I don't know if
@OCC7N is genuinely asking a question, or is he making a statement, I am going to assume he is genuinely asking a question.
The issue of latency recently came up on ASR with hugo2, and a member actually measured latencies of various DACs, including Hugo2.
On ASIO USB input , for stability, the tester had to use 128 frames buffer minimum (a very low figure) and his results were these:
You can see that latencies vary from about 15ms to 0ver 40ms for other DACs and Hugo2 being highest at over 60ms.
So if
@OCC7N needs no more than 10ms, USB is out of question! USB is Asynchronous and there is an ASIO buffer involved.
No luck there.
The tester, then, did another test with SPDIF interface. Optical (or coax) are synchronous interfaces, so there are no USB buffers involved. the results were:
Here you get no latency with chip-based DACs (Topping or RME), but Hugo2 is at 35ms.
Two reasons for this latency:
- Chord uses a small buffer on SPDIF to make sure it locks to the signal and reclock the signal (my understanding, correct me).
- Chord uses a long tap reconstruction filter, it has to look into the
future and
past to do its calculations to generate the inner samples, so it has to buffer.
The more taps, the larger the latency, as much as 660ms on mScaler, and 100ms on Dave.
Again, bear in mind that a software music player plus USB latencies together can be close to 150ms alone, on default settings. This is in addition to any latencies a DAC may have.
So, if
@OCC7N requires no higher latency than 10ms (for any reason) , then the answer is clear:
Use a chip-based DAC with SPDIF connection,
period.
Of course, this is not a shortcoming of Chord DACs.
Its like saying, A Porsche is out of question, I need to carry 2 tons of bricks with my car!
The Porsche is made for a different use.