If you are looking for Technical Support, please refer to the 3 articles below.
Managing High CPU Usage:
Purchasing a new computer:
Optimizing your OS for Audio:
For a synopsis on improving your computer performance, please read the article below.
Studio One: how to get the most out of your system performance
This article aims to provide insights into how system components, tweaks, and your song setup can impact performance. It does not go into detail on tweaks or performance settings themselves.
Obviously, the CPU and other components affect the performance more than anything else. Generally, a modern CPU with many fast cores is usually going to offer the most performance. Studio One supports full multithreading, which means it can use any number of cores which the system has to offer, including dual Xeon CPUs. Some rules of thumb:
· Core speed is essential. The faster the base clock speed of a CPU, the better it will be for low latency (recording) performance.
· Pick a CPU with a fast base clock. The Turbo Speed indication of an Intel CPU can be misleading, as a CPU with a low base clock but a very high Turbo speed may still be weak for mixing. Intel CPUs work with so-called Turbo bins, which specify at what speed cores run. The highest Turbo speeds are only available on 1 or 2 cores at the same time. DAWs typically trigger the “all core Turbo” mode, the lowest Turbo speed, but on all cores. So, try to find a good base clock (> 3.5 GHz) with a high Turbo value, and you should be good. Newer CPU models will still have a speed on all cores which is close to the Turbo Speed if the UEFI is configured correctly. AMD CPU documentation is more direct regarding their speed.
· As stated above, many cores will allow larger songs. But due to the nature of multithreading, more cores typically also mean lower core speed, as CPUs have a maximum heat capacity which needs to be balanced. If lowest latency is essential, higher core speed is preferred. If you tend to have many tracks (> 50) but not necessarily many plug-ins per track, many cores may be an advantage.
· Still, higher core speed has the larger benefit: it offers the potential to handle more channels on a single core, while more cores cannot “bundle together” to handle tasks of a single heavy channel. Audio channels cannot be parallel processed. Also, with low buffers, the performance draw can be very dynamic, so processing overhead is required. Only high core speeds offer this.
· If only one core is overloading, meaning it cannot perform its tasks in time, it will cause a dropout, no matter how many other cores have nothing to do. The buffer will not completely be filled with processed audio as it simply was not finished in time. This is why the Task Manager in Windows can say “10% use” but your audio output sounds completely distorted due to dropouts.
· Consulting audio benchmarks from Scan UK or dawbench.com can provide insights into performance differences between CPU models and generations. Synthetic benchmarks are not a good measure as they reflect rendering performance, not audio realtime performance, which is very sensitive for any core overloading.
Does RAM improve performance?
It depends on your system. Until a few years ago, RAM speed did not matter at all. However, some Ryzen generations have better performance with faster RAM. Intel CPUs also started to benefit from faster RAM. It is recommended to use the fastest possible supported RAM (DDR5 if supported) for recent CPU generations to get the most out of your system. For those using a lot of large sample libraries, using DDR5 is a significant improvement on any system supporting DDR5.
The amount of RAM does not directly have an impact on performance. However, if you do not have enough RAM, your system will bottleneck and start using drive space as virtual RAM. This can slow down the system. Look at Windows Task Manager or Mac Activity Monitor to see how much RAM you use while your song is running.
Regular audio tracks and software instruments do not use a lot of RAM. However, sample libraries such as for Kontakt, Opus or Vienna Instruments/Synchron Player can use large amounts. So, make sure to estimate how much you may need. Consider 16GB a solid starting point if you only use a handful of normal libraries, but 64 or even more if you are doing large orchestral mockups. Look at your system RAM use when loading your largest songs and you will get an idea.
In the end, the buffer is where the limit of your system will show itself. When the buffer cannot be processed in time by the system, a dropout will occur. As stated above, only one core is required to cause dropouts. By using larger buffers, the system gets more time to process. So, instead of just using the CoreAudio (MacOS) or ASIO (Windows) device driver buffers, many DAWs, including Studio One, use extra software buffers. In Studio One, this is called Dropout Protection.
Large device buffers cause high latency during recording, which is a problem when monitoring through the software. The delayed monitor signal is too far off, and the performance is compromised. Dropout Protection offers an extra playback buffer, while the device input buffer can be set low. Monitored inputs with Green Z active (which means when the Dropout Protection buffer is higher than the device buffer and the “Z” symbol on the main bus is clicked to turn green) only use the device buffer, while all playback tracks additionally use the extra Dropout Protection buffer. Set a low device buffer of 64 or 128 samples and a high Dropout Protection value (high or maximum) and you will have optimal performance. When mixing, you can raise the device buffer to 1024 as well, which in most cases will offer the most power. Avoid the 2048 samples device buffer, as with most audio interfaces, it will give worse performance. Try it out yourself with a song, what settings are optimal for your setup.
Macs have pretty good power settings, there is not much to do. In system settings, there is a “Battery” tab on Macbooks, which has an impact, depending on whether you have a power adapter connected. Set the “low power” option to never or “only on battery” if possible, for optimal performance.
On Windows, it has become harder over the years to get good performance.
In short, please refer to the articles at the end of this article for optimal settings.
Power settings are handled both by the UEFI (BIOS on older systems) and by Windows itself. They sometimes are limited in settings or the actual response to these settings, especially by off-the-shelf systems and laptops. Energy saving is a good thing to do, and Windows systems typically are set by default to do this. However, realtime audio is extremely sensitive to interruptions, so use the above links to get the best out of your system.
In modern UEFIs, it depends on the generation and motherboard what settings should be used. Disabling C-States (energy saving modes for thr CPU) usually improves realtime performance.
However, do make sure to set your Windows power scheme to “High Performance”.
Also, make sure to disable the energy saving on your USB hubs in Device Manager.
Many users like to use laptops for mobility. However, please note that they do have significant drawbacks:
· They are not as powerful as their desktop counterparts. Even with the same CPU model number, laptops have less potential power due to their lower TDP rating. This means they need to slow down earlier to avoid overheating. Desktops can provide 95 Watts or higher, while laptops typically have limits of 25 or 40 Watts.
· This also goes hand in hand with fan noise. Significant power requires significant cooling. So the fan will turn up, and it is right in front of you.
So, unless you need to be mobile, consider using a desktop system instead. You get more power for the price and you can more easily make it silent with decent coolers and fans.
Apple Macbooks with ARM CPUs (M1, M2, M3) are not affected by the above points, as the ARM chips are designed for mobile use and hardly get hot. They have almost identical power compared to the desktop units using these chips. Once Windows laptops get ARM chips with similar power, they will most likely be a good alternative.
There are high power Windows laptops available, in this case it is highly recommended to purchase these at a pro audio DAW builder. Only then you will be sure they have a solid realtime audio performance, and you will not need to do any configuring or tweaking yourself. Of course, such companies also offer desktops.
While hardware and OS settings have a strong influence on performance, there actually is another area which has possibly even more effect on the performance of your system: how you set up your song! This is good news, as it offers flexibility to get more headroom for your songs.
As stated, Studio One is a multithreading application. Roughly, each channel with its inserts is handed to the OS for processing, which will hand the processing to a core. This is done for each track, until all cores are used. Then the OS will typically assign the next track to the core with the most headroom, and so on. This is done on a per-channel basis, as the inserts on a channel are a serial process: insert 2 needs to wait for insert 1 to be finished.
But what about FX channels, which run in parallel? And bus inserts? And master bus inserts? Indeed, this is one of the major reasons why many users run into performance issues.
FX and bus channels are fed by audio or instrument channels, so they must wait until these channels have finished processing. And only then the FX and buses can be processed. Even if they are processed on another core, they still must be finished within the same buffer. So, FX and bus channels with many and/or heavy plug-ins cause additional stress. By planning how you can set these plug-ins up and whether you need them, you may be able to optimize your performance. In some cases, it may be more effective to add a reverb or delay directly as an insert in a channel, using its dry/wet parameter.
Master bus inserts are “the worst” performance killers because they get added on top of the channel, FX channel and bus channel consuming the most CPU. It is like always standing in the wrong line at the supermarket, which takes the longest. That is exactly how the main bus inserts are processed.
So, if you like to use really good mastering chain plug-ins, with saturation, non-linear EQ, multiband compression, glue compression, limiting, loudness maximizing and some other fancy processors – all with extra oversampling of course (which more or less increases the load of that plug-in with the factor of the oversampling), you are inviting performance issues, especially if your song load is already heavy for your system.
If you bypass the master chain during recording and arranging, possibly also during mixing, you will both save CPU but also tend to create a more balanced mix, as you will not have plug-ins automatically compensating peak dynamics for you. Switch them on regularly to listen to how it sounds “mastered”, then continue.
Also, when you are satisfied with certain tracks, you can transform the track to rendered audio, saving CPU.
How to solve a performance issue?
If you hear dropouts or crackling, look at the Studio One performance meter. It gives you the best indication whether it is a performance issue. If the red light peaks, it is a performance issue. It means at least one core is getting too much load. Now, verify the following:
· First: if you use a laptop, make sure the power adapter is plugged in.
· Make sure Plug-In Nap is activated, so inactive plug-ins do not consume CPU processing.
· Then, bypass all master inserts. Verify the load. Then, bypass bus/FX channels.
· Or deactivate the plug-ins with the highest load; you can check the “Devices” checkbox in the performance meter to see all plug-ins. Click in the CPU tab to sort to the highest load.
· If this does not help, or the system should be able to handle this song, verify that the High-Performance Power Plan is activated (Windows)
· Verify that no USB devices can sleep (Windows, Device Manager USB devices)
· Deactivate channel insert chains one by one. It may be a single plug-in causing mayhem.
· Also verify hidden channels!
· If you use many UAD2 DSP plug-ins: only their DSP is processed on the hardware. The handling of these plug-ins also cause a load on the system, and this all is done on one core. So if you have >100 UAD plug-ins, this core may cause performance issues.
Some additional tips:
· Don’t install all kinds of tuning and tweaking tools, unless you really know what they do and that they actually do improve audio performance. They can do more damage than good.
· If you buy an off-the-shelf system with much software preconfigured, consider installing a clean Windows version and only installing the drivers for that system. All this so-called bloatware often compromises the performance of your system and even the pre-installed support tools from the vendor can negatively impact the realtime performance.
· Not all tweaks shown in YouTube videos are valid or recommended. Many of them do nothing, are outdated, or put your system at risk in some form. Do not disable standard services on Windows. By default, they do not interrupt or draw performance.