Jump to content


Please note: You can easily log in to MPN using your Facebook account!

Keyscape, Mainstage, MacBook Pro Compatibility


Recommended Posts

Last year I started a thread that many of you helped me out on. For reference: https://forums.musicplayer.com/ubbthreads.php/topics/2931655/1

 

The problem was I had what I thought was a pretty souped up MacBook Pro and yet I was having a lot of issues running Keyscape (and a couple other AU's like Arturia Piano, DX7 and Pigments) inside Mainstage. We tried a multitude of software and setting solutions but nothing totally solved my problems. It took me a long time to come to some kind of conclusion but I think I have and I'm posting here to hopefully save some of you the same trouble I had.

 

Mainly, the 13 inch MacBook pro. Models may vary but I don't think that you can get one with a dedicated graphics card. They frequently use what's called integrated graphics, which means that the processor does a lot of the graphics work. So my pretty expensive 2015 13 inch MBP with a 3.1GHZ i7, 16GB RAM and 1TB flash drive was not fast enough and would suffer from CPU usage spikes that would generate loud pops and glitches.

 

Weirdly, these same resource demanding AU's would run smoothly on my iMac which had both a slower processor and slower drive...but a dedicated graphics card.

 

I talked to some guys who build Mac's at a place called FastMacs.net (I don't work for them and this isn't an attempt to advertise). They hooked me up with a very pricey hot-rodded 15 inch MBP with a 2.9GHZ i9 chip and 32GB RAM and a dedicated graphics card and voila; everything is running smoothly. I can run multiple instances of Keyscape with 32 voices, no thinning, pedal realism on, without bussing the reverbs or changing buffers. It just cruises right along.

 

Anyway, I posted this for you guys so you could see that for some 3rd party instruments, don't just assume that a recent or expensive MBP has the horsepower to run your program and make sure it has a dedicated graphics card.

 

There's been a lot of talk about solid state vs spinning drives but I actually don't think that makes much of a difference. My 13 inch MBP had the fastest drive available yet that didn't seem to make a difference. My iMac had a much slower fusion drive and it works just fine with Keyscape/Mainstage. I had messaged with other dudes who had spinning hard drives who said their systems worked fine.

 

FastMacs guys were easy to work with but their stuff is NOT cheap. Their website has many different models of varying capability and prices but they cater to super high end video or music production.

 

You want me to start this song too slow or too fast?

 

Forte7, Nord Stage 3, XK3c, OB-6, Arturia Collection, Mainstage, MotionSound KBR3D. A bunch of MusicMan Guitars, Line6 stuff

Link to comment
Share on other sites



  • Replies 20
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Yeah, as much as I love MainStage, one of the very weird things about it is that for some reason it implodes with integrated graphics cards. Even if you remove all the 3D widgets (which are the only things that would benefit from a GPU), it sometimes still suffers. It's a total bug. I just bought a new MBP a few weeks ago, after 8 years on the same 13" machine. I really like the 13" footprint for stage use, but they don't make one with a dedicated GPU, so I had to move up to a 15". Now it blazes along like a dream, though!

 

Check your processor speed though, they don't make a MBP with a 2.9Ghz chip, their top-of-the-line machine is an 8-core i9 with a 2.4Ghz CPU, which is what I got in mine. Maybe last years model was a 6-core 2.9Ghz though, that would make sense (they often scale back processor speed with more cores). You likely have the same machine as mine: 32GB RAM, 8-core 2.4Ghz i9, RadionPro 560X, 1TB SSD.

 

Got mine from Apple, not cheap.

Puck Funk! :)

 

Equipment: Laptop running lots of nerdy software, some keyboards, noise makersâ¦yada yada yadaâ¦maybe a cat?

Link to comment
Share on other sites

I was puzzled how music apps, VSTS, etc would put any load on the integrated graphics card. I suppose the MS 3d widget is the culprit.

 

I have the older 13" MBP, 16 GB ram, 2.7 GHz, Core i7. Running High Sierra 10.13.

 

I will soon be pushing it with Keyscape and Omnisphere. I would be surprised if I run into issues.

 

I don't have any need for MainStage. It is not on my MBP

 

I don't use my old MBP for Internet surfing , shopping etc etc. Its my dedicated machine for my music production.

 

I have a new Mac mini [ 2013 model crapped out] for all the daily computer, Internet, email and app tasks But I can see how brute processing

is needed to blast thru all the Internet video ads and assorted crap.

 

This might be a good comparison since I will not run MainStage on my old MBP[ which is maxed out on RAM and has a 2nd SSD].

If I run into glitchiness , will report back

 

Why fit in, when you were born to stand out ?

My Soundcloud with many originals:

[70's Songwriter]

Link to comment
Share on other sites

If you are mainly using Keyscape/Omnisphere and other third party AUs consider:

https://gigperformer.com/ Instead of MainStage?

 

That's what I did. I haven't done any A/B comparison tests, but GP definitely feels a little more "relaxed" on my 2015 MBP (13"-2.9GHz-8GB RAM). More importantly, it made instant sense to me, whereas Mainstage always felt less intuitive. YMMV, JM$0.02 etc.

 

On a different note, the v1.0.0 of Keyscape was pretty inefficient, but it has gotten a lot better with the updates.

 

A few years ago, somebody at Spectra told me in an email: "MainStage is OK, but it's a memory and CPU hog (which Omnisphere can be too, so the two of them is really pushing it)." quote truncated for relevance.

 

So maybe GP is a better fit if your primary goal is to run Omnisphere/Keyscape. Or use them standalone, of course, if you don't need anything else.

 

 

Link to comment
Share on other sites

thx, I will keep that option in mind.

 

To clarify, I don't gig, I record. I don't noodle much ;)

 

I have some options, I can use Omnisphere in Stand-Alone, sort of like a sound module.

 

Likely I will run Omnis/Keyscape as VST instruments in Cubase

Why fit in, when you were born to stand out ?

My Soundcloud with many originals:

[70's Songwriter]

Link to comment
Share on other sites

Yeah, as much as I love MainStage, one of the very weird things about it is that for some reason it implodes with integrated graphics cards. Even if you remove all the 3D widgets (which are the only things that would benefit from a GPU), it sometimes still suffers. It's a total bug. I just bought a new MBP a few weeks ago, after 8 years on the same 13" machine. I really like the 13" footprint for stage use, but they don't make one with a dedicated GPU, so I had to move up to a 15". Now it blazes along like a dream, though!

 

Check your processor speed though, they don't make a MBP with a 2.9Ghz chip, their top-of-the-line machine is an 8-core i9 with a 2.4Ghz CPU, which is what I got in mine. Maybe last years model was a 6-core 2.9Ghz though, that would make sense (they often scale back processor speed with more cores). You likely have the same machine as mine: 32GB RAM, 8-core 2.4Ghz i9, RadionPro 560X, 1TB SSD.

 

Got mine from Apple, not cheap.

You're right, they don't make one...unless you get it from https://www.fastmacs.net/ . Came with factory warranty and apple care. It's not hardly more expensive than Apple. Plus they make super high end stuff that apple doesn't really offer. Check their website or their ebay store.

You want me to start this song too slow or too fast?

 

Forte7, Nord Stage 3, XK3c, OB-6, Arturia Collection, Mainstage, MotionSound KBR3D. A bunch of MusicMan Guitars, Line6 stuff

Link to comment
Share on other sites

I was puzzled how music apps, VSTS, etc would put any load on the integrated graphics card. I suppose the MS 3d widget is the culprit.

 

I have the older 13" MBP, 16 GB ram, 2.7 GHz, Core i7. Running High Sierra 10.13.

 

I will soon be pushing it with Keyscape and Omnisphere. I would be surprised if I run into issues.

 

I'll be very interested to see how that works as I had a similar machine (2015 3.1GHZ i7 16GB RAM, SSD) and I really couldn't run Keyscape even stand alone without pops and glitches (limited voices, thinning, pedal realism off, buffer adjusted, everything else shut down, low resolution mode, nothing really fixed it).

 

With my new computer I can run with ALL that stuff on and it just hums along with multiple instances of Keyscape running like it was nothing. It's honestly made Mainstage much much more usable.

 

Anyway, let me know how it works. My prediction is it won't but I'd love to be wrong.

You want me to start this song too slow or too fast?

 

Forte7, Nord Stage 3, XK3c, OB-6, Arturia Collection, Mainstage, MotionSound KBR3D. A bunch of MusicMan Guitars, Line6 stuff

Link to comment
Share on other sites

Guys, it surprises me no one has mentioned the audio IO they are using and the impact different IO drivers have on your overall CPU overheads.

 

In a sense the ability any audio driver has to deliver low latency performance eventually comes down to the CPU overhead the driver imposes. Clicks and pops start when your CPU starts to not keep up.

 

Secondary to this the converter latency all AD/ DA converters have. Newer ones add less latency that older ones.

 

So guys please share which audio interface you are using with your MS rigs to add some context here.

 

The other thing users might consider is dumping their laptops for a headless mac and use a controller that gets enough feedback from the software that you dont need a computer screen on stage. No graphics card issues then. Im happy to share what 5 years of headless MainStage work has taught me.

 

Kind regards

Paul Najar

Jaminajar music production

www.jaminajar.com

Link to comment
Share on other sites

You're right, they don't make one...unless you get it from https://www.fastmacs.net/ . Came with factory warranty and apple care. It's not hardly more expensive than Apple. Plus they make super high end stuff that apple doesn't really offer. Check their website or their ebay store.

 

Apple solders their CPUs, RAM, SSD, and Video Card to the motherboard, so it's not actually possible to create a different spec than what Apple provides. This seemed a little suspicious, so I looked into it. What I noticed is that the 2.9Ghz model, which lists the processor as "Intel 8950KH", is an 8th-gen 6-core CPU, which Apple used on last year's highest spec model. This year's highest-spec system drops the native clock speed back to 2.4Ghz, but with 8-cores, and is of the Intel 9th-gen class, which due to upgrades in architecture is benchmarked (by third party testers) to be about twice as fast as last years 6-core model. Both last year's and this years models have the same 4.8Ghz turbo speed boost, so basically the only difference is that this year's model runs slower, cooler, and uses less power under normal conditions, but will match last year's clock speed when it needs to... and with two more 9th-gen cores.

 

What also concerns me is that FastMacs supposedly offers a 13" model with a Radeon card. This is technically "impossible" for an officially sanctioned Apple computer. The only possibility is that it's actually a gutted mac with Hackintosh internals, though with pretty much the same components Apple uses but not pre-soldered. This doesn't really bother me, as I disagree with Apple on this policy, but FastMacs should be more upfront about this, I can't find any explanation for this on their website. Apple doesn't have ANY deals with any third-party suppliers either. They also watermark their systems, so if you take it to an AppleStore or certified technician, they contractually won't be able to work on it. Your computer, on the other hand, is of a spec set that Apple does offer (or did), so it is likely an officially sanctioned system.

 

Note that the 8-core system has only been available for 6 weeks, so if you bought your 6-core system before then, it technically WAS the fastest MacBook on the market at the time. That said, FastMacs has yet to release an 8-core package, and their prices are still higher than Apple's, so they're currently offering last year's models at a higher price :/

Puck Funk! :)

 

Equipment: Laptop running lots of nerdy software, some keyboards, noise makersâ¦yada yada yadaâ¦maybe a cat?

Link to comment
Share on other sites

Guys, it surprises me no one has mentioned the audio IO they are using and the impact different IO drivers have on your overall CPU overheads.

 

In a sense the ability any audio driver has to deliver low latency performance eventually comes down to the CPU overhead the driver imposes. Clicks and pops start when your CPU starts to not keep up.

 

Secondary to this the converter latency all AD/ DA converters have. Newer ones add less latency that older ones.

 

I purposefully only use the internal audio card on the MacBook. The problem is an external interface could potentially become unplugged, and would cause a full reboot of the CoreAudio system (and VIs), up to 2-4mins of reloading on my larger MS concerts. This would be a gig killer. Fortunately (and somewhat surprisingly), Apple provides one of the most low-latency and highest quality DACs on the market, with a faster roundtrip than most third-party interfaces. Keep in mind that the audio interface (either external or internal) has its own sub-processor, so the mainboard CPU is never handling the nitty-gritty of D/A conversion. If you're getting audio glitches, it's always the fault of the main CPU trying to keep up with software synthesis, not low-level audio generation. Conversely, ALMOST no audio interfaces handle VI/Effects processing. The rare exceptions are proprietary systems like ProTools TDM or a Universal Audio Accelerator, but nothing we're discussing here pertains to those.

 

Now, non-apple computers could come equipped with similarly low-latency interfaces, but some like Dells are supposedly pretty laggy, so I can understand why some Windows laptop users would resort to using external interfaces, but it's not really necessary with a mac. That said, if you need more than 2-channel output or audio input, then an external interface is your own option.

Puck Funk! :)

 

Equipment: Laptop running lots of nerdy software, some keyboards, noise makersâ¦yada yada yadaâ¦maybe a cat?

Link to comment
Share on other sites

Generally speaking the internal audio card on any MacBook that I have owned over the years has been sufficient. But you can definitely get better low latency performance with the added bonus of pro grade connectors from many USB and Thunderbolt interfaces. Keep in mind that you can kick out the 3.5mm audio cable, yank the MacBook with it, and damage the jack and or cable in the process too. But you are right, MainStage shouldnt flinch with the removal of a 3.5mm cable. With removal of interface on OSX you should get a message that the assigned audio interface is no longer available - I havent purposefully yanked out a USB or Thunderbolt interface lately so I am not sure what may fail - but it should not require a reboot of the computer.

 

I like the internal card because its one less box, one less cable to carry and setup - but I wind up bringing a direct box with a 3.5mm input and 1/4 out to my monitor anyway. An interface would replace the direct box if I select wisely.

 

 

Yamaha CP88, Casio PX-560

Link to comment
Share on other sites

not surprising, its not 1 size fits all.

 

In my case and many others, Kronos refused to play nice with Mac USB audio. This went on for 2 years as more folks got puzzled ' WTH ".

 

The Korg theory was that Apple removed audio packets

It happened when Apple went from El Cap or thereabouts.

 

I bought a UR44 Ai in self defense.

 

I can't be certain, but I believe the audio got resolved later with Mojave . I would have to dig to get the exact

chronology. Again, it was us Kronos owners.

Why fit in, when you were born to stand out ?

My Soundcloud with many originals:

[70's Songwriter]

Link to comment
Share on other sites

Generally speaking the internal audio card on any MacBook that I have owned over the years has been sufficient. But you can definitely get better low latency performance with the added bonus of pro grade connectors from many USB and Thunderbolt interfaces. Keep in mind that you can kick out the 3.5mm audio cable, yank the MacBook with it, and damage the jack and or cable in the process too. But you are right, MainStage shouldnt flinch with the removal of a 3.5mm cable. With removal of interface on OSX you should get a message that the assigned audio interface is no longer available - I havent purposefully yanked out a USB or Thunderbolt interface lately so I am not sure what may fail - but it should not require a reboot of the computer.

 

I like the internal card because its one less box, one less cable to carry and setup - but I wind up bringing a direct box with a 3.5mm input and 1/4 out to my monitor anyway. An interface would replace the direct box if I select wisely.

^ This.

 

I should specify, I don't mean that you have to reboot the entire computer, but the CoreAudio system does get reloaded any time there's a hot-swap in the audio hardware, this means that all VIs and effects get reloaded. This is an effect of the CoreAudio OS system. It figures that a different interface will likely have a different sample or bitrate, and whole-sale restarts the audio system, even if the sample/bit rates are the same. During normal operation, this happens instantaneously, however, VIs see this as "oh, we have to reload to account for a possible change of specs", and all open synths in a project get fully reloaded.

 

WARNING: Unfortunately beginning with Mojave (or maybe High Sierra, I skipped that one), MacOS now sees "headphones" and "internal speakers" as two different audio devices. This means that if you unplug the jack, it reboots the CoreAudio system. This is really worrisome. To help counteract this, as well as making the connection physically safer, I purchased a small "L" connector that I keep semi-permanently in the jack. As long as it's connected, macOS sees it as the headphones being the output. Most likely, if anyone kicked the cable, it would pull out of the adaptor, not out of the jack itself, but it's still a design flaw IMO.

Puck Funk! :)

 

Equipment: Laptop running lots of nerdy software, some keyboards, noise makersâ¦yada yada yadaâ¦maybe a cat?

Link to comment
Share on other sites

Hi Everyone,

 

Yes I get the idea of cords coming unplugged! This and other vulnerabilities of a laptop floating up around me on stage is what led to the idea of going with a powerful cheaper mac without a screen, tucked in a box with the audio interface. And if I could just be vein for a second I dont like the look of the laptop up there either.

 

To EricBarker, your suggestion that Apple makes/ uses low latency and good sounding converters may be true compared to what other computer makers use but compared to even a small professional audio interface the theory falls over. And that goes for sound quality, round trip latency AND the CPU overhead Apples Core Audio driver imposes on any system. Of course theres always good enough and if theyre good enough then great.

 

The thing that got me chiming in on this thread was the idea of getting all the desired CPU performance out of a MainStage Mac with as much CPU cycles in reserve as possible because the less hard the computer has to work the more stable the system will always be.

 

So when mainstage is running at a show the CPU is used by all the stuff MS has to do quickly and then add the audio driver CPU overhead on top of that. If you want to work at the lowest possible latency then the less time the CPU gets to do all its processes. Eventually as we all know the CPU can run out of time to process the required number of tasks thrown at it, and then errors in the system start to occur. This is why CPU overhead of the audio driver matters. Less overhead effectively means the CPU can process more tasks before errors occur. This is true for every audio driver including Apples own core audio driver.

 

If you want to gain a clue whether a given interface has a chance of bettering the the performance of Apples core audio drivers then check whether the interface installs its own drivers or uses Apples built in ones.

 

Some non Apple audio drivers add less CPU overhead at small audio buffer settings than many others including Apples own. The gold standard are all the interfaces from RME. Some newcomers like the Presonus Quantum are making big claims but even though they claim very low latency at 44.1Khz - like 2.1 - 2.5ms without testing its not clear whether at that low audio buffer that theres much CPU cycles left over for general computing. Its also not clear without testing whether those figures include converter delays. I know with my own testing on multiple RME interfaces, and multiple different macs as far back as 2008 I can reliably get 3.0 - 3.3ms round trip including converter delays. This is as good as it gets at 44.1Khz. Youre getting even better performance if you dont plug a mic in for vocoder work etc and just play your MIDI controllers as its only the output latency that is in play. For RME thats below 1.5MS.

 

Ive tested a few different mac models built in converters for RTL and the best I have ever seen (with a proper audio loop back test) is around 4.5MS with 2.X for output latency.

 

Also, some audio drivers wont impose much CPU overhead but they also wont perform near industry standards for low latency.

 

MainStage in its audio preferences reports on round trip latency figures. Here again this is not clear because as I have discovered those figures are up to the audio driver to report (interface maker) not MainStage itself. I have tested interfaces that claim very low latency but when I do a loop back test and physically see what the RTL is its more that whats reported. In other cases whats reported is verified by my ow RTL test.

 

Other benefits of an external interface are +4dB balanced audio outputs. As ElmerJFudd points out one less box but then one more box to get that dodgy 3.5mm plug into the PA/ monitoring.

 

My own system is 2012 Quad core 2.3Ghz Mac Mini Server with 2 1TB SSDs and 16GB ram running Mojave. The small audio interface I use most is a Zoom TAC-2 connected via thunderbolt which tests tell me its RTL is 3.3ms. I can also get similar performance out of a Focusrite Scarlett 6i6 gen 2 for when I need more IO. These are not expensive interfaces. The entry level for RME is the Babyface Pro. Its 2-3 times the cost. I love RME and have used them in my studio for years. Maybe one day Ill treat myself and get one for the road.

 

Because I get less glitches when playing through my RME interface and most glitches playing through the Zoom this tells me the Zoom has the highest CPU overhead of the three devices I mention.

 

Kind regards

 

 

Paul Najar

Jaminajar music production

www.jaminajar.com

Link to comment
Share on other sites

There is no mainboard CPU overhead imposed by using the internal soundcard, period. That's handled by a separate sub-processor off the main CPU. A USB interface is not going to save you any CPU power, because the CPU itself is never doing any D/A conversion, so you can take that out of the equation. Now, the software drivers bussing the datastream TO the interface are relevant. However, drivers for a hardwired bus are going to be inherently simpler than those bussing through USB/Thunderbolt, and it's unlikely that they would cause any quantifiable CPU overhead. USB software drivers used to be inherently laggy which is why the best interfaces tended to be Firewire, but programmers have long since solved that bag of worms, so I think we can pretty much take everything before the the interface itself out of the equation.

 

These days, most of the latency is going to be caused at the OS-level to buffer against complex DSP, and that's obviously going to be the same no matter what interface you're using. The only other realistic consideration is latency at the interface itself by the DAC. In my experience, the onboard one is "quite good". Obviously you can always eek out a bit more, but I personally would toss it in the "good enough for jazz" category.

 

To clarify, "CoreAudio" is the name of the MacOS-level audio API well above any interface drivers running at the kernel level. CoreAudio handles things like AU/VST and DSP processing. You're going through CoreAudio no matter whether you're using an Apogee Duet, an RME Fireface, or the internal soundcard.

 

As for sound quality? That is up for grabs. And physical jack quality & wear is definitely a consideration. I'm willing to concede that interfaces have their advantages. However, being a professional means weighing quality of product over potential for failure. IMO, the internal soundcard on a MacBook is solidly in the "good enough" category, and outweighs the potential risks from adding another complexity to the chain.

 

Also, I don't use "one more box" than I would with an interface. I typically gig with a second hardware board and a mic, so I have a small A&H mixer right at my feet which also doubles as a DI (balanced XLR outs). If I were using an external interface, I'd run it a 1/4" cable. As it is, I just use a 3.5mm -> 1/4" cable, so no more pieces to the puzzle than I would otherwise. Contrary to popular opinion, the size of audio plug doesn't inherently effect the quality of the audio. Though, larger plugs don't suffer as much from wear & tear and are more secure. It's also technically a USB interface, and I could use it as such. But then I'd be running a USB cable into it instead of an audio cable. If for some reason the audio cable gets unplugged, I can plug it in or grab a spare within a few seconds, if the USB cable unplugs, that's a 3min reload of all my patches.

Puck Funk! :)

 

Equipment: Laptop running lots of nerdy software, some keyboards, noise makersâ¦yada yada yadaâ¦maybe a cat?

Link to comment
Share on other sites

EricBarker,

 

it's only a guess on my part but I get a nagging feeling from your previous post that you took my previous post as a personal attack. If that's true I am sorry and it was not at all intended.

 

I only intended to reference something you had said in one paragraph alone. I probably could have been clearer even though I referenced another user in the post as well.

 

That said I disagree with your first three paragraphs of the post before this one almost 100%

 

So for example read this article from Focusrite

 

I'll quote one of the paragraphs you'll want to look at carefully.

____________________

"The driver and related software are critically important to achieving good low-latency performance. Well-written driver code manages the systems resources more efficiently, allowing the buffer size to be kept low without imposing a heavy load on the computers central processing unit. The importance of drivers means its not possible to simply say that one type of computer connection is always better than another for attaching audio interfaces. Any technical advantage that, say, Thunderbolt has over USB is only meaningful in practice if the manufacturer can exploit it in their driver code."

_____________________

You'll also find similar more technically written articles at rme-audio.de. I would also hazard a guess and say that Apogee, Universal Audio, motu.com, Antelope Audio also has some informative reading in a similar vein.

 

Finally in your last paragraph you say the size of the audio plug doesn't inherently effect the audio quality. In a physical sense that is correct, however given that a three point mini jack can't carry 2 channels of balanced audio and rarely if ever are 3.5mm audio outs referenced to a +4 level there are clear electrical restrictions that relate to sound quality when comparing the mini jack to something like 2 XLR connectors or 2 x6.25mm TRS connectors.

 

Kind regards

Paul Najar

Jaminajar music production

www.jaminajar.com

Link to comment
Share on other sites

not surprising, its not 1 size fits all.

 

In my case and many others, Kronos refused to play nice with Mac USB audio. This went on for 2 years as more folks got puzzled ' WTH ".

 

The Korg theory was that Apple removed audio packets

It happened when Apple went from El Cap or thereabouts.

 

I bought a UR44 Ai in self defense.

 

I can't be certain, but I believe the audio got resolved later with Mojave . I would have to dig to get the exact

chronology. Again, it was us Kronos owners.

 

That was kind of the point I was making. That if you are going to try an external interface it has to be the right one or you'll have issues like you mention and/ or you might end up with less useable CPU cycles and worse latency. Or if you get it right you end end up with the opposite. More usable CPU cycles and lower usable latency.

Paul Najar

Jaminajar music production

www.jaminajar.com

Link to comment
Share on other sites

You guys are obviously way smarter than me about computers...but that's not gonna stop me!

 

Playing live with mainstage vs recording applications are two different things. I have a feeling that the two uses of VST's or AU's are getting confused in the argument.

 

My degree is in Accounting...not EE. But it seems to me that there is no true "round-trip" if playing live. Midi signals are input directly into the computer via USB, the VST turns them into digital signals that are sent to the I/O device via USB. Inside that I/O device they are turned into analog signals.

 

So while I'm well aware, versed, cursed and experienced with latency when recording, I rarely see it as a problem when performing with Keyscape, Pigments or Arturia DX7 Live. My problem has always been pops, hiccups and glitches.

 

I have used several different I/O devices over the years, and latency has changed as that hardware has changed, but with my 13 inch mac, nothing could get rid of the glitchy poppy-ness. M-Audio, Steinberg, Presonus and the onboard headphone jack all produced the same result.

You want me to start this song too slow or too fast?

 

Forte7, Nord Stage 3, XK3c, OB-6, Arturia Collection, Mainstage, MotionSound KBR3D. A bunch of MusicMan Guitars, Line6 stuff

Link to comment
Share on other sites

You guys are obviously way smarter than me about computers...but that's not gonna stop me!

 

Playing live with mainstage vs recording applications are two different things. I have a feeling that the two uses of VST's or AU's are getting confused in the argument.

 

My degree is in Accounting...not EE. But it seems to me that there is no true "round-trip" if playing live. Midi signals are input directly into the computer via USB, the VST turns them into digital signals that are sent to the I/O device via USB. Inside that I/O device they are turned into analog signals.

 

So while I'm well aware, versed, cursed and experienced with latency when recording, I rarely see it as a problem when performing with Keyscape, Pigments or Arturia DX7 Live. My problem has always been pops, hiccups and glitches.

 

I have used several different I/O devices over the years, and latency has changed as that hardware has changed, but with my 13 inch mac, nothing could get rid of the glitchy poppy-ness. M-Audio, Steinberg, Presonus and the onboard headphone jack all produced the same result.

 

I reckon you're on the right track here, but live and studio are not completely different. Live is a subset of recording when talking about triggering virtual instruments from a MIDI controller. To quote from my own earlier post I allude to what you're thinking. That triggering virtual instruments is not round trip, it is only the output side of that full RTL. it's not always half of it either. Sometimes it's a bit more that half the RTL sometimes a bit less.

 

I know with my own testing on multiple RME interfaces, and multiple different macs as far back as 2008 I can reliably get 3.0 - 3.3ms round trip including converter delays. This is as good as it gets at 44.1Khz. Youre getting even better performance if you dont plug a mic in for vocoder work etc and just play your MIDI controllers as its only the output latency that is in play. For RME thats below 1.5MS.

 

Your issues may simply be that you are asking too much of your system at the latency you wish to perform at. How much RAM? What CPU? What hard drive'/s have you got. What Audio buffer are you working at?

 

OTOH the audio interface makers you mention are not known for their low latency, low CPU overhead performance. So there may be some wiggle room it it for you but more info please.

Paul Najar

Jaminajar music production

www.jaminajar.com

Link to comment
Share on other sites

"The driver and related software are critically important to achieving good low-latency performance. Well-written driver code manages the systems resources more efficiently, allowing the buffer size to be kept low without imposing a heavy load on the computers central processing unit. The importance of drivers means its not possible to simply say that one type of computer connection is always better than another for attaching audio interfaces. Any technical advantage that, say, Thunderbolt has over USB is only meaningful in practice if the manufacturer can exploit it in their driver code."

Actually, I'm completely in agreement with that, and I thought I pretty much said the same in my last post, but maybe it got lost. Drivers are everything. USB drivers used to be a lot more prone to latency than firewire (and maybe thunderbolt, but I can't say I know much about TB), but beginning in about 2010, I think it was Apogee or maybe RME made some breakthroughs that decimated the trouble with USB audio handoffs, which is one of the reasons Firewire interfaces quickly became unnecessary, even before computer manufacturers started dropping FW ports.

 

That said, and I realize I'm making a big assumption here, the drivers required to stream data within the motherboard itself are likely a lot less complex than those that have to pass it first to a USB/TB/Fire bus, as they don't have to switch protocols along the way. Hardwired protocols typically don't require as much error correction either, since there's much less chance of packet failure from the physical connection. I'll admit I don't know much about internal mainboard-level data bussing, but I have to assume it's far more streamlined than USB, Fire, or Thunderbolt.

 

Anyway, I think we're actually pretty much in agreement about this, but it's complicated enough a lot of it is getting "lost in translation", so to speak!

 

Finally in your last paragraph you say the size of the audio plug doesn't inherently effect the audio quality. In a physical sense that is correct, however given that a three point mini jack can't carry 2 channels of balanced audio and rarely if ever are 3.5mm audio outs referenced to a +4 level there are clear electrical restrictions that relate to sound quality when comparing the mini jack to something like 2 XLR connectors or 2 x6.25mm TRS connectors.

 

Absolutely, and if I were doing long runs, I certainly wouldn't want to use unbalanced 3.5mm. However, active line level signals can go a bit of distance before they pick up any noticeable noise. Consider that even weaker, "instrument level" from guitar often runs 15-20ft unbalanced without any perceived degradation. Laptops are technically "headphone level", which are even hotter (too hot for most boards, which is why they should only be turned up to 75%), an 8ft unbalanced run is going to collect a negligible amount of noise at those levels.

 

But it seems to me that there is no true "round-trip" if playing live. Midi signals are input directly into the computer via USB, the VST turns them into digital signals that are sent to the I/O device via USB. Inside that I/O device they are turned into analog signals.

You're absolutely right, and I realized later that "round trip" was a bit misleading. There is a multi-stage trip, but half of it is a MIDI data stream. BTW: you'd be surprised how crummy a lot of USB midi data handling is. I ran some tests, and most drivers will "quantize" incoming midi to about 2-4ms. That means that if you play two notes 3ms apart, there's a good chance it will send them to the VI at the same time. The handling varies greatly by manufacturer. That said, the latency and overhead caused by MIDI input pales in comparison to CoreAudio imposed latency. And really, everything we've been talking about here is small fries compared to that.

 

Likely, if you're having audio dropouts and perceivable latency issues, it's a software DSP issue, not a midi or audio interface issue... the mainboard CPU. swapping out Midi or Audio interfaces is not likely to solve any major issues, though there is always a chance of bad drivers grinding your system to a halt.

Puck Funk! :)

 

Equipment: Laptop running lots of nerdy software, some keyboards, noise makersâ¦yada yada yadaâ¦maybe a cat?

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Unfortunately, your content contains terms that we do not allow. Please edit your content to remove the highlighted words below.
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...