iTunes 2 is wayyy too processor hungry

124»

Comments

  • Reply 61 of 76
    sinewavesinewave Posts: 1,074member
    [quote]Originally posted by bradd:

    <strong>So I read this whole board and I have one conclusion to make about OS X and mp3s at this point: the software isn't optimized/mature enough at this point. I realize that iTunes is based on SounjamMP but it still isn't where Winamp is on the PC. I have both a Titanium 667 with 512MB RAM and a PC. The PC is a homebuilt 750MHz Athlon with 384MB of RAM. Under Windows 2000 service pack 2, Winamp version 2.77 takes up about 1-2% of the CPU. My total overall usage under Windows 2000 (idling with AIM, WinAmp and a few services open) is less than 5%. I just think that with some more time and development, MacOS X 10.x and it's mp3 abilities will improve. With enough people complaining about the iTunes performance, it will have to change. A third of the CPU time taken up by mp3 decoding is simply unexcusable, even with the eye candy effects enabled. Winamp does scrolling titles and has an equilizer. I don't use the visuals because I don't have screen space, but even so, its a far cry from the 20% iTunes uses on average on my machine.



    I can post screenshots from my Windows 2000 machine with Winamp running to illustrate my point if anyone so desires.



    --bradd</strong><hr></blockquote>



    1 - 2% or your CPU? I am wondering if it's because of your processor speed. My roomates 350mhz PII took just as much processor time running WinAMP only as my 350 G3 took up using iTunes. Playing the same song. That was between 4 and 25%



    However the WinAMP averaged out at 10% and iTunes averaged out at 11%
  • Reply 62 of 76
    [quote]Originally posted by Sinewave:

    <strong>



    1 - 2% or your CPU? I am wondering if it's because of your processor speed. My roomates 350mhz PII took just as much processor time running WinAMP only as my 350 G3 took up using iTunes. Playing the same song. That was between 4 and 25%



    However the WinAMP averaged out at 10% and iTunes averaged out at 11%</strong><hr></blockquote>



    The Athlon is a pretty unbelievable CPU, and WinAMP likes it. It's what the G4 should've been...comparing a 667MHz G4 to a 750MHz Athlon will always result in the G4's being blown away unless SIMD comes into heavy play (i.e. Photoshop). Even so, I'd bet that the Athlon XP series would stomp the G4 simply because of the Athlon XP's raw processing power in addition toits support of MMX, SSE, and SSE2. While the Athlon was plagued by some dodgy chipsets, the new nForce and KT266 chipsets are fantastic.



    For a comparison between the Athlon 750 and the G4 867 in Quake 3, which doesn't benefit a whole lot from SIMD, look at these two articles.



    <a href="http://www.anandtech.com/cpu/showdoc.html?i=1092&p=7"; target="_blank">http://www.anandtech.com/cpu/showdoc.html?i=1092&p=7</A>;



    <a href="http://www.xlr8yourmac.com/systems/G4_867mhz_performance.html"; target="_blank">http://www.xlr8yourmac.com/systems/G4_867mhz_performance.html</A>;



    While the tests are conducted in different demos, it's pretty clear that the G4 is crushed by the Athlon. The numbers are even worse if you look at the Q3A tests in this article, which uses the same demo as the AnandTech article:



    <a href="http://www.xlr8yourmac.com/Graphics/geforce3/G4_733_GeForce3_tests.html"; target="_blank">http://www.xlr8yourmac.com/Graphics/geforce3/G4_733_GeForce3_tests.html</A>;



    Note that the only test that matters for CPU comparison's sake is the 640x480x32 test, since that test puts little load on the GF3 or any other modern video card. Basically, it's a pretty pure test of CPU power, and floating point performance in particular. It just so happens that FP performance is important when it comes to multimedia apps such as playing MP3s.



    It's why I keep whining for Apple to ditch IBM and Moto and work something out with AMD. It'd help both Apple and AMD, and it would result in some simply pimped-out Macs that could double as room heaters in a pinch.



    Upon further reflection, I could've started a whole new thread with this post. Oh well.
  • Reply 63 of 76
    sinewavesinewave Posts: 1,074member
    Games are better on PCs for the same reason Photoshop is better on Macs. It's called ports. When you port a game to a different platform that game wont be as optimized. Esp when talking about Frame rates. Which even though you run at a low resolution.. the video card STILL comes into play.



    I just hope if Apple DOES go to AMD that the chips aren't as temperamental as a lot of AMDs. If your fan ever goes bad you don't have much time before that sucker is gone. I have had a few friends that fried one just replacing it.. and he is a certified tech. The quality from chip to chip isn't up to par IMHO.
  • Reply 64 of 76
    maniamania Posts: 104member
    ok boyz answer me this



    i have a dual 800 with 640 megs of ram, os 10.1.1, latest i-tunes, mint audio, mpg123 - whatever - and when using the net via DSL (basically always), my mp3s skip (not much but enough to make me mad ).



    when i turned off one processor - no skipping no matter what i do or what player. turn it back on and skipping in every mp3 app just browsing the web.



  • Reply 65 of 76
    cosmocosmo Posts: 662member
    My guess is that os X is splitting the mp3 decoding between the processors, and there is a bottlneck somewhere that isnt' allowing htem to communicate fast enough (or somethign to that effect).

    Do mp3s skip under OS 9? Most likely not, because os 9 doesn't use both processors in way that X does.



    I think there is a way to make an app run on a secific processor (via the terminal)

    anyone care to elaborate?
  • Reply 66 of 76
    xmogerxmoger Posts: 242member
    In defense of winamp, I didn't think I had seen these sort of processor usages before, so I clocked my duron down to 500Mhz(as low as I could) and tested it on win2k with 256MB. Playing a 128kbit song averaged less &lt; 1% in winamp 2.76. Windows media player 7 averaged 11% for the same song.



    The funny thing is that if WMP had played a song before or was currently playing, winamp's usage shot up to around %4.

    [quote]I just hope if Apple DOES go to AMD that the chips aren't as temperamental as a lot of AMDs. If your fan ever goes bad you don't have much time before that sucker is gone. I have had a few friends that fried one just replacing it.. and he is a certified tech. The quality from chip to chip isn't up to par IMHO.<hr></blockquote>

    The current motherboards don't support the athlon's thermal diode, however AMD did demo a regular board with a little circuit soldered on that can shut it down. In the future they might integrate their powernow features with the thermal protection, which would be cool. I've repaired hundreds systems at work and have mostly found quality problems with (older) athlon chipsets and not the cpu's.
  • Reply 67 of 76
    applenutapplenut Posts: 5,768member
    [quote]Well considering me and another person did the SAME tests and we BOTH came out with different data than you.. why should we suddenly believe you when we know better?<hr></blockquote>



    considering several others have tested WinAMP and the finder and gotten different results why should I believe you?



    [quote]Yes but I never use column mode. BTW Finder takes up just about as much CPU time. Go figure.<hr></blockquote>



    stop lying.



    [quote]Got water in your ears? This isn't even subjective.<hr></blockquote>



    quicktime has a better codec. better low frequency. iTunes has been criticized for its poor sound. the sound enhancer and equalizer help it a bit.



    [quote]Hmm, on my system (a Pismo 400 with OS X 10.1.1 and 256MB of RAM), QT uses up a bit more CPU time than iTunes, with the EQ and Sound Enhancer in iTunes off. The MP3 being played is Hungarian Rhapsody #2, encoded at 44.1@192 using Fraunhofer. Sound quality in QT is somewhat muddy; the percussive bass and phase variances in the song aren't as sharp as they are in iTunes. Also, QT skips when I start Word up while the song is playing. I'm listening on AIWA HP-X311 headphones. Not the best, but good enough for me to hear differences between decoders.<hr></blockquote>



    that doesn't make sense. perhaps QT has a problem with certain encoders.

    :confused:



    [quote]Incedentally, the Finder consumes the same amount of processor time as QT does when playing the same MP3. Not surprising, since it uses QT.<hr></blockquote>



    yea and its less than iTunes. significantly less



    [quote]And here I thought that you were holding WinAMP up as the paragon of all goodness. By the by, Sinewave pointed out that on a P2 400, WinAMP was taking 5%-55% of CPU time. That's not terribly effecient for a system with a processor that is roughly equivalent to a G3-300. Still, WinAMP is a good player, in my experience, but it is no more effecient than iTunes.<hr></blockquote>



    sinewave great.. other people in this thread have claimed the opposite of him.



    [quote]No, that's what you want the point of the thread to be. You simply want everyone to fall in line with you and piss on Apple for something that they have little control over. Howe<hr></blockquote>



    uh... no that was what my original post said.



    [quote]ou simply want everyone to fall in line with you and piss on Apple for something that they have little control over. H<hr></blockquote>



    they have little control over the efficiency of their software?







    [quote]However, people are smarter than that, and they look for the reasons behind the percieved problems.<hr></blockquote>



    yes, you should follow your own advice rather than making up reasons like MP3s are hard to decode and macs don't have the memory bandwidth to do it more efficient.



    [quote]Some things can't be helped, such as the effeciency of MP3 as a format, but other problems, such as the scrolling text bug, have been revealed, and hopefully Apple will fix those. <hr></blockquote>



    the "scrolling text bug" apparantly doesn't affect iTunes so what does that have to do with anything.



    [quote]Bah. It seems that you lack the basic facilities to read, comprehend, and then post. That, or you're a troll. Either way, I'm done with you. <hr></blockquote>



    it seems you lack the basic facilities to understand a basic arguement and accept the fact that something Apple made may not be perfect.



    I'm a troll? yea





    [quote]yeah Applenut I am making up the figures WTF would I care what CPU time WinAMP uses?<hr></blockquote>



    I don't know. seems to amuse you just trying to "win" an arguement with me. I didn't think you would stoop low enough to make up numbers but based on the results of others I guess you would.



    [quote]One other thing that I noticed is that iTunes' CPU usage scores for G4s are pretty much equivalent to those of G3s running at the same clock speed, and DP G4s score the same as SP. Is it possible that the Altivec and SMP enhancements in iTunes were removed or broken in 2.0? <hr></blockquote>



    the only altivec/MP optimizations in iTunes are in the encoder and visualizers I believe.



    what I am more surprised by is the consistant percentage of CPU time iTunes takes on different clocked processors. why is that?



    [quote]1 - 2% or your CPU? I am wondering if it's because of your processor speed. My roomates 350mhz PII took just as much processor time running WinAMP only as my 350 G3 took up using iTunes. Playing the same song. That was between 4 and 25%<hr></blockquote>



    Apple sold that 350 Mhz G3 has equivelant to a 450-500 Mhz P2 so I would hope you aren't using that as a basis that iTunes is inline with WinAmp's performance.



    [quote]Games are better on PCs for the same reason Photoshop is better on Macs. It's called ports. When you port a game to a different platform that game wont be as optimized. Esp when talking about Frame rates. Which even though you run at a low resolution.. the video card STILL comes into play. <hr></blockquote>



    it's not because of the ports... it's because the hardware. Quake 3 is possibly the most optimized mac game available and it can't come close to Quake 3 on the PC. It's not because its a port its because the hardware doesn't match up unfortunately.



    PS is the same.



    but ports do of course play a role sometimes... cough SimCity 3000..





    [quote]ok boyz answer me this



    i have a dual 800 with 640 megs of ram, os 10.1.1, latest i-tunes, mint audio, mpg123 - whatever - and when using the net via DSL (basically always), my mp3s skip (not much but enough to make me mad ).



    when i turned off one processor - no skipping no matter what i do or what player. turn it back on and skipping in every mp3 app just browsing the web.<hr></blockquote>



    I think that's a bug. I think Mike from xlr8yourmac has been complaining about the same thing.
  • Reply 68 of 76
    sinewavesinewave Posts: 1,074member
    [quote]Originally posted by applenut:

    <strong>considering several others have tested WinAMP and the finder and gotten different results why should I believe you?

    </strong><hr></blockquote>

    Just like other people have gotten different iTunes results too and you don't believe them? shocking :eek:

    [quote]<strong>

    stop lying.

    <hr></blockquote></strong>

    Ok lets follow your "others have tested" reasoning. Others have said the Finder takes just as much or more as iTunes. I guess your wrong.

    [quote]<strong>

    quicktime has a better codec. better low frequency. iTunes has been criticized for its poor sound. the sound enhancer and equalizer help it a bit.

    <hr></blockquote></strong>

    It has? News to me!

    <strong> [quote]

    Apple sold that 350 Mhz G3 has equivelant to a 450-500 Mhz P2 so I would hope you aren't using that as a basis that iTunes is inline with WinAmp's performance.

    </strong><hr></blockquote>

    "Apple sold that 350 Mhz G3 has equivelant to a 450-500 Mhz P2"

    really can you show me where Apple stated this? Actually Apple claimed it to be TWICE as fast... and if you believe that I got a nice bridge to sell ya.

    <strong> [quote]

    it's not because of the ports... it's because the hardware. Quake 3 is possibly the most optimized mac game available and it can't come close to Quake 3 on the PC. It's not because its a port its because the hardware doesn't match up unfortunately.

    </strong><hr></blockquote>

    Was Quake 3 wrote from the ground up to run on a Mac?



    Seems to me and most of the people in this forum think your being a little troll. Your just making false accusations and calling people liars because they don't agree with you.



    If anything this proves that not every one gets the same amount of CPU scores cause every system is different. And when I mean every system is different I am not just talking about hardware. ram and the like.. but also what services your running.
  • Reply 69 of 76
    Quote:

    Originally posted by applenut:

    <strong>

    quicktime has a better codec. better low frequency. iTunes has been criticized for its poor sound. the sound enhancer and equalizer help it a bit.</strong><hr></blockquote>



    Criticised by whom? iTunes uses Fraunhofer, which is second-only to LAME or not-LAME in quality. However, codecs matter most during encoding insofar as sound quality goes, and QT doesn't do MP3 encoding. As far as effeciency of decoding goes, some codecs are better than others, and sound quality varies slightly.



    Sound quality can also be affected by the sound processing that a player does, but QT, so far as I know, simply decodes and plays with some primitive bass and treble controls.



    If you want some good analysis, go to <a href="http://r3mix.net"; target="_blank">http://r3mix.net</A>;



    <strong>



    that doesn't make sense. perhaps QT has a problem with certain encoders.</strong>[/quote]



    The sound differences were pretty slight, not really noticable without careful listening. It's possible that QT has problems with higher bitrates, as I didn't hear any differences between QT and iT with 128Kb/s MP3s.



    <strong> [quote]sinewave great.. other people in this thread have claimed the opposite of him.</strong><hr></blockquote>



    On signifigantly more powerful systems, yes. An Athlon in one case, and a Duron (e.g. an Athlon with less L2 cache, but still quite powerful) in another. Sinewave was talking about a P2 450, a pretty weak processor.



    <strong> [quote]

    they have little control over the efficiency of their software?</strong><hr></blockquote>



    They have little control over the effeciency of MP3 decoding in general. Read -&gt; comprehend -&gt; post.



    <strong> [quote]

    yes, you should follow your own advice rather than making up reasons like MP3s are hard to decode and macs don't have the memory bandwidth to do it more efficient.</strong><hr></blockquote>



    MP3's aren't hard to decode, as, after all, it's just an algorithm. However, it does take up a pretty good amount of CPU power to do so, just as it takes up a certain amount of CPU power to process any kind of data. MP3 happens to not be terribly effecient, for reasons related to the basic format itself. More powerful processors can do the work faster. Very simple computing stuff here, applenut. If you want to prove me wrong, then do so, but don't sit around and just say I'm wrong.



    By the by, I said that iMacs have crap memory bandwidth, not Macs in general. PowerMacs are very good because of the 64-bit I/O.



    <strong> [quote]

    the "scrolling text bug" apparantly doesn't affect iTunes so what does that have to do with anything.</strong><hr></blockquote>



    It's been demonstrated in this very thread that the bug does affect iTunes. Read -&gt; comprehend -&gt; post.



    <strong> [quote]

    it seems you lack the basic facilities to understand a basic arguement and accept the fact that something Apple made may not be perfect.</strong><hr></blockquote>



    Funny, since I've grumbled about the iMac's inability to push data fast, the scrolling text bug, the G4 being a weak processor, and other things.



    As for you being a troll, you may be. But you may just be someone who can't base opinions on solid ground and back them up with fact, due to laziness or lack of mental prowess. Either way, I don't really care.
  • Reply 70 of 76
    Playing a mp3 is like unstuffit a sit image.. while playing music at the same time. It takes up a lot of CPU to do that.







    Uhhh... I used to run MP3's, a couple dozen programs, servers, etc. etc. on my 6400/200 without any problems.



    Even with Audion on my Pismo it never takes up THAT much CPU (well, actually in X it does... ).
  • Reply 71 of 76
    sinewavesinewave Posts: 1,074member
    [quote]Originally posted by The Toolboi:

    Uhhh... I used to run MP3's, a couple dozen programs, servers, etc. etc. on my 6400/200 without any problems.[/QB]<hr></blockquote>



    As do I on my 350mhz G3 running OS X

    <strong> [quote]

    Even with Audion on my Pismo it never takes up THAT much CPU (well, actually in X it does... ).</strong><hr></blockquote>



    Actually unstuffing or stuffing a .sit file can take up more CPU than playing a mp3 in iTunes.



    [ 11-28-2001: Message edited by: Sinewave ]</p>
  • Reply 72 of 76
    Well I have to agree that iTunes 2 is very CPU inefficient. But to those that refuse to believe it, use REALbasic as a simple mp3 player. Just add the movieplayer control to a window and see what happens. Heh on my Ti 500 it uses less than 10%. iTunes 2 just suxors. =(
  • Reply 73 of 76
    kukukuku Posts: 254member
    There are ways to lower cpu usage for mp3s, but it's more smoke and mirrors.



    iTune on portables(as well as soundjam among others in classic) can off load to memory and lessen cpu usage.



    Though this is really marginal on average as the initial spike in cpu usages makes up it.



    ~Kuku
  • Reply 74 of 76
    sinewavesinewave Posts: 1,074member
    [quote]Originally posted by discstickers:

    <strong>Well I have to agree that iTunes 2 is very CPU inefficient. But to those that refuse to believe it, use REALbasic as a simple mp3 player. Just add the movieplayer control to a window and see what happens. Heh on my Ti 500 it uses less than 10%. iTunes 2 just suxors. =(</strong><hr></blockquote>



    And it sounds like shit.
  • Reply 75 of 76
    An important observatIon:



    iTunes takes up the same CPU time for me whether I'm playing an mp3 or a CD. Thus, the problem is not mp3 decoding.



    Try it. Play a CD in iTunes and you'll see.



    There is no reason for iTunes to use so much CPU power for playing a CD.



    I also noticed that CPU usage is lowest with both iTunes windows minimized (even compared to hiding the app, or to docking it.. Very odd.



    [ 11-29-2001: Message edited by: Junkyard Dawg ]</p>
  • Reply 76 of 76
    cosmocosmo Posts: 662member
    I tried playing a CD as well, it didn't use the same amount of the CPU as playing an mp3 but the drop was only around 1-2%. Playing a CD shoudl use a whole lot less CPU than decoding an mp3.



    very strange indeed
Sign In or Register to comment.