We have created a new survey that asks questions about the latency performance of your personal recording setup, your recording use cases and requests. You would help us a lot if you would take 5 minutes of your time to complete the survey.
I wrote something stupid in the survey…
I wrote that My Steinberg Interface at 256 has a higher roundtrip latency than my RME Fireface at 128. Of course its the other way round: The Steinberg has a HIGHER latency at 128 than the RME at 256…
I hope this survey will lead to improvements, IMO latency is the final frontier in recording through a DAW. My Skylake system is plenty fast enough for anything I throw at it when mixing with a 128 buffer, but recording/jamming with a couple of VI’s and virtual guitar amps is a problem.
I really need to be careful if I want to jam with 1 Pianoteq 6 instance, 1 Superior Drummer 3, 2 channels with guitar amps and a vocal channel strip with a low buffer(64/88.2 kHz gives the best results latency wise). I tried several audio interfaces, some have lower latency because they have faster AD and/or DA converters, but the stress on the computer is more or less the same.
Cubase can just about cope with such a session. Logic Pro X is a little bit better for this, but not much.
That’s… what ASIO Guard is. It’s been around since 2012. Steinberg isn’t behind in this area, Cubase was in fact one of the very first DAWs to implement the idea, while SO didn’t have anything like it at all until a few months ago. This is very, very far from being a “Studio One philosophy”.
The point of the survey is to get info on which configurations and workflows are working and which are causing Cubase to underperform. This info allows the development team to better assess what should be done to improve Cubase’s performance, and is extremely valuable.
Yes we all know what ASIO Guard is and what it does. However, S1 does it better at this point.
This survey would not show up if Steinberg felt everything was at an optimum level dont ya think? All of us would like to be able to run a couple plugins whilst tracking without delay. A 'separate buffer for input and output is all Im sayin. I did the survey, you’re welcome, haha.
I have given up on trying to track WITH plugins anyway. I use the MR Console software or Direct Monitoring and just deal with it. Playback is easily fixed with cranking the buffer.
I did the survey, so how do you know what the round trip latency figure is?
I actually use Adobe Premiere for voice overs and it seems to work differently to Cubase when recording.
I want to test recording with Cubase soon but I mainly just use it (cubase) for creating stems of VST instruments that I export into Premiere where I have captions etc., (pity cubase or Nuendo can’t do captions or I’d just use that).
No, it’s not. S1’s implementation on mac at least is terrible.
The low latency tracks work ok but when it shuffles it to the internal higher buffer, Cubase thrashes it for polyphony count.
I can prove this on 3 macs with 3 different interfaces… I am talking about Cubase getting DOUBLE the amount of fx and VI’s, not a small difference.
Not sure if it;s different on windows though.
Asio guard works very, very well… It just goes to show how some people don’t even understand what it is… saying, “cubase needs a dual buffer system” LOL!
Pro tools has it, Logic has it, Cubase has it, S1 was one of THE very last to have it.
Cubase’s issue is monitoring tracks under 128 buffer. That’s the issue, as others seem to be confirming all the time.
For some reason, S1 can play live tracks even at 16 or 32 buffer. I will admit that.
i’d be curious to exchange tests with you and see how we fare…
At 128 i can’t get S1 anywhere near Cubase’s performance for effect and VI instance count.
if there are NO tracks armed anywhere, i.e, everything is on the internal higher buffer, pro tools beats them all for me, including cubase.
It’s the only DAW where I can literally see the real cpu usage of each core at 90% and everything spread to perfection. It all changes if i arm a track of course at low buffer, but for playback performance for me, it’s PT, closely followed by CB, then Logic, then reaper, then S1… Ableton/Bitwig/DP untested in regards to benchmarks.
For low latency performance, S1 seems to be quite decent with a quantum, Logic thrashes absolutely everything including S1 and reaper.
Interestingly though, Logic can only spread the live load amongst half the total cores… so if you have 8 cores for example, logic will only use 4 for armed tracks… so because of this, if you are working at buffers of 128 or higher, cubase and PT beats it for live recording performance as well.
but at 32 and 64 Logic thrashes everything, even many of the DAWs on windows. I understand Apple’s “logic” here (pun totally intended lol)…they split the load between live tracks and keep half the processing power for playback tracks… I guess in real world situations it works, i only broke it because i made unrealistic benchmarks (for example, I armed 64 external channels at 32 buffer and added multiple fab filter pro R reverb plugins on every single live track LOL… this is how i noticed logic could not go beyond 50% of the cores for the live monitored tracks).
I can play Diva at 32 samples and 96K in Logic, i can’t play it at 32 samples and 44k in cubase… just as a comparison…I need to go to 128 in Cubase…
As far as DAW’s go though, besides just a couple of tempo features of Logic, Cubase is so far ahead it’s not even funny. To me it thrashes Logic completely and actually has working PDC.
I answered to the survey. I made a mistake for the question ‘Have you experiment a better latency and on what system’ I should answer yes instead of no because I thought to an interface audio but in Pro Tools I experiment better performance for the latency.