Hi stijn, I'm interested in using the HQPlayer NAA setup too, using something like a BBB to get header pin output to avoid using USB. What software are you running on the BBB? Are you using an external clock/isolation interface such as the Acko SO3?
I've been wondering if it is possible to add Signalyst NAA to my existing BBB/Botic but miero isn't sure it will work.
Ray
IS2 is a big part of what makes the Non-dac dac work (tick).
That's what makes the idea so simple, yet so exciting.
Basically treating (using) the digital domain as analogue…
I like the idea so if there's a way to do the same with the Raspi and get the pure DSD straight from I2S, I would like to know how.
I read this interesting thread straight-through last night and am very impressed with the outcomes to date! Congratulations on your successes!
I too am interested in what can be accomplished with headless system-on-chip development boards and am messing with a re-clocked BBB. For each of us with a bit of thirst for innovation, we focus on what we perceive to be the weak link(s) in the audio chain. I would greatly appreciate your perspectives on which link or links are most affected by this setup, and your conclusions or 'lessons learned' to date. Thanks in advance for your thoughts!
It might take some thinking 'outside' this thread's 'box' for me to include some of these innovations into my multi-amp analog setup.
One slightly OT comment if I may... I have noticed in the past that improving the performance of a DAC, by adding very clean power to the digital side, sounded exactly like a very different tweak in the analog chain - namely, treating transducer cones to cancel distortion. So I wonder how much the 'non-DAC' works by removing a source of noise, and how much it succeeds by better accommodating the needs/limitations of analogue transducers (and amps?).
Keep up the good work!
I too am interested in what can be accomplished with headless system-on-chip development boards and am messing with a re-clocked BBB. For each of us with a bit of thirst for innovation, we focus on what we perceive to be the weak link(s) in the audio chain. I would greatly appreciate your perspectives on which link or links are most affected by this setup, and your conclusions or 'lessons learned' to date. Thanks in advance for your thoughts!
It might take some thinking 'outside' this thread's 'box' for me to include some of these innovations into my multi-amp analog setup.
One slightly OT comment if I may... I have noticed in the past that improving the performance of a DAC, by adding very clean power to the digital side, sounded exactly like a very different tweak in the analog chain - namely, treating transducer cones to cancel distortion. So I wonder how much the 'non-DAC' works by removing a source of noise, and how much it succeeds by better accommodating the needs/limitations of analogue transducers (and amps?).
Keep up the good work!
DSD out of the BBB via I2S, a good place to start is here:
http://www.diyaudio.com/forums/twis...-botic-cape-beaglebone-black.html#post4272641
I'm sure the latest RPI has enough grunt to play DSD, I've just not looked into it yet.
@Ray, Sorry I refered to NAA in a general sense, I didn't realise it was a specific feature in HQ Player that was being refered to.
Last edited:
I don't feel your comment is OT at all. I was thinking the same thing.
Previously, in the DDDAC development we made big gains by using oversized unregulated (& seperate) PS's on the digital source side.
No worries Stijn.
I've played upto DSD256 via my BBB/Botic/Acko SO3 combo - a wired network connection was essential.
Ray
The Sonore uRendu will perhaps be a better solution.
I've searched and found people using it and I2S with PCM only I think.
It looks OK if you want a pre-packaged consumer product but I would opt for the BBB/Botic based solution (TPA or Acko) and pocket the extra flexibility/capability/leftover cash. Of course, YMMV
The Sonore web page says it only goes to DSD128; IMO that is the minimum DSD rate you should consider for this 'nonDAC' DSD playback approach.
Ray
I have now received the correct PCBs for my simple direct DSD filter and mute module for my JL Sounds USB board. I have made a couple up, minus the mute relay, and installed one for testing.
I have played a native DSD256 album (dsf files) via HQPlayer - sounded v good. Obviously DSD128 (converted off line from FLAC CD rips) played very well too.
Not surprisingly, my Atom D610 equipped computer didn't have the grunt to do any on the fly upampling from 44.1KHz FLAC files to DSD.
I installed a mute relay and silence - the outputs are continuously grounded. There is something very strange going on with the relay that I don't understand so if anyone can shine a light; basically, I can measure the voltage change across the relay coil and can hear the click of the relay but the contacts don't seem to change. Here's the circuit;
and I'm using one of these relays (4.5V coil, DPCO, non latching version);
http://www.farnell.com/datasheets/1683361.pdf
The pin annotations on the schematic relate to the relay pins.
With no music playing I measure 0VDC across the relay coil and 5VDC when music is playing, which is what I expect.
Ray
I have played a native DSD256 album (dsf files) via HQPlayer - sounded v good. Obviously DSD128 (converted off line from FLAC CD rips) played very well too.
Not surprisingly, my Atom D610 equipped computer didn't have the grunt to do any on the fly upampling from 44.1KHz FLAC files to DSD.
I installed a mute relay and silence - the outputs are continuously grounded. There is something very strange going on with the relay that I don't understand so if anyone can shine a light; basically, I can measure the voltage change across the relay coil and can hear the click of the relay but the contacts don't seem to change. Here's the circuit;
An externally hosted image should be here but it was not working when we last tested it.
and I'm using one of these relays (4.5V coil, DPCO, non latching version);
http://www.farnell.com/datasheets/1683361.pdf
The pin annotations on the schematic relate to the relay pins.
With no music playing I measure 0VDC across the relay coil and 5VDC when music is playing, which is what I expect.
Ray
The thing with the BBB/Botic approach is that these are still too general-purpose devices for me if you really want to reach great audiophile-quality results. The Sonore uRendu will be that audiophile-specific small circuit motherboard some of us look for.
Hoping one day Quad-DSD is available. That's the best I can get with my existing DAC in native mode.
That's not what it is for. Its maintaining the isolation between the two parts of the JLSounds USBtoI2S board. The codec trigger signal for the mute circuit is on the 'dirty' side of the board but the output mute is on the 'clean' side.
This may be an item on which people must agree to disagree. Stand-alone with generalized protocols, BBB is not impressive as an audio appliance. Implemented with audio-specific OS, clean clocks and power, plus isolated outputs - the unit gains much in terms of simplicity, clarity, (and similarity IMO) at about 10% of the cost of the Sonore uRendu. A full order of magnitude difference! Enjoy your choice!
I don't get that, the BBB is a general purpose device that has been repurposed for a very specific task, just as the Rendu will contain a general purpose processor that has been repurposed to a specific task. Yes the BBB/Botic is a little less elegant, requiring additional boards but that is what gives it the edge with respect to flexibility as you have fifo data reclocking available, you can experiment with different clock devices and the software is exposed should you wish to experiment. I use a BBB/Botic with an Acko SO3 reclocker feeding I2S (PCM and DSD) to a Buffalo IIIse DAC - the whole chain is run synchronously from high quality clocks on the SO3.
I've got no axe to grind with the Rendu, I'm sure it is a fine device, but don't underestimate just how good a BBB/Botic based implementation can be or dismiss it as general purpose.
I guess we'll have to agree to disagree as Francolargo suggests.
If you're interested in a neat integrated PCB option you may find the Edel renderer of interest.
Network Audio Renderer | ABC PCB
One other thing to consider with a DSD 'noDac' project is the player you'll be using. HQPlayer seems to be the top of the pile for DSD playback and upsampling options, optimally used with it's own proprietary NAA software on a remote renderer. Should you adopt HQPlayer I don't think you can use the Rendu as the NAA.
BTW, I do not dismiss the 'general purpose' criticism as invalid per se, and I thought of an example that contrasts 'general purpose' with 'specific purpose' hardware. When I was extracting I2S from a PCi sound card in a desktop PC mobo, it was always important to 'tune' the memory use. If audio buffers were too small/short, under runs (etc.) would create artifacts and DAC noise. If buffers were too large/long, signals were proportionally degraded in terms of 'general musicality'. Regardless of the software, the trick was to have just enough data buffered and no extra. The BBB is different. On the BBB I have messed with all kinds of buffer settings both in Squeezelite and MPD. I have listened critically and can find no difference in sound between the shortest possible and longest possible buffer settings. ...pretty encouraging, because it means that buffers can be large enough to accommodate all possible sample rates. Data in the lower range of rates can sit longer in the buffer without audible degradation.
So, I really look forward to reports concerning the suitability of the open-source SoC development boards as renderers of asynchronous DSD.
The uRendu will function as a NAA. It will run the same Sonic Orbiter software.
I do not want anything as complex as the BBB as an NAA: I want the simplest circuit that will do the necessary buffering and handling over of the data to the DAC.
Note here that I am talking of the microRendu (or uRendu), which is a forthcoming product and not an existing previous version of the Rendu devices.
It just is for me: for an NAA, there's a lot of stuff on the BBB motherboard that I don't want.
That's precisely what I am talking about: HQ Player on a beefy server computer and the simplest audiophile circuit as the NAA.
Nautibuoy, you must be confused since the uRendu (microRendu) will definitely have the ability to be used as an NAA with HQ Player.