[WesternE]

has anyone rolled the wima coupling caps for something better? Seems more logical to start there than spend $300 on a pair of tubes.

 

[songmic]

Quote:

So yes I have - have you? Not a cheap USB 'jellybean' chip implementation, but one of these very well implemented solutions (including most XMOS async engines, etc.). If you haven't, and judging by your distain for USB you probably haven't, you should. What I'm talking about IS one of the newer incarnations. I believe there's a very good chance it'll blow your mind. Or at the very least, you'll find it a very acceptable alternative to SPDIF (which honestly, against these newer converters, has higher jitter, more noise and less bandwidth).

 
Agreed. People who claim that USB is inherently inferior to S/PDIF ain't heard USB these days. Despite the trend in computer audio, it's true most computers (including my PC and iMac) doesn't have coaxial SPDIF, which is why many digital audio companies nowadays compete for better USB implementations that can rival S/PDIF. I use a Stello U3 (USB-to-S/PDIF converter that uses XMOS async engine with capability up to 196/24) between my iMac and DAC, and I was surprised by its results when I first heard it. Hopefully we may even see DAC's that implement USB 3.0 or Thunderbolt sometime in the near future, but for now USB 2.0 is the real deal.

 

[HK_sends]

What they said... 

 
Cheers!

-HK sends

 

[bleudeciel16]

Just hooked up the Lyr with the 6BZ7 tubes.  (no burn in done yet)
 
Good lord they sound great.  I was using my temporary Fiio E7 as my amp/dac.  I didn't think I would get much of an upgrade but man was I wrong.  The biggest thing I notice I gained was clarity at lower levels.  It felt like I had to crank the Fiio to get any good sound out of it.  The soundstage also feels wider now.  I can't wait for burn-in to complete.  I also have on hand the other two sets of tubes Schiit sells for the Lyr.  Excited to try those out as well.
 
ASUS Xonar Essense SXT > (SPDIF Digital Coax) > Schiit Bifrost > Schiit Lyr (6BZ7 tubes) > HD650

 

[dailydoseofdaly]

Man the more searching i do, it seems like every company is jumping on the USB converter bandwagon. It's there a thread im missing on here or else where that has good comparisons of the different ones?

Sorry about the untube related question

 

[bleudeciel16]

Quote:

Man the more searching i do, it seems like every company is jumping on the USB converter bandwagon. It's there a thread im missing on here or else where that has good comparisons of the different ones?
Sorry about the untube related question

 
I haven't been able to find solid info on this after lots of searching.  People have generally been saying that all three are very close, so that it shouldn't matter, and that USB is a lot better than it used to be.
 
Schiit themselves said Coax > Optical > USB.  I still don't know to be honest, so I just went with their recommendation. (plus their USB addon is $100. screw that noise)
 
Back on subject though, I'm sad that I have plans tonight, because I would rather stay home and get some time on my 650's. (will be at work all day)

 

[IEMCrazy]

Quote:

 
Agreed. People who claim that USB is inherently inferior to S/PDIF ain't heard USB these days. Despite the trend in computer audio, it's true most computers (including my PC and iMac) doesn't have coaxial SPDIF, which is why many digital audio companies nowadays compete for better USB implementations that can rival S/PDIF. I use a Stello U3 (USB-to-S/PDIF converter that uses XMOS async engine with capability up to 196/24) between my iMac and DAC, and I was surprised by its results when I first heard it. Hopefully we may even see DAC's that implement USB 3.0 or Thunderbolt sometime in the near future, but for now USB 2.0 is the real deal.

 
No coax SPDIF?  I haven't had a PC since the 90's that didn't have either coax or optical S/PDIF, and usually both. 
 
Of course if folks are talking about the S/PDIF implementation from onboard sound interfaces versus a high-end USB setup, that's understandable.  S/PDIF is only as good as the transmitter, and onboard is usually garbage.  On the other hand, onboard USB is often garbage too, both have interference out the wazoo. 
 
A dedicated sound card from a good audio interface grade manufacturer feeding S/PDIF out should certainly sound great.  And on that note a dedicated USB interface would hopefully have less noise issues than an onboard USB port for using an external USB DAC as well. 
 
I imagine they can both be equal given the proper conditions.  It's just that USB is less often given optimal conditions.  (I say that as someone that's been using an EMU-0404 USB on the PC setup for a while for various reasons unrelated to USB vs. S/PDIF SQ comparisons.)
 
Quote:

Just hooked up the Lyr with the 6BZ7 tubes.  (no burn in done yet)
 
Good lord they sound great.  I was using my temporary Fiio E7 as my amp/dac.  I didn't think I would get much of an upgrade but man was I wrong.  The biggest thing I notice I gained was clarity at lower levels.  It felt like I had to crank the Fiio to get any good sound out of it.  The soundstage also feels wider now.  I can't wait for burn-in to complete.  I also have on hand the other two sets of tubes Schiit sells for the Lyr.  Excited to try those out as well.
 
ASUS Xonar Essense SXT > (SPDIF Digital Coax) > Schiit Bifrost > Schiit Lyr (6BZ7 tubes) > HD650

 
Hehe, yeah the GE's are pretty good tubes considering the price.  I could have lived with those forever in bliss had I not had recurring "tube whining" issues that forced me to look elsewhere. And of course, once you taste the good stuff...

 

[bleudeciel16]

Quote:

 
No coax SPDIF?  I haven't had a PC since the 90's that didn't have either coax or optical S/PDIF, and usually both. 
 
Of course if folks are talking about the S/PDIF implementation from onboard sound interfaces versus a high-end USB setup, that's understandable.  S/PDIF is only as good as the transmitter, and onboard is usually garbage.  On the other hand, onboard USB is often garbage too, both have interference out the wazoo. 
 
A dedicated sound card from a good audio interface grade manufacturer feeding S/PDIF out should certainly sound great.  And on that note a dedicated USB interface would hopefully have less noise issues than an onboard USB port for using an external USB DAC as well. 
 
I imagine they can both be equal given the proper conditions.  It's just that USB is less often given optimal conditions.  (I say that as someone that's been using an EMU-0404 USB on the PC setup for a while for various reasons unrelated to USB vs. S/PDIF SQ comparisons.)
 
 
Hehe, yeah the GE's are pretty good tubes considering the price.  I could have lived with those forever in bliss had I not had recurring "tube whining" issues that forced me to look elsewhere. And of course, once you taste the good stuff...

 
any tube suggestions? (incase I run into similar issues, or incase I decide to stock up on extra tubes incase there's a tube-eating zombie apocalypse?)

 

[MickeyVee]

The zombie tube eating apocalypse has already started.. try to find a set of Lorenz Stuttgarts now at any reasonable price 

Quote:

 
any tube suggestions? (incase I run into similar issues, or incase I decide to stock up on extra tubes incase there's a tube-eating zombie apocalypse?)

 

[SeaHawk]

Quote:

 
No coax SPDIF?  I haven't had a PC since the 90's that didn't have either coax or optical S/PDIF, and usually both.

 
Laptops, brother!  It's what I get to use at work, and that's where my rig is for music bliss to help numb the pain from powerpoint agony, so I'm stuck with USB for the time being...
 
From a geek standpoint, I'm not sure what all the fuss over USB is, as it's merely data transfer - if you don't starve the buffer, the onus is on the receiving end rather than the transmit end to spew the individual bits in a timely fashion.
 
Of course, drivers are an invented mechanism to screw this otherwise simple process.  Speaking of which, I may start a new thread for Bifrost USB driver settings (which report odd things back to Foobar2000 when trying to choose 24bit depth, etc - there's no way to verify if it worked or not

 

[HK_sends]

Matsush**a/National 6922s are pretty good all around and not overly expensive.
 
http://www.tubemonger.com/Matsu****a_Japan_6922_E88CC_Mullard_Tooling_NOS_p/591.htm
 
I have a couple of sets and they are my general, everyday listening tube.
 
Cheers!

-HK sends

 

[HK_sends]

Quote:

 
Laptops, brother!  It's what I get to use at work, and that's where my rig is for music bliss to help numb the pain from powerpoint agony, so I'm stuck with USB for the time being...
 
From a geek standpoint, I'm not sure what all the fuss over USB is, as it's merely data transfer - if you don't starve the buffer, the onus is on the receiving end rather than the transmit end to spew the individual bits in a timely fashion.
 
Of course, drivers are an invented mechanism to screw this otherwise simple process.  Speaking of which, I may start a new thread for Bifrost USB driver settings (which report odd things back to Foobar2000 when trying to choose 24bit depth, etc - there's no way to verify if it worked or not

I just installed the WASAPI plug-in and set the output to WASAPI and 24bit.  So far, I haven't had a problem.  What version of Foobar are you using?  I am using it on a Dell XPS laptop plugged into my Bifrost.
 
Cheers!

-HK sends

 

[IEMCrazy]

Quote:

 
Laptops, brother!  It's what I get to use at work, and that's where my rig is for music bliss to help numb the pain from powerpoint agony, so I'm stuck with USB for the time being...
 
From a geek standpoint, I'm not sure what all the fuss over USB is, as it's merely data transfer - if you don't starve the buffer, the onus is on the receiving end rather than the transmit end to spew the individual bits in a timely fashion.
 
Of course, drivers are an invented mechanism to screw this otherwise simple process.  Speaking of which, I may start a new thread for Bifrost USB driver settings (which report odd things back to Foobar2000 when trying to choose 24bit depth, etc - there's no way to verify if it worked or not

 
You may want to look back a few months at the Bifrost thread in the source forums or the last day or two in the HD650 appreciation thread for some digital audio discussion.  Unfortunately I didn't get to finish that conversation in the thread since I'm not sure if it was that discussion or another one that got the thread locked for a day.  Short story: there is no buffer.  It just throws bits in series, fire and forget.  If it doesn't arrive in order, it's discarded.  If it arrives damaged, it's just left wrong.  Same for S/PDIF.  It doesn't transfer data in a data/error-control/buffered sense.  It's still just sending a PCM stream and happens to use a noise-infested USB cable for doing so with some different packet overhead.  It's not data transfer, it's still PCM streaming. Take coax S/PDIF, shove it in the wire with a +5V power line, don't shield it, feed it around the back of a PC power supply over normally poorly location interference prone USB jacks, and you've got USB-Audio

  It takes a darned good DAC interface to try to piecemeal anything sensible out of the mess that can sometimes arrive.   Generally USB and S/PDIF are doing the same thing.  USB is just a far more hostile environment in which to try to do it. 
 
 
Quote:

Matsush**a/National 6922s are pretty good all around and not overly expensive.
 
http://www.tubemonger.com/Matsu****a_Japan_6922_E88CC_Mullard_Tooling_NOS_p/591.htm
 
I have a couple of sets and they are my general, everyday listening tube.
 
Cheers!

-HK sends

 
HK: Do you know how those rate compared to real Mullard 6922s and compared to the PCC88 version of the National?  And what would you consider "better" than them (Lorenz and CCa aside

)  I've been looking at those for a while.

 

[WesternE]

the only thing I don't love about this amp is 6DJ8 tubes. Years ago a swore off any design that used these because they are expensive, their are copies (fakes), the tone varies significantly by brand/construction, and used ones are prone to testing "strong" while being noisy. For the price when authentic quiet ones can be found in good shape I much prefer Amperex 60's 6DJ8, but finding a pair that is real and not noisey it difficult. Second choice at the time was Amperex '70's Orange PQ with dimpled getters. These weren't being copied at the time and a person had a resonable chance of finding an authentic noise free pair (Their design is intrinsically quieter than the 60's designs). Luckily I held onto a pair that I am using now in the Lyr. They sound perfect to me. Compared to 70's tesla, Sovteks, they just have better tone to my ears. Problem is I need a backup pair. One thing I have noticed is that ebay seems to have many more 70's Amperex 6DJ8's for sale at reasonable pricing than they did 5 years ago. Are there now fake orange PQ Amperex on the market? Also I am not letting go of a cap upgrade. The two red .47uf WiMA caps are the coupling caps in this design that block DC from the tube stage to the mosfet stage. Most of my better caps in that size are too big, surely someone has tried an upgrade that would fit ? Seriously believe the Wimas are holding the tube section of the amp back a bit, makes little sense IMHO to go all out on expensive german tubes till those caps are upgraded.

 

[SeaHawk]

Take a look: http://www.beyondlogic.org/usbnutshell/usb3.shtml
 
I'm by no means authoritative on the subject since I haven't done a USB implementation, but I am a network engineer and generally understand the layers of protocols at work.  If someone can point to a better reference than I can find that refutes me, I'm all ears.  I live for this sort of stuff!

 
USB passes packets, so this bitstreaming/jitter/no buffer nonsense is right out.  I have USB-based data acquisition devices that run at much higher sampling rates and are very accurate that make most of the audio discussion a poor joke.  Stop it.. Please
 
There *must* be at least a buffer to hold a packet, likely many more than a single packet, otherwise the USB device in question would not be able to determine if the incoming data is addressed to it or another device on the bus.  There is certainly more than enough bandwidth for robust error-correction (including re-ordering packets, using sliding receive-window mechanisms that have been in TCP/IP since the late 80s).
 
I can accept that the higher layer protocol that our devices use is poorly designed, but there's absolutely no reason the industry cannot address this moving forward (even if it requires proprietary drivers).  Wasn't this whole new USB 2.0 "Asynchronous" (meditate strongly on that word for a moment) mode supposed to resolve the buffer underrun issues?
 
USB Audio class 1 isochronous mode is limited to 24bit/96KHz as I don't believe the standard embraces the improved bandwidth of USB 2.0.  However, it reserves bandwidth on the USB bus with a certain size window (a buffer!) that the host computer has to keep filled.  Luckily, Windows has a driver for this built-in and does a fairly decent job of it.  There can be buffer underrun conditions causing temporal distortion in the audio stream, the worst occurring when the bus clock is not perfectly synced with the DAC clock, causing bits to go missing every so often.  This isn't precisely jitter, however.  True jitter encountered in this mode is solely the result of a poorly engineered design (e.g./READ: it's actually quite prevalent since most USB DAC implementations use a synthesized clock which creates lots of jitter all on its own accord - but you get very little latency!).
 
Keep in mind that this isochronous stream is being reassembled from the payload of packets - it is not a continuous stream of only audio bits falling out the end of your USB cable.
 
USB Bulk mode transfer (excuse me.. "asynchronous" mode) doesn't guarantee bandwidth on the bus or latency for that matter, but generally with USB 2.0+ this isn't an issue unless you have other devices on the same bus.  Error correction and guaranteed delivery (meaning out-of-order packets aren't an issue) are easily available if they choose implements it (basically setting a flag to "yes" since it's already in the protocol).  Unfortunately I can't find an easy reference to how C-Media does it (there is no Windows driver, so the manufacturer must develop their own and can hide their "secret sauce"), but if they didn't, they took the lazy way out and would be receiving UPS boxes full of poo from me if I find out if it is true.
 
...and now we're WAAAAY off-topic... I looked but didn't really find an appropriate thread to move this to as of yet
 
As far as my foobar2000 issues - using whatever version is marked current/stable.  When I go into the ASIO configuration and see the device config, I click it, it opens and says 16 bit.  I select 24 bit and close it.  If I reopen it, it says 16 bit again, leaving me the impression that 1. the driver Schiit provided sucks (yes, they just redistribute the C-Media driver), 2. the driver doesn't seem to store the state of my selection correctly, at least when using Foobar2000 and 3. Neither it or I know if it's really putting out 16 or 24 bits.
 
And now it's Friday evening, relaxing while eating dinner with the wife and contemplating which part of Scotland to visit for my single malt journey shortly