Lynx PCIe AES interfaces: AES16 and E22

Same restrictions with the Thunder 2.

Lynx card should work on the Thunder 3, then.

 

Guess the vx222e was a good deal for a reason haha

 

Not a PC but if you don't need it right now the Raspberry Pi + the SPD2 hat (has BNC and XLR outs I believe) seems worth keeping an eye on.

http://www.superbestaudiofriends.org/index.php?threads/raspberry-pi-i2s-to-spdif-hat.1990/

Unfortunately the production version doesn't seem very imminent.

 

I've got a raspberry pi 2 but don't have the hat. Perhaps worth a try to compare vs. the aes card.

 

Do you guys use rate lock or synchrolock on the lynx mixer?
Still tinkering...

 

So, I find myself temporarily with a surfeit of multibit:



Before I sell the Gungnir Multibit on, I thought I'd run the comparo between Yggdrasil/USB and Gungnir Multibit/SPDIF. Both DACS were well warmed-up. Chains were:

Foobar 2000 -> USB -> Yggdrasil -> Klein & Hummel O300D
Foobar 2000 -> RME9632 -> SPDIF -> Gungnir Multibit -> Klein & Hummel O300D

It's not completely cut and dried.

- Yggdrasil over USB has a really terrible spitty top end. Imaging is quite compromised. The fizz up top makes the bass seem pretty lean. Everything sounds a bit flat. However, it is still clear that you are listening to a better DAC than the Gungnir Multibit. The huge macrodynamics and ultra-snappy transient response are still there. Blacker background. You can still sense the throbbingly good tonality and plankton retrieval though it is obscured by the sea of tizz.

- Gungnir Multibit SPDIF is way more listenable and less offensive up top. Imaging is better and more bass presence (though some of this seems to be coming from the slightly warmer character of the Gungnir Multibit). However, the sound is greyer, leading edges are a bit rounder, and everything is generally more polite.

Personally, I wouldn't recommend Gungnir Multibit + SPDIF over Yggdrasil + USB just because there are cheap and easy ways to get an acceptable (if maybe not ultimate) non-USB signal into Yggdrasil and at that point there is no competition.

One other note. Running both from the RME, the Gungnir Multibit still has a slightly fatiguing etch in the top end compared to the Yggdrasil. I don't know if this is an inherent characteristic or if it means that the Yggdrasil has more robust jitter rejection. In any case, I see this as another reason to not dink around too much with sources below Yggdrasil, because it seems that a proportion of those problems are solved for you by Yggdrasil.

 

I found this thread from some early adopters getting digital out to (pre-USB, decent quality) DACs:

http://www.audioasylum.com/forums/pcaudio/messages/8/80107.html

I don't know if there is much new information, but it supports what I already suspected, that the RME cards and the Lynx cards are on a similar footing as a digital source. Note these are the PCI cards, but it seems reasonable to generalise to the (fairly similar) PCIe cards. In short, if an RME card comes up at a nice price, it's probably worth a shot.

 

I actually just grabbed a RME HDSPe RayDAT card to put in place of the digigram. Looks like it should work in the akitio tb3 chassis. A bit pricier than I wanted, but I have the digigram to offload to hopefully make it less painful

 

OK, I finally got in an AES breakout for my RME 9632 (previously using SPDIF). The RME outputs SPDIF and AES simultaneously, so I now can test:

PC > RME 9632 > SPDIF vs AES > Yggdrasil > Klein & Hummel O300

It's clear: the AES is better. If USB vs SPDIF was 20%, SPDIF vs AES is something like 5%. Pretty certain I could A/B this. I am not conscious of the remaining digititus on the SPDIF when used alone but in direct comparison it is clear. Most improvements are along the same axes as before: closely spaced transients are not temporally smeared together, reduced hash above 10khz, imaging is more precise. Some new ones too, though: background is blacker, and I think soundstage is slightly wider. I feel like I'm converging on something!

Now for some extended listening to see how this plays out in practice...

EDIT: The plot thickens! My two breakout cables sound quite different. The new one with AES and SPDIF has a fair bit more energy up top and slightly scooped mids compared to the stock one. So while there is less digititus over AES the tonal balance is off (in fact "ow my ears" off). The new one I got off ebay and looking at it I suspect it may be a dodgy DIY job, maybe with something egregiously wrong with its impedance. Will try getting someone local to make me up a breakout cable with some Mogami 3080 and report back...

 

I still get sporadic episodes of choppy clippy sounding audio. But it is totally sporadic.
So I don't know if it the cable of my abyss, lynx break out cable, power issues with ragnarok, driver issues with my gaming PC which I use as a source.. Or something else.. Any advice is always appreciated.

 

I took delivery of the E22 today. Been listening for a few hours and it's clearly better that USB. I have been using the PRO3Z which was a very good upgrade especially running of batteries but this is better. The most impressive aspect is the layering as each note is so well placed in the soundstage. Bass also seems to have more impact and the high end seems sweeter perhaps? Might be because of the lack of any sort of sibilance or harshness which I did observe using just USB.
I'm using Yggdrasil->Moon 600i/Stratus->Abyss/HD650.
No clicks or pops so far but a lot of my channels have disappeared from the mixer. Somehow I'm only left with 1 AOut 2 and Digital Out but at least those the mandatory ones.

At this point I don't care if the RedNet is better as this is so convenient especially since it does automatic sample rate changes. Since my music is a mix of 44.1 and 48 the RedNet would be a pain to manually switch every time.

 

Note that Rednet's latest software can follow sample rate changes, but that is still a large part of why I kept the Lynx E22 and sold the Rednet 3--the Rednet is bulky and a shitload more finicky to keep working 100% perfectly.

 

Is USB really that bad? I find that my Geek pulse xfi actually sounds quite good with a LPS and a split power cable disconnecting the power leg after handshake.

 

"Quite good" versus AES on the Pulse? Or versus something else?

 

Yes the USB was quite good compared to my Singxer SU-1 AES out to the Pulse DAC, i would say the differences where subtle rather than night and day.
It could just be that the USB input of the Pulse DAC is just as good compared to the SU-1, I've read somewhere that the USB input uses femto clocks with sub 2uv noise.
Maybe i should try to get my hands on the lynx/mutec/rednet etc and compare.

 

Yes, I imagine USB implementations are improving - and your PC may also have a relatively good implementation.

 

I've had the RME9632 in my system for a little over 2 weeks now and I've been able to use it in lieu of USB on both the Modi Multibit and my new NFB-28. In both cases the improvements were readily noticeable, but I felt the Modi Multibit benefited more from moving off USB than the NFB. I used the included breakout cable to output coax SPDIF.

Pretty much just parroting what everyone else has said at this point but I've noticed a blacker background, delineation of instruments/vocals, less smearing in complex passages and the elimination of any traces of digital hash that I can pick up on. The biggest improvement however and the reason why I will feel uneasy ever going back to USB is a complete sense of ease and effortlessness to the presentation that comes from being able to distinguish nuanced changes in volume.

Thanks to @Garns for recommending this card and answering my stupid questions. It's significantly cheaper than the Lynx ones (which I think are a bit overkill for this application), and I've seen it go for as low as $105 on eBay which is a steal when you compare it to some of the USB de-fuckifying solutions people chase.

 

Got all my parts and cables tonight and things are up and running into the Gungnir Multibit. Listening to Beck's Morning Phase, the reverb seems to be way more noticeable and the background really seems blacker. Overall it just sounds less "gritty" and more "flowing". Sorry, I know I am piss-poor at describing sound.

My chain is:

AES16 -> AES1604 Breakout Cable -> Neutrik Transformer AES to BNC -> Blue Jean Cable BNC -> Gungnir Multibit.

Reading through the manual, I see the BNC connection on the breakout cable can be an output with 75ohms. I'm sure I'm missing something but out of curiosity, why the need for the transformer?

 

The BNC on the AES16 is used for clocking purposes only, not data output.

 

Right! Thank you Gunner.

Currently having an issue with higher res files playing properly. 44.1/48/96 all work but when I get into 24/192 and DSD files(downsampled for the Gungnir Multibit), it's a computerized gargling.

Actually what Foobar is doing is playing the song in FFWD! Only when using the Lynx on 24/192 formats. When using the USB for the same format, it plays as it should...