Can anyone explain the advantages of a "direct to I2S" USB DAC? I understand that with this configuration there is no conversion to SPDIF stage?
Can someone elaborate on advantages ?
Can someone elaborate on advantages ?
I guess you could look at this two ways.
1) Any conversion can increase the chance of distortion, so if its unnecessary, don't use it.
2) SPDIF is self clocking, it embeds the clock information within the one signal, with the data. I2S has seperate signals for clocks and data. Accurately rebuilding the clock from the SPDIF signal, to create an I2S stream, can introduce a lot of jitter into the system.
1) Any conversion can increase the chance of distortion, so if its unnecessary, don't use it.
2) SPDIF is self clocking, it embeds the clock information within the one signal, with the data. I2S has seperate signals for clocks and data. Accurately rebuilding the clock from the SPDIF signal, to create an I2S stream, can introduce a lot of jitter into the system.
The above plus :
"direct I2S" USB DAC:
PC usb - usb cable - usb receiver - i2s - dac
USB SPDIF DAC:
PC usb - usb cable - usb receiver - spdif transmitter - spdif cable - spdif receiver - i2s - dac
"direct I2S" USB DAC:
PC usb - usb cable - usb receiver - i2s - dac
USB SPDIF DAC:
PC usb - usb cable - usb receiver - spdif transmitter - spdif cable - spdif receiver - i2s - dac
So, why do DAC designers use USB to SPDIF to I2S? Because of stupidity and laziness i think.
The USB input is often an afterthought, an ugly additional convenience to a DAC to increase sales. As modern spdif receivers have several inputs and allow for easy switching and 95% of USB chips are PCM27xx which have SPDIF built-in, it's very tempting to route the spdif from the usb to one of the receiver inputs. This slightly simplifies the input switching compared to I2S where you have to switch 3 lines. One of the justifications for such an obvious compromise is that the I2S signals will be reclocked before the dac anyway, as if this automatically removes any trace of jitter.
Conclusion: if USB is important to you don't buy into such a compromise. Even better, avoid USB solutions which use the 27xx type of chips or isochronous chips in general.
The USB input is often an afterthought, an ugly additional convenience to a DAC to increase sales. As modern spdif receivers have several inputs and allow for easy switching and 95% of USB chips are PCM27xx which have SPDIF built-in, it's very tempting to route the spdif from the usb to one of the receiver inputs. This slightly simplifies the input switching compared to I2S where you have to switch 3 lines. One of the justifications for such an obvious compromise is that the I2S signals will be reclocked before the dac anyway, as if this automatically removes any trace of jitter.
Conclusion: if USB is important to you don't buy into such a compromise. Even better, avoid USB solutions which use the 27xx type of chips or isochronous chips in general.
I know I'm showing ignorance, but from a DIY point of view - can you not use some of those PCM27* chips as they output I2S as well as SPDIF?
For a while I've toyed with the idea of getting a cheap USB DAC, with one of these chips, and sniffing of I2S to feed my DAC.
Hurdles being: maybe they don't output I2S; I need to find a cheap USB DAC with a PCM27* fitted.
For a while I've toyed with the idea of getting a cheap USB DAC, with one of these chips, and sniffing of I2S to feed my DAC.
Hurdles being: maybe they don't output I2S; I need to find a cheap USB DAC with a PCM27* fitted.
- Status
- Not open for further replies.