CPU/speed differences between Compressor settings - H.264 versus Mpeg-2

After setting up my 2.26 8-core Mac Pro for distributed processing for Compressor, I was able to get nice speeds using mpeg-2, but the h.264 codec seems to throttle back the speed of the compression.
For mpeg-2, Activity Monitor shows around 55% of the CPU being used for the process overall. For h.264, it shows around 15% CPU use, which is around the amount that would be used without distributed processing. I am using 8 instances in the qmaster setup.
Any ideas? I've tried some other settings and the mpeg ones (mpeg-1, mpeg-4) seem to take advantage of the speed while others (h.264, Uncompressed 8-bit) don't.

This sounds like its I/O bound. Open Activity Monitor, and check data read/sec and data written/sec in the Disk Activity tab. If the number is over like 100MB/sec, that could be whats holding you back, depending on the speed of your disk. If you are using a network volume, check the network tab. 1 instance seems to be good for 2 physical cores when doing H.264. 8 instances should max you out, unless your source is something like Uncompressed 10 bit and the reads are bottlenecking processing. In essence you have 8 instances trying to read data as fast as it can, which means no process reads it fast. So the instances are waiting around for data.

Similar Messages

  • AppleScript speed difference between 10.4 & 10.5

    This post is related to a question I posed in March 2009 (https://discussions.apple.com/message/9226100) regarding the speed difference between AppleScript text handling in OS X 10.5 compared with 10.4.11.
    I have recently written a script to search through a text file which is an xml file exported from WorkGroup Manager. The file, which is approximately 1.8 MB, contains all the user names on the school server (just over 1720 of them), their settings and any other information we have added. After adapting some script ideas (using text delimiters and offsets) which I found on the internet to assist with counting the number of users and searching through the information of each one, one at a time, I have a script which works far in excess of my expectations in terms of speed when run under OS X 10.4, but runs slower than molasses in winter under 10.5.
    The script counts the number of accounts, extracts each user's log-in name and also their actual name (which we have entered under a 'general info' or 'comments' field - I can't remember exactly where, but it is stored in the file), makes up a complete list of all account names (with the actual name, if present) and also makes up a subset of this information into a separate list which contains only the log-in names of those whose actual name has not yet been recorded on the server, then saves both lists to separate text files.
    My original attempts, using some rather clumsy text searching and comparison techniques, took about 1-2 minutes to go through the whole procedure. After some internet searching, head-scratching and a bit more more work, the finished script, when run on my G5 under OS X 10.4.11, takes less than 3 seconds to do all I have described above, including the writing to text files. To say I was pleased would be an understatement!
    Because the school/work environment, where the script would be run, is all OS X 10.5 or higher, I thought I should test it out under at least OS X 10.5. The script was developed on my DP 2.5GHz G5 with 8GB RAM running OS X 10.4.11. Using a fresh, clean installation of 10.5 I ran the script and was extremely disappointed. After letting it run for more than 20 minutes and still not finishing, I force quit AppleScript and reduced the number of accounts in the file. 40 accounts took 3 seconds, 60 accounts took 6 seconds, 80 took 11 seconds, 100 took 19 seconds and 120 took 27 seconds. I eventually let the script run right through - total time taken to process 1720 users was 65 minutes, compared to 3 seconds under 10.4.11!
    Given that it was tested on the same Mac, with the same amount of RAM, using the same script and the same original text file, the only variable left (that I can think of) is the change of OS version - 10.4.11 vs 10.5.
    As was pointed out to me in my previous post, AppleScript in OS X 10.5 handles all text as Unicode and has a greater overhead in processing time as a result. I have implemented the various bits of advice offered by respondents to my original post and changed the way I handled lists, etc. Obviously, if the script works so speedily under 10.4.11, I must have done something right in terms of code optimization/efficiency. It's when that exact same script is run under OS X 10.5 that it slows down incredibly. I even tried it on a 2.66GHz Intel Core 2 Duo iMac running 10.5.8 - it still took 40 minutes and Activity Monitor showed CPU usage by Script Editor constantly above 80% (often well above 90%). Saving the script as an application made no difference in the time taken.
    Does anybody have any knowledge about what makes AppleScript Unicode-only text handling so slow? In this case, it is 1300 times slower! Is there any way of coercing/restricting the text handling/parsing to ASCII? If I really have to, I will set up a humble eMac at school with 10.4 on it just to handle large text files quickly with AppleScript, but I would prefer to be able to do it on the normal work Macs which have 10.5 or later on them. As mentioned, under 10.4.11 the script processes the 1.8MB text file with 1720 users and writes results to two files - all within 3 seconds, so I'm not really looking for coding suggestions unless they are directly related to what has changed in AppleScript under OS X 10.5. Without sounding too smug, the script works properly and speedily (at least in 10.4). I really would like to learn about the changes in AppleScript in OS X 10.5 and how to cope with or work around those changes.

    Hi,
    text 3 thru 14 in largeText
    character 3 thru 14 in largeText
    text 3  in largeText
    character 3  in largeText
    These lines will be slow on Leopard
    Getting some text in a variable that contains more than 60000 characters will be slower on Leopard,
    but I don't know why it's slower
    Here a test script.
    script o
        property t_text : ""
    end script
    set a to "abcdefghij"
    set tResult to ""
    repeat with i from 1 to 3
        set o's t_text to a
        repeat (item i of {17, 15, 14}) times
            set o's t_text to o's t_text & o's t_text -- add  characters in the variable
        end repeat
        set StartTime to current date
        -- test
        repeat 20 times
            set b to text 3 thru 14 in o's t_text -- get a text in the variable
        end repeat
        -- end test
        set EndTime to current date
        set TimeTaken to EndTime - StartTime
        set tResult to tResult & " Getting text in the variable which contains " & length of o's t_text & ", (20 times) = " & TimeTaken & " seconds." & return
    end repeat
    tResult
    Here the result on my old G5, 2 x 1.8 GHZ
    Getting text in the variable which contains 1310720 characters, (20 times)  = 50 seconds.
    Getting text in the variable which contains   327680 characters, (20 times)  = 12 seconds.
    Getting text in the variable which contains   163840 characters, (20 times)  =   6 seconds.
    The result on the same machine on Tiger is always less of one second.
    Also, I try with 20 millions characters on Tiger, the result : getting text in the variable which contains 20971520 characters, (20 times)  = 0 seconds.
    The solution ( text item in a list)
    Here the script
    script o
        property my_List : {}
    end script
    set OldDelims to text item delimiters
    set RecordDelimiter to "::::::::::::::"
    set LengthOfRecordDelimiter to length of RecordDelimiter
    set o's my_List to findAll(read (choose file), RecordDelimiter)
    on findAll(str, findString)
        set Oldtid to text item delimiters
        try
            set text item delimiters to findString
            if str does not contain findString then return {"Nothing found"}
            set t to str's text items
            set text item delimiters to Oldtid
            return t
        on error eMsg number eNum
            set text item delimiters to Oldtid
            error "Can't findAll: " & eMsg number eNum
        end try
    end findAll
    set NumberOfrecords to (count o's my_List)
    display dialog "There are " & NumberOfrecords & " accounts."
    set StartTime to current date
    set text item delimiters to ":"
    set FullUserList to {}
    set ListOfUnnamedUsers to {}
    -- first user needs to be done separately as it is not preceded by RecordDelimiter
    set EndOfHeader to "Standard:URL"
    set LengthOfEndOfHeader to length of EndOfHeader
    set EndOfHeaderOffset to the offset of EndOfHeader in (item 1 of o's my_List)
    set OffsetToApply to EndOfHeaderOffset
    set TextBeingChecked to text (OffsetToApply + LengthOfEndOfHeader + 1) thru -1 of (item 1 of o's my_List)
    tell TextBeingChecked to set {UserName, NameForInfo} to {text item 1, text item 7}
    if NameForInfo = "" then set end of ListOfUnnamedUsers to UserName & return
    set end of FullUserList to (UserName & tab & NameForInfo & return)
    -- now do all the others
    repeat with CounterG from 2 to (NumberOfrecords - 1)
        set TextBeingChecked to item CounterG of o's my_List
        tell TextBeingChecked to set {UserName, NameForInfo} to {text item 1, text item 7}
        if NameForInfo = "" then set end of ListOfUnnamedUsers to UserName & return
        set end of FullUserList to (UserName & tab & NameForInfo & return)
    end repeat
    set o's my_List to {}
    set text item delimiters to OldDelims
    -- write results to file
    -- 1). full user list
    set TargetFile1 to (path to desktop folder as string) & "FullUserList1.txt"
    try
        open for access file TargetFile1 with write permission
    on error
        close access file TargetFile1
        open for access file TargetFile1 with write permission
    end try
    set EndTime to current date
    beep
    set InfoToBeWrittenToFile to FullUserList as text
    write InfoToBeWrittenToFile to file TargetFile1
    close access file TargetFile1
    -- 2). list of users without NameForInfo
    set TargetFile2 to (path to desktop folder as string) & "UnnamedUsersList1.txt"
    try
        open for access file TargetFile2 with write permission
    on error
        close access file TargetFile2
        open for access file TargetFile2 with write permission
    end try
    set InfoToBeWrittenToFile to ListOfUnnamedUsers as text
    write InfoToBeWrittenToFile to file TargetFile2
    close access file TargetFile2
    beep 3
    set TimeTaken to EndTime - StartTime
    set TimePerAccount to TimeTaken / NumberOfrecords
    display dialog (NumberOfrecords & " accounts took " & TimeTaken & " seconds." & return & return & "That equals " & TimePerAccount & " seconds per account.") as string

  • Speed difference between comressor 1 & 2

    I've ben working on a contract that requires me to use Compressor (2). I usally set up a batch of 3 timelines (with DVD SP markers) to for a total of 1 hour of content. Using Compressor's setting for 90 min. fast encode on my DP 1 GIG desktop this takes approx 7 hours and on my G4 1.25 laptop it takes 10 hours. A friend did a simillar compression on a single QS 867 mhz with 1.5 GIG of ram and this took 5 hours. Is it possible there is so much difference between Compressor 1 vs. 2?
    Jeff
    G4 DP 1 GIG G4 laptop 1.25   Mac OS X (10.4.6)  

    Hi Hanumang/ John
    I'm haveing a hard time understanding this myself. The settting I used in Compressor 2 was the fast encode 90 min which is 6.2mbps 1 pass & Dolby 2.0. The compreson I told my friend to use (using Compressor 1) was also a preset This was the 60 min. fast encode @ 6.0mbps. Although his was doing video only, then running the audio through Apack which would only take 20 mins. at most. These are very simple timelines with no effects and jus a few DVD SP markers. If it is a speeed difference of at least 30% my purchase of a MBP 1.83 (for added speed, may not be fully realized when using Compresor.
    Jeff
    G4 DP 1 GIG G4 laptop 1.25, MBP 18   Mac OS X (10.4.6)  

  • Any speed difference between Apple Wifi Router(air express) and Tp link(or others) wifi router)

    Any speed difference between Apple Wifi Router(air express) and Tp link(or others) wifi router?
    I'm using a tp link router and sometimes it takes a long time delay for my 2 mac computers to connect each other. I don't know this is because of the router compability or the computers.

    Anyone knows?

  • What's the difference between Compressor & Roxio's Popcorn?

    What's the difference between apps like Compressor & Roxio's Popcorn?
    Can they both be used to perform similar tasks?

    Can anyone recommend the very best quality settings within compressor 1.2 so i can squeeze a 5.5 GB .vob (non commericial without css) to a standard 4.7GB 4x DVD-R apple media?
    It's not the size of the original VOB file that counts, it's the runtime of the video itself (one hour, two hours?). In any case, you'll probably want to use Dolby/AC-3 audio instead of uncompressed PCM/AIFF. After you de-multiplex the VOB into separate audio and video and check the runtime you'll have a better idea of what Compressor preset to use.

  • Speed difference between i7 quad 3rd gen 2.6ghz and i7 quad 4th gen 3.1ghz?

    How much speed difference in Logic X will I notice between an i7 quad 3rd gen 2.6ghz and i7 quad 4th gen 3.1ghz?
    both loaded with 16gb 1600 ram... 256gb ssd

    Kappy wrote:
    All other things being equal 4-cores are twice as fast as 2-cores. This is regardless of what you are doing. The only way a 2-core processor would be as fast is if it were run at twice the clock speed. But the 2-core CPU is only clocked around 10% or so faster. Now, for all that you claim you will be doing the relevant question is do you need 4-cores. My answer is that you don't. But the 2-core machine will be slower - all things being equal.
    Iris is a far better GPU than HD4 or 5000. But not as fast as a discreet GPU with better performance measures. Again, you don't really need high-power GPU. Iris should meet your needs adequately for now.
    But remember you want to keep the computer for 6 or 7 years. No one knows what your needs will be then. You may find whatever you buy today that meets today's needs will be inadequate for tomorrow's.
    Your comments on graphics pretty much confirm the majority of what I've been reading on other sites.
    The conflict is with cores. About half the people are saying for low CPU intensive tasks 4 cores are totally unnecessary, while the other half says they'll make things faster no what the task (one thing all seem to have in common is they're totally against the 2014s).
    So it seems I'm left with the decision between faster performance vs better graphics.
    I appreciate your input.

  • Difference between jpeg settings

    hi,
    i want to know the purpose of jpeg quality setting in publish
    settings box. what's the purpose of that setting? and whats the
    difference between this setting and bitmap properties settings in
    library? the livedocs on macromedia states : NOTE
    JPEG Quality settings that you select in the Publish
    Settings dialog box do not specify a quality setting for imported
    JPEG files. You must specify a quality setting for imported JPEG
    files in the Bitmap Properties dialog box.
    i am a little confused because of this.
    any help would be grately appreciated.
    thanks
    gaurav

    There are many options and yes it is confusing.
    Assuming Bitmap properties calls for JPEG/Lossy Compression:
    1. Publish window - whatever you put in here is the default.
    That number is used if nothing else overrides it.
    2. Bitmap Properties setting - if you put a number in there,
    it will override the default.
    3. Use imported data: You imported a JPG, Flash won't mess
    with it at all, it wil just export the same data.
    Note that it is a bad idea to import a JPG and then
    recompress it a different way for export. You get double artifacts
    that way.

  • Real world performance/speed difference between 3ghz and 3.2

    Hello,
    I have a 2008, (not clovertown) Harpertown 3.0ghz Mac Pro. I wanted to know if its feasible to upgrade the cpus to the 3.2 ones? Also, what is the REAL WORLD speed/performance difference between the 3.0 and 3.2? I am so stressed out over this that I really need to have an answer to this.
    Not that I am going to buy the 3.2 processors, just wanting to see what I am missing here in terms of percentage overall between my mac pro and a 3.2ghz mac pro from 2008.
    Thank you,

    You realize that by now you can probably guess what some of our answers might be?
    Real world... well, in the real world you drive to work stuck in traffic most of the time, too.
    Spend your money on a couple new solid state drives.
    You want this for intellectual curiosity, so look at your Geekbench versus others.

  • Difference between capture settings...?

    OK, there are two settings I would consider using for Final Cut Express to grab video from my Sony DCR-TRV140:
    -DV-NTSC 32 kHz
    -DV-NTSC 32 kHz Firewire Basic
    I understand that one uses the Firewire basic protocol and the other does not (although I don't know what the other uses if it's not using Firewire basic)...
    So what's the difference, and is there any reason for me to choose one over the other if I can't (initially, at least) tell any difference between the two when it comes to capturing video?
    Many thanks in advance,
    Tim

    Also, you only need to use one of the 32KHz setups +if you recorded your tapes in 12-bit audio.+
    Check a few of your tapes to see what the audio is ... you should be able to see this somewhere on the LCD at the beginning of playback or in the menus when you have a recorded tape in your camera.
    Your camera (the DCR-TRV140) supports both 12-bit *AND 16-bit audio*.
    You should set your camera to record in 16-bit audio from now on. It presents fewer potential problems for editing your videos in FCE.
    The following Easy Setups are available for 16-bit audio:
    DV-NTSC and
    *DV-NTSC FireWire Basic*

  • Is there much speed difference between the iMacs?

    Just wondering, if there is any noticeable performance difference between the 1.83 - the 2.0 and the 2.16?
    I'd be surprised to learn that there was (and also, keenly interested to learn).
    My experience is that a dual processor is noticeably more capable than a single processor (regardless of minor increments at the end of the processor number).
    Anyone?

    You are right that it is more about futher differentiating the models than anything else. The speed difference is negligible, but a nice plus is you choose a larger display. Typically you also get a larger standard HD and maybe a better video card and if you add all those things up, the price difference is worth it.

  • Any data on speed differences between 2.4 and 2.93 ghz MBP?

    Has anyone seen any performance tests showing the difference between 2.4 and 2.93 ghz MBP? I've googled around and haven't found anything.

    Check Bare Feats for benchmarks and do some Google searching. Logically, the difference is not likely more than the difference in processor speed, plus or minus.

  • Speed Differences Between 2.66ghz i7 15" and 17" models

    The specs on the top of the line 15&17" 2.66ghz i7 Macbook Pros appear to be identical.
    Aside from cosmetic differences, (like express card vs SD, or the presence of extra ports), can someone please tell me if there are performance differences between these two models?
    Would the 15" be as fast as the 17", for instance, in rendering a quicktime movie to h.264 mpeg-4?
    On a second not, I know they are also apples and oranges, but how fast would these new laptops be compared to my 2008 2.8 ghz dual quad core Mac Pro? Is my Mac Pro still faster than these laptops? I know I would get blazed by the new i7 iMac.

    The specs are identical, unless you consider it useful to have an ExpressCard slot to connect eSATA drives and various other devices, or to have 30% more screen real estate than the hi-res 15" model (or 78% more than the lo-res version). Performance-wise, the machines should be almost exactly alike.
    Your Mac Pro will still outperform either.
    Message was edited by: eww

  • Speed difference between hotspot and cellular

    Looking at an iPad mini, I have a hot spot on my phone, so I'm wondering how much of a speed difference there might be using a Mini with an iPhone 5 hotspot with AT&T and AT&T cellular on the Mini.

    Download an app such as Speedtest to your iPad.  Test both ways and see the results for yourself.

  • Differences between import settings

    Can someone explain the different import settings in iTunes? All I am finding is very general information. I want excellent sound, but I am concerned about using too much disk space. However, if I can hear the difference, I will use more space.
    Am I correct that once a song is imported with a given setting, the only way to change it would be to reimport the song?
    Thanks,
    Ben

    *I want excellent sound, but I am concerned about using too much disk space. However, if I can hear the difference, I will use more space.*
    Of the various options iTunes gives, AAC and MP3 are compressed formats so you lose some quality but gain space. AIFF (CD) and WAV are uncompressed so you gain quality at the expense of space. Apple Lossless falls somewhere in between, doesn't compress as much as AAC and MP3 so you gain some space but not as much, however it is claimed to give a quality which is the equal to CD. Of the compressed formats AAC is less widely used at present, if you have a non Apple music player or a newer in-car CD player then MP3 would be the most compatible compressed format and AIFF the most compatible uncompressed one: iTunes Encoding Strategies
    *Am I correct that once a song is imported with a given setting, the only way to change it would be to reimport the song?*
    No, songs can be converted in iTunes but it' is preferable where possible to re-rip the songs if they were previously imported and compressed from an original uncompressed form such as CD. In fact converting from one lossy format to another (MP3 to AAC for example) can result in even more quality loss. It's a bit like photocopying a photocopy, the farther you get away from the original the lower the quality.
    Have a look at this article, you'll get an extensive comparison of the AAC and MP3 formats and bit rates benchmarked against CD format: Planet of Sound - Quality of AAC audio and MP3

  • Combined PDF page display speed difference between Acrobat 8 Pro and Acrobat X Pro

    I have encountered a display speed issue with regards to combining Tiffs with Acrobat X Pro.  The Tiffs are between 200 KB and 600 KB.  Tested the issue by using a series of Tiffs originally combined using Acrobat 8 Pro but used Acrobat X Pro to combine them.  When I open the combined PDF files using X, I can cycle through the pages of the one created using 8 significatinly faster (i.e. about as fast as I can click) but the file created using X is significately slower.  Note: PDF file size is smaller using X (30.9 MB) than same combined files using 8 (46.1 MB).  Any ideas on what might be causing this issue?

    Found a Tips & Tricks document for Acrobat 8 which indicated the default Monochrome Compression setting for Acrobat 8 is CCITT G4.  Recreated the combined PDF file from the original Tiffs using X Pro with the modified setting and works like a charm.  Thanks for the point in the right direction.

Maybe you are looking for