Sadly many manufacturers using per se great ICs don't allow them to perform optimally because they just wanted to save a few $$$$ manufactruing costs.
For example the best audio CODEC IC is useless if there are power supply noise or sampling clock jitter issues. Of course some parts used in top notch instruments are custom designed, which contributes to high prices.ĭoes someone know if the current best off-the-shelf audio chips or chipsets, if sorted carefully in the lab, can match audio quality of high-end recorders if the overall design is done very carefully?Īs often mentioned, it's not just about ICs with amazing datasheet specs (measured under ideal test lab conditions with high-end equipment), overall performance is limited by the weakest points of the design. Older designs are possibly based on non-audio-specific semiconductors, typically parts used more for test and measurement equipment as well as some higher-end industrial (non-audio) instrumentation. My bet is that recent designs mostly rely on very good common off-the-shelf (COTS) audio chips, possibly carefully individually hand-picked based on extensive lab screening tests.
Still wondering about the probably-not-so-outstandingly-breathtakingly-magical ADCs and DACs used high end professional (not audiophool) equipment like (in no specific order) Sonosax, Nagra (Audio Technology Switzerland), Sound Devices, Aaton.
Now I could imagine cryogenically cooled electronics for audiophools. Noise levels will ultimately be limited by physical laws and the only way to reduce them would be to lower temperatures. Also it helps avoiding rounding error issues. Processing 24-bit data doesn't make sense, I expect it to be handled internally as 32-bit with 8 padded bits. Typically DSPs and algorithms are optimized to handle data of 2^n bits length. But there are still only a handful manufacturers, I mean not that many.ģ2-bit sounds (pun intended?) logical for computations, historically data processing is based on 8, 16, 32, 64.