Agreed. I run the Pi2AES as a bridge for my stereo. Seriously thinking about using a bridge for my desktop as well.
Ran across this the other day, his summary is that the Pi2AES measured better than a Denafrips Hermes and the only explanation he could come up with is that the Pi2AES never touched a USB connection that then has to be converted to I2S before the reclocker does it’s job.
EDIT. funny just saw @A_Jedi had shared this.
That’s how I’ve got my stereo connected at the moment.
So much work instead of just using a USB or Optical cable lol
Does the Pi2AES introduce noticeable audio delay when watching videos?
Man, I’m sittin here thinking the same thing! When I did the initial setup here on the desk, I used a crap USB cable and it sounded good. I quickly started improving things, including the investment into a 70 dollar USB cable. And I had to go with 3M length to properly place everything. Well I’ll be darned, that cable did make a difference. I improved the cabling throughout the system, and I have had a good result.
Having said that, I will not be doing this exercise of eliminating USB in the chain. Not anytime soon, anyway! For me, I don’t think my ears will benefit from the extra work and expense.
It’s not about the distance. It’s about removing USB from the chain. The wireless has error checking unlike USB and there is no cable to potentially degrade sound. PI2AES has really good clocks onboard. See the video I posted.
Not sure. I only use it for audio.
Same here. Will likely get another one soon.
Kinda sorta.
Depend entirely on the protocol used in the network stack.
Only has one job, converting the files it receives over wifi into AES. I hit play the music starts to play.
What software are you using? Volumio, Ropieee, something else?
With anything you are likely to use for streaming, it’ll be on top of TCP, or UDP with correction logic (CIF) so yes it’s error corrected.
I think posts like this just confuse the issue, while yes raw Ethernet, IP and UDP packets have no error correction, in this case the poster is factually correct. Though in this case it’s likely few if any errors are being corrected anyway.
Just switched to RoPieee because it had the native drivers for the Pi2AES. (for whatever that’s worth) I also wasn’t using Volumio as a player so figured the RoPieee client is easier on the Pi resources. I didn’t notice a difference, but hey, theory. lol
I run both (2 of each as it happens), I keep meaning to move everything to Ropieee, but only because it’s lighter weight
Thanks for the confirmation. I’ve been genuinely impressed with the Pi2AES. In the process now of building a Roon Core dedicated PC.
Stripped the Windows 10 OS of EVERYTHING it didn’t need. Don’t even have the onboard audio enabled. Removed ports and even the wifi board and so the only thing Windows is doing is running Roon. Still not up yet, moving things around and I’m having a hard time getting RDP working so I’m forced to physically switch machines for the time being. I’ll figured it all out eventually.
Pretty sure most streaming services (VoIP, IPTV, nearly all computer games) just drop faulty packages.
Honestly, cable talk on this forum never stays on subject for more than 2 days.
Can I use the AES BNC output to connect to the Coax input on my DAC or should I just use the SPDIF Coax output?
Yes most streaming video services will choose what they correct.
Most audio streaming protocols, airplay, DLNA, RAAT are going to correct errors, because they are designed to run on reliable networks.
In terms of audio streaming services, I know Amazon and Qobuz and I’d expect all the others to use reliable protocols with a large buffer, latency isn’t critical, and you need the data intact.
Games do what they have to, TCP/IP doesn’t make sense for latency sensitive communication over the internet, the way the early Online Madden games worked was utterly insane, and they relied on correcting what was sent, usually by sending it more than once to reduce the chance of an error. Most modern network games aren’t really streaming in any useful sense of the word, they are using a combination of projection and just correcting errors when they get data.
I also looked at the Raspberry pi.
The idea is great and everything.
Nir what keeps me away is just the supposed power supply it has and is supposed to undo the whole thing.
If you dig a little deeper on the net, you’ll find a bit about it.
And honestly, if I have to spend money on a better power supply again, I’m not going to buy it.
Then I can just go for an audio interface from Singxer or Matrix Audio.
And I can be sure that everything is clean.
The 20-30$/€ would then no longer matter to me.
I would have the possibility to connect I2S and will think about it again, but I still have some doubts whether it really brings something.
AFAIK Airplay relies on RTP for data transport. RTP is stacked on top of UDP, so no error correction.
DLNA is a mess on all fronts. If the “certification” reflects the software side of it, I would guess it just relies on hope & prayers.
Streaming is sending data that is not written to disk, so it is streaming.
That is what I find so funny about this “cheap does it”-trend. A good power supply costs a bit. So, yeah you saved yourself from the evil USB clock jitter (which should be a non-issue to begin with) but now have the cheapest 5V PSU china could make in your chain.