Glassman
Headphoneus Supremus
- Joined
- Mar 22, 2003
- Posts
- 1,830
- Likes
- 11
Quote:
why are you calling that rolloff a problem? it is a design decision, half of the rollof is on output and half on input, when you do loopback on the same card it sums to some half dB at 20kHz which isn't noticable at all.. they are probably using slower filters like Bessel that have constant group delay and minimal phase shift but not that steep attenuation, that's why it starts falling rather early but goes down rather long up to the stopband, maybe that's the reason for the card to sound that good, using Bessel instead of Butterworth filters..
Xilinx is FPGA or CPLD chip, that means programable logic array, used mostly for signal routing and such, you can find it on most modern cards today.. ADC and DAC seems to be connected to that chip so I suppose it's kind of logic array..
| Originally Posted by richx Oh so the rolloff is still there? How bad is it when measured with "not-RMAA"? Thanks for showing/telling because I'm also considering the ESI Juli@ which doesn't have the rolloff problem. Edit: What would the Xilinx chip be used for huh? 
|
why are you calling that rolloff a problem? it is a design decision, half of the rollof is on output and half on input, when you do loopback on the same card it sums to some half dB at 20kHz which isn't noticable at all.. they are probably using slower filters like Bessel that have constant group delay and minimal phase shift but not that steep attenuation, that's why it starts falling rather early but goes down rather long up to the stopband, maybe that's the reason for the card to sound that good, using Bessel instead of Butterworth filters..
Xilinx is FPGA or CPLD chip, that means programable logic array, used mostly for signal routing and such, you can find it on most modern cards today.. ADC and DAC seems to be connected to that chip so I suppose it's kind of logic array..
