Overclock.net › Forums › Components › Other Peripherals › Does a usb hub introduce more jitter in usb audio? why are decrapifiers needed?
New Posts  All Forums:Forum Nav:

Does a usb hub introduce more jitter in usb audio? why are decrapifiers needed?

post #1 of 8
Thread Starter 
When I read threads like this one and this one, I wonder what is going on in the usb signal. The whole point of a digital signal is so that the values are quantized and there is no possibility of confusion. You either get a 'one' or a 'zero'. Noise in the line is excluded by definition. If the usb receiver chip gets, 4.9 v or 5.05 v or some transient, that is irrelevant. The data is interpreted as a 'one'. So the signal amplification done by a usb hub is good - usb cables can be long, or usb chips in the motherboard may be weak. But then, the issue is whether the timing of the signals is affected.

When the dac gets the digital signal, it expects the data to come in at a specific rate, for usb that is 60 mB/s (for usb3 it is 640 mB/s). If there is jitter in the signal, doesn't the dac automatically assume the data is coming over at the specified rate? At least this is what happens w USB type 1. But w USB type 2 (async USB or ASIO) the packet gets sent only when the device asks for it. So how would jitter happen w ASIO?

I realize that if you put an oscilloscope at the end of a chain of a bunch of usb hubs, a signal coming out of the last usb hub will have more electrical noise than one coming out of the motherboard. But why does that matter?

What are the devices like the singxer su-1 and the schiit wyrd doing that the usb chips in the dac DON'T ALREADY DO?
post #2 of 8
Quote:
The Cure for Noise and Glitches
Have you ever heard strange noises from your USB DAC? Or have you had it “drop out” after the computer goes to sleep? Or does your computer have trouble recognizing your DAC at all? It could be due to noisy USB power, or USB port power management. Wyrd has you covered—Wyrd is the only USB decrapifier that includes a linear power supply (not a noisy switcher) and precision 2.5uV LM723 regulators with discrete pass elements for up to 200,000x less noise than what comes out of your USB port, up to 500mA—the full USB 2.0 spec.

From the Schiit wyrd product page.
I have experienced all the issues described in that paragraph with my Audio interface (USB 2.0 Native instruments audio kontrol 1) but never frequently enough to worry about it.
If i was cashed up to the eye-balls I'd probably buy one of these to see if it did infact stop all those things from happening,
but at the end of the day, the solution was either just to unplug and re-plug the interface or just move it to a root hub by itself, which is what I did and have never seen another issue.


edit: here's a review of the wyrd, might be biased, dunno, have a read anyway, explains a few things
http://www.digitalaudioreview.net/2015/09/schiit-wyrd-usb-decrapifier-review/
Edited by spinFX - 10/5/16 at 7:41pm
Ol'Faithful
(11 items)
 
  
CPUMotherboardGraphicsRAM
Intel Xeon x5660 P6X58D-E XFX AMD Radeon R9 280X Corsair Vengeance 24 GB Triple Chanel 6x4GB 
Optical DriveCoolingOSMonitor
Lite-On DVD-RW Noctua DH-14 CPU Cooler Windows 7 Pro x64 Samsung 24" LED 1080p 5ms DVI (S23A300B) 
PowerMouseAudio
Thermaltake Toughpower 750W Gold Corsair M95 Onboard :S 
  hide details  
Reply
Ol'Faithful
(11 items)
 
  
CPUMotherboardGraphicsRAM
Intel Xeon x5660 P6X58D-E XFX AMD Radeon R9 280X Corsair Vengeance 24 GB Triple Chanel 6x4GB 
Optical DriveCoolingOSMonitor
Lite-On DVD-RW Noctua DH-14 CPU Cooler Windows 7 Pro x64 Samsung 24" LED 1080p 5ms DVI (S23A300B) 
PowerMouseAudio
Thermaltake Toughpower 750W Gold Corsair M95 Onboard :S 
  hide details  
Reply
post #3 of 8
Quote:
Originally Posted by bwana View Post

When I read threads like this one and this one, I wonder what is going on in the usb signal. The whole point of a digital signal is so that the values are quantized and there is no possibility of confusion. You either get a 'one' or a 'zero'.

That is true, but if a one is sent a zero may be received instead, same when a one is sent.

Noise in the line is excluded by definition. If the usb receiver chip gets, 4.9 v or 5.05 v or some transient, that is irrelevant. The data is interpreted as a 'one'. So the signal amplification done by a usb hub is good - usb cables can be long, or usb chips in the motherboard may be weak. But then, the issue is whether the timing of the signals is affected.

USB does not perform any signal amplification. It just supplies power to devices and transmits data..



When the dac gets the digital signal, it expects the data to come in at a specific rate, for usb that is 60 mB/s (for usb3 it is 640 mB/s). If there is jitter in the signal, doesn't the dac automatically assume the data is coming over at the specified rate? At least this is what happens w USB type 1. But w USB type 2 (async USB or ASIO) the packet gets sent only when the device asks for it. So how would jitter happen w ASIO?

I realize that if you put an oscilloscope at the end of a chain of a bunch of usb hubs, a signal coming out of the last usb hub will have more electrical noise than one coming out of the motherboard. But why does that matter?

What are the devices like the singxer su-1 and the schiit wyrd doing that the usb chips in the dac DON'T ALREADY DO?

When the dac gets the digital signal, it expects the data to come in at a specific rate,

It kind of doesn't work that way.
You're assuming the the data terminals in the USB pinout above get either a positive signal (a one) or a negative signal (a zero) and that those signals are sent (or received) at the USB's data rate (USB 1, USB 2, etc.). That's not the way USB transfers work.

USB data is in "packets" and the packets are sent at the USB's data rate. What's in the packets, a lot of stuff.

The Universal Serial Bus: How it Works and What it Does .

Digital music, and hence the packets, does not contain ones and zeros like this
a 1, then another 1, then a 0, then a 1, then a 0, then a 1 then a 0 then a 1, etc.
They contain ones and zeros in "numbers" like this
1101010, then 10110100, then 110110011, then 10011001, then 11100111, etc.
The number of digits in each "number" is determined by the "bit depth" at which the music was converted to digital
That's typically 4 bits (1010), 8 bits, 16 bits, or 32 bits (100110010001000111100011100).
And how many "numbers" there are per second of sound (music) is determined toy the "sampling rate" at which the music was converted to digital
That's typically 44,100 parts, 48,000 parts, 96,000 parts, etc.

So, how do the "numbers" in a digital file get converted into packets?
Software does that - Windows Media Player, iTunes, Foobar, etc.
And it's the software that is a big source for jitter.
11001110 can become 01001110, or 8 digits can become just 7 digits in the "number". Or samples can be completely missed. Instead of 44.100 samples per second, a particular second may have only 43,999 samples.

So, why use USB anyway? There's also Ethernet, Toslink, Coax, Firewire, XLR, etc.
Some would argue that a digital connection (just transmitting the "numbers", not in "packet" form) would avoid a lot of errors in digital playback.
1,) That doesn't avoid software errors.
2.) Any electrical transmission can (and is) effected by EMF, RF, magnetism, etc.

So why is USB the best? Because, as opposed to any of the above, even though it is also subject to EMF, RF, magnetism, etc. USB is an asynchronous transmission. Each packet received is checked against what was sent. If errors are found the packet is resent.
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
FX6300 Black M5A99X EVO R2.0 Nvidia GTS450 Team Vulcan PC3 12800 
Hard DriveOptical DriveCoolingOS
Samsung 840 PRO Asus DRW-1608P (x2) Custom Water Cooling Win7 (Ult), Win 8.1 & Win Server 2012 R2 
MonitorKeyboardPowerCase
2 X Samsung 915N Ducky Shine III, Blue Cherry/Blue LEDs PCP&C 1kw Lian Li PC-71 (W/Window) 
MouseAudio
Logiteck G400s none 
  hide details  
Reply
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
FX6300 Black M5A99X EVO R2.0 Nvidia GTS450 Team Vulcan PC3 12800 
Hard DriveOptical DriveCoolingOS
Samsung 840 PRO Asus DRW-1608P (x2) Custom Water Cooling Win7 (Ult), Win 8.1 & Win Server 2012 R2 
MonitorKeyboardPowerCase
2 X Samsung 915N Ducky Shine III, Blue Cherry/Blue LEDs PCP&C 1kw Lian Li PC-71 (W/Window) 
MouseAudio
Logiteck G400s none 
  hide details  
Reply
post #4 of 8
Thread Starter 
thank you both for your replies.
Since I asked this question I have done some reading and Id like to add an explanation I read that rationalizes how analog noise carried on the digital cable is the culprit. Apparently analog noise is carried on the ground (the shield on the outside of the USB cable can transmit noise from the pc as well as pick up em interference like an antenna). This noise is 'injected' into the dac. And even tho the dac has circuits and chips intending to correct these issues, they do not work all the time. One solution is to use an optical usb cable. Another is to use something like a jitterbug or a REGEN.
post #5 of 8
My Native Instrument Audio 2 DJ sometimes have issues and if i connect it to the front panel the audio just crash randomly rolleyes.gif
Quote:
Originally Posted by bwana View Post

thank you both for your replies.
Since I asked this question I have done some reading and Id like to add an explanation I read that rationalizes how analog noise carried on the digital cable is the culprit. Apparently analog noise is carried on the ground (the shield on the outside of the USB cable can transmit noise from the pc as well as pick up em interference like an antenna). This noise is 'injected' into the dac. And even tho the dac has circuits and chips intending to correct these issues, they do not work all the time. One solution is to use an optical usb cable. Another is to use something like a jitterbug or a REGEN.

Please can you share the links with the info ? thumb.gif
Lili-chan
(18 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7 4790K - 4.5Ghz - Undervolted - WC ASUS Z97A EVGA Geforce GTX 1070 Founders Edition + EKWB F... Corsair Vengance 2x8 2400Mhz DDR3 
Hard DriveHard DriveHard DriveHard Drive
WD Green 1TB WD Blue 1TB Samsung EVO 840 250GB Intel 600p 1TB 
Optical DriveCoolingOSMonitor
No they are just horrible Custom Watercooling Windows 10 LG 29UM58-P 
KeyboardPowerCaseMouse
Kingston Hyper X FPS Seasonic M12 750W LIAN LI PC A05N HyperX Pulsefire 
Mouse PadAudio
A few Native Instrument Audio 2 DJ 
  hide details  
Reply
Lili-chan
(18 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7 4790K - 4.5Ghz - Undervolted - WC ASUS Z97A EVGA Geforce GTX 1070 Founders Edition + EKWB F... Corsair Vengance 2x8 2400Mhz DDR3 
Hard DriveHard DriveHard DriveHard Drive
WD Green 1TB WD Blue 1TB Samsung EVO 840 250GB Intel 600p 1TB 
Optical DriveCoolingOSMonitor
No they are just horrible Custom Watercooling Windows 10 LG 29UM58-P 
KeyboardPowerCaseMouse
Kingston Hyper X FPS Seasonic M12 750W LIAN LI PC A05N HyperX Pulsefire 
Mouse PadAudio
A few Native Instrument Audio 2 DJ 
  hide details  
Reply
post #6 of 8
Thread Starter 
post #7 of 8
Quote:
Originally Posted by lexer View Post

My Native Instrument Audio 2 DJ sometimes have issues and if i connect it to the front panel the audio just crash randomly rolleyes.gif
Please can you share the links with the info ? thumb.gif

PC cases often have horrible connectors and wiring going to the front panel.
SUPERPWN
(12 items)
 
  
CPUMotherboardGraphicsRAM
Core i7 7820x @ 4.7 GHz Asrock x299 OC Formula MSI GTX 1080 Aero 32GB Gskill DDR4-4000 
Hard DriveOSMonitorMonitor
4TB Seagate Windows 8 x64 Overlord x270 OC HP ZR27 
MonitorKeyboardPowerAudio
Asus PG279Q G-Tune Topre Realforce 800w Benchmark Dac2 | Audeze LCD-X 
  hide details  
Reply
SUPERPWN
(12 items)
 
  
CPUMotherboardGraphicsRAM
Core i7 7820x @ 4.7 GHz Asrock x299 OC Formula MSI GTX 1080 Aero 32GB Gskill DDR4-4000 
Hard DriveOSMonitorMonitor
4TB Seagate Windows 8 x64 Overlord x270 OC HP ZR27 
MonitorKeyboardPowerAudio
Asus PG279Q G-Tune Topre Realforce 800w Benchmark Dac2 | Audeze LCD-X 
  hide details  
Reply
post #8 of 8
And the USB power supplied to the USB ports is, well, not suitable for even mid range audio.
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
FX6300 Black M5A99X EVO R2.0 Nvidia GTS450 Team Vulcan PC3 12800 
Hard DriveOptical DriveCoolingOS
Samsung 840 PRO Asus DRW-1608P (x2) Custom Water Cooling Win7 (Ult), Win 8.1 & Win Server 2012 R2 
MonitorKeyboardPowerCase
2 X Samsung 915N Ducky Shine III, Blue Cherry/Blue LEDs PCP&C 1kw Lian Li PC-71 (W/Window) 
MouseAudio
Logiteck G400s none 
  hide details  
Reply
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
FX6300 Black M5A99X EVO R2.0 Nvidia GTS450 Team Vulcan PC3 12800 
Hard DriveOptical DriveCoolingOS
Samsung 840 PRO Asus DRW-1608P (x2) Custom Water Cooling Win7 (Ult), Win 8.1 & Win Server 2012 R2 
MonitorKeyboardPowerCase
2 X Samsung 915N Ducky Shine III, Blue Cherry/Blue LEDs PCP&C 1kw Lian Li PC-71 (W/Window) 
MouseAudio
Logiteck G400s none 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Other Peripherals
Overclock.net › Forums › Components › Other Peripherals › Does a usb hub introduce more jitter in usb audio? why are decrapifiers needed?