1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Logic 8 How to benchmark with Logic?

Discussion in 'Logic 8' started by afavreau, Mar 21, 2009.

  1. afavreau

    afavreau Senior member


    Everybody knows http://www.barefeats.com is the best place to get the cold facts about the performance of your MAC. All kinds of tests are made to find out how (and if) you favorite app will better perform on a new Mac. Unfortunately, Logic is normally not evaluated.

    So I reached Rob-ART, the mad-scientist from Barefeats to ask him if Logic could be evaluated too in the future, so we would know if a new Mac is really worth the buck. He responded to me that he is ready to invest and do it, but since he doesn't know Logic, he needs some ideas on benchmarking it.

    Rob-ART really knows Macs and LUG is where there's the most knowledge about Logic. So I suggested to Rob to start a discussion here. He is ready to join in. If this works, we could do ourselves (Logic Users) a big favor by being able to refer to a site before our next MAC purchases.

    So the question is: How can we best benchmark Logic and evaluate the performance of the Mac running it?

    Happy Discussion!

    André Favreau
  3. Orren Merton

    Orren Merton Logic Samurai / Administrator Staff Member

    Years ago, there was an Italian (I believe) PC Logic user who had a site on the internet that offered some standard files and had people report their systems and the results of the test running platinumverbs and spectral gates.

    If I recall correctly, the German site www.logicusers.de has a standard test song they use for tests—Mark Cahill is an administrator there, and I know that another member here Putte posts on that forum, so hopefully they'll chime in.

    Of course, we could always come up with our own, and post it in the Downloads section of www.logic-users.com...

  4. Markdvc

    Markdvc Administrator Staff Member

    I like the idea of benchmarking Logic!

    The test Orren refers to is really quite simple. It requires playback of a Logic song which can be found here:


    this has one instance each of the es2 and platinum reverb.

    The idea is to determine how many tracks Logic can playback by copying region, instrument and reverb to more and more tracks/channel strips. The number of tracks which the mac is still able to playback and remain responsive without core audio overloads determines the benchmark.

    However, this test was created over 6 years ago, when the G4 was the fastest Mac, and things have moved on a lot since then. The results are now exceeding 150 tracks, and that requires an awful lot of copying! As a benchmark usually means testing reasonably current macs it seems that we should come up with a more appropriate benchmark which will take the power of current Macs into account.

    A simple suggestion would be to replace the ES2 with Sculpture, and platinum reverb with Space Designer. Both require considerably more CPU.

    We could also ask members of the German forum to participate and cross reference results.

    kind regards

  5. Markdvc

    Markdvc Administrator Staff Member

    The other thing that could be interesting would be a benchmark trying to compare different DAWs on the same Mac. While it is easy enough to compare track counts (and Logic 8 has improved hugely in this area, especially with high sample rate audio) CPU power is more difficult to measure as it requires finding a common denominator, but few of us all have one and the same AUs and other DAWs at hand.

    A combination such as Zebra and Altiverb springs to mind, but not everyone has these - even if they really should IMO ;) :D

    If there is some interest, I don't mind posting results comparing Nuendo running these as VSTs with Logic running them as AUs.

    kind regards

  6. Pepe Ortega

    Pepe Ortega New Member

  7. Eli

    Eli Senior member

    Good benchmark test. But maybe they should have used a longer impulse response in the Space Designer settings?

    I started maxing out at about 60 tracks. I'm on a 8 core 2.8 Intel tower with 6 GB of ram. I had a couple of other apps open at the time.....

    Oh yeah - and I had my buffer at 32
  8. barefeats

    barefeats New Member

    Thanks for the ideas on benchmarking with Logic.

    I'll give them a try and report back.

    Rob-ART @ BareFeats.com
  9. leytonnz

    leytonnz Senior member

    bump... did the logic benchmarking idea get any further..i was just thinking of it because of the new MBPs.
  10. afavreau

    afavreau Senior member

    No, Rob-ART needs a version of Logic

  11. leytonnz

    leytonnz Senior member

    so we need 50 peeps to put in $10.. ill do that to get good benchmarks..

    what do the mods think? group buy for rob.. ie we as a group buy him logic studio?
  12. Peter Ostry

    Peter Ostry Administrator Staff Member

    I need just 10 minutes to find out that my MBP is crap in the meantime ;) Oops :redface:

    Not really a good idea, I think. The "donation buy" itself is thinkable but let us look at the known facts:

    • The response to this subject wasn't really great.
    • Most people talked about other ways to do benchmarks.
    • To the $500 you have to add the money transfer costs.
    • If there are not enough people finally, sending the money back to each person costs another round of transfer costs.
    • If we (the owners of the forum) tell a programmer to make the software, we will spend our own money from the moment he starts typing.

    If there were better chances, you could just start a poll with a new thread, if it is satisfying post your Paypal account, collect the money and send it to Rob-ART. Are you brave enough?

    I would rather follow Mark's suggestion, were we put a test together that fits to the current machines and maybe share it with other groups. This does not help if you don't know if you want to use Cutools, Prodesigner or Digibase but is quite usable to learn how Logic performs on other Mac models. We would make and maintain a kind of performance-table with the results.

    So, who of you volunteers to make the test project?
    Logic 9 exclusively or not?

Share This Page