4 ms·

If USB-C has taught me anything it’s that higher signal integrity in more expensive cables is actually really a thing. I have usb-c devices which are so tempera

by paulmd 3y ago

If USB-C has taught me anything it’s that higher signal integrity in more expensive cables is actually really a thing. I have usb-c devices which are so temperamental (razer kiyo pro) that they won’t run stable without actual thunderbolt cables. Even high-end fasgear 10gbps/100w cables aren’t good enough signal integrity or the camera will eventually freeze and crash. The extra thunderbolt shielding/better termination really does make a difference even if the pairs are not being used. Hdmi and DP cables also have a range of shielding and termination quality, and pushing more data through them will show the difference, especially in longer lengths that are at the limit of the spec. Everything looks good when you are only doing 1080p30, try doing 2160p120 on a tv or UWQHD 160 hz on a pc, with a max-length (2m) cable and tell me every cable is just as good. Even every DisplayPort 1.4 cable or hdmi 2.1 cable is not “as good” let alone a rando lower-tier cable with the older shielding. Oculink folks know damn well that not every cable is just as good either. Copper oculink cables are a fucking mess. I’m not saying Monster wasn’t ripping you off for some gold-plated connectors, especially back in 2000 when hdmi stated with 1.x in the name. But the reality is that if nothing else we have manifested the world where digital cable quality matters. And really the “hurr durr digital signals” thing has always been a little bit of an oversimplification, even as someone who’s said it before too. There isn’t a correlation with price in the high end, going from a $30 cable to a $150 cable doesn’t get you anything, but going from a $5 cable to a $30 cable you certainly do get better shielded cable and better termination if nothing else, and yes, it does make a difference when you are actually pushing the spec to the limit. You will definitely find out that not even all 10gbps usb-c or hdmi 2.1 UHBR cables are created equal let alone ye olde hdmi 1.3 bullshit cable.

Kirby64 3y ago
Two points: Digital audio, no matter how high quality and how many channels, uses vastly less bandwidth than even the lowest quality video signal these days. At absolute best, lossless audio with 16 channels is still less total bandwidth than even USB2.0 speeds. It’s just not fast. Although signal integrity matters, signal integrity in digital signals is almost always not silent degradation. And often is extremely obvious degradation. So if your cable has problems, you know almost immediately (or, when it happens you know). Unlike analog audio, it either works or it doesn’t. In the marginal case you don’t just get worse sounding audio… you just don’t have audio periodically which sounds super obvious to the ear.
- vvillena 3y ago
  You're about an order of magnitude wrong. USB2 can transport close to 200 channels at studio quality (24bit 96 Khz).
  - Kirby64 3y ago
    Why so aggressive? I didn’t make any claims other than less than USB2. And for reference, I was assuming 384 kHz, 32 bit depth, which would be 196 Mbps at 16 channels. You cannot even do 32 channels over USB2 which would be the next logical step.
    - vvillena 3y ago
      Sorry you got offended, that wasn't my intention. Pointing out inaccuracies is not aggressive, and in this case, I was reinforcing your claim, by saying USB2 is not only capable of tens of audio channels, but hundreds. I was just adding context, and I'm happy you provided even more. 32/384 is indeed at the extreme top of digital audio signals, beyond what many people consider to be useful. Even in such cases, USB2 can still do the job, but most modern devices will use USB3 or Thunderbolt, simply to avoid issues and ensure enough headroom on the digital bus.
    - CyberDildonics 3y ago
      384 kHz, 32 bit depth This is nonsense for recording or playback. Maybe someone works with this internally but there is no reason to output this over usb.
      - paulmd 3y ago
        > there is no reason to output this over usb some inexpensive SDR dongles sidestep the question of getting a really good high-bandwidth ADC (still not cheap even today!) by just hooking up an intermediate frequency stage which downmixes RF to AF, and then you just hook it to one of the innumerable sound cards on the market to pull it into the PC for processing etc. Wider bandwidth is wider bandwidth, the faster your sample rate the wider the range that you can sample. Iirc it's nyquist rate, so, with 384kbps rate you can sample a 192khz window.
adhesive_wombat 3y ago
Right, but in the context of an audio stream, it's not going to matter these days. Even a hugely overkill uncompressed 8-channel, 32 bit, 192kHz sample rate audio stream[1] is under 50 megabits/second. Any digital cable and transceiver from USB 2 (480Mbps) or 100BASE-T Ethernet (100Mbps) onwards will handily do that, with headroom for so much forward error correction that you probably won't expect a single bit error in your lifetime. [1]: good luck getting a clock with low enough sample aperture jitter to actually record that in the first place, it's probably under a femtosecond.
- CyberDildonics 3y ago
  it's probably under a femtosecond. What are you talking about here? A femtosecond is a quadrillionth of a second. 10gbs would mean one 10 billionth of a second.
  - adhesive_wombat 3y ago
    If the sample clock edges aren't very (very very very) regular, on a sample-and-hold ADC, the waveform isn't sampled evenly and that manifests as noise that swamps the detail provided by the higher bit depth. This is called "sample aperture jitter". Requirements scale linearly with frequency and exponentially with bit depth. These a calculator here: https://www.analog.com/en/design-center/interactive-design-tools/jitter-calculator.html https://www.analog.com/en/design-center/interactive-design-t... https://www.analog.com/en/technical-articles/aperture-jitter-calculator-for-adcs.html https://www.analog.com/en/technical-articles/aperture-jitter... Sure enough, 32 bits sampling a 96kHz signal, which is the Nyquist frequency of 192kHz sampling rate, is 0.3fs. At 24 bits, it's more like 100fs, which is much more doable, but still not easy. Which is why audio bit depths usually don't go to 32 bits, despite formats like FLAC supporting that. The practical upshot of this and other noise sources is that higher audiophile-grade bit depths and sampling frequencies are quite likely to have at least some of those bits swamped out by noise on real hardware. This is just getting the audio recorded. Playing it back as physical sound waves adds something between quite a bit and radically more noise to the signal, even if there's never any lossy compression.
kevin_thibedeau 3y ago
High speed digital signaling over differential pairs requires precise impedance control to handle multi-gigahertz data rates. You're paying for that in a high quality USB3 cable. Audio signals running below 20 kHz are in a completely different electrical regime that doesn't require special manufacturing attention. It's easily doable with 19th century technology.
Dalewyn 3y ago
>UWQHD I agree, sending Alphabet Soup down those cables isn't going to end well. Those things aren't designed to carry liquids.
Dah00n 3y ago
> I have usb-c devices which are so temperamental (razer kiyo pro) that they won’t run stable without actual thunderbolt cables. That is a device problem, not a "there is a need for better cables" problem.

by paulmd 3y ago

Kirby64 3y ago

Two points: Digital audio, no matter how high quality and how many channels, uses vastly less bandwidth than even the lowest quality video signal these days. At absolute best, lossless audio with 16 channels is still less total bandwidth than even USB2.0 speeds. It’s just not fast. Although signal integrity matters, signal integrity in digital signals is almost always not silent degradation. And often is extremely obvious degradation. So if your cable has problems, you know almost immediately (or, when it happens you know). Unlike analog audio, it either works or it doesn’t. In the marginal case you don’t just get worse sounding audio… you just don’t have audio periodically which sounds super obvious to the ear.

vvillena 3y ago
You're about an order of magnitude wrong. USB2 can transport close to 200 channels at studio quality (24bit 96 Khz).
- Kirby64 3y ago
  Why so aggressive? I didn’t make any claims other than less than USB2. And for reference, I was assuming 384 kHz, 32 bit depth, which would be 196 Mbps at 16 channels. You cannot even do 32 channels over USB2 which would be the next logical step.
  - vvillena 3y ago
    Sorry you got offended, that wasn't my intention. Pointing out inaccuracies is not aggressive, and in this case, I was reinforcing your claim, by saying USB2 is not only capable of tens of audio channels, but hundreds. I was just adding context, and I'm happy you provided even more. 32/384 is indeed at the extreme top of digital audio signals, beyond what many people consider to be useful. Even in such cases, USB2 can still do the job, but most modern devices will use USB3 or Thunderbolt, simply to avoid issues and ensure enough headroom on the digital bus.
  - CyberDildonics 3y ago
    384 kHz, 32 bit depth This is nonsense for recording or playback. Maybe someone works with this internally but there is no reason to output this over usb.
    - paulmd 3y ago
      > there is no reason to output this over usb some inexpensive SDR dongles sidestep the question of getting a really good high-bandwidth ADC (still not cheap even today!) by just hooking up an intermediate frequency stage which downmixes RF to AF, and then you just hook it to one of the innumerable sound cards on the market to pull it into the PC for processing etc. Wider bandwidth is wider bandwidth, the faster your sample rate the wider the range that you can sample. Iirc it's nyquist rate, so, with 384kbps rate you can sample a 192khz window.

adhesive_wombat 3y ago

Right, but in the context of an audio stream, it's not going to matter these days. Even a hugely overkill uncompressed 8-channel, 32 bit, 192kHz sample rate audio stream[1] is under 50 megabits/second. Any digital cable and transceiver from USB 2 (480Mbps) or 100BASE-T Ethernet (100Mbps) onwards will handily do that, with headroom for so much forward error correction that you probably won't expect a single bit error in your lifetime. [1]: good luck getting a clock with low enough sample aperture jitter to actually record that in the first place, it's probably under a femtosecond.

CyberDildonics 3y ago
it's probably under a femtosecond. What are you talking about here? A femtosecond is a quadrillionth of a second. 10gbs would mean one 10 billionth of a second.
- adhesive_wombat 3y ago
  If the sample clock edges aren't very (very very very) regular, on a sample-and-hold ADC, the waveform isn't sampled evenly and that manifests as noise that swamps the detail provided by the higher bit depth. This is called "sample aperture jitter". Requirements scale linearly with frequency and exponentially with bit depth. These a calculator here: https://www.analog.com/en/design-center/interactive-design-tools/jitter-calculator.html https://www.analog.com/en/design-center/interactive-design-t... https://www.analog.com/en/technical-articles/aperture-jitter-calculator-for-adcs.html https://www.analog.com/en/technical-articles/aperture-jitter... Sure enough, 32 bits sampling a 96kHz signal, which is the Nyquist frequency of 192kHz sampling rate, is 0.3fs. At 24 bits, it's more like 100fs, which is much more doable, but still not easy. Which is why audio bit depths usually don't go to 32 bits, despite formats like FLAC supporting that. The practical upshot of this and other noise sources is that higher audiophile-grade bit depths and sampling frequencies are quite likely to have at least some of those bits swamped out by noise on real hardware. This is just getting the audio recorded. Playing it back as physical sound waves adds something between quite a bit and radically more noise to the signal, even if there's never any lossy compression.

kevin_thibedeau 3y ago

High speed digital signaling over differential pairs requires precise impedance control to handle multi-gigahertz data rates. You're paying for that in a high quality USB3 cable. Audio signals running below 20 kHz are in a completely different electrical regime that doesn't require special manufacturing attention. It's easily doable with 19th century technology.

Dalewyn 3y ago

>UWQHD I agree, sending Alphabet Soup down those cables isn't going to end well. Those things aren't designed to carry liquids.

Dah00n 3y ago

> I have usb-c devices which are so temperamental (razer kiyo pro) that they won’t run stable without actual thunderbolt cables. That is a device problem, not a "there is a need for better cables" problem.