Because errors do happen in digital data transmission, but there are very effective error correction and detection coding algorithms used in making up the packets, using more bits that would be needed if errors couldn't happen. 10101 doesn't get sent as 10101 really, it is encoded. Those make the likelihood of having undetected errors in the final stream extremely low, but if the coding finds that it didn't successfully decode a packet, it doesn't really know which bit or bits are wrong, just that the packet has an error. An all or nothing kind of thing, which is the maybe the main point of digital in the first place.
Sure, there's a lot of incompetent designs floating around.
But the subject was bits falling over due to bad USB cables.
Jan
I don't know anything more than the most rudimentary description of digital so can someone help me out here? I thought that jitter was purely a clock related phenomenon. Why the talk of dropout? Does the question of good or bad USB connection have anything to do with timing?
Thanks
Written by a journalist who cannot actually find a single 'rotted' file so hand waves about them. Sorry never had this 'bit rot' problem from HDD. Cassette based computer games maybe. But HDD work 100% or die. If lucky you get a days warning, but I have had catastropic failures.
You've not done any serious FMEA by the sound of it. RAID helps if a single disk fails on its own but
-lightning strike taking out server
-Fire
-Theft
- rm-r *
etc
All result in total data loss, whereas an off site backup means data is kept. In my case with a friend and I hold his backup.
Well the clock at the DAC pins is, in many USB DACs, generated by the USB receiver chipset. In some USB receivers (synchronous and adaptive), the audio clock has a (a more or less direct) link to the timing of the USB packets. And, as in all clock generation processes, the power supply quality somewhat defines the performance of the clock.
But, as always, the audiophile angle is to overstress actual problems which can be (and usually are) adressed to a satisfactory extent.
Any measurements of a DAC showing this effect? Only serious boutique seems to measure badly these days.
Yes, and that's crystal-controlled. If that aspect is done wrong, performance suffers, but it's unrelated to the interface, whether USB, Firewire, whatever. And perhaps not coincidentally, every DAC in my house has jitter performance far better than audible thresholds.
$149 buys a USB DAC, in a box that has performance that is, by any known metric transparent. As you go down from that performance will go down and at some point you will reach some cheap unit that might have a problem related to the USB side. But you'll be at happy meal prices by then.
Above $150 you are paying for features and a nicer box. Above $1500 you are buying a story.
If you are going to make millions of something, you tend to try and make sure you wont get 50% returns so they tend to be better considered than one might think.
Above $150 you are paying for features and a nicer box. Above $1500 you are buying a story.
If you are going to make millions of something, you tend to try and make sure you wont get 50% returns so they tend to be better considered than one might think.
Have a look at the c-media chipsets measured by NwAvGuy here: NwAvGuy: FiiO D5 DAC
They're pretty low end chipsets integrating everything so it's not very surprising.
You can also search the web for measurements of the pcm2707's audio clock depending on implementation (and particularly power supply), when it spits out I2S. It's easily found. The interesting point is that a barebone pcm2707 still achieves about 2 - 3ns of jitter which might not be audible.
. If I take the kids to McD its a lot more!
Yes, but add in that this doesn't seem to appear in anything other than the very cheapest and the very most expensive units. It is about number 1,476,352 on the list of actual things to worry about.
A few years back SY and I were testing some USB>SPDIF convertors. It's documented here somewhere. We tested some cheap and cheerful, home recording types and modified "better" units. Most tested just fine for noise and low jitter. I did find some very cheap ones I got form eBay with CMedia ships that were not so great. Everything else did fine. As was mentioned above, I think it was the chipset an the implementation that was to blame, not the USB cable.
What I find interesting in this thread is that problems with USB cables tend to get described in Digital or Analog terms.
Digital: It works or it doesn't. Dropouts.
Analog: There is a veil or bad bass or something.
IME with SPDIF and AES digital audio, it's not like that. Bad cables can cause problems, but they aren't "all or nothing" or in any way analog sounding. Yes, I've gotten clicks and pops and a few drop outs (bad noise on the ground line) but there IS an in-between area. It sounds awful. A sour sound is how I would describe it. It doesn't sound like an analg fault, but it's not right. Can't remember having that problem with USB, tho.
Don't fall into the trap of thinking that digital faults will sound "digital" or "analog". It's something else, and it isn't pleasant.
What I find interesting in this thread is that problems with USB cables tend to get described in Digital or Analog terms.
Digital: It works or it doesn't. Dropouts.
Analog: There is a veil or bad bass or something.
IME with SPDIF and AES digital audio, it's not like that. Bad cables can cause problems, but they aren't "all or nothing" or in any way analog sounding. Yes, I've gotten clicks and pops and a few drop outs (bad noise on the ground line) but there IS an in-between area. It sounds awful. A sour sound is how I would describe it. It doesn't sound like an analg fault, but it's not right. Can't remember having that problem with USB, tho.
Don't fall into the trap of thinking that digital faults will sound "digital" or "analog". It's something else, and it isn't pleasant.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.