MDAC First Listen (part 00111000)
- Thread starter PhilCTTE
- Start date
- Status
-i have invested 1270£ into this project, if i get offered a full refund+interest i would probably take it despite needing an ADC.
-my plan was to focus on designing speakers.
-the ess version of mdac2 seams to be pretty close to ideal when it comes to audio performance, i only miss a few features
I have been thinking about striking a deal with lakewest audio to design my own dac similar to the mdac2 but i first need to start my own business and make some money.
btw: i generally ignore what people write about how electronics sound, when i read reviews i skip directly to the measurements section(if there is any).
On the matter of measurements, it should be said those of you getting a streamer built in are in for a treat.
I've recently discovered just how much of a PITA USB can be when it comes to getting ultimate sound quality, so the fact you don't have to deal with this is a big big plus.
Infact, I'd go as far as saying its possible some part of the anti-digital brigades complaints about digital audio could be partly due to USB....
I've recently discovered just how much of a PITA USB can be when it comes to getting ultimate sound quality, so the fact you don't have to deal with this is a big big plus.
Infact, I'd go as far as saying its possible some part of the anti-digital brigades complaints about digital audio could be partly due to USB....
To be fair didnt JohnW himself say "v1 will be with ESS silicon" ? With the option later to upgrade to the discrete DAC daughter board when its ready. Which seems sensible as a path.
sq225917
Bit of this, bit of that
"Any"? Not really. Some USb interfaces block 5v noise and are galvanically isolated, and some like the Mdac weren't. There are good and bad implementations out there, saying they're all shit is miles wide of the mark.
On a lighter note, my Mdac is fixed. Still has a crap USB interface though.
adamdea
You are not a sound quality evaluation device
JohnW
pfm member
Not 100% true
- the interface between streamer and Mainboard is I2S - but between CM3 and audio coprocessor its USB as the CM3 cannot support DSD or any data rate above 192KHz, but the "fatal" issues is that the I2S interface of the CM3 does not allow the DAC to be "Clock Master". All these issues are solved with USB.The issue with USB is not the interface protocol itself, rather the external noise impact to the system - with embedded "clock locked" USB this is no longer a concern.
The Digital and Analogue domains of the MDAC2 are electrically isolated - and with onboard single stage "Detoxing" I hope that the negative effects of external USB will have less of an impact.
Julf
Facts are our friends
The real issue is that USB was originally defined as a mismatch of differential and single-end signaling that makes it hard to fully isolate the connection.
fusion5
on the dark side of hifi...
IIRC a friend of mine has successfully slaved the CM3 to an ADAU1701 DSP
JohnW
pfm member
I doubt the CM3 is "slave" or the DSP is performing ASRC as the BCM SOC cannot support external I2S Audio Master clock. If I'm mistaken (and it be good to know the EXACT implementation) then still it does not get around the 192KHz / DSD limitation of the RPi I2S port.
Also a big factor is "Driverless" audio support which USB allows - I have been so burnt with Beagle-bone and the H3 board in the past.
Also a big factor is "Driverless" audio support which USB allows - I have been so burnt with Beagle-bone and the H3 board in the past.
What about get allo USBridge instead of RPi?
Would it be possible to just get it without the external hq psu? (saves money and i would not need to have an external PSU). Besides what we have been promised so far with the mdac2 i would also like to have High quality DSP and support for samplingrates above 768KS/s.
SInce i plan to design speakers i might actually need the fdac too, it's mostly the DSP i am interested in.
I might only need a DSP for the subwofer and if that is the case i can just use minidsp.
A=connect your dac of choice directly to the speakers
B=add the mdac2 to the loop(AD and DA conversion and level matching).
This allows you to test the transparency of the mdac2, it's highly unlikely it wont be transparent 100% of the time. I can construct a test like that once i get the mdac2 and the rest of the system.
SInce i plan to design speakers i might actually need the fdac too, it's mostly the DSP i am interested in.
I might only need a DSP for the subwofer and if that is the case i can just use minidsp.
Since the mdac2 also has an ADC it should be very easy to contstruct an ABX test where
A=connect your dac of choice directly to the speakers
B=add the mdac2 to the loop(AD and DA conversion and level matching).
This allows you to test the transparency of the mdac2, it's highly unlikely it wont be transparent 100% of the time. I can construct a test like that once i get the mdac2 and the rest of the system.
JohnW
pfm member
The MDAC2 cannot be operated without its PSU - the PSU is not an option. For those who can DIY a PSU then I can provide information, but its a 6/7 Rail PSU and thus far more complex then usual - also care must be taken in its construction as the Digital / Analogue domains are galvanically isolated.
The big step in sound quality comes when going from PCM to DSD - there's little sonic improvement going from 96KHz, 192KHz and then 384KHz - but a big difference when the Decimation / oversampling filters are removed from the signal chain (with DSD).
You keep pushing for higher then 768KHz but this is crazy (768KHz is already "madness"), I know of no recordings or even Audio grade ADC that samples this fast - forget DSP at this rate.
IMO, for sound quality only native DSD brings a us closer.
Hey John, so we've nearly come full circle. I was just looking back to when we first chatted back in 2008 over on DIY Audio. My Dacapo had just died and you described what the new DaCapo 2 would be like. From that moment it's been hook,line and sinker for me. The sound of the Dacapo was incredible and very, very special. The way it communicated music was palpable.
Just promise me two things, 1, No NIPPON CHEMI CON anywhere and No potted DAC
All the very best John, keep up the great work! I'm in it for the long run and it would appear i can wait(a little longer) for the FDAC/Dacpo 2.
Steven.
Just promise me two things, 1, No NIPPON CHEMI CON anywhere and No potted DAC
All the very best John, keep up the great work! I'm in it for the long run and it would appear i can wait(a little longer) for the FDAC/Dacpo 2.
Steven.
It's really not that simple. Sure go ahead and measure. Infact, on paper the current DAC I have is supposedly 10db less SNR than the previous, and yet here's me defending a discrete DSD implementation which sounds far more natural than its predecessor ever did.
- Status