Life after Yggdrasil: Watering the Ash

Thanks for the write up Darren. Your words sum up exactly what the Yggdrasil does for me.

 

Would you mind terribly to elaborate a bit on the AOIP product you are referring to, please?

 

I can't talk about it until it is announced. Hopefully that will be soon!

 

I wonder if any AOIP could sound better than a rpi with the 502dac shield. I plug in 3 wires, ethernet from a switch ,ifi power to the 502dac, and spdif (AES is available). Isn't this what the AOIP products provide, albeit with a USB initial hookup? I suppose a direct ethernet to dac might have possibilites but I don't know of any dacs that have ethernet inputs for audio.

 

I have no idea what "a rpi with the 502dac shield. I plug in 3 wires, ethernet from a switch ,ifi power to the 502dac, and spdif (AES is available)" is, but my AOIP setup has no need for any connection to USB.
Nor is there a switch as there is a direct connection from my Mac to the RedNet3, albeit thru a pair of FMC's (Fibre Media Converters).

 

No, they are different, in principle. With Raspberry Pi combo, the ethernet send music from your source (computer/NAS) to RPi via file transfer (just like you copy this FLAC file from one folder to another). In RPi, these files are decompressed, decoded to PCM signal. PCM signal is sent to 502DAC via i2s interface, and outputed to DAC via AES/EBU or SPDIF.

In AIOP, your source is responsible for all decompression and decode to PCM signal. PCM signal is transfered to AIOP devices via Ethernet with AES67 protocol. AIOP devices receive this signal and send it to DAC via AES/EBU or SPDIF.

Compare to RPI+502DAC or another music player (Auralic Aries), AIOP has advantages in flexibility. With the source is your computer, you can do many thing to your sound (upsampling and room-correction with resource-hunger filters,...), without being restrained by a relatively weak computing power of Pi.
About sound quality difference, I think it downs to implementation. In my short experiences with AIOP (rednet 3), a full force Pi streamer (not based one like RPi+502DAC+iFi iPower) can rival AIOP.

 

Yggdrasil give me goosebumps.

To be honest this audiophile hobby can be a kind of curse. I am somewhat envious of those who enjoy listening on their Beats headphones, and never had a thought about the gear. Unfortunately I can't help but hear digital sheen, glare, and fuzz. It annoys me, it distracts me, and I find myself listening to that rather than the music. The best compliment I can give Yggdrasil is that I find myself forgetting about it, and get lost listening.

I have a skeptical mind, so am skeptical of major improvements without explanation, but whatever Yggdrasil is doing, it's doing something different from all the CD player DACs, sound card DACs, even the Modi 1, even the Bifrost Multibit. It might just be a case of the whole is greater than the sum of the parts, but in total, Yggdrasil sounds musical. Could just be that combination of good power supplies, good re-clocking, a very good analog section, a filter that doesn't suck, the precision R2R DACs (once warmed up so the resistors are operating at spec), and maybe that balanced further reduces noise. Not sure, but it really does make a dramatic difference in the gear chain.

 

Me too, Mac--Ethernet--unDAES-O--AES--Yggdrasil.

JJ, what fibre-optic links are you using? I tried a pair of TP-Links running from my router/switch for Tidal streaming, but didn't have time to really evaluate before one died.

 

@gefski, when you say Mac to the unDAESO, is it a Mac Mini? Thanks

 

I've not had a chance to try Pi 2 Design's 502 DAC board, though I'll get around to it shortly, but at least compared to other similar setups, yes there are NAI (Network Audio Interface) and AOIP solutions that will outperform them.

No USB at all with AOIP/NAI solutions, which is part of the point. The originally these products were created to allow for freedom of placement (more important in their intended professional applications) and to support multi-channel/multi-device operation with both DACs and ADCs.

Merging NADAC, Auralic Altair, Auralic Aries Mini, Linn's entire DS line, PS Audio DirectStream all have Ethernet inputs and support either AES67, or some other industry standard.

 

No, it's a hand-me-down iMac from my wife. With A+, it sounds as good to me as my old Windoz/Jplay rig. Thinking about a Mac Mini with SSD though.

 

The Aries Mini is a streamer only though...?

There's been quite a few more (other than those listed by Torq) like the suave looking Resolution Audio Cantata and the Moon MiND, but I think most use UPnP style rather than AES67/Ravenna style solutions.

 

Nope, the Aries Mini also incorporates an ESS 9018 DAC with the expected analog outputs and can be used as a complete end-point feeding an amplifier.

 

Ack, I trusted their site:

Should have checked other links, but I didn't think a manufacturer wouldn't list that...

 

I trusted my memory, since I've played with the Aries Mini.

It's not in their specs, but it is mentioned in the basic product description:

"ARIES MINI is a wireless streaming node designed for connecting to your existing home audio system. It not only has shared all software and hardware functions coming from original ARIES, but also equipped with additional high quality analog output and optional hard drive slot for anyone whom don’t want to invest for a dedicate DSD DAC and NAS drive."​

 

I'm using MC-210CS TP-Link FMC's with the 10' dual fiber cable.
And I'm running the 'downstream' FMC on a 'spare' LPS (along with a fan) to eliminate yet another SMPS.

JJ

 

PS Audio DirectStream/Jr are releasing another firmware update like Torreys, which was described by @Torq in this thread. Huron details are here:
http://www.psaudio.com/pauls-posts/get-ready-for-the-weekend/

 

I'm glad they're still providing updates but I hate descriptors like "you can hear a whole additional octave of extension on the cymbals!" I mean, come on.

 

Wow, that just sounds bright to me. While it's nice to see a company continuing to support their products, doing things like firmware updates to tweak the sound endlessly is just confusing. Shouldn't they have gotten it right to begin with? It's like they thought the firmware was good enough and will polish it in later updates. It makes me lack confidence in their sound approach.

 

Every single time I build something, be it software or hardware, I learn something new that lets me do a better job next time.

With embedded systems especially, there has been more than one occurrence of me finding a more efficient way to utilize a platform that means I can push limits that previously were considered "fixed" or "maxed out". I'd rather take advantage of that when it occurs, than either never release anything (which is the ONLY way to be "sure") or say "this is as much work as I'm going to do, even if I find ways to do it better in future".

Chord takes the latter approach. PS Audio and Linn the former.

And while it's worth noting that not every firmware (filter) update that Linn has released has been positively received ... the audible and measured performance of the newer series of firmwares is notably different to that of earlier releases (which at this point spans TEN YEARS).

It's one thing if we're wringing more performance from a platform that lets us do a better job than before, rather than simply tweaking FR via hidden EQ. All the serious FPGA-based/updateable products I know of, which includes the PS Audio stuff, is pushing the hardware harder via progressive adoption of "learned tricks" and not just f'ing around with "tweaking the sound".

--

It's really no different to a company releasing a newer version of an existing amplifier. Why didn't they do it right the first time? If you keep taking what you learn as you build something, and apply that before you release it, you'll never ship anything (my perspective being that of someone that's been shipping major software and hardware for more than thirty years).