We are in agreement! Still, the OP’s computer seem (to him anyways) behave differently upsampling a FLAC and AIFF file respectively.
I cannot see any reason for that to cause audible differences either.
I believe the upsampling is more destructive to sound quality than any two lossless formats could ever contribute.
Sorry, I was trying irony…That obviously didn’t fly! 
1 Like
My apologies. I see it now 
Yes, your analogy is much like mine that my post could read differently on different people’s screens…
1 Like
I think I didn’t clearly make one of my points. The process of uncompromising a lossless file increases processor load, and may impact (since it is a dynamically changing ratio) the read into a buffer. None of these ought to be much of an issue, but they might. I will say its quite clear that as load on a streamer (PC/laptop) increases, the ground and power noise, rise, and this has a small but real impact on sound quality. I can speculate why, but cannot prove anything (i doubt anyone can).
So i would not entirely write off the potential for impact on a noisy or low-powered system.
I personally have no such issues, nor do i have them in my eval lab.
While unspoken, this is the implied reason that Roon suggests a 3-layer architecture too.
'later, done splitting nits. And nits they are - there are WAY better places to focus efforts.
1 Like
Huh? The USB music interface standard is Asynchronous, unlike the original profile (1?) in SPDIF. The receiving element (DAC) times the processing of the bits. So the USB segment itself is NOT about timing.
Edit/add: Although, going back and reading the entire original post with the above quote in context - i pretty much agree with the idea if not the specific wording…
1 Like
Quite agree - and yes, I am quite sure a noisy USB bus or other components can have an impact. The actual raw data file, not so much but then I am sure we agree about that also
And yes, 100% agree there are better areas to focus time and effort in chasing perfection; particularly with modern DACs etc.
I am correct.
You are correct too, in that USB Audio is isochronous which is about timely delivery of data. But that is a QoS (quality of service) mechanism, ensuring that if there is a lot of other traffic that wants to go over the USB cable, Audio can get guaranteed bandwidth.
But that has nothing to do audio sample timing, where USB is asynchronous.
USB Audio is both isochronous and asynchronous, but those words are not in conflict, they are totally unrelated. Look at the words: the prefix “iso” means equal, and the base word “chronous” is about time. But for asynchronous, the prefix “a” is a negation, but the base “synchronous” is not about time, it’s about synchronization.
Wikipedia, in the Audio section of the USB write up:
USB provides three isochronous (fixed-bandwidth) synchronization types,[67] all of which are used by audio devices:[68]
- Asynchronous – The ADC or DAC are not synced to the host computer’s clock at all, operating off a free-running clock local to the device.
- Synchronous – The device’s clock is synced to the USB start-of-frame (SOF) or Bus Interval signals. For instance, this can require syncing an 11.2896 MHz clock to a 1 kHz SOF signal, a large frequency multiplication.[69][70]
- Adaptive – The device’s clock is synced to the amount of data sent per frame by the host[71]
The isochronous QoS feature has nothing to do with the timing of the audio samples, it just guarantees that the DAC has enough samples to work with and doesn’t fall behind.
So therefore, by the definition of jitter, a USB Audio signal cannot introduce jitter because jitter is defined as the deviation from strict periodicity in a signal that is presumed periodic. When the signal doesn’t try to be periodic, it can’t fail to be periodic.
Are there audible differences between different file formats and different USB connections? I take no position on that here. But if there are, they are not caused by jitter in the asynchronous USB connection because that is a contradiction in terms.
2 Likes
I just had a totally unplanned blind experiment
Went to reference system, in which i have some experimental hardware configurations (nothing to do with Roon). I had been listening to the Tidal MQA version of American Beauty, and put it on. Listened. Not happy. What have i done wrong? Why is this clearly superior hardware not sounding right?
Walk over to Laptop acting as Roon bridge and remote too. For some reason Roon decided to switch to my local MP3 version. QED.
My DAC is not an MQA renderer. So basically i was comparing MP3 (192k) to MQA which is compressed more than FLAC is (although still effectively lossless, but that’s a murky subject so let’s sidestep it shall we?).
Bottom line: blatantly, impossible-to-miss differences. Poor bass, poorer imaging. lack of detail. Some harshness int he upper mids. In other words, exactly what one would expect of the perceptively-coded Mp3 methods, which among other things, are aging (this is MP3 not AAC).
Just pointing out that there are hugely relevant things that impact sound. FLAC vs uncompressed redBook is simply not one of them
1 Like
I would like to compare diffrent file formats, so I have now about 11 different files of a song. I have also collected other songs in different file formats.
All files are on my Roon streamer under folder Music File Formats.
Now I need Roon to look specific in that folder and let me chose different file qualities to play. I don’t see how this is possible in Roon. Any clues?
Create a playlist…
More nonsense, is my first reaction. But that’s too harsh.
What we’re witnessing is the gradual death and entombment of an entire hobby, hardware audiophilia. These superb digital formats don’t need us to build shelves for them or wipe them carefully with a lint-free cloth. The reproduction chain is trivial to build or buy, and doesn’t really require much much of an outlay; $1000 will buy a good streamer, DAC, and amp. Excellent class-D amplifiers make them tiny, so there’s no imposing racks of equipment to construct and improve. The last frontier is speakers and room correction, and WAF concerns limit much of what can be done there. Even DSP room correction is quite possible with simply Roon and a $100 microphone, though perhaps not as straightforward as Dirac Live. The only things left are meaningless snake-oil hardware tweaks: audiophile fuses and fancy cables and the like.
Really, I’m expecting an obit in the New York Times any day now. Little wonder, then, that people are aimless, restless, constrained by the pandemic, and casting around for some way to occupy their time. Comparing equivalent audio formats is harmless, after all, and does give them an outlet.
There are more constructive things to do, of course. Children to be tutored over Zoom, books to write, paintings to paint, etc.
2 Likes