So for me actually the worst problem currently is Kontakt. Some libraries perform really awfully on my machine, but also in the standalone version, so it is not a Cubase problem.
Apart from that, my ASIO problems are not spikes, but ASIO load in general. ASIO load is very volatile, it seems to be different every time I play a song. If you wanted to declare ASIO load as a physical measurement, you would have to specifiy a ± 30% error at least. Especially when jumping around in a track during playback ASIO load kind of builds up. It somewhat behaves like a big ship, once you have some speed in one direction it takes time to change to another direction.
I would not know how to describe that to the support. And if they would at all accept this as a bug or an issue. And of course they could say: buy a faster PC.
In the end I raised my buffer to 2048 samples. Then I can work with acceptable performance. But it is no solution really, because with such a latency I can't record fast notes from my MIDI keyboard because I get all thrown out of the rhythm if playback is so delayed. So I have to change the buffer settings depending on whether I want to work on the mix or record midi takes. This especially sucks as my M-Audio driver does not allow me to change the buffer while the driver is in use, I have to close Cubase to do it.
It's ridiculous, I was recording Midi with 5 or 10 ms latency back in the nineties, today I have to deal with 25 ms latency.
Things I hate about... - Cubase 7
-
- KVRAF
- 1758 posts since 15 Mar, 2013 from Germany
-
- KVRAF
- Topic Starter
- 14739 posts since 19 Oct, 2003 from Berlin, Germany
More than 25ms.
Actually, sample related instruments are known to cause issues with the ASIO Guard. Which is why I was constantly spammed with the info "turn off the guard for these plugins individually".
Might be a solution for you, still not acceptable though. Another hoop to jump through.
Actually, sample related instruments are known to cause issues with the ASIO Guard. Which is why I was constantly spammed with the info "turn off the guard for these plugins individually".
Might be a solution for you, still not acceptable though. Another hoop to jump through.
-
- KVRAF
- 1758 posts since 15 Mar, 2013 from Germany
ASIO Guard is off individually for all samplers that I use multitimbral, i.e. Kontakt, Independence, ... Really, I have also done a lot of research and optimization with my system to get better performance (to no avail though). There isn't anything I can further optimize.
BTW, ASIO Guard helps me big time with amp sims and the like. So it is indeed useful. Pity that it can't be used on multitimbral instruments.
BTW, ASIO Guard helps me big time with amp sims and the like. So it is indeed useful. Pity that it can't be used on multitimbral instruments.
-
- KVRAF
- Topic Starter
- 14739 posts since 19 Oct, 2003 from Berlin, Germany
Then I can only emphasize that once more:
File a bug report.
Can't stress that enough.
File a bug report.
Can't stress that enough.
-
- KVRist
- 259 posts since 8 May, 2010 from US-Washington
Long time cubase user here.
Cable channel - you might want to think about upgrading your m-audio card. They are not known for making the most bulletproof interfaces. Try something like RME or Lynx. I've been using an RME 9632 for years. Works great and latency is low. And RME has some of the best drivers out there.
Cable channel - you might want to think about upgrading your m-audio card. They are not known for making the most bulletproof interfaces. Try something like RME or Lynx. I've been using an RME 9632 for years. Works great and latency is low. And RME has some of the best drivers out there.
-
- KVRAF
- 1758 posts since 15 Mar, 2013 from Germany
Yeah I saw a ranking somewhere on the net, RME and Lynx were scoring >9 points, my M-Audio had 6.7 or so. So not as good as RME, but not too bad either. Its an 8 in 8 out with Mic Preamps, cost 500 Euros, and its only one or two years old, so I want to stick with it if possible. And if buying an RME would gain me only about 30% more ASIO power, that's not the solution of my problems.
-
- Banned
- 22457 posts since 5 Sep, 2001
[DELETED]
-
- KVRAF
- Topic Starter
- 14739 posts since 19 Oct, 2003 from Berlin, Germany
What saddens me though... I've read your posts over at the Steinberg boards, and nobody gave any brown matter.
It really looks to me now as if Cubase 7/Nuendo 6 is a complete failure (code wise, not financially it seems), and Steinberg wants to cover it up by remaining silent on a lot of ends!
But instead of lowering prices, they go up. Instead of actual fixes, only minor stuff every four months. Instead of finally confirming certain bugs, it's constantly a third party issue.
Seriously?
Regarding RME and being rock solid:
Yes, that hardware is rock solid. The drivers are also rock solid.
But... not if it's the most current ones, and not if we talk about Cubase 7. Because they seem to interfere with each other.
It really looks to me now as if Cubase 7/Nuendo 6 is a complete failure (code wise, not financially it seems), and Steinberg wants to cover it up by remaining silent on a lot of ends!
But instead of lowering prices, they go up. Instead of actual fixes, only minor stuff every four months. Instead of finally confirming certain bugs, it's constantly a third party issue.
Seriously?
Regarding RME and being rock solid:
Yes, that hardware is rock solid. The drivers are also rock solid.
But... not if it's the most current ones, and not if we talk about Cubase 7. Because they seem to interfere with each other.
-
- Banned
- 22457 posts since 5 Sep, 2001
[DELETED]
-
- Banned
- 22457 posts since 5 Sep, 2001
[DELETED]
-
- Banned
- 22457 posts since 5 Sep, 2001
[DELETED]
-
- KVRian
- 540 posts since 8 Jul, 2009
Cubase 7 works great here. Better than Cubase 6. I have to say that I work on a Windows system. And I don't use ASIO guard.
And I switched from a M-Audio firewire audio-interface to a RME usb audio-interface. That resulted in a more stable system.
And I switched from a M-Audio firewire audio-interface to a RME usb audio-interface. That resulted in a more stable system.
-
- KVRAF
- Topic Starter
- 14739 posts since 19 Oct, 2003 from Berlin, Germany
The thing is Theo, you're on Mac. I am on Windows (iMac 11.3 equivalent). That's a different engine and driver topography alltogether.TheoM wrote:that's weird with RME, but now it's causing problems with core audio itself, confirms to me something has changed with it's asio behaviour in version 7. But that said, it is not qualified for mavericks, and i only upgraded cause it's better on battery power for the laptop with the app nap (which is of course turned off for cubase). Oh i tested both 64 and 32 bits as well, no change.
From a friend (and client) I know how crazy or well a Mac can behave compared to Windows. And he's using the RME UFX (Firewire). Then again, he's using Logic, not Cubase. He thought about switching, but after reading along he's definitely not interested as of this moment.
Thing is, we seem to have different issues, but they ultimately bolt down to the same root.TheoM wrote:Steinberg DO care about fixing it as i have spoken to them and you are wrong about that and they know the issue exists. A hybrid buffer is a tricky thing to implement correctly and this is not just some arbitrary fix - it will be a big job to fix so i don't expect it before another 12 months to be honest.
In my case, the ASIO drivers and Cubase ASIO engine. Which in turn could mean that it's indeed the new "hybrid buffer" system they implemented, or ASIO 2.3 which both the most current RME drivers and C7 utilizes. And that (I was assured of that) is already forwarded to the German development section. But I have no results so far.
I'm still testing for them, which is one-sided. I wish I could finally get a definite answer other than "you're right, something is wrong - certain drivers interfere with C7, please roll back drivers as interim solution" or "we're working on it, but couldn't reproduce yet". Then again, I reasked again on Thursday - let's see what the next week offers.
But else... I'm on one of RME's most stablest systems: HDSPe. And the most current drivers from like last half to 1 year messes with Cubase? Uhm... okay~!
Standard procedure with a lot of developers as of late - "not our problem, it's your rigs fault". Can't hear that anymore.TheoM wrote:Emapple have never fixed it and blamed plug in developers, but obviously hybrid buffers are tricky to get 100% right.
BTW, I don't think that this issue will be fixed by C8 - they don't even admit it yet. And that means for me... sitting out C8, but maybe jumping again on C9 and paying 499 as Upgrade while still sitting on a non fluent working host (C7)!
As freelancer, that pisses me off big time. Because I invested a load of money, and I run into one issue after another. While working(!).
And here's the main thing that ticks me off: focus on Macintosh only.TheoM wrote:Steinberg are at least trying to improve cubase performance especially on OSX. Before AG it was a massive issue on mac and with it disabled, it is bordering on useless for many of us that have demanding projects. That's an osx core audio issue itself, and one can get away without hybrid buffers and staying on low latency in most cases on modern windows machines. I am sure they only did this to help mac users mostly and i didn't expect a bulletproof first revision of asio guard.
A lot of users have Windows engines. Yet it seems to me that the most issues show up on Windows. And here the standard answer is "your rig is too old, upgrade" or "it's not our problem, it's the fault by the 3rd party developer". Apparently, Macs also have issues, but I do not hear/read any such wording.
Looks to me as if Steinberg wants to abandon the Windows users again. For the third time by now (first with VST4 which was Mac exclusive, then with C4 IIRC - hints were there at Musikmesse back in the days, now with C7).
-
- Banned
- 22457 posts since 5 Sep, 2001
[DELETED]
-
- KVRAF
- Topic Starter
- 14739 posts since 19 Oct, 2003 from Berlin, Germany
I read that you had issues with an USB connected speaker system, but I had the impression that your RME rig runs fine.
Must have misread something.
Either way - still not acceptable for such a pricey host.
Must have misread something.
Either way - still not acceptable for such a pricey host.