{ "mode": "man", "parameter": "ppmtompeg", "section": "1", "url": "https://www.chedong.com/phpMan.php/man/ppmtompeg/1/json", "generated": "2026-10-01T17:37:30Z", "synopsis": "ppmtompeg [options] parameter-file", "sections": { "NAME": { "content": "ppmtompeg - encode an MPEG-1 bitstream\n\n", "subsections": [] }, "SYNOPSIS": { "content": "ppmtompeg [options] parameter-file\n\n", "subsections": [] }, "DESCRIPTION": { "content": "This program is part of Netpbm(1).\n\nppmtompeg produces an MPEG-1 video stream. MPEG-1 is the first great video compression\nmethod, and is what is used in Video CDs (VCD). ppmtompeg originated in the year 1995. DVD\nuses a more advanced method, MPEG-2. There is an even newer method called MPEG-4 which is\nalso called Divx. I don't know where one finds that used.\n\nThere's technically a difference between a compression method for video and an actual file\n(stream) format for a movie, and I don't know if it can be validly said that the format of\nthe stream ppmtompeg produces is MPEG-1.\n\nMencoder from the \u001b]8;;http://www.mplayerhq.hu\u001b\\Mplayer package\u001b]8;;\u001b\\ is probably superior for most video format generation\nneeds, if for no other reason than that it is more popular.\n\nThe programming library \u001b]8;;http://pm2v.free.fr\u001b\\PM2V\u001b]8;;\u001b\\ generates MPEG-2 streams.\n\nUse \u001b]8;;http://www.mplayerhq.hu\u001b\\Mplayer\u001b]8;;\u001b\\ (not part of Netpbm) to do the reverse conversion: to create a series of PNM\nfiles from an MPEG stream.\n\nparamfile is a parameter file which includes a list of input files and other parameters.\nThe file is described in detail below.\n\nTo understand this program, you need to understand something about the complex MPEG-1 format.\nOne source of information about this standard format is the section Introduction to MPEG in\nthe \u001b]8;;http://www.faqs.org/faqs/compression-faq/\u001b\\Compression FAQ\u001b]8;;\u001b\\ .\n\n", "subsections": [] }, "OPTIONS": { "content": "The -gop, -combinegops, -frames, and -combineframes options are all mutually exclusive.\n\n\n", "subsections": [ { "name": "-stat stat_file", "content": "This option causes ppmtompeg to append the statistics that it write to Standard Output\nto the file statfile as well. The statistics use the following abbreviations: bits\nper block (bpb), bits per frame (bpf), seconds per frame (spf), and bits per second\n(bps).\n\nThese statistics include how many I, P, and B frames there were, and information about\ncompression and quality.\n\n\n" }, { "name": "-quiet _", "content": "causes ppmtompeg not to report remaining time more often than every numseconds sec‐\nonds (unless the time estimate rises, which will happen near the beginning of the\nrun). A negative value tells ppmtompeg not to report at all. 0 is the default (re‐\nports once after each frame). Note that the time remaining is an estimate and does\nnot take into account time to read in frames.\n\n" }, { "name": "-realquiet", "content": "causes ppmtompeg to run silently, with the only screen output being errors. Particu‐\nlarly useful when reading input from stdin. The equivalent of the -quiet common op‐\ntion of most other Netpbm programs.\n\n" }, { "name": "-no_frame_summary", "content": "This option prevents ppmtompeg from printing a summary line for each frame\n\n" }, { "name": "-float_dct", "content": "forces ppmtompeg to use a more accurate, yet more computationally expensive version\nof the DCT.\n\n" }, { "name": "-gop _", "content": "causes ppmtompeg to encode only the numbered GOP (first GOP is 0). The parameter file\nis the same as for normal usage. The output file will be the normal output file with\nthe suffix .gop.gopnum. ppmtompeg does not output any sequence information.\n\n" }, { "name": "-combine_gops", "content": "causes ppmtompeg simply to combine some GOP files into a single MPEG output stream.\nppmtompeg inserts a sequence header and trailer. In this case, the parameter file\nneeds only to contain the SIZE value, an output file, and perhaps a list of input GOP\nfiles (see below).\n\nIf you don't supply a list of input GOP files is used, then ppmtompeg assumes you're\nusing the same parameter file you used when you created the input (with the -gop op‐\ntion) and calculates the corresponding gop filenames itself. If this is not the case,\nyou can specify input GOP files in the same manner as normal input files -- except in‐\nstead of using INPUTDIR, INPUT, and ENDINPUT, use GOPINPUTDIR, GOPINPUT, and\nGOPENDINPUT. If no input GOP files are specified, then the default is to use the\noutput file name with suffix .gop.gopnum, with gopnum starting from 0, as the input\nfiles.\n\nThus, unless you're mixing and matching GOP files from different sources, you can sim‐\nply use the same parameter file for creating the GOP files (-gop) and for later turn‐\ning them into an MPEG stream (-combinegops).\n\n\n" }, { "name": "-frames _", "content": "This option causes ppmtompeg to encode only the frames numbered firstframe to\nlastframe, inclusive. The parameter file is the same as for normal usage. The out‐\nput will be placed in separate files, one per frame, with the file names being the\nnormal output file name with the suffix .frame.framenum. No GOP header information\nis output. (Thus, the parameter file need not include the GOPSIZE value)\n\nUse ppmtompeg -combineframes to combine these frames later into an MPEG stream.\n\n\n" }, { "name": "-combine_frames", "content": "This option causes ppmtompeg simply to combine some individual MPEG frames (such as\nyou might have created with an earlier run of ppmtompeg -frames) into a single MPEG\nstream. Sequence and GOP headers are inserted appropriately. In this case, the para‐\nmeter file needs to contain only the SIZE value, the GOPSIZE value, an output file,\nand perhaps a list of frame files (see below).\n\nThe parameter file may specify input frame files in the same manner as normal input\nfiles -- except instead of using INPUTDIR, INPUT, and ENDINPUT, use FRAMEINPUTDIR,\nFRAMEINPUT, and FRAMEENDINPUT. If no input frame files are specified, then the de‐\nfault is to use the output file name with suffix .frame.framenum, with framenum\nstarting from 0, as the input files.\n\n\n\n" }, { "name": "-nice", "content": "ity. (This is relevant only if you are running ppmtompeg in parallel mode. Other‐\nwise, there are no remote processes). See 'man nice.'\n\n" }, { "name": "-max_machines _", "content": "This option causes ppmtompeg to use no more than nummachines machines as slaves for\nuse in parallel encoding.\n\n" }, { "name": "-snr", "content": "tistics. Prints SNR (Y U V) and peak SNR (Y U V) for each frame. In summary, prints\naverages of luminance only (Y). SNR is defined as 10*log(variance of original/vari‐\nance of error). Peak SNR is defined as 20*log(255/RMSE). Note that ppmtompeg runs a\nlittle slower when you use this option.\n\n" }, { "name": "-mse", "content": "tomatically reports the quality of the images, so there is no need to specify -snr\nthen.\n\n" }, { "name": "-bit_rate_info _", "content": "This option makes ppmtompeg write bit rate information into the file ratefile. Bit\nrate information is bits per frame, and also bits per I-frame-to-I-frame.\n\n" }, { "name": "-mv_histogram", "content": "This option causes ppmtompeg to print a histogram of the motion vectors as part of\nstatistics. There are three histograms -- one for P frame, one for forward B frame,\nand one for backward B frame motion vectors.\n\nThe output is in the form of a matrix, each entry corresponding to one motion vector\nin the search window. The center of the matrix represents (0,0) motion vectors.\n\n" }, { "name": "-debug_sockets", "content": "This option causes ppmtompeg to print to Standard Output messages that narrate the\ncommunication between the machines when you run ppmtompeg in \u001b]8;;#parallel\u001b\\parallel mode\u001b]8;;\u001b\\ .\n\n" }, { "name": "-debug_machines", "content": "This option causes ppmtompeg to print to Standard Output messages that narrate the\nprogress of the conversion on the various machines when you run ppmtompeg in \u001b]8;;#parallel\u001b\\parallel\nmode\u001b]8;;\u001b\\ .\n\n\n\n" } ] }, "PARAMETER FILE": { "content": "The parameter file must contain the following lines (except when using the -combinegops or", "subsections": [ { "name": "-combine_frames", "content": "PATTERN pattern\nThis statement specifies the pattern (sequence) of I frames, P frames, and B frames.\npattern is just a sequence of the letters I, P, and B with nothing between. Example:\n\nPATTERN IBBPBBPBBPBBPBB\n\n\nSee\n\u001b]8;;#ipb\u001b\\I Frames, P Frames, B Frames\u001b]8;;\u001b\\\n.\n\n\nOUTPUT output file\nThis names the file where the output MPEG stream goes.\n\n\nINPUTDIR directory\nThis statement tells where the input images (frames) come from.\nIf each frame is in a separate file, directory is the directory\nwhere they all are. You may use . to refer to the current\ndirectory. A null directory refers to the root directory of the\nsystem file tree.\n\nTo have ppmtompeg read all the frames serially from Standard\nInput, specify\nINPUTDIR stdin\n\n\n\nINPUT This line must be followed by a list of the input files (in display order) and then\nthe line ENDINPUT.\n\nThere are three types of lines between INPUT and ENDINPUT. First, a line may simply\nbe the name of an input file. Second, the line may be of the form singlestarexpr\n[x-y]. singlestarexpr can have a single * in it. It is replaced by all the numbers\nbetween x and y inclusive. So, for example, the line tennis*.ppm [12-15] refers to\nthe files tennis12.ppm, tennis13.ppm, tennis14.ppm, tennis15.ppm.\n\nUniform zero-padding occurs, as well. For example, the line football.*.ppm [001-130]\nrefers to the files football.001.ppm, football.002.ppm, ..., football.009.ppm, foot‐\nball.010.ppm, ..., football.130.ppm.\n\nThe third type of line is: singlestarexpr [x-y+s], where the line is treated exactly\nas above, except that we skip by s. Thus, the line football.*.ppm [001-130+4] refers\nto the files football.001.ppm, football.005.ppm, football.009.ppm, football.013.ppm,\netc.\n\nFurthermore, a line may specify a shell command to execute to generate lines to be in‐\nterpreted as described above, as if those lines were in the parameter file instead.\nUse back ticks, like in the Bourne Shell, like this:\n\n`cat myfilelist`\n\n\nIf input is from Standard Input (per the INPUTDIR statement), ppmtompeg ignores the\nINPUT/ENDINPUT block, but it still must be present.\n\n\nBASEFILEFORMAT {PPM | PNM | YUV |\nJPEG | JMOVIE} ppmtompeg must convert all input files to one of the following\nformats as a first step of processing: PNM, YUV, JPEG(v4), or JMOVIE. (The conversion\nmay be trivial if your input files are already in one of these formats). This line\nspecifies which of the four formats. PPM is actually a subset of PNM. The separate\nspecification is allowed for backward compatibility. Use PNM instead of PPM in new\napplications.\n\n\nINPUTCONVERT conversioncommand\nYou must specify how to convert a file to the base file format. If no conversion is\nnecessary, then you would just say:\n\nINPUTCONVERT *\n\n\nOtherwise, conversioncommand is a shell command that causes an image in the format\nyour specified with BASEFILEFORMAT to be written to Standard Output. ppmtompeg exe‐\ncutes the command once for each line between INPUT and ENDINPUT (which is normally,\nbut not necessarily, a file name). In the conversion command, ppmtompeg replaces each\n'*' with the contents of that line.\n\nIf you had a bunch of gif files, you might say:\nINPUTCONVERT giftopnm *\n\n\nIf you have a bunch of separate a.Y, a.U, and a.V files (where\nthe U and V have already been subsampled), then you might say:\n\nINPUTCONVERT cat *.Y *.U *.V\n\n\nInput conversion is not allowed with input from stdin, so use\n\nINPUTCONVERT *\n\n\nas described above.\n\n\nSIZE widthxheight\n\nwidth and height are the width and height of each frame in pixels.\n\nWhen ppmtompeg can get this information from the input image files, it ignores the\nSIZE parameter and you may omit it.\n\nWhen the image files are in YUV format, the files don't contain dimension information,\nso SIZE is required.\n\nWhen ppmtompeg is running in parallel mode, not all of the processes in the network\nhave access to the image files, so SIZE is required and must give the same dimensions\nas the input image files.\n\n\nYUVSIZE widthxheight\nThis is an obsolete synonym of SIZE.\n\n\nYUVFORMAT {ABEKAS | PHILLIPS | UCB |\nEYUV | pattern} This is meaningful only when BASEFILEFORMAT\nspecifies YUV format, and then it is required. It specifies the sub-format of the YUV\nclass.\n\n\n\nGOPSIZE n\nn is the number of frames in a Group of Pictures. Except that because a GOP must\nstart with an I frame, ppmtompeg makes a GOP as much longer than n as it has to to\nmake the next GOP start with an I frame.\n\nNormally, it makes sense to make your GOP size a multiple of your pattern length (the\nlatter is determined by the PATTERN parameter file statement).\n\nSee \u001b]8;;#gop\u001b\\Group Of Pictures\u001b]8;;\u001b\\ .\n\n\nSLICESPERFRAME n\nn is roughly the number of slices per frame. Note, at least one MPEG player may com‐\nplain if slices do not start at the left side of an image. To ensure this does not\nhappen, make sure the number of rows is divisible by SLICESPERFRAME.\n\n\nPIXEL {FULL | HALF}\nuse half-pixel motion vectors, or just full-pixel ones It is usually important that\nyou use half-pixel motion vectors, because it results in both better quality and bet‐\nter compression.\n\n\n\nRANGE n\nUse a search range of n pixels in each of the four directions from a subject pixel.\n(So the search window is a square n*2 pixels on a side).\n\n\nPSEARCHALG {EXHAUSTIVE | TWOLEVEL |\nSUBSAMPLE | LOGARITHMIC} This statement tells ppmtompeg what kind of search\ntechnique (algorithm) to use for P frames. You select the desired\ncombination of speed and compression. EXHAUSTIVE gives the\nbest compression, but LOGARITHMIC is the fastest.\nTWOLEVEL is an exhaustive full-pixel search, followed by a\nlocal half- pixel search around the best full-pixel vector (the\nPIXEL option is ignored for this search technique).\n\n\nBSEARCHALG {SIMPLE | CROSS2 | EXHAUSTIVE}\nThis statement tells ppmtompeg what kind of search\ntechnique (algorithm) to use for B frames. SIMPLE means\nfind best forward and backward vectors, then interpolate.\nCROSS2 means find those two vectors, then see what backward\nvector best matches the best forward vector, and vice versa.\nEXHAUSTIVE does an n-squared search and is\nextremely slow in relation to the others (CROSS2\nis about half as fast as SIMPLE).\n\n\nIQSCALE n\nUse n as the qscale for I frames.\nSee \u001b]8;;#qscale\u001b\\Qscale\u001b]8;;\u001b\\ .\n\n\nPQSCALE n\nUse n as the qscale for P frames.\nSee \u001b]8;;#qscale\u001b\\Qscale\u001b]8;;\u001b\\ .\n\n\nBQSCALE n\nUse n as the qscale for B frames.\nSee \u001b]8;;#qscale\u001b\\Qscale\u001b]8;;\u001b\\ .\n\n\nREFERENCEFRAME {ORIGINAL | DECODED}\nThis statement determines whether ppmtompeg uses the original images or the decoded\nimages when computing motion vectors. Using decoded images is more accurate and\nshould increase the playback quality of the output, but it makes the encoding take\nlonger and seems to give worse compression. It also causes some complications with\nparallel encoding. (see the section on parallel encoding). One thing you can do as a\ntrade-off is select ORIGINAL here, and lower the qscale (see QSCALE if the quality is\nnot good enough.\n\nOriginal or Decoded? (Normalized)\n────────────────────────────────────────────────────────────────────\nReference Compression Speed Quality I Quality P Quality B\nDecoded 1000 1000 1000 969 919\nOriginal 885 1373 1000 912 884\n\n\n\n\n\n\nThe following lines are optional:\n\n\n\n\nFORCEENCODELASTFRAME\nThis statement is obsolete. It does nothing.\n\nBefore Netpbm 10.26 (January 2005), ppmtompeg would drop trailing B frames from your\nmovie, since a movie can't end with a B frame. (See \u001b]8;;#ipb\u001b\\I Frames, P Frames, B Frames\u001b]8;;\u001b\\ .)\nYou would have to specify FORCEENCODELASTFRAME to stop that from happening and get\nthe same function that ppmtompeg has today.\n\n\n\nNIQTABLE\nThis statement specifies a custom non-intra quantization table. If you don't specify\nthis statement, ppmtompeg uses a default non-intra quantization table.\n\nThe 8 lines immediately following NIQTABLE specify the quantization table. Each line\ndefines a table row and consists of 8 integers, whitespace-delimited, which define the\ntable columns.\n\n\nIQTABLE\nThis is analogous to NIQTABLE, but for the intra quantization table.\n\n\nASPECTRATIO ratio\nThis statement specifies the aspect ratio for ppmtompeg to specify in the MPEG output.\nI'm not sure what this is used for.\n\nratio must be 1.0, 0.6735, 0.7031, 0.7615, 0.8055, 0.8437, 0.8935, 0.9157, 0.9815,\n1.0255, 1.0695, 1.0950, 1.1575, or 1.2015.\n\n\nFRAMERATE rate\nThis specifies the frame rate for ppmtompeg to specify in the MPEG output. Some play‐\ners use this value to determine the playback rate.\n\nrate must be 23.976, 24, 25, 29.97, 30, 50, 59.94, or 60.\n\n\nBITRATE rate\nThis specifies the bit rate for Constant Bit Rate (CBR) encoding.\n\nrate must be an integer.\n\n\nBUFFERSIZE size\nThis specifies the value ppmtompeg is to specify in the MPEG output for the Video\nBuffering Verifier (VBV) buffer size needed to decode the sequence.\n\nA Video Verifying Buffer is a buffer in which a decoder keeps the decoded bits in or‐\nder to match the uneven speed of the decoding with the required constant playback\nspeed.\n\nAs ppmtompeg encodes the image, it simulates the decoding process in terms of how many\nbits would be in the VBV as each frame gets decoded, assuming a VBV of the size you\nindicate.\n\nIf you specify the WARNVBVUNDERFLOW statement, ppmtompeg issues a warning each time\nthe simulation underflows the buffer, which suggests that an underflow would occur on\nplayback, which suggests the buffer is too small.\n\nIf you specify the WARNVBVOVERFLOW statement, ppmtompeg issues a warning each time\nthe simulation overflows the buffer, which suggests that an overflow would occur on\nplayback, which suggests the buffer is too small.\n\n\nWARNVBVUNDERFLOW\n\nWARNVBVOVERFLOW\nSee BUFFERSIZE.\n\nThese options were new in Netpbm 10.26 (January 2005). Before that, ppmtompeg issued\nthe warnings always.\n\n\n\n\nThe following statements apply only to parallel operation:\n\n\n\n\nPARALLEL\nThis statement, paired with END PARALLEL, is what causes ppmtompeg to operate in par‐\nallel mode. See \u001b]8;;#parallel\u001b\\Parallel Operation\u001b]8;;\u001b\\ .\n\n\nEND PARALLEL\nThis goes with PARALLEL.\n\n\nPARALLELTESTFRAMES n\nThe master starts off by measuring each slave's speed. It does this by giving each\nslave n frames to encode and noting how long the slave takes to finish. These are not\njust test frames, though -- they're real frames and the results become part of the\noutput. ppmtompeg is old and measures time in undivided seconds, so to get useful\ntimings, specify enough frames that it will take at least 5 seconds to process them.\nThe default is 10.\n\nIf you specify FORCEIALIGN, ppmtompeg will increase the test frames value enough to\nmaintain the alignment.\n\nIf there aren't enough frames for every slave to have the indicated number of test\nframes, ppmtompeg will give some slaves fewer.\n\n\n\nPARALLELTIMECHUNKS t\nWhen you specify this statement, the master attempts to feed work to the slaves in\nchunks that take t seconds to process. It uses the speed measurement it made when it\nstarted up (see PARALLELTESTFRAMES) to decide how many frames to put in the chunk.\nThis statement obviously doesn't affect the first batch of work sent to each slave,\nwhich is the one used to measure the slave's speed.\n\nSmaller values of t increase communication, but improve load balancing. The default\nis 30 seconds.\n\nYou may specify only one of PARALLELTIMECHUNKS, PARALLELCHUNKTAPER, and PARAL‐\nLELPERFECT. PARALLELCHUNKTAPER is usually best.\n\n\nPARALLELCHUNKTAPER\nWhen you specify this statement, the master distributes work like with PARAL‐\nLELTIMECHUNKS, except that the master chooses the number of seconds for the chunks.\nIt starts with a large number and, as it gets closer to finishing the job, reduces it.\nThat way, it reduces scheduling overhead when precise scheduling isn't helpful, but\nstill prevents a slave from finishing early after all the work has already been handed\nout to the other slaves, and then sitting idle while there's still work to do.\n\nYou may specify only one of PARALLELTIMECHUNKS, PARALLELCHUNKTAPER, and PARAL‐\nLELPERFECT. PARALLELCHUNKTAPER is usually best.\n\n\n\nPARALLELPERFECT\nIf this statement is present, ppmtompeg schedules on the assumption that each machine\nis about the same speed. The master will simply divide up the frames evenly between\nthe slaves -- each slave gets the same number of frames. If some slaves are faster\nthan others, they will finish first and remain idle while the slower slaves continue.\n\nThis has the advantage of minimal scheduling overhead. Where slaves have different\nspeeds, though, it makes inefficient use of the fast ones. Where slaves are the same\nspeed, it also has the disadvantage that they all finish at the same time and feed\ntheir output to the single Combine Server in a burst, which makes less efficient use\nof the Combine Server and thus can increase the total elapsed time.\n\nYou may specify only one of PARALLELTIMECHUNKS, PARALLELCHUNKTAPER, and PARAL‐\nLELPERFECT. PARALLELCHUNKTAPER is usually best.\n\n\nRSH remoteshellcommand\nppmtompeg executes the shell command remoteshellcommand to start a process on an‐\nother machine. The default command is rsh, and whatever command you specify must have\ncompatible semantics. ssh is usually compatible. The command ppmtompeg uses is one\nlike this: ssh remote.host.com -l username shellcommand.\n\nBe sure to set up .rhosts files or SSH key authorizations where needed. Otherwise,\nyou'll have to type in passwords.\n\nOn some HP machines, rsh is the restricted shell, and you want to specify remsh.\n\n\nFORCEIALIGN\nThis statement forces each slave to encode a chunk of frames which is a multiple of\nthe pattern length (see PATTERN). Since the first frame in any pattern is an I frame,\nthis forces each chunk encoded by a slave to begin with an I frame.\n\nThis document used to say there was an argument to FORCEIALIGN which was the number\nof frames ppmtompeg would use (and was required to be a multiple of the pattern\nlength). But ppmtompeg has apparently always ignored that argument, and it does now.\n\n\nKEEPTEMPFILES\nThis statement causes ppmtompeg not to delete the temporary files it uses to transmit\nencoded frames to the combine server. This means you will be left with a file for\neach frame, the same as you would get with the -frames option.\n\nThis is mostly useful for debugging.\n\nThis works only if you're using a shared filesystem to communicate between the\nservers.\n\nThis option was new in Netpbm 10.26 (January 2005).\n\n\n\n\n" }, { "name": "Parameter File Notes", "content": "If you use the -combinegops option, then you need to specify only the SIZE and OUTPUT val‐\nues in the parameter file. In addition, the parameter file may specify input GOP files in\nthe same manner as normal input files -- except instead of using INPUTDIR, INPUT, and\nENDINPUT, use GOPINPUTDIR, GOPINPUT, and GOPENDINPUT. If you specify no input GOP\nfiles, then ppmtompeg uses by default the output file name with suffix .gop.gopnum, with\ngopnum starting from 0, as the input files.\n\nIf you use the -combineframes option, then you need to specify only the SIZE, GOPSIZE, and\nOUTPUT values in the parameter file. In addition, the parameter file may specify input frame\nfiles in the same manner as normal input files -- except instead of using INPUTDIR, INPUT,\nand ENDINPUT, use FRAMEINPUTDIR, FRAMEINPUT, and FRAMEENDINPUT. If no input frame\nfiles are specified, then the default is to use the output file name with suffix\n.frame.framenum, with framenum starting from 0, as the input files.\n\nAny number of spaces and tabs may come between each option and value. Lines beginning with #\nare ignored. Any other lines are ignored except for those between INPUT and ENDINPUT. This\nallows you to use the same parameter file for normal usage and for -combinegops and -com‐\nbineframes.\n\nThe file format is case-sensitive so all keywords should be in upper case.\n\nThe statements may appear in any order, except that the order within a block statement (such\nas INPUT ... END INPUT) is significant.\n\nppmtompeg is prepared to handle up to 16 B frames between reference frames when encoding with\ninput from stdin. (To build a modified ppmtompeg with a higher limit, change the constant\nBFRAMERUN in frame.c and recompile).\n\n" } ] }, "GENERAL USAGE INFORMATION": { "content": "", "subsections": [ { "name": "Qscale", "content": "The quantization scale values (qscale) give a trade-off between quality and compression. Us‐\ning different Qscale values has very little effect on speed. The qscale values can be set\nseparately for I, P, and B frames.\n\nYou select the qscale values with the IQSCALE, PQSCALE, and BSCALE parameter file statements.\n\nA qscale value is an integer from 1 to 31. Larger numbers give better compression, but worse\nquality. In the following, the quality numbers are peak signal-to-noise ratio, defined as:\nsignal-to-noise formula where MSE is the mean squared error.\n\n\nFlower garden tests:\n" }, { "name": "Qscale vs Quality", "content": "────────────────────────────────────────\nQscale I Frames P Frames B Frames\n1 43.2 46.3 46.5\n6 32.6 34.6 34.3\n11 28.6 29.5 30.0\n16 26.3 26.8 28.6\n21 24.7 25.0 27.9\n26 23.5 23.9 27.5\n31 22.6 23.0 27.3\n" }, { "name": "Qscale vs Compression", "content": "────────────────────────────────────────\nQscale I Frames P Frames B Frames\n1 2 2 2\n6 7 10 15\n11 11 18 43\n16 15 29 97\n21 19 41 173\n26 24 56 256\n31 28 73 330\n\n\n" }, { "name": "Search Techniques", "content": "There are several different motion vector search techniques available. There are different\ntechniques available for P frame search and B frame search. Using different search techniques\npresent little difference in quality, but a large difference in compression and speed.\n\n\nThere are 4 types of P frame search: Exhaustive, TwoLevel, SubSample, and Logarithmic.\n\n\nThere are 3 types of B frame search: Exhaustive, Cross2, and Simple.\n\nThe recommended search techniques are TwoLevel and Logarithmic for P frame search, and Cross2\nand Simple for B frame search. Here are some numbers comparing the different search methods:\n" }, { "name": "P frame Motion Vector Search (Normalized)", "content": "─────────────────────────────────────────────────────────────────────────────────\nTechnique Compression 1 Speed 2 Quality 3\n⟨#smallbetter⟩ ⟨#largefaster⟩ ⟨#largebetter⟩\nExhaustive 1000 1000 1000\nSubSample 1008 2456 1000\nTwoLevel 1009 3237 1000\nLogarithmic 1085 8229 998\n" }, { "name": "B frame Motion Vector Search (Normalized)", "content": "────────────────────────────────────────────────────────────────────────────────\nTechnique Compression 1 Speed 2 Quality 3\n⟨#smallbetter⟩ ⟨#largefaster⟩ ⟨#largebetter⟩\nExhaustive 1000 1000 1000\nCross2 975 1000 996\nSimple 938 1765 991\n\n1Smaller numbers are better compression.\n\n2Larger numbers mean faster execution.\n\n3Larger numbers mean better quality.\n\nFor some reason, Simple seems to give better compression, but it depends on the image se‐\nquence.\n\nSelect the search techniques with the PSEARCHALG and BSEARCHALG parameter file statements.\n\n\n" }, { "name": "Group Of Pictures (GOP)", "content": "A Group of Pictures (GOP) is a roughly independently decodable sequence of frames. An MPEG\nvideo stream is made of one or more GOP's. You may specify how many frames should be in each\nGOP with the GOPSIZE parameter file statement. A GOP always starts with an I frame.\n\nInstead of encoding an entire sequence, you can encode a single GOP. To do this, use the" }, { "name": "-gop", "content": "tompeg with the -combinegops command option.\n\n\n" }, { "name": "Slices", "content": "A slice is an independently decodable unit in a frame. It can be as small as one macroblock,\nor it can be as big as the entire frame. Barring transmission error, adding slices does not\nchange quality or speed; the only effect is slightly worse compression. More slices are used\nfor noisy transmission so that errors are more recoverable. Since usually errors are not such\na problem, we usually just use one slice per frame.\n\n\nControl the slice size with the SLICESPERFRAME parameter file statement.\n\nSome MPEG playback systems require that each slice consist of whole rows of macroblocks. If\nyou are encoding for this kind of player, if the height of the image is H pixels, then you\nshould set the SLICESPERFRAME to some number which divides H/16. For example, if the image\nis 240 pixels (15 macroblocks) high, then you should use only 15, 5, 3, or 1 slices per\nframe.\n\n\nNote: these MPEG playback systems are really wrong, since the MPEG standard says this doesn't\nhave to be so.\n\n\n\n" }, { "name": "Search Window", "content": "The search window is the window in which ppmtompeg searches for motion vectors. The window\nis a square. You can specify the size of the square, and whether to allow half-pixel motion\nvectors or not, with the RANGE and PIXEL parameter file statements.\n\n" }, { "name": "I Frames, P Frames, B Frames", "content": "In MPEG-1, a movie is represented as a sequence of MPEG frames, each of which is an I Frame,\na P Frame, or a B Frame. Each represents an actual frame of the movie (don't get confused by\nthe dual use of the word \"frame.\" A movie frame is a graphical image. An MPEG frame is a\nset of data that describes a movie frame).\n\nAn I frame (\"intra\" frame) describes a movie frame in isolation -- without respect to any\nother frame in the movie. A P frame (\"predictive\" frame) describes a movie frame by describ‐\ning how it differs from the movie frame described by the latest preceding I or P frame. A B\nframe (\"bidirectional\" frame) describes a movie frame by describing how it differs from the\nmovie frames described by the nearest I or P frame before and after it.\n\nNote that the first frame of a movie must be described by an I frame (because there is no\nprevious movie frame) and the last movie frame must be described by an I or P frame (because\nthere is no subsequent movie frame).\n\nBeyond that, you can choose which frames are represented by which types. You specify a pat‐\ntern, such as IBPBP and ppmtompeg simply repeats it over and over throughout the movie. The\npattern affects speed, quality, and stream size. Here is a chart which shows some of the\ntrade-offs:\n" }, { "name": "Comparison of I/P/B Frames (Normalized)", "content": "────────────────────────────────────\nFrame Type Size Speed Quality\nI frames 1000 1000 1000\nP frames 409 609 969\nB frames 72 260 919\n\n(this is with constant qscale)\n\n\nA standard sequence is IBBPBBPBBPBBPBB.\n\n\nSelect the sequence with the PATTERN parameter file statement.\n\nSince the last MPEG frame cannot be a B frame (see above), if the pattern you specify indi‐\ncates a B frame for the last movie frame of the movie, ppmtompeg makes it an I frame instead.\n\nBefore Netpbm 10.26 (January 2005), ppmtompeg instead drops the trailing B frames by default,\nand you need the FORCEENCODELASTFRAME parameter file statement to make it do this.\n\nThe MPEG frames don't appear in the MPEG-1 stream in the same order that the corresponding\nmovie frames appear in the movie -- the B frames come after the I and P frames on which they\nare based. For example, if the movie is 4 frames that you will represent with the pattern\nIBBP, the MPEG-1 stream will start with an I frame describing movie frame 0. The next frame\nin the MPEG-1 stream is a P frame describing movie frame 3. The last two frames in the\nMPEG-1 stream are B frames describing movie frames 1 and 2, respectively.\n\n\n" }, { "name": "Specifying Input and Output Files", "content": "Specify the input frame images with the INPUTDIR, INPUT, ENDINPUT, BASEFILEFORMAT, SIZE,\nYUVFORMAT and INPUTCONVERT parameter file statements.\n\nSpecify the output file with the OUTPUT parameter file statement.\n\n\n" }, { "name": "Statistics", "content": "ppmtompeg can generate a variety of statistics about the encoding. See the -stat, -snr," }, { "name": "-mv_histogram -quiet -no_frame_summary -bit_rate_info", "content": "" } ] }, "PARALLEL OPERATION": { "content": "You can run ppmtompeg on multiple machines at once, encoding the same MPEG stream. When you\ndo, the machines are used as shown in the following diagram. We call this \"parallel mode.\"\n", "subsections": [ { "name": "ppmtompeg-par.gif", "content": "To do parallel processing, put the statement\n\nPARALLEL\n\n\nin the parameter file, followed by a listing of the machines, one machine per line, then\n\nENDPARALLEL\n\n\nEach of the machine lines must be in one of two forms. If the machine has filesystem access\nto the input files, then the line is:\n\nmachine user executable\n\nThe executable is normally ppmtompeg (you may need to give the complete path if you've built\nfor different architectures). If the machine does not have filesystem access to the input\nfiles, the line is:\n\nREMOTE machine user executable parameter file\n\nThe -maxmachines command option limits the number of machines ppmtompeg will use. If you\nspecify more machines in the parameter file than -maxmachines allows, ppmtompeg uses only\nthe machines listed first. This is handy if you want to experiment with different amounts of\nparallelism.\n\nIn general, you should use full path file names when describing executables and parameter\nfiles. This includes the parameter file argument on the original invocation of ppmtompeg.\n\nAll file names must be the same on all systems (so if e.g. you're using an NFS filesystem,\nyou must make sure it is mounted at the same mountpoint on all systems).\n\nBecause not all of the processes involved in parallel operation have easy access to the input\nfiles, you must specify the SIZE parameter file statement when you do parallel operation.\n\nThe machine on which you originally invoke ppmtompeg is the master machine. It hosts a \"com‐\nbine server,\", a \"decode server,\" and a number of \"i/o servers,\" all as separate processes.\nThe other machines in the network (listed in the parameter file) are slave machines. Each\nhosts a single process that continuously requests work from the master and does it. The\nslave process does the computation to encode MPEG frames. It processes frames in batches\nidentified by the master.\n\nThe master uses a remote shell command to start a process on a slave machine. By default, it\nuses an rsh shell command to do this. But use the RSH parameter file statement to control\nthis. The shell command the master executes remotely is ppmtompeg, but with options to indi‐\ncate that it is to perform slave functions.\n\nThe various machines talk to each other over TCP connections. Each machine finds and binds\nto a free TCP port number and tells its partners the port number. These port numbers are at\nleast 2048.\n\nUse the PARALLELTESTFRAMES, PARALLELTIMECHUNKS, and PARALLELPERFECT parameter file\nstatements to control the way the master divides up work among the slaves.\n\nUse the -nice command option to cause all slave processes to run \"nicely,\" i.e. as low prior‐\nity processes. That way, this substantial and long-running CPU load will have minimal impact\non other, possibly interactive, users of the systems.\n\n" } ] }, "SPEED": { "content": "Here is a look at ppmtompeg speed, in single-node (not parallel) operation:\n", "subsections": [ { "name": "Compression Speed", "content": "───────────────────────────────────────\nMachine Type Macroblocks per second1\nHP 9000/755 280\nDEC 3000/400 247\nHP 9000/750 191\nSparc 10 104\nDEC 5000 68\n1A macroblock is a 16x16 pixel square\n\nThe measurements in the table are with inputs and outputs via a conventional locally attached\nfilesystem. If you are using a network filesystem over a single 10 MB/s Ethernet, that con‐\nstrains your speed more than your CPU speed. In that case, don't expect to get better than 4\nor 5 frames per second no matter how fast your CPUs are.\n\nNetwork speed is even more of a bottleneck when the slaves do not have filesystem access to\nthe input files -- i.e. you declare them REMOTE.\n\nWhere I/O is the bottleneck, size of the input frames can make a big difference. So YUV in‐\nput is better than PPM, and JPEG is better than both.\n\nWhen you're first trying to get parallel mode working, be sure to use the -debugmachines op‐\ntion so you can see what's going on. Also, -debugsockets can help you diagnose communica‐\ntion problems.\n\n\n" } ] }, "AUTHORS": { "content": "• Kevin Gong - University of California, Berkeley, keving@cs.berkeley.edu\n\n\n• Ketan Patel - University of California, Berkeley, kpatel@cs.berkeley.edu\n\n\n• Dan Wallach - University of California, Berkeley, dwallach@cs.berkeley.edu\n\n\n• Darryl Brown - University of California, Berkeley, darryl@cs.berkeley.edu\n\n\n• Eugene Hung - University of California, Berkeley, eyhung@cs.berkeley.edu\n\n\n• Steve Smoot - University of California, Berkeley, smoot@cs.berkeley.edu\n", "subsections": [] }, "DOCUMENT SOURCE": { "content": "This manual page was generated by the Netpbm tool 'makeman' from HTML source. The master\ndocumentation is at\n\nhttp://netpbm.sourceforge.net/doc/ppmtompeg.html\n\nnetpbm documentation 23 July 2006 Ppmtompeg User Manual(1)", "subsections": [] } }, "summary": "ppmtompeg - encode an MPEG-1 bitstream", "flags": [ { "flag": "", "long": null, "arg": null, "description": "This option causes ppmtompeg to append the statistics that it write to Standard Output to the file statfile as well. The statistics use the following abbreviations: bits per block (bpb), bits per frame (bpf), seconds per frame (spf), and bits per second (bps). These statistics include how many I, P, and B frames there were, and information about compression and quality." }, { "flag": "", "long": null, "arg": null, "description": "causes ppmtompeg not to report remaining time more often than every numseconds sec‐ onds (unless the time estimate rises, which will happen near the beginning of the run). A negative value tells ppmtompeg not to report at all. 0 is the default (re‐ ports once after each frame). Note that the time remaining is an estimate and does not take into account time to read in frames." }, { "flag": "", "long": null, "arg": null, "description": "causes ppmtompeg to run silently, with the only screen output being errors. Particu‐ larly useful when reading input from stdin. The equivalent of the -quiet common op‐ tion of most other Netpbm programs." }, { "flag": "", "long": null, "arg": null, "description": "This option prevents ppmtompeg from printing a summary line for each frame" }, { "flag": "", "long": null, "arg": null, "description": "forces ppmtompeg to use a more accurate, yet more computationally expensive version of the DCT." }, { "flag": "", "long": null, "arg": null, "description": "causes ppmtompeg to encode only the numbered GOP (first GOP is 0). The parameter file is the same as for normal usage. The output file will be the normal output file with the suffix .gop.gopnum. ppmtompeg does not output any sequence information." }, { "flag": "", "long": null, "arg": null, "description": "causes ppmtompeg simply to combine some GOP files into a single MPEG output stream. ppmtompeg inserts a sequence header and trailer. In this case, the parameter file needs only to contain the SIZE value, an output file, and perhaps a list of input GOP files (see below). If you don't supply a list of input GOP files is used, then ppmtompeg assumes you're using the same parameter file you used when you created the input (with the -gop op‐ tion) and calculates the corresponding gop filenames itself. If this is not the case, you can specify input GOP files in the same manner as normal input files -- except in‐ stead of using INPUTDIR, INPUT, and ENDINPUT, use GOPINPUTDIR, GOPINPUT, and GOPENDINPUT. If no input GOP files are specified, then the default is to use the output file name with suffix .gop.gopnum, with gopnum starting from 0, as the input files. Thus, unless you're mixing and matching GOP files from different sources, you can sim‐ ply use the same parameter file for creating the GOP files (-gop) and for later turn‐ ing them into an MPEG stream (-combinegops)." }, { "flag": "", "long": null, "arg": null, "description": "This option causes ppmtompeg to encode only the frames numbered firstframe to lastframe, inclusive. The parameter file is the same as for normal usage. The out‐ put will be placed in separate files, one per frame, with the file names being the normal output file name with the suffix .frame.framenum. No GOP header information is output. (Thus, the parameter file need not include the GOPSIZE value) Use ppmtompeg -combineframes to combine these frames later into an MPEG stream." }, { "flag": "", "long": null, "arg": null, "description": "This option causes ppmtompeg simply to combine some individual MPEG frames (such as you might have created with an earlier run of ppmtompeg -frames) into a single MPEG stream. Sequence and GOP headers are inserted appropriately. In this case, the para‐ meter file needs to contain only the SIZE value, the GOPSIZE value, an output file, and perhaps a list of frame files (see below). The parameter file may specify input frame files in the same manner as normal input files -- except instead of using INPUTDIR, INPUT, and ENDINPUT, use FRAMEINPUTDIR, FRAMEINPUT, and FRAMEENDINPUT. If no input frame files are specified, then the de‐ fault is to use the output file name with suffix .frame.framenum, with framenum starting from 0, as the input files." }, { "flag": "", "long": null, "arg": null, "description": "ity. (This is relevant only if you are running ppmtompeg in parallel mode. Other‐ wise, there are no remote processes). See 'man nice.'" }, { "flag": "", "long": null, "arg": null, "description": "This option causes ppmtompeg to use no more than nummachines machines as slaves for use in parallel encoding." }, { "flag": "", "long": null, "arg": null, "description": "tistics. Prints SNR (Y U V) and peak SNR (Y U V) for each frame. In summary, prints averages of luminance only (Y). SNR is defined as 10*log(variance of original/vari‐ ance of error). Peak SNR is defined as 20*log(255/RMSE). Note that ppmtompeg runs a little slower when you use this option." }, { "flag": "", "long": null, "arg": null, "description": "tomatically reports the quality of the images, so there is no need to specify -snr then." }, { "flag": "", "long": null, "arg": null, "description": "This option makes ppmtompeg write bit rate information into the file ratefile. Bit rate information is bits per frame, and also bits per I-frame-to-I-frame." }, { "flag": "", "long": null, "arg": null, "description": "This option causes ppmtompeg to print a histogram of the motion vectors as part of statistics. There are three histograms -- one for P frame, one for forward B frame, and one for backward B frame motion vectors. The output is in the form of a matrix, each entry corresponding to one motion vector in the search window. The center of the matrix represents (0,0) motion vectors." }, { "flag": "", "long": null, "arg": null, "description": "This option causes ppmtompeg to print to Standard Output messages that narrate the communication between the machines when you run ppmtompeg in \u001b]8;;#parallel\u001b\\parallel mode\u001b]8;;\u001b\\ ." }, { "flag": "", "long": null, "arg": null, "description": "This option causes ppmtompeg to print to Standard Output messages that narrate the progress of the conversion on the various machines when you run ppmtompeg in \u001b]8;;#parallel\u001b\\parallel mode\u001b]8;;\u001b\\ ." } ], "examples": [], "see_also": [], "tldr": { "source": "official", "description": "Encode an MPEG-1 stream.", "examples": [ { "description": "Produce an MPEG-1 stream using the parameter file to specify inputs and outputs", "command": "ppmtompeg {{path/to/parameter_file}}" }, { "description": "Encode the GOP with the specified number only", "command": "ppmtompeg {{-g|-gop}} {{gop_num}} {{path/to/parameter_file}}" }, { "description": "Specify the first and last frame to encode", "command": "ppmtompeg {{-fr|-frames}} {{first_frame}} {{last_frame}} {{path/to/parameter_file}}" }, { "description": "Combine multiple MPEG frames into a single MPEG-1 stream", "command": "ppmtompeg -combine_frames {{path/to/parameter_file}}" } ] } }