The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Do Today's High End PCs Compete or Exceed Performance of My Mac Pro DAW Software
Old 20th July 2014
  #31
Gear Addict
 

I haven't geeked out in a while, and I'm feelin' the need. Let me see if I can't get some empirically measured and thoroughly accurate AD/software/DA latency numbers on a trashcan on both Logic and Cubase. I'll probably use an Apollo via Thunderbolt as the I/O.

Probably be a week or so before I can get to it, but if/when I do, I'll start a thread. If I do measure, it will be a properly scientific and documented affair. No use doing it otherwise. I would measure indicated latency vs measured latency, etc. Logic has some poorly understood modifiers for latency in it's settings. Might as well measure what they really do. It would be a nice itch to scratch to have those numbers, even though I'm not a Logic user. I think I can borrow a trashcan and TB Apollo for a couple of days, and it's been too long since I did something like this. The trashcan I would borrow is being used for testing and has all variants of the CPU available to swap.

Which of course means, if you do buy a trashcan MB, you can swap/upgrade the CPU at will. Apple has in no way crippled that, which is nice of them, considering their former ways.
Old 20th July 2014
  #32
Lives for gear
 

Nice! That'd be very interesting, I think a lot of people would be interested.

AFAIK Logic uses a hybrid buffer, it internally syncs higher-buffered playback tracks with the low-buffered "record-enabled" tracks, to create an illusion of handling all the tracks at a low buffer. Cubase 7 introduced something similar called VST guard (located in the Device Setup menu), I'm not sure how it works - probably in a similar fashion - but it is turned on by default and seems to affect low-latency performance.

The fact you can upgrade the nMP processor down the road is very good news. I plan on getting the 6-core model, stuffing it full of RAM, with 3 monitors and SSD's..and a few years later, I could look into replacing the 6 core with an 8 or 12 to get even more years out of it..

But yeah, we use Nuendo and Cubase. No plans of switching to Win so I'm hoping the trashcan suits our needs.
Old 20th July 2014
  #33
Gear Addict
 

The only real way to truly test all of this cross platform OS speculation would be to get somebody with a Hackintosh and good audio hardware like RME to run a test using both Windows and OSX natively on the same exact hardware using the most current drivers and Cubase versions on both. Otherwise, we're all still speculating to a degree.

The end.

But as far as a Logic vs Cubase test on the most currently powerful Mac, that should help out some Mac users wondering about Cubase performance. I'll see if I can get that done.
Old 20th July 2014
  #34
Gear Addict
 

You could add another Mac or PC using Vienna Ensemble Pro while still being able to get a few more years use out of your 2010 Mac Pro. It might not be practical for your situation but it's at least something to look into.
Old 20th July 2014
  #35
Gear Addict
 

Quote:
Originally Posted by StereoPari View Post
I know that a PC is the cheaper "pro" option and that Logic is more efficient w/ low latencies on Mac than Cubase. However I'm leaning towards a 6 or 8 core trashcan because it's small, very silent and can easily run 3 hi-res monitors that I need for scoring to film..
I'm sure you saw the post above. Hopefully, I can find the time (it takes a while) to do a proper test and put the wives tales to rest. I embrace reality, whether it suits my predispositions or not!

On another note, it's very simple, and for DAW use, pretty inexpensive to run 3 Hi-res monitors on a PC. I'm running 3 now on my main DAW, including one 2560 x 1600 30" plus a pair of 1920 x 1080. Cost me an extra $60 for a video card. Frankly, you could run 4, 5 or 6 monitors on a PC if you so desired.
Old 20th July 2014
  #36
Lives for gear
 

Quote:
Originally Posted by Getalife2 View Post
I'm sure you saw the post above. Hopefully, I can find the time (it takes a while) to do a proper test and put the wives tales to rest. I embrace reality, whether it suits my predispositions or not!

On another note, it's very simple, and for DAW use, pretty inexpensive to run 3 Hi-res monitors on a PC. I'm running 3 now on my main DAW, including one 2560 x 1600 30" plus a pair of 1920 x 1080. Cost me an extra $60 for a video card. Frankly, you could run 4, 5 or 6 monitors on a PC if you so desired.
Actually, you can run up to 8 off two sub-$300 video cards, and get 4k.
Old 20th July 2014
  #37
Gear Addict
 

Quote:
Originally Posted by brianbotkiller View Post
Actually, you can run up to 8 off two sub-$300 video cards, and get 4k.
OK, just go ahead and 2-up me on the monitor count. Ha.

I have 6 monitors in the room and people think I'm extreme. Never seen anybody use 8 in the same control room, much less on a single computer.

But, yes, you could do that.
Old 20th July 2014
  #38
Lives for gear
 

Quote:
Originally Posted by Getalife2 View Post
OK, just go ahead and 2-up me on the monitor count. Ha.

I have 6 monitors in the room and people think I'm extreme. Never seen anybody use 8 in the same control room, much less on a single computer.

But, yes, you could do that.
oh, no one up - just pointing out that you can easily get multi-multi out on a windows machine for very little. I couldn't say what I'd do with eight monitors, three is enough for now ;P
Old 20th July 2014
  #39
Gear Addict
 

Quote:
Originally Posted by brianbotkiller View Post
oh, no one up - just pointing out that you can easily get multi-multi out on a windows machine for very little. I couldn't say what I'd do with eight monitors, three is enough for now ;P
Hey, just jokin' with you.

Honestly, the flat, reflective surface of big screen monitors is terrible in a control room, acoustically. The more you have, the worse it is. Lots of early reflection garbage.

One tip. Angling the monitors just slightly can have a positive impact on the undesirable reflections. Slightly counterintuitive, but often, angling them down a bit can help a lot. Not enough to affect the viewing angle, just a few degrees. Try moving the angle of your computer monitors with your eyes closed. Might be surprised how that can clean up the monitoring environment, depending on the room and control surfaces.

When you have a bunch of screens, you figure this stuff out in self defense!
Old 20th July 2014
  #40
Lives for gear
 

Quote:
Originally Posted by Getalife2 View Post
Hey, just jokin' with you.

Honestly, the flat, reflective surface of big screen monitors is terrible in a control room, acoustically. The more you have, the worse it is. Lots of early reflection garbage.

One tip. Angling the monitors just slightly can have a positive impact on the undesirable reflections. Slightly counterintuitive, but often, angling them down a bit can help a lot. Not enough to affect the viewing angle, just a few degrees. Try moving the angle of your computer monitors with your eyes closed. Might be surprised how that can clean up the monitoring environment, depending on the room and control surfaces.

When you have a bunch of screens, you figure this stuff out in self defense!
You're totally right, which is why I've not gone to really huge monitors - I have enough issues treating the average room, already!
Old 20th July 2014
  #41
Lives for gear
 
BobTheDog's Avatar
 

Quote:
Originally Posted by Getalife2 View Post
The only real way to truly test all of this cross platform OS speculation would be to get somebody with a Hackintosh and good audio hardware like RME to run a test using both Windows and OSX natively on the same exact hardware using the most current drivers and Cubase versions on both. Otherwise, we're all still speculating to a degree.

The end.

But as far as a Logic vs Cubase test on the most currently powerful Mac, that should help out some Mac users wondering about Cubase performance. I'll see if I can get that done.
You don't need a hackintosh as apple hardware can run windows. I did some tests on a MacPro 8 core a few years ago with Cubase 6 and it ran much better on windows than OSX, they have improved the performance greatly in 7/7.5 on OS X though.
Old 20th July 2014
  #42
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by Getalife2 View Post
The only real way to truly test all of this cross platform OS speculation would be to get somebody with a Hackintosh and good audio hardware like RME to run a test using both Windows and OSX natively on the same exact hardware using the most current drivers and Cubase versions on both. Otherwise, we're all still speculating to a degree.

The end.
Unfortunately these are not up to date but they give a general idea of the differences between Win and OS-X:







More info here: DAW Bench : DAW Performance Benchmarking

Alistair
Old 20th July 2014
  #43
Lives for gear
 
BobTheDog's Avatar
 

Not up to date and not a good benchmark!
Old 20th July 2014
  #44
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by BobTheDog View Post
Not up to date and not a good benchmark!
Why is it not a good benchmark?

Alistair
Old 20th July 2014
  #45
Lives for gear
 
T_R_S's Avatar
Quote:
Originally Posted by 12tone View Post
The limited expansion possibilities of the new Mac Pro's are a real turn-off for me.
There are far more expansion possibilities with my 2013 MP I have 2 Expansion Chassis (6 pcie slots total) a 30TB Thunderbolt 2 drive and Black Magic Thuderbolt dock with 4 SSD Drives. Plus support for 3 high resolution displays
the new Mac Pro's are far more expansion possibilities that any previous mac ever built. Having Thunderbolt gives access to a lot more expansion possibilities.
I had a overclocked 4Ghz Quad core i7 PC and I can do 4X as many native plug-ins on the 12 core MP. A quad core i7 is no match for server class E5 Xeon.
If they were better nobody would buy E5's
Old 20th July 2014
  #46
Lives for gear
 
T_R_S's Avatar
Quote:
Originally Posted by UnderTow View Post
Why is it not a good benchmark?

Alistair
those benchmarks are 4 years old
Old 20th July 2014
  #47
Gear Maniac
 

Quote:
Originally Posted by T_R_S View Post
those benchmarks are 4 years old
...and was flawed and useless already then.
Old 20th July 2014
  #48
Lives for gear
Quote:
Originally Posted by kalle789 View Post
...and was flawed and useless already then.
Can you elaborate on why they were flawed?
Old 20th July 2014
  #49
Gear Addict
 

Quote:
Originally Posted by T_R_S View Post
Having Thunderbolt gives access to a lot more expansion possibilities.
This is NOT an attempt to sound like a smart**s. Obviously, adding TB adds options. Great.

But would having a couple of PCIe slots included on the new MP reduce the expansion options? I think that's most people's point. Your system sounds impressive. But just getting access to PCIe cost you a nice 500 module or three. That's great if you're in that place, but not all of these guys are, ya know?

Eliminating PCIe from the new MP did not make it better. It's obviously steered mostly towards video pros, hence the focus on video card options from Apple. Music folk will just have to pony up the extra dollars to play. Like you did, and I'm happy for you. Sounds like a monster rig.
Old 21st July 2014
  #50
Lives for gear
 

Quote:
Originally Posted by BobTheDog View Post
You don't need a hackintosh as apple hardware can run windows. I did some tests on a MacPro 8 core a few years ago with Cubase 6 and it ran much better on windows than OSX, they have improved the performance greatly in 7/7.5 on OS X though.
Good point. I sometimes forget that I can run Windows on a Mac..so one could settle for the nMP and run Cubase on either platform if needed. This approach also brings less potential problems than having it the other way around (personally I'm not even considering that).

Currently the trashcan Pro just feels like it'd be a good solution. I don't mean price/performance ratio or "raw" performance, but the whole approach.

The hi-res monitors were only one part of the equation, I know I can buy a cheap GC for PC that runs them, also the iMac can run 3 HD monitors...but cards are usually equipped with a fan, and even an iMac probably gets louder when driving 3 monitors (and it can't drive 2560 x 1440 or 4Ks, only standard HD..). I just like the idea that it's designed to be very quiet and they've thought the heatsink so well it only has one single fan. I agree it feels easily expandable, can't imagine needing more than 6 Thunderbolt 2 connections for my devices, ever. If you can even upgrade the processor later on, even better.
Old 21st July 2014
  #51
Sure a mac can run windows,thru parallels or whatever,but can it run a windows DAW?

That's what I wan't to know,how many have successfully run a Win Daw like Sonar on a mac?
Old 21st July 2014
  #52
Lives for gear
 

Quote:
Originally Posted by rksguit View Post
Sure a mac can run windows,thru parallels or whatever,but can it run a windows DAW?

That's what I wan't to know,how many have successfully run a Win Daw like Sonar on a mac?
Yes. Total success. You don't use Parallels. You use boot camp. The Mac then becomes a total pc. Or a total Mac... all via a click of the reboot switch onscreen.

In my case, several Mac Minis running windows 7 64 bit and all my normal pc programs such as Cubase.

Via boot camp, you have two totally independent computers in one. I also set my Minis to always boot up into windows as a default. I sometimes forget they're even Macs as I rarely.... and I mean rarely... switch over to boot up into osx 10 for running the rare Mac-only programs I sometimes need to run.

When you first use boot camp, it prompts you for setting up the new windows partition (no rocket science involved) , after which you pop in your windows 7 or 8 dvd and install... and then pop in your Cubase/Nuendo/Sonar dvds (or whatever) and install....and any other windows programs you want......... just like you would do on a Dell... or a whatever. Absolutely no learning curve and no glitching.

Boot Camp is my favorite Apple invention.
Old 21st July 2014
  #53
Lives for gear
 
BobTheDog's Avatar
 

Quote:
Originally Posted by J-S-Q View Post
Can you elaborate on why they were flawed?
Basically they test a certain set of plugins that just happen to run better on windows than osx.

So really the test is benchmarking certain plugins rather that the Daw/Operating system.
Old 21st July 2014
  #54
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by T_R_S View Post
those benchmarks are 4 years old
Is there anything that demonstrates that OS-X is now better at low latency performance?

Quote:
Originally Posted by BobTheDog View Post
Basically they test a certain set of plugins that just happen to run better on windows than osx.

So really the test is benchmarking certain plugins rather that the Daw/Operating system.
Is there a set of plugins that run better on OS-X? How do you know the differences are not typical? Developers have been complaining about OS-X's low latency performance ever since OS-X exists. Steinberg, Coco's and even IBM even criticized Apple for how inefficient the OS-X is/was.

It seems you are just dismissing this test because you don't like what it says about OS-X. If you have a better test that demonstrates that OS-X performs as well or even better than Windows then please provide a link. Until then, I think this offers the best data on the subject.

Alistair
Old 21st July 2014
  #55
Lives for gear
 
BobTheDog's Avatar
 

No I am dismissing it as I have done my own benchmarks.
Old 21st July 2014
  #56
Gear Guru
 
UnderTow's Avatar
Quote:
Originally Posted by BobTheDog View Post
No I am dismissing it as I have done my own benchmarks.
And can you share any more info on this? For instance which plugins you used for your testing?

Alistair
Old 21st July 2014
  #57
Lives for gear
 
BobTheDog's Avatar
 

The tests were done with a VST passthrough plugin, this is a plugin that copies the audio input data unchanged to the audio output data.

A specially written host was used to set up N threads of audio processing, the host then gradually added plugins to each thread until the thread could not process the data through the chain within the required time-slice.

I can't go into the results too deeply as it was paid work but the results are a lot closer than DawBench would have you believe.
Old 21st July 2014
  #58
Lives for gear
 
Timur Born's Avatar
 

Main problem with OS X is the aggressive CPU power-saving behavior that cannot optionally be disabled. On Windows and with some PC EFIs I got full control over that.

With Core2Duo CPUs I disabled the corresponding OS X kext and used extra software to control the CPU clock myself. That software doesn't work with "I" type CPUs, so sometimes one has to create artificial CPU load (free Systemload tool) to get better low latency performance out of OS X. That doesn't exactly help to keep temps and fan noise down, of course.
Old 21st July 2014
  #59
Lives for gear
 
norbury brook's Avatar
 

Quote:
Originally Posted by BobTheDog View Post
The tests were done with a VST passthrough plugin, this is a plugin that copies the audio input data unchanged to the audio output data.

A specially written host was used to set up N threads of audio processing, the host then gradually added plugins to each thread until the thread could not process the data through the chain within the required time-slice.

I can't go into the results too deeply as it was paid work but the results are a lot closer than DawBench would have you believe.
The point of the DAW bench test was to use plugins and virtual instruments that most of us use. 99% of my contemporaries run Kontakt so it's performance is very pertinent in a test rather than a VST pass through plugin.


I switched to windows for the very situation DAW bench highlighted.

Most of the major V.I's ran great and at the time fully x64 on windows;

Kontakt
EW play
Halion
All Spectrasonics products
Superior drummer
BFD


The gap has closed over the last few years I agree, but it would have to surpass widows performance for me to switch back.

Something I'd have no problem doing if that were the case, it's just a tool.



MC
Old 21st July 2014
  #60
Lives for gear
 
BobTheDog's Avatar
 

The test I described tests a machine/OS for VST throughput.

The Daw Bench test tests a DAW/particular Plugin (mostly plugin) performance.

All I was trying to say is that OSX is not flawed for low latency VST performance, now the version of cubase and the version of Kontakt you used obviously are. But that is nothing to do with the operating system, it is to do with the programmers who created that software.

As I said in an earlier thread post Cubase runs much better on a Windows than OSX, this is a Cubase issue not an OSX issue.

I can easily write a VST plugin that uses SIMD on OSX and the FPU on Windows, so now we benchmark the 'same' VST on windows and OSX and low and behold OSX is performing much better than Windows. This has nothing to do with the underlying OS and everything to do with the way I coded that VST.
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump