In Reply to: RE: Does ASRC (upsampling) hurt pass band ripple in general? posted by dave789 on July 24, 2015 at 06:55:05:
Especially for that bit of "noise" shown in the response of the Cambridge DAC in the 350-2000 Hz range.
I don't see why the digital filter used in the upsampling process would be flat in the pass-band when the output sample rate is a power of 2 multiple of the input sample rate (oversampling), but then have ripple in the pass-band when the output rate is a non-integer multiple of the input rate (ASRC). It could be true, but I just don't understand what digital filter design method would produce that and why that method would only be used in the ASRC case.
I don't think we can conclude much from 3 samples, especially considering the iPhone is measured differently than the Cambridge DacMagic and the Benchmark DAC1. The iPhone has no power supply connection. Inherently, it has a floating ground, and in the test setup it's ground reference is provided by the probe of the audio analyzer. So there's no possibility of ground loops like there is with the other DACs. Also, the iPhone doesn't require an external transport. The others must have a transport included in the test setup. Finally, all three products have different implementations of volume control.
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- I'm not convinced ASRC is the culprit though - Dave_K 04:27:05 07/26/15 (6)
- passband ripple specs in +/- dB - dave789 06:56:42 07/26/15 (5)
- So? - Dave_K 10:03:35 07/26/15 (4)
- RE: So? - dave789 11:03:19 07/26/15 (3)
- Benchmark DAC1 has AD1896, and Cambridge DacMagic Plus has Anagram upsampler - Dave_K 11:31:49 07/26/15 (2)
- RE: Benchmark DAC1 has AD1896, and Cambridge DacMagic Plus has Anagram upsampler - dave789 20:09:24 07/26/15 (1)
- RE: Benchmark DAC1 has AD1896, and Cambridge DacMagic Plus has Anagram upsampler - Dave_K 02:28:36 07/27/15 (0)