPE</constant>, see
632<xref linkend="mpeg-control-id" />.</entry>
633 </row>
634 </tbody>
635 </tgroup>
636 </table>
637 </section>
638
639 <section id="pixfmt-reserved">
640 <title>Reserved Format Identifiers</title>
641
642 <para>These formats are not defined by this specification, they
643are just listed for reference and to avoid naming conflicts. If you
644want to register your own format, send an e-mail to the linux-media mailing
645list &v4l-ml; for inclusion in the <filename>videodev2.h</filename>
646file. If you want to share your format with other developers add a
647link to your documentation and send a copy to the linux-media mailing list
648for inclusion in this section. If you think your format should be listed
649in a standard format section please make a proposal on the linux-media mailing
650list.</para>
651
652 <table pgwide="1" frame="none" id="reserved-formats">
653 <title>Reserved Image Formats</title>
654 <tgroup cols="3" align="left">
655 &cs-def;
656 <thead>
657 <row>
658 <entry>Identifier</entry>
659 <entry>Code</entry>
660 <entry>Details</entry>
661 </row>
662 </thead>
663 <tbody valign="top">
664 <row id="V4L2-PIX-FMT-DV">
665 <entry><constant>V4L2_PIX_FMT_DV</constant></entry>
666 <entry>'dvsd'</entry>
667 <entry>unknown</entry>
668 </row>
669 <row id="V4L2-PIX-FMT-ET61X251">
670 <entry><constant>V4L2_PIX_FMT_ET61X251</constant></entry>
671 <entry>'E625'</entry>
672 <entry>Compressed format of the ET61X251 driver.</entry>
673 </row>
674 <row id="V4L2-PIX-FMT-HI240">
675 <entry><constant>V4L2_PIX_FMT_HI240</constant></entry>
676 <entry>'HI24'</entry>
677 <entry><para>8 bit RGB format used by the BTTV driver.</para></entry>
678 </row>
679 <row id="V4L2-PIX-FMT-HM12">
680 <entry><constant>V4L2_PIX_FMT_HM12</constant></entry>
681 <entry>'HM12'</entry>
682 <entry><para>YUV 4:2:0 format used by the
683IVTV driver, <ulink url="http://www.ivtvdriver.org/">
684http://www.ivtvdriver.org/</ulink></para><para>The format is documented in the
685kernel sources in the file <filename>Documentation/video4linux/cx2341x/README.hm12</filename>
686</para></entry>
687 </row>
688 <row id="V4L2-PIX-FMT-SPCA501">
689 <entry><constant>V4L2_PIX_FMT_SPCA501</constant></entry>
690 <entry>'S501'</entry>
691 <entry>YUYV per line used by the gspca driver.</entry>
692 </row>
693 <row id="V4L2-PIX-FMT-SPCA505">
694 <entry><constant>V4L2_PIX_FMT_SPCA505</constant></entry>
695 <entry>'S505'</entry>
696 <entry>YYUV per line used by the gspca driver.</entry>
697 </row>
698 <row id="V4L2-PIX-FMT-SPCA508">
699 <entry><constant>V4L2_PIX_FMT_SPCA508</constant></entry>
700 <entry>'S508'</entry>
701 <entry>YUVY per line used by the gspca driver.</entry>
702 </row>
703 <row id="V4L2-PIX-FMT-SPCA561">
704 <entry><constant>V4L2_PIX_FMT_SPCA561</constant></entry>
705 <entry>'S561'</entry>
706 <entry>Compressed GBRG Bayer format used by the gspca driver.</entry>
707 </row>
708 <row id="V4L2-PIX-FMT-SGRBG10">
709 <entry><constant>V4L2_PIX_FMT_SGRBG10</constant></entry>
710 <entry>'DA10'</entry>
711 <entry>10 bit raw Bayer, expanded to 16 bits.</entry>
712 </row>
713 <row id="V4L2-PIX-FMT-SGRBG10DPCM8">
714 <entry><constant>V4L2_PIX_FMT_SGRBG10DPCM8</constant></entry>
715 <entry>'DB10'</entry>
716 <entry>10 bit raw Bayer DPCM compressed to 8 bits.</entry>
717 </row>
718 <row id="V4L2-PIX-FMT-PAC207">
719 <entry><constant>V4L2_PIX_FMT_PAC207</constant></entry>
720 <entry>'P207'</entry>
721 <entry>Compressed BGGR Bayer format used by the gspca driver.</entry>
722 </row>
723 <row id="V4L2-PIX-FMT-MR97310A">
724 <entry><constant>V4L2_PIX_FMT_MR97310A</constant></entry>
725 <entry>'M310'</entry>
726 <entry>Compressed BGGR Bayer format used by the gspca driver.</entry>
727 </row>
728 <row id="V4L2-PIX-FMT-OV511">
729 <entry><constant>V4L2_PIX_FMT_OV511</constant></entry>
730 <entry>'O511'</entry>
731 <entry>OV511 JPEG format used by the gspca driver.</entry>
732 </row>
733 <row id="V4L2-PIX-FMT-OV518">
734 <entry><constant>V4L2_PIX_FMT_OV518</constant></entry>
735 <entry>'O518'</entry>
736 <entry>OV518 JPEG format used by the gspca driver.</entry>
737 </row>
738 <row id="V4L2-PIX-FMT-PJPG">
739 <entry><constant>V4L2_PIX_FMT_PJPG</constant></entry>
740 <entry>'PJPG'</entry>
741 <entry>Pixart 73xx JPEG format used by the gspca driver.</entry>
742 </row>
743 <row id="V4L2-PIX-FMT-SQ905C">
744 <entry><constant>V4L2_PIX_FMT_SQ905C</constant></entry>
745 <entry>'905C'</entry>
746 <entry>Compressed RGGB bayer format used by the gspca driver.</entry>
747 </row>
748 <row id="V4L2-PIX-FMT-MJPEG">
749 <entry><constant>V4L2_PIX_FMT_MJPEG</constant></entry>
750 <entry>'MJPG'</entry>
751 <entry>Compressed format used by the Zoran driver</entry>
752 </row>
753 <row id="V4L2-PIX-FMT-PWC1">
754 <entry><constant>V4L2_PIX_FMT_PWC1</constant></entry>
755 <entry>'PWC1'</entry>
756 <entry>Compressed format of the PWC driver.</entry>
757 </row>
758 <row id="V4L2-PIX-FMT-PWC2">
759 <entry><constant>V4L2_PIX_FMT_PWC2</constant></entry>
760 <entry>'PWC2'</entry>
761 <entry>Compressed format of the PWC driver.</entry>
762 </row>
763 <row id="V4L2-PIX-FMT-SN9C10X">
764 <entry><constant>V4L2_PIX_FMT_SN9C10X</constant></entry>
765 <entry>'S910'</entry>
766 <entry>Compressed format of the SN9C102 driver.</entry>
767 </row>
768 <row id="V4L2-PIX-FMT-SN9C20X-I420">
769 <entry><constant>V4L2_PIX_FMT_SN9C20X_I420</constant></entry>
770 <entry>'S920'</entry>
771 <entry>YUV 4:2:0 format of the gspca sn9c20x driver.</entry>
772 </row>
773 <row id="V4L2-PIX-FMT-STV0680">
774 <entry><constant>V4L2_PIX_FMT_STV0680</constant></entry>
775 <entry>'S680'</entry>
776 <entry>Bayer format of the gspca stv0680 driver.</entry>
777 </row>
778 <row id="V4L2-PIX-FMT-WNVA">
779 <entry><constant>V4L2_PIX_FMT_WNVA</constant></entry>
780 <entry>'WNVA'</entry>
781 <entry><para>Used by the Winnov Videum driver, <ulink
782url="http://www.thedirks.org/winnov/">
783http://www.thedirks.org/winnov/</ulink></para></entry>
784 </row>
785 <row id="V4L2-PIX-FMT-TM6000">
786 <entry><constant>V4L2_PIX_FMT_TM6000</constant></entry>
787 <entry>'TM60'</entry>
788 <entry><para>Used by Trident tm6000</para></entry>
789 </row>
790 <row id="V4L2-PIX-FMT-YYUV">
791 <entry><constant>V4L2_PIX_FMT_YYUV</constant></entry>
792 <entry>'YYUV'</entry>
793 <entry>unknown</entry>
794 </row>
795 <row id="V4L2-PIX-FMT-Y4">
796 <entry><constant>V4L2_PIX_FMT_Y4</constant></entry>
797 <entry>'Y04 '</entry>
798 <entry>Old 4-bit greyscale format. Only the least significant 4 bits of each byte are used,
799the other bits are set to 0.</entry>
800 </row>
801 <row id="V4L2-PIX-FMT-Y6">
802 <entry><constant>V4L2_PIX_FMT_Y6</constant></entry>
803 <entry>'Y06 '</entry>
804 <entry>Old 6-bit greyscale format. Only the least significant 6 bits of each byte are used,
805the other bits are set to 0.</entry>
806 </row>
807 </tbody>
808 </tgroup>
809 </table>
810 </section>
811
812 <!--
813Local Variables:
814mode: sgml
815sgml-parent-document: "v4l2.sgml"
816indent-tabs-mode: nil
817End:
818 -->