What is the real benefit of Hi-Res Audio support?
About the Hi-Res Audio Certificate
Today, many products carry the high resolution logo, which stands for high resolution support. For sources such as portable players, sound cards, and USB headphones, this means support for sample rates above 44/48 kHz.
But if you don’t use special audiophile recordings in high resolution formats and you don’t touch on the controversial issue of audibility of frequencies above 20 kHz, is there any benefit to this support for normal use? For example, when watching YouTube videos or sound in games?
As practice shows, there is also a very notable one.
Hi-Res Audio certification is awarded to devices capable of reproducing sound at frequencies above 20 kHz. For headphones with an analog connection, this certificate is advertising tinsel, because all headphones are capable of reproducing frequencies above 20 kHz; only some models play them silently, while others are very quiet. The criteria for the threshold of this “silence” does not have a Hi-Res Audio certificate (or supposedly exists, but is not subject to disclosure). Consequently, absolutely any headset can get it.
For digital sources, the reproduction of frequencies above 20 kHz is dependent on the sample rates supported by the DAC, and consequently all devices containing a modern DAC are Hi-Res Audio certified.
The system mixer is to blame.
The fact is that in modern devices, both on a computer and on a smartphone, all sound passes through the system mixer. He is in charge of mixing all the audio streams of different programs. All separate sounds from YouTube, video player, Skype, music player and other programs need to be converted to stereo broadcast.
The mixer always runs at a specific sample rate.
All incoming audio streams are converted to the frequency at which the system mixer operates. The higher the frequency, the less distortion will go into the audible frequency range.
Where does the distortion come from when the sample rate is increased?
It’s like taking a photo of a checkerboard and zooming in slightly from 8×8 pixels to 15×15 pixels. Obviously, with multiple magnifications, you can’t just double pixels to preserve the original image. And in a multiple magnification, there will be a question, the new pixels should simply double the adjacent ones or contain an intermediate color between the original pixels.
The best option will depend on the type of image. But the higher the resolution of the final image relative to the source, the less visible the artifacts of resizing will be.
In our case, as the pixel resolution increases, each pixel will be smaller. Those. By increasing the image in pixels, we are essentially increasing the pixel density for the same visual image size.
Similarly, with an increase in sample rate, from an increase in sample rate, we do not get fundamentally new sounds, tonality measurements, or playback speed. But at the same time, by changing the sample rate, we get additional distortions in the sound. The higher the sample rate of the system mixer, the more distortion will fall in the inaudible high-frequency range.
If there is only one sound reproduction source, then the system mixer is not needed in the signal path. But for the stability of the whole system, it processes the audio stream regardless of whether the sound is reproduced by only one program or ten.
For those who want to listen to high-quality music, it makes sense to use the sound output bypassing the system mixer.
This is compatible with some Windows and Android players (and professional sound processing software). This is not possible in games, browsers, or instant messaging. For the Android operating system, the RAA project conducts separate tests for software players, identifying players with optimal settings and smartphones on which it works.
In games on low-power systems, excessively high frequency can reduce overall performance; Here it is worth making a reasonable compromise between quality and performance (if possible).
Sound enhancement at high sample rates
Quality can significantly depend on the conversion algorithm.
