Hi Soekris
That means is to use FPGA to convert file format to PCM then decode? This is the only way to do ? Can not be native decode the dsd format? Ok if yes , is possible to upsampling to DXD format first to decode (optional choice on firmware).
Thanks for your prompt reply.
DSD is a one bit delta sigma modulated signal at 2.822, 5.644 Mbs or higher, afaik, you just feed it do a 50-70 Khz lowpass filter to get the analog signal back.
On a parallel DAC like mine you have to do an equivalent digital lowpass filter, then feed it in parallel form to the R-2R network, not really any conversion as such....
On a parallel DAC like mine you have to do an equivalent digital lowpass filter, then feed it in parallel form to the R-2R network, not really any conversion as such....
Hi Soekris
If I am nit misunderstood that means we just only need a digital filter to get analog signal to feed in the r2r dac only? If yes this is a extra benefit to all user .
Thanks
No, DSD is digital serial signal, a digital lowpass filter in the FPGA convert it to a PCM digital parallel signal. See Direct Stream Digital - Wikipedia, the free encyclopedia for more info on DSD.
Great design. Please sign me up for one. Also, is there any provision onboard for a direct USB input from a pc?
I wanted to limit my workload, already having the STM32 and FPGA firmware, so USB input will be by separate board, any with I2S output will do, like the XMOS based ones, got a couple here for testing.
If we connect an external analog stage to J7, the DAC would need to drive onboard opamp and external analog stage concurrently. Would it degrade the sound? Should we add a 0 Ohm resistor or jumper just before the opamp so that we can disconnect it when using external analog stage?
Attachments
Hi,
I personnally don´t believe in a gross sonic superiority of R2R DAC-Cores, as in our experiments they performed no better against bypassed Sigma-Delta-DAC-Cores (PCM1704x8, MSB-DAC against PCM179x).
I don´t know of a valid comparison as comparing DACs doesn´t mean to compare the DAC-Cores only, but also different digital frontends, different supplies and different analog stages at the same.
I rather believe that the surroundings are more important for the acoustic end result.
Especially the digital front end side can make a difference.
Standard oversampling to 192kHz is certainly efficient regarding Hardware effort, but the required uneven upsampling is imho far from optimal.
Why trying to extract Audio bit-correct from CD to a HDR if complete frames are dismissed when 44.1kHz synchronizes to 192kHz (factor 4.353741497......)
Which means You are doing ´linear´, or better said ´even´, upsampling (times 1, 2,4,8, etc.) in the FPGA and run the DAC-Core NOS?
That´d be great and I´m certain that it´d result in considerable sonic improvement, as it also improves the outcome with Sigma-Delts-Cores.
If You do it rather the common way of standard uneven upsampling to a single clock rate, You may waste a thought about the alternative 😉
jauu
Calvin
I personnally don´t believe in a gross sonic superiority of R2R DAC-Cores, as in our experiments they performed no better against bypassed Sigma-Delta-DAC-Cores (PCM1704x8, MSB-DAC against PCM179x).
I don´t know of a valid comparison as comparing DACs doesn´t mean to compare the DAC-Cores only, but also different digital frontends, different supplies and different analog stages at the same.
I rather believe that the surroundings are more important for the acoustic end result.
Especially the digital front end side can make a difference.
Standard oversampling to 192kHz is certainly efficient regarding Hardware effort, but the required uneven upsampling is imho far from optimal.
Why trying to extract Audio bit-correct from CD to a HDR if complete frames are dismissed when 44.1kHz synchronizes to 192kHz (factor 4.353741497......)
From this quote in #34 I assume, that You switch the clock rate of the DAC, depending on the clock of the incoming signal.
Which means You are doing ´linear´, or better said ´even´, upsampling (times 1, 2,4,8, etc.) in the FPGA and run the DAC-Core NOS?
That´d be great and I´m certain that it´d result in considerable sonic improvement, as it also improves the outcome with Sigma-Delts-Cores.
If You do it rather the common way of standard uneven upsampling to a single clock rate, You may waste a thought about the alternative 😉
jauu
Calvin
Me too for 0.002 or 0.001 🙂
Firmware is crucial to implement a good filter, which is really what makes a difference. See the Chord FPGA for example.
-R
Firmware is crucial to implement a good filter, which is really what makes a difference. See the Chord FPGA for example.
-R
I think I need some update on implementing filters using FPGA.. 🙁
Anyone a good suggestion on what to read?
Calvin,
We can agree that the best Sigma-Delta DAC's are not bad at all, but I'm in pursuit of the absolute best, like so many other here 🙂
The R-2R DAC don't really care about the clock frequency, so any oversampling will of course be even, no reason to mess with things, but as it also will be very configurable you can select any even oversampling rate you want (up to hardware limits), or none....
We can agree that the best Sigma-Delta DAC's are not bad at all, but I'm in pursuit of the absolute best, like so many other here 🙂
The R-2R DAC don't really care about the clock frequency, so any oversampling will of course be even, no reason to mess with things, but as it also will be very configurable you can select any even oversampling rate you want (up to hardware limits), or none....