users-guide.ps

linux下将各类格式图片转换工具

0 F
3.56 (This makes all slave processes run nicely. This) 85.04 560.64 P
0.8 (means that interactive users take precedence, so they don\325t) 56.69 549.64 P
-0.2 (feel like they\325re running in molasses. If you want to be mean) 56.69 538.64 P
(to them, don\325t use this option. :-\051) 56.69 527.64 T
0 12 Q
(3.1 Architecture Overview) 56.69 500.31 T
0 10 Q
1.3 (Figure 1 shows a diagram of the system architecture. The) 56.69 474.64 P
-0.03 (slaves exist on the different slave machines which you spec-) 56.69 463.64 P
1.5 (ify \050see Section 3.2\051. The server processes all live on the) 56.69 452.64 P
(machine you run the encoder on.) 56.69 441.64 T
0 12 Q
(3.2 Specifying Slave Machines \050both\051) 56.69 412.31 T
0 10 Q
0.03 (You specify the slave machines in the parameter file.) 85.04 386.64 P
-0.2 (For each slave you must specify the username to use, as well) 56.69 375.64 P
0.06 (as the executable mpeg_encode program. If a slave does not) 56.69 364.64 P
1.52 (have NFS access, then it is REMOTE and you must also) 56.69 353.64 P
(specify where the parameter file is.) 56.69 342.64 T
(Usage:) 85.04 316.64 T
2 F
(PARALLEL) 85.04 303.64 T
(slave_specification) 85.04 290.64 T
(END_PARALLEL) 85.04 277.64 T
(slave_specification) 85.04 251.64 T
0 F
( can be either:) 198.98 251.64 T
2 F
(machine username executable) 85.04 238.64 T
0 F
(or) 85.04 225.64 T
2 F
2.9 (REMOTE machine username executable) 85.04 212.64 P
(param_file) 56.69 201.64 T
0 F
0.76 (You must have an account with the given username) 85.04 175.64 P
0.06 (on each machine, and you must place your machine/login in) 56.69 164.64 P
(the appropriate .rhosts files.) 56.69 153.64 T
1.59 (To make it easier to run experiments with varying) 85.04 127.64 P
1.35 (numbers of processors, there is a command-line argument) 56.69 116.64 P
(which limits the number of slave machines.) 56.69 105.64 T
(Usage:) 85.04 79.64 T
2 F
(-max_machines num_machines) 85.04 66.64 T
314.36 54.99 552.76 735.31 R
7 X
V
0 F
0 X
0.98 (This means that the encoder will use no more than) 342.71 728.64 P
2 F
(num_machines) 314.36 717.64 T
0 F
( machines as slaves.) 386.32 717.64 T
0 12 Q
(3.3 Remote Shell \050parameter \336le\051) 314.36 677.31 T
0 10 Q
0.98 (To run processes on the slave machines, mpeg_en-) 342.71 651.64 P
1.52 (code uses the remote shell command. On most machines) 314.36 640.64 P
-0.09 (this is the command) 314.36 629.64 P
2 F
-0.2 (rsh) 396.76 629.64 P
0 F
-0.09 (. On HP machines, however, rsh is) 414.75 629.64 P
0.39 (the restricted shell; on HP machines, the right command to) 314.36 618.64 P
(use is) 314.36 607.64 T
2 F
(remsh) 339.35 607.64 T
0 F
(, rather than) 369.33 607.64 T
2 F
(rsh) 419.84 607.64 T
0 F
(.) 437.83 607.64 T
(Usage:) 342.71 594.64 T
2 F
(RSH <rsh command>) 342.71 568.64 T
0 12 Q
(3.4 Scheduling Algorithms \050parameter \336le\051) 314.36 528.31 T
0 10 Q
0.87 (The encoder provides 3 different scheduling algorithms to) 314.36 502.64 P
(schedule which processors get which frames.) 314.36 491.64 T
2.03 (The first scheduling algorithm simply assigns N/P) 340.16 478.64 P
0.28 (frames to each processor, where N is the number of frames) 314.36 467.64 P
0.72 (and P is the number of processors. This has the advantage) 314.36 456.64 P
-0.01 (of minimal overhead, but only works well when all the pro-) 314.36 445.64 P
0.3 (cessors run at nearly the same speed. Also, since most pro-) 314.36 434.64 P
0.37 (cessors will finish at about the same time, you will have to) 314.36 423.64 P
1.41 (wait at the end while the Combine Server gathers all the) 314.36 412.64 P
(frame files together.) 314.36 401.64 T
(Usage:) 340.16 388.64 T
(PARALLEL_PERFECT) 340.16 375.64 T
3.6 (The second scheduling algorithm first assigns S) 340.16 349.64 P
-0.09 (frames to each processor. When a processor is finished, it is) 314.36 338.64 P
1.49 (assigned T seconds of work \050the scheduler estimates this) 314.36 327.64 P
1.97 (based on previous performance\051. S should be at least 3,) 314.36 316.64 P
0.37 (preferably at least 5 or 6, to insure a good estimate of each) 314.36 305.64 P
(processor\325s speed.) 314.36 294.64 T
(Usage:) 340.16 281.64 T
(PARALLEL_TEST_FRAMES S) 340.16 268.64 T
(PARALLEL_TIME_CHUNKS T) 340.16 255.64 T
0.79 (The third scheduling algorithm is probably the best.) 340.16 229.64 P
2.43 (It also first assigns S frames to each processor. Subse-) 314.36 218.64 P
3.44 (quently, however, whenever a processor finishes, it is) 314.36 207.64 P
1.1 (assigned enough work to keep it busy until almost every-) 314.36 196.64 P
2.04 (thing is done. Effectively, a processor is assigned many) 314.36 185.64 P
-0.06 (frames, and then fewer and fewer frames as more work gets) 314.36 174.64 P
2.92 (done. This insures good load balancing, while limiting) 314.36 163.64 P
(scheduling overhead.) 314.36 152.64 T
(Usage:) 340.16 139.64 T
(PARALLEL_TEST_FRAMES S) 340.16 126.64 T
(PARALLEL_CHUNK_TAPER) 340.16 113.64 T
0 12 Q
(3.5 Parallel problems \050parameter \336le\051) 314.36 86.31 T
0 10 Q
8.16 (There are some unsupported features using) 340.16 73.64 P
-0.18 (REMOTE to specify slaves: The \324command\324 form of gener-) 314.36 62.64 P
FMENDPAGE
%%EndPage: "8" 7
%%Page: "7" 7
612 792 0 FMBEGINPAGE
57.97 26.65 554.03 40.82 R
7 X
0 K
V
0 10 Q
0 X
(7) 303.5 34.15 T
56.69 54.99 297.64 477.36 R
7 X
V
0 X
(you can specify frame files to combine.) 56.69 470.69 T
1.11 (The parameter file may specify input frame files in) 85.04 457.69 P
0.94 (the same manner as normal input files -- except instead of) 56.69 446.69 P
7.11 (using INPUT_DIR, INPUT, and END_INPUT, use) 56.69 435.69 P
28.3 (FRAME_INPUT_DIR, FRAME_INPUT, and) 56.69 424.69 P
1.83 (FRAME_END_INPUT. If no input frame files are speci-) 56.69 413.69 P
0.16 (fied, then the default is to use the output file name with suf-) 56.69 402.69 P
(fix \322.frame.<frame_num>\323 starting from 0 as the input files.) 56.69 391.69 T
0 12 Q
(2.15 Stats and Other Options \050command line\051) 56.69 364.36 T
0 10 Q
0.43 (There are several options for printing or suppressing) 85.04 338.69 P
(useful information.) 56.69 327.69 T
0.75 (The encoder always prints \050to stdout\051 parameter file) 85.04 314.69 P
2.87 (information and statistics about how many I, P, and B) 56.69 303.69 P
0.8 (frames there were, and information about compression and) 56.69 292.69 P
2.31 (quality. You can send these statistics, in addition to the) 56.69 281.69 P
(screen, to a file.) 56.69 270.69 T
(Usage:) 85.04 257.69 T
2 F
( -stat stat_file) 110.83 244.69 T
0 F
2.56 (This appends the parameter file info and stats to) 85.04 231.69 P
2 F
(stat_file) 56.69 220.69 T
0 F
0.87 (Normally, the statistics do not include any informa-) 85.04 194.69 P
1.77 (tion about quality. This is because computing the quality) 56.69 183.69 P
2.02 (takes a little more time. If you wish to have the quality) 56.69 172.69 P
1.17 (included in the statistics, use the -snr command line argu-) 56.69 161.69 P
(ment.) 56.69 150.69 T
(Usage:) 85.04 137.69 T
2 F
(-snr) 85.04 124.69 T
0 F
1.78 (This prints the signal-to-noise ratio \050snr\051 and peak) 85.04 111.69 P
(snr.) 56.69 100.69 T
3.69 (An additional statistic measure is mean squared) 85.04 74.69 P
0.78 (error. If you wish to see the per-block mean squared error,) 56.69 63.69 P
314.36 54.99 552.76 477.36 R
7 X
V
0 X
(use the -mse command line flag \050sets -snr as a side effect\051.) 314.36 470.69 T
(Usage:) 342.71 457.69 T
2 F
(-mse) 342.71 444.69 T
0 F
(This prints the MSE for each block encoded) 342.71 431.69 T
1.02 (Another set of data which can be useful is a histo-) 342.71 418.69 P
0.44 (gram of the motion vectors. The encoder can keep track of) 314.36 407.69 P
-0.16 (P-frame motion vectors and forward and backward B-frame) 314.36 396.69 P
0.4 (motion vectors. The output is in the form of a matrix, each) 314.36 385.69 P
0.98 (entry corresponding to a motion vector in the search win-) 314.36 374.69 P
0.49 (dow. The center of the matrix represents \0500,0\051 motion vec-) 314.36 363.69 P
(tors.) 314.36 352.69 T
(Usage:) 342.71 339.69 T
2 F
(-mv_histogram) 342.71 326.69 T
0 F
1.47 (During normal execution, the encoder outputs two) 342.71 300.69 P
0.58 (kinds of information. It prints a single line for each frame,) 314.36 289.69 P
0.86 (summarizing block type and time info. It also prints, after) 314.36 278.69 P
-0.07 (each frame, an estimate of the remaining running time. You) 314.36 267.69 P
(can modify how often or if this information is to be shown.) 314.36 256.69 T
(Usage:) 342.71 243.69 T
2 F
( -quiet num) 342.71 230.69 T
( -no_frame_summary) 342.71 217.69 T
(-realquiet) 349.8 204.69 T
0 F
-0.13 (If) 342.71 191.69 P
2 F
-0.31 (num) 351.74 191.69 P
0 F
-0.13 ( is negative, the time estimate is never shown;) 369.73 191.69 P
1.94 (otherwise, it reports a time estimate no more often than) 314.36 180.69 P
1.22 (every) 314.36 169.69 P
2 F
2.92 (num) 340.27 169.69 P
0 F
1.22 ( seconds \050unless the time estimate rises, which) 358.26 169.69 P
1 (will happen near the beginning of the run\051. The default is) 314.36 158.69 P
2 F
(num) 314.36 147.69 T
0 F
( = 0, which means report after every frame.) 332.35 147.69 T
1.08 (If) 342.71 134.69 P
2 F
2.59 (-no_frame_summary) 352.94 134.69 P
0 F
1.08 ( is given, then informa-) 454.89 134.69 P
(tion about each frame is not printed.) 314.36 123.69 T
2 F
10.05 (-realquiet stops all printing,) 342.71 110.69 P
(other than error messages.) 314.36 99.69 T
0 F
1 (Another nice feature is that the encoder can output) 342.71 73.69 P
0.75 (the bit rate, on both a frame-to-frame scale, and also an I-) 314.36 62.69 P
385.51 518.46 470.55 575.15 R
7 X
V
0.5 H
0 Z
0 X
N
141.05 721.49 294.05 766.49 18 RR
7 X
V
0 X
N
1 18 Q
(Master Server) 165.05 738.43 T
102.05 623.49 181.05 660.49 R
7 X
V
0 X
N
(Slave) 121.05 637.49 T
452.05 625.49 531.05 662.49 R
7 X
V
0 X
N
(Slave) 471.05 639.49 T
333.05 624.49 412.05 661.49 R
7 X
V
0 X
N
(Slave) 352.05 638.49 T
218.05 623.49 297.05 660.49 R
7 X
V
0 X
N
(Slave) 237.05 637.49 T
147.47 523.06 300.47 568.06 18 RR
7 X
V
0 X
N
(Combine Server) 165.31 539.33 T
498.64 673.25 501.04 661.49 492.77 670.19 495.71 671.72 4 Y
V
472.44 709.73 470.04 721.49 478.31 712.79 475.37 711.26 4 Y
V
495.71 671.72 475.38 711.26 2 L
7 X
V
0 X
N
137.07 670.41 129.04 661.49 131.12 673.31 134.09 671.86 4 Y
V
149.02 710.57 157.04 719.49 154.97 707.67 151.99 709.12 4 Y
V
134.1 671.86 152 709.12 2 L
7 X
V
0 X
N
164.04 661.55 152.04 661.49 162.18 667.9 163.11 664.73 4 Y
V
344.03 719.43 356.03 719.49 345.89 713.08 344.96 716.26 4 Y
V
163.12 664.72 344.98 716.25 2 L
7 X
V
0 X
N
392.55 667.26 382.02 661.5 387.89 671.96 390.22 669.61 4 Y
V
430.5 714.73 441.02 720.49 435.15 710.03 432.82 712.38 4 Y
V
390.24 669.6 432.85 712.37 2 L
7 X
V
0 X
N
277.89 664.38 266.04 662.49 275.09 670.37 276.49 667.37 4 Y
V
377.19 717.6 389.04 719.49 379.98 711.61 378.59 714.61 4 Y
V
276.5 667.37 378.6 714.61 2 L
7 X
V
0 X
N
241.78 672.73 248.04 662.49 237.31 667.86 239.55 670.29 4 Y
V
192.3 709.25 186.04 719.49 196.78 714.12 194.54 711.69 4 Y
V
239.55 670.29 194.54 711.69 2 L
7 X
V
0 X
N
353.94 670.32 363.03 662.49 351.17 664.31 352.56 667.32 4 Y
V
245.13 712.66 236.04 720.49 247.9 718.67 246.52 715.67 4 Y
V
352.57 667.31 246.52 715.66 2 L
7 X
V
0 X
N
470.93 667.95 481.05 661.49 469.05 661.61 469.99 664.78 4 Y
V
287.16 714.03 277.05 720.49 289.05 720.37 288.1 717.2 4 Y
V
469.99 664.78 288.1 717.2 2 L
7 X
V
0 X
N
336.05 720.49 489.05 765.49 18 RR
7 X
V
0 X
N
(Decode Server) 360.05 737.43 T
134.71 613.56 130.39 624.76 139.99 617.55 137.35 615.56 4 Y
V
168.59 579.26 172.91 568.07 163.31 575.27 165.95 577.27 4 Y
V
137.35 615.56 165.96 577.26 2 L
7 X
V
0 X
N
234.18 617.55 243.77 624.76 239.45 613.56 236.82 615.56 4 Y
V
210.85 575.27 201.26 568.07 205.58 579.26 208.21 577.27 4 Y
V
236.82 615.56 208.22 577.26 2 L
7 X
V
0 X
N
352.39 622.81 364.23 624.77 355.23 616.83 353.81 619.82 4 Y
V
255.6 570.03 243.76 568.08 252.77 576.01 254.18 573.02 4 Y
V
353.83 619.81 254.2 573.01 2 L
7 X
V
0 X
N
465.63 624.68 477.63 624.76 467.5 618.33 466.56 621.51 4 Y
V
291.21 568.14 279.21 568.07 289.34 574.49 290.28 571.32 4 Y
V
466.57 621.51 290.28 571.31 2 L
7 X
V
0 X
N
393.94 616.24 385.49 624.76 397.17 622.01 395.55 619.12 4 Y
V
440.83 597.86 449.27 589.33 437.59 592.09 439.21 594.98 4 Y
V
395.57 619.11 439.23 594.97 2 L
7 X
V
0 X
N
281.34 617.08 272.12 624.76 284.01 623.14 282.67 620.11 4 Y
V
376.28 582.84 385.5 575.16 373.61 576.78 374.94 579.81 4 Y
V
282.68 620.1 374.96 579.8 2 L
7 X
V
0 X
N
168.9 618.38 158.74 624.76 170.74 624.73 169.82 621.56 4 Y
V
375.34 567.36 385.51 560.98 373.51 561 374.43 564.18 4 Y
V
169.82 621.55 374.43 564.18 2 L
7 X
V
0 X
N
312.01 543.5 300.47 546.8 312.01 550.11 312.01 546.8 4 Y
V
373.98 550.11 385.51 546.8 373.98 543.5 373.98 546.8 4 Y
V
312.01 546.8 373.98 546.8 2 L
7 X
V
0 X
N
302.9 498.61 259.93 498.61 2 L
V
1.14 H
N
1 12 Q
(Figure 1) 259.93 499.8 T
385.51 498.49 302.9 498.49 2 L
V
0.59 H
N
0 F
(: Network Model) 302.9 499.8 T
495.43 619 505.95 624.77 500.08 614.3 497.76 616.65 4 Y
V
473.96 588.02 463.44 582.25 469.3 592.72 471.63 590.37 4 Y
V
497.79 616.64 471.66 590.35 2 L
7 X
V
0.5 H
0 X
N
1 18 Q
(Disk) 410.52 530.49 T
7 X
90 450 42.52 14.17 428.03 575.15 G
0 X
90 450 42.52 14.17 428.03 575.15 A
7 X
180 270 42.94 21.26 428.45 521.28 G
2 Z
0 X
180 270 42.94 21.26 428.45 521.28 A
7 X
270 360 42.94 21.26 427.68 521.61 G
0 X
270 360 42.94 21.26 427.68 521.61 A
FMENDPAGE
%%EndPage: "7" 6
%%Page: "6" 6
612 792 0 FMBEGINPAGE
57.97 26.65 554.03 40.82 R
7 X
0 K
V
0 10 Q
0 X
(6) 303.5 34.15 T
56.69 54.99 297.64 735.31 R
7 X
V
0 X
3.03 (number\051. This field entry can be changed by adding a) 56.69 728.64 P
1.27 (USER_DATA parameter whose vale is the name of a file) 56.69 717.64 P
(continuing the data to add to the header.) 56.69 706.64 T
(Usage:) 85.04 693.64 T
(USER_DATA ./user_data.txt) 85.04 667.64 T
0 12 Q