Discussion in 'Digital: DACs, USB converters, decrapifiers' started by Rotijon, Jul 17, 2017.
An upgrade from USB, or an upgrade from other USB to SPDIF converters?
Read the long impressions from the first page or two.
Better than all other USB -> S/PDIF converters of even a reasonable price.
Not better than a Lynx PCIe card (E22/AES16e)
Probably also means it's not technically better than Rednet
Way cheaper than both of those.
This is the only sub-$200 SPDIF option that even makes sense apart perhaps from an ESI [email protected] XTe (PCIe card, not as good as Lynx)
Glad i only paid €375,- for my Aes16e-50. Doesn't sting as much now. Buying a Lynx new is a complete waste of money.
Some curiosities I have if someone has time to experiment
RPi player + Eitr with linear vs switching power on the RPi
RPi + USB DVD-Drive + Eitr vs Audiophile CD transport
It'll inevitably depend on your exact configuration - I've not tested this with a broad selection of other DACs.
In general I'd choose either Eitr or USB Gen 5 based on what you have and how you want to run it/what your future DAC/source plans are and just go with that. The nervosa is misplaced at this point.
I don't think Rednet (etc.) is a good solution for most users as it has significant hassles and limitations compared to purpose-built solutions and doesn't really sound any better than things like the much-simpler/more convenient Aries or Eitr/UG5 anyway. Same with things like the Lynx cards. I won't be re-testing this (don't have a way to do that anymore) , but they didn't beat other solutions when I did test against them and which are equaled by the Eitr (etc.), so you won't find me recommending them. Other's mileage has varied here and said other's here can, have and no doubt will continue to make such direct comparisons.
For me, I don't want a computer in my direct replay chain - largely for reasons of convenience, so USB isn't very relevant to me (excepting my day-job office rig) at all.
Yes, to the former. And, while I've not heard them all, generally yes to the latter. Though if I already had a solution that bypassed (or properly isolated) USB then I probably wouldn't bother changing it unless I could recover enough money to make the time to do so worthwhile (which is unlikely).
TL;DR ... if you're using a computer as a source then I'd simply not bother looking further than Eitr or, if you have Schiit DAC, the USB Gen 5 option, with a standard Schiit USB cable, and not spend any more time thinking about it.
Thanks so much for all the time you put into this. I've got another question about use in conjunction with the the microRendu, specifically about the power supply:
I want to keep the microRendu in my system for the networking capacity, but I've only got the basic power supply (iFi iPower) while saving up to buy the UltraCap LPS-1. Now, however, I'm wondering if I even need to upgrade the power supply when going microRendu > Eitr > Yggdrasil. What power supply did you use in your tests? Did you try different power supplies? And did you use the same power supply for the Sonicorbiter SE and the microRendu?
Thanks! I'll be getting an Eitr at some point for my Modi Multibit and will eventually get a Gungnir Multibit or Yggdrasil with Gen 5 in it.
So is the difference between Gungnir Multibit and Yggdrasil (with USB5) now smaller or larger than the difference used to be between Gungnir Multibit and Yggdrasil (with USB2 and USB 3)?
Save your money.
But then I'd also say that even without Eitr in the mix.
When I originally tested the microRendu I ran it with several PSUs, including an iFi Power, a battery supply, a Teddy Pardo unit, some $2 Chinese regulated wart, and an extremely quiet (and horrifically expensive) semi-custom specialty multi-mode lab PSU that makes a complete mockery of all these so-called "audiophile" PSUs (think in terms of the kind of thing you want supplying your test gear/instrumentation when you're specifically looking for infinitesimal signals).
I never found a useful/reliably detectable difference in the audible performance of the microRendu beyond the iFi Power supply.
Again, this wasn't done with a huge array of DACs ... principally Yggdrasil and Spring.
From memory, I don't think I've ever compared Gungnir MB via USB Gen 2 to Yggdrasil via USB Gen 3, largely since I do not generally use USB in my audio chain in the first place (and certainly not when auditioning DACs unless I need to include high-rate DSD or get an I2S feed).
Nor have I had a USB Gen 5 board in a Gungnir MB ... only Bifrost and Yggdrasil.
So it's pure speculation to say that I'd expect the difference to be smaller than it was, but not to the degree it would in anyway change which DAC I personally opted for.
Thanks so much, Torq. Well, this is certainly going to save me some money. Think I'm approaching my endgame.
My read -
Gen5 - less boxes, all in one solution. Cool!
Eitr - one more box, and one more wall plug, but ... you could potentially use it with other DACs if you ever care; a bit more but not necessarily if you factor in shipping costs.
Both solutions isolate GND and VCC. The difference is how close to the DAC box this isolation occurs.
USB is already differential data signaling (so common noise rejection on the data lines), but USB is packets, so one way or another has to be converted to a stream of PCM data words.
Eitr moves the USB packet->SPDIF stream conversion outside of the DAC, and I suppose, USB processing generated noise. if any, outside the DAC box, but whether either matters, no idea. Being outside the shielded DAC case appeals to me, but again, does it actually matter?
On the flip side, the connection from Eitr -> DAC is COAX SPDIF. Again GND from the source is isolated if using Eitr, so no ground loop concerns. There is no common noise rejection on the COAX cable (like AES), or EMI rejection (like TOSLINK), but again, does it matter? Use a short cable and don't plop it down next to noisy sources if concerned.
@zerodeefex : time to update the first post of the USB nervosa thread.
I hope this puts an end to the USB nervosa, at least for eight months. Gives me a headache and a sunken stomach feeling every time I hear of strange USB chains with exotic power supplies.
Though I do wonder what six Wyrds chained up in row sound like.
I note that the description mentions 44.1 and 48K input re-clocking. Works for me, because I don't up-sample on my PC, I send at 44.1.
It's not clear what either solution does if you send at higher sampling rates.
Sounds like a very busy power strip.
It's talking about the native time bases (clocks) for valid and relevant audio sample rates, which are always integer multiples (scalars) of one those two base frequencies.
Tried that. (okay, admittedly, I'm just that crazy )
It's like really... muddy and somewhat digital sounding.
Wyrd w/ exotic caps on power supply is what I'd settle with, but now that Eitr is out, I guess it's time to let Wyrd regulate my USB 2.0 hub.
Eitr sounds cool, still waiting to see if Schiit ever offers a solution to bypass "Adapticlock" via I2S or something else.
My M7 improved by strides bypassing their own version of adapticlock with i2s. I see Torq also mentioning Holo Spring improving with i2s, not sure if it bypasses the internal clock correction there tho?
Don't forget the Cthulhu floor wart. For my Vali 2/Modi MB stack, I'd get an Eitr and a floor wart so I don't have to eat up another power strip port. I wish the floor wart could power the Vali 2 as well, but it requires funky voltages.
Separate names with a comma.