Would forum members chip in $$ for a USB analyzer

[Miska]

2 minutes ago, jabbr said:

Right @The Computer Audiophile. Let's assume for the sake of argument, that the devices that claim to "clean up" the USB signal, do just that. Ok, so what effect does this really have on the DAC is the question that interests me more. This then raises the question: how is the DAC measurably affected such that an "improvement" is measurable?

 

...that's why I always prefer sticking to measure the DAC output instead of any intermediate step... I can replace things from DAC upstream and see how it affects (or not) things that come out from the DAC. That's in the end what matters. And quite many of the interactions actually happen inside the DAC, even though some upstream equipment's involvement is needed for that thing.

[Miska]

5 hours ago, Teresa said:

I agree with Speed Racer perhaps new DACs should move away from USB. Maybe something like Thunderbolt.

 

Ethernet?

 

[Miska]

8 hours ago, The Computer Audiophile said:

The DAC designers I've talked to about Thunderbolt all say it's a terrible interface compared to USB because it's so high speed that the noise is off the charts. 

 

Well, it is pretty much just PCIexpress on a cable. Companies like RME, Lynx and many others have been making sound cards that perform better than many audiophile DACs and sit inside the computer connecting to the PCIe...

 

Focusrite, Universal Audio and many others already have Thunderbolt interfaces for studios. It allows low latency and huge number of channels at the same time.

 

Of course audiophiles can keep worrying about the noise and all that, while not knowing that the music they've been listening to has been recorded using Thunderbolt connected ADC...

 

[Miska]

8 hours ago, jabbr said:

 

The Ethernet->I2S/DSD interfaces will generally run ARM perhaps with an associated FPGA (e.g. Zynq), but not necessarily (e.g. rPi). Or another SoC such as ClearFog (Armada). Lots of options. Each DAC needs an ALSA driver. A remote network protocol such as netJACK->ALSA or @Miskas NAA would connect. The interface would run some version of Linux, typically. Presumably one could do the same type of thing for Windows IoT.

 

Smallest device I run NAA on is 400 MHz ARM9 with 64 MB RAM. That is good up to stereo 768/32 PCM and DSD256. Power consumption is 750 mW peak.

 

[Miska]

3 minutes ago, The Computer Audiophile said:

 

Just because it's done doesn't mean it's done for the best sound quality. 

 

It also doesn't mean that the devices are technically bad... In fact, there are many great performing pro-audio devices out there.

 

Your USB and Ethernet interfaces are connected to the PCIexpress bus inside your computer anyway. So either USB and Ethernet are then bad too, or then if someone claims PCIe is bad they are incompetent in using it correctly.