1. Keep the operating system fresh with the number of running apps at a minimum
The more software you load on your PC, the slower it runs. Each process eats resources. You may think that it’s nothing for your latest PC with high-speed CPU and lots of RAM, but the truth is anything that’s running on your PC during audio playback has a bad impact on sound quality. The best thing you can do is to set up a dedicated operating system with nothing else installed on it except for a USB DAC driver and a software player or media server software (if it’s a music server). For Windows PCs it is also a good idea to make a fresh OS installation once a year or so as Windows registry gets bigger and fragmented over time which lowers performance as well.
2. Adjust latency in USB DAC’s control panel
This is one of the things that have the biggest effect on sound quality in a computer audio setup. Not all DAC drivers come with a buffer size/latency adjustment, but if yours does, I highly recommend setting it to the lowest value. The values are often shown in milliseconds, but can be also presented descriptively; e.g. normal, safe, minimal. You want the 1ms or the minimum latency setting. It will sound smoother, more dynamic, more detailed and with a better control.
3. Fine-tune software player’s settings
It doesn’t matter which player are you using, switching between various output methods (i.e. DirectSound, WASAPI, Kernel Streaming, ASIO) will produce different sound, even though all methods provide bit-perfect output when volume is set to maximum – I don’t recommend reducing volume in the system mixer or in a software player as this will lower resolution (unless JPLAY BitPerfect volume control is used).
The choice of the output method, as well as the buffer size (among other less significant factors) on a software player side, alters the sound. It’s not just about bits. I recommend Kernel Streaming as it is the lowest level audio interface available in Windows (for MacOS I will prepare a separate article). In JPLAY, KS is the most optimized output with the lowest latency and lowest CPU utilization. It offers the best sound too. You may find this article helpful if you use JPLAY.
4. Separate the OS from the media library
Having a dedicated drive (physical, not a partition) just for the operating system and essential software (drivers, audio player, basic utilities) is very important for the performance of our computer audio setup. This way Windows or any other OS – this is not a Windows-specific behavior – does not have to compete for bandwidth when accessing system resources and in a result runs faster.
Except for maximum OS responsiveness, which translates to a better sound, we also gain higher data security in case the system crashes. We can then perform a clean system installation, knowing that our music files sit safely on a different drive.
Media library doesn’t have to be on the same computer. It can be on an external drive, on a NAS or on another PC connected to your home network. I personally have a second drive installed in my PC for all my music.
5. Optimize OS
If you’re a skilled computer user, you may try tweaking Windows yourself: set power scheme to high performance, disable power saving features (for USB, network, drives etc), disable unnecessary services, disable unused devices, adjust visual effects for best performance, disable virtual memory (if you have at least 4GB RAM), disable logs & performance counters, clear autorun. There are literally hundreds of things you can do to make your OS faster and more lightweight with each and every tweak more or less improving performance. However, you really have to know what you’re doing as you may end up with your PC unable to boot. That’s why it might be a good idea to use optimization tools which do the job well and will save you many hours of work. Good examples of Windows optimizers for audio are:
– CAD scripts (Windows and Mac – FREE)
– Fidelizer (Windows – free and paid options)
6. Use wired Ethernet connection instead of Wi-Fi
Wi-Fi is not the way to go with high-end audio playback in mind. Using wired Ethernet connection between your network audio devices not only is more reliable – packets are much less likely to get lost – but most importantly achieving high-quality sound is easier, especially with audio-grade LAN cables and specialized network devices optimized for digital audio playback (e.g. NET Card FEMTO). 1Gbit connection speed is better than 100Mbit too.
7. Use high-quality power cord for your PC music server.
Treat your music server/computer transport as if it was a regular hi-fi component. It seems that many people forget that this is a part of an audio system and ignore all computer audio-related devices when it comes to cabling – they use whatever is available and expect great sound because it’s bit-perfect. But guess what? It won’t sound great, as any digital audio playback device wouldn’t sound any good with generic cables. Try listening to a high-end DAC or CD transport with a generic power cord – it will make your ears bleed. The same is with computer-based transports/servers or even doubly so because these things are very sensitive to all power supply related tweaks. As an audiophile, you probably have a spare power cord, but never thought of using it on a music server. Do it now.
8. Shut down your LCD screen during playback
LCD displays are incredibly noisy. The high-frequency noise they generate not only gets to your AC mains but is radiated from the screen. So keep your LCD away from your audio components and switch it off when music is playing. Now the funny part: a better power cord to an LCD screen makes quite a difference in sound too.
9. Make your system headless
If you’re using a dedicated desktop PC as a music server/computer audio transport, it’s a good idea to make it headless once all software configurations are finished and the PC can be controlled remotely from a tablet or smartphone. By headless I mean removing a keyboard, a mouse and an LCD display from the system – you leave only a network (LAN) connection and a USB cable to your USB DAC. This not only saves space but most importantly leads to a better sound quality, because the fewer devices are connected the more resources a PC has to focus on what’s important: digital audio playback.
10. Treat your PC for vibrations.
PC music server/computer audio transport being a digital playback device is as sensitive for vibrations as any hi-fi digital audio component such as Digital-to-Analog Converter (DAC) or CD transport. Thinking that PC is just a PC and whatever happens on its side doesn’t matter as long as playback is bit-perfect or not corrupted is wrong. It is one of the most common myths on computer audio. A music server is a part of a hi-fi system and must be treated for vibrations in order to achieve high sound quality. Place your PC server on a hi-fi rack or on an anti-vibration platform, install isolation footers underneath the enclosure or at least under hard drives and you will hear the difference.
Shop the full JCAT collection here
A recognized high-end audio consultant. Computer audio pioneer. Co-creator of the award-winning software, JPLAY. Manufacturer, curator, and distributor of high-quality audio products. Constantly strives for perfection of sound. Foundations for his breakthroughs finds in areas that others ignore.
This article is very useful not only for Jplay users, but also for computer audiophiles.
I have a question. The section 2 shows JLsounds’ control panel can adjust latency and buffer size.
JLsounds is my favorite DDC maker. But I can’t change latency etc. I’d like to know whether your DDC is special specification or not.
Thank you! JLSounds board with some modifications is installed in my APL DSD-MR DAC. The driver is licensed to APL (version 4.13). You might want to contact JLSounds if the ‘full’ driver is available for JLsounds owners.
Thank you for your useful information.
I could understand why JLSounds could use the full version driver.
Your information is very precise.
As you wrote, the driver is licensed to APL, not for JLSounds.
JLSounds tried to make a contract with Thesycon last year. But not yet.
Yes, that would be about right: last year I used Mystique v3 DAC from Mojo Audio for a while. It also had JLSounds USB input board, but latency/buffer options in the driver were missing and because of that I couldn’t use the most extreme output settings in JPLAY. I contacted JLSounds and the owner informed me that they will have a new driver with the options unlocked soon. Not sure if this is available already.
Best regards,
Marcin
The newest Mac Mini has 8GB of RAM, standard. How much RAM do you recommend for computer audio playback? I was thinking to add another 8GB of RAM, totaling 16GB.
Thanks in advance.
Hi Sean,
For computer audio playback I recommend 8GB RAM. It is optimal unless you use some heavy processing which may require additional memory.
Regards,
Marcin
Great tips.
In your USB DAC Control panel, you have aiso buffers set at 64.
Should this setting be highest possible or lowest possible.
hi Brian,
Thanks! It depends if you prefer high buffer or low. I prefer low, so I aim at the lowest possible latency.
Regards,
Marcin
To get the better sound quality in your computer you need to run only few apps while playing music otherwise your system would get hang in between of it moreover you can also use high configuration system so to avoid from this types of problems.
Exactly!
Great and useful tips. Thanks a lot for sharing. But in my opinion you have forgotten the most important tweak to do first:
Switch off “clock spreading” in UEFI/BIOS! By default it is activated on every kind of computer due to european EMV-regulations.
Oh by the way I am very satisfied by your SATA-cables I have installed some months ago.
Hi Carsten,
I’m glad you’re enjoying the cables! Yes, clock spreading should be disabled as well as many other features in BIOS, but this is a topic for another article 🙂
Best regards,
Marcin Ostapowicz
JPLAY & JCAT Founder
thanks for the information
A useful article – thank you. I’m just starting to try computer audio and want to understand the reasons for the BIG differences I am hearing. I have JRiver MC running on a Microsoft Surface Pro and also on a Raspberry Pi running JRiver ‘id’ (an idPi). The idPi sounds delicate & detailed but sterile, and high frequencies sound unnaturally ‘metallic’. The Surface Pro is overly warm and a bit congested, rather like compressed FM radio can be.
Both are connected by USB to the same Benchmark DAC2 – so why on earth do they sound so incredibly different. Neither is fully acceptable as they are. I prefer the idPi’s presentation but the top end is simply ‘wrong’. It also gives me the IMPRESSION that I am missing some mid and lower frequency information – rather like DAB discards subtle details as ‘noise’.
Hi Martin,
There could be many reasons, but the 2 are obvious:
1) different hardware platforms
2) different operating system
Both define the sound.
Best regards,
Marcin Ostapowicz
JPLAY & JCAT Founder
Hello Marcin,
I want to improve the audio quality of YouTube videos played through a PC. I don’t use the computer for streaming from Amazon or Tidal. I have a dedicated streamer for that. Using a JCAT network bridge care, can I stream the audio from Youtube playing on my computer to my streamer, then to a DAC? If not, what is the most cost-effective route to improve the Computer to Dac interface?
Thanks
Dan
Hi Dan,
If you have a USB DAC then the USB Isolator would be my choice: https://jcat.eu/product/usb-isolator-usb-audio-galvanic-isolator/
Regards,
Marcin