Discussion in 'Digital: DACs, USB converters, decrapifiers' started by Luckbad, Apr 1, 2016.
spotify never sounded so good, great update
not to be contrarian, but didn't notice major diff in my chain with PCM high res files, or 16/44.
Mutec3+USB AES --> Yggdrasil SE --> EC Uber 7 --> HE1K.
maybe a bit more bass kick at times, but can't confirm this across every track just yet.
I don't have my Yggdrasil yet, but I've heard the improvement across 2Qute, Bifrost Multibit and Modi Multibit through a Magni 2U and Ether C. I have misterrogers' uber USB source kit plus microrendu and Rednet chains. I've been listening primarily with Modi Multibit/m2u today and then updated the Mutec right before I posted. All DACs through the m2u improved from their respective baseline signature. I will definitely listen more through Asgard 2 and ZDS. I can flash the older firmware, but it's readily apparent so far.
To put things into perspective, my system has been relatively immune to the digital chain upgrades. It wasn't until I got the schiit DACs to experience the positive contributions to sound made possible through the two digital chains. Previously 2Qute saw no apparent improvement compared to MBP direct USB. Today's update seems to have been the breakthrough in sound I've been hoping for.
congrats for having a solid USB chain, but I dropped off at "Ether C".
wait until you get more positive contributions to the sound with a different transducer.
ps. nothing personal. i haven't heard Eth-Closed v2.0. but v1.0 was crap. simple.
Here's the thing. My Ether C's are my most lackluster cans. For me to hear the improvement I did is no small feat. I will listen more with modded 650's and stock 800S.
The ETHER-C is very resistant to and unrevealing of source chain updates, based on personal experience. This is very frustrating in fact. So the fact that audible differences from firmware were realized through the ETHER-C is pretty interesting actually - like the ETHER-C or not.
Now as to why firmware makes difference for PCM, I'd love to hear the explanation. Sorry, but I just don't buy that source interface clock jitter and recovery is likely to be significant enough by itself so as to be the answer.... unless they had a bug in the algorithm in the first place.
They changed the algorithm (according to their site). Since that's pretty much what this device does in PCM, I'd expect changing the algorithm would change the sound.
It's like fixing the timing of your turntable. Sure it's pretty accurate, but it could always be more.
Don't think I see the turn table analogy as being of like kind.... A device like Mutec isn't really "reducing" jitter, but more so reclocking and adding its own jitter. If I have digital data pattern 10100011 sitting in a file on the computer, then as long as 10100011 ends up in my buffer on the DAC side, it should not matter how much jitter accuracy was obtained in between. If bits are flipping somewhere in the chain, then that's a bug. On the other hand, if their "algorithm" is doing some DSP, then that's a completely different ball game. And electrical isolation, now that's where things could really get interesting especially with those DACs that have no isolation of their own.
Digital Audio is as much about the timing as it is about the data. I bet this algorithm change makes the bits line up more accurately to the clock, hence the better sound. I'm not really sure it has too much to do with what's coming in and I'd be shocked if this is doing any DSP. Getting bits to line up properly to a clock is hard, even with a really good clock.
But if USB was not a robust interface for data transfer due to difficulty in properly clocking data streams, then we'd see data corruption on our thumb drives in all sorts of documents all the time. I'm not aware of an error correction scheme. And even non-isochronous USB doesn't have a retry mechanism that I'm aware of either. Also, USB doesn't really have a clock per se, at least not in USB 2.0. The first block in a USB PHY chip is a CDR - Clock & Data Recovery. Bits will flip or shift if CDR isn't doing its job, but that shouldn't be a firmware problem. I'd be surprised to learn firmware influenced the USB physical interface. I'd also doubt there is any DSP.
This is where USB audio messes with those of us who are in the bits are bits crowd (myself included). USB Audio isn't just transferring bits to your DAC and the DAC deals with them and puts them in order with correct timing, USB Audio is streaming bits (much like UDP) which is pretty much unlike how USB is used to put bits on a thumb drive at all. It essentially is a quirky protocol similar in idea to the way SPDIF works. It has a clock for the protocol, USB is just the connector. USB Audio tends to suck because the clock of your computer for audio tends to be not amazing. This is why asynchronous USB DACs became a thing.
Not sure what magic mojo Mutec has done with latest firmware but I'm really enjoying the new sonic changes.
Mutec MC3+USB > Lifatec Glass Optical > Yggdrasil(XLR) > Audio-gd Master 9 > Yamaha HS8 is really fooling my brains with this song(PCM version via Musicbee ASIO, not playback from youtube):
It is as if I am sitting inside a Cathedral, and Sarah McLachlan is singing live to me. The firmware seems have added a sense of height to the sound. It may seem absurb to some here, but I am just describing what I am hearing.
to me, the update added a certain liveliness to the music, the notes sound fuller and more real. I wouldn't call it a drastic change, but it's certainly more enjoyable.
Yeah, so the physical interface isn't the issue. I.e. cycle to cycle or bit to bit jitter of data in flight. If this were an issue, then we'd not be talking about sound quality - we'd be saying it either works or it doesn't work. So I think it's safe to simplify it for now and say that really two things have to be accomplished here: (1) Correct data needs to be received by the device and (2) it needs to be received at the right time. By "the right time", what I'm really thinking of is isochronous channels, i.e. guaranteed bandwidth. But what I think is very important that you touched on above is the comment about it "isn't just transferring bits to your DAC and the DAC deals with them and puts them in order with correct timing". So what we're *really* talking about is sample to sample jitter. And I had implicitly assumed that an audio-oriented USB implementation like an XMOS Audio Class 2 device would actually design in appropriate measures to mitigate or avoid it. For e.g., what prevents the receiver from introducing some latency by buffering up enough data at the interface to avoid underflow and then tightly controlling the sample clock on the device side, rather than relying on absolute timely and consistent arrival of the data stream itself? It seems as though a store and forward architecture would go a long way here...
Well, blow me down. The Mutec 3+USB firmware upgrade is really a discernable improvement. More detail, less smoothness, more energy, but still pleasant. I think it does sound quite a bit better, certainly more nuanced in volume. I ran one on the old software for two hours, ran one with the new software for two hours, but I only really needed 10 minutes to be sure that it was. This elevates the 3+ by quite a bit, to the point where I can let go of the Berkeley, well maybe. It serves poorly recorded music well, because everything is more pleasant, but the upgrade pulls back in the make everything sound smoother shtick, and puts some microdynamics back.
I think it still fleshes out the midrange more than is in the recording at times, thereby losing timbre and some clarity (not much) but it is appealing to the ear. In Michael Hedges 'Aerial Boundaries' I hear the decay less with it than with the BADA, and less space between the notes. But it reduces glare on bad digital like the Bangle's 'Manic Monday'.
And it works well with the mircrorendu, because that tends towards flattening dynamics and initial attacks as well while drying the sound a little. Is it accurate? Nah. But its impressive sounding, the kind that prevents earbleeds and fatigue.
Don't buy it if you think this is end of game, its not. Buy it if you like the sound enough and have stuff that's digtal only. Then forget about it. For a while.
Just wondering has anyone here tried to pair an external 10MHz word clock to the Mutec? Would like to hear your impressions.
Mutec's upcoming word clock would be quite interesting for mutec MC3 users.
Link to upcoming Mutec clock?
It's called Mutec REF10
sorry for the stupid question, but wouldn't it be better to just get a master clock and make the digital signal go through it before getting into the DAC ? I'm not sure I get the purpose of say the MC3+ without the clock generator.... what is the purpose of each one of those?
The MC3+ can act as a word (master) clock generator. Part of the above video is describing an additional higher precision/higher quality 10Mhz external clock (the big long box) which you can feed into a MC3+ to improve this already existing functionality. The 10Mhz clock is useless by itself.
Separate names with a comma.