I am using a Grimm Audio MU1 which is a digital to digital converter, Roon streamer and FPGA upsampler that sends digital audio to the LS60 via S/PDIF. This is a MU-1 is a high-end device that produces astonishingly musical audio. However, when using it with the LS60, I don’t seem to get the same benefit through the coax SPDIF input . The LS60 always seems to sound better when direct streaming via Ethernet from Roon. Why would this be the case? The same is true with the analog RCA inputs on the LS60. Analog audio from a Mola Mola Tambaqui DAC into the RCA inputs on the LS60 does not sound as good as The Roon Ethernet input and it should sound MUCH better.
I would truly like to understand what it is about the LS60 and Roon that would favor the Ethernet input. Have others noticed that there is marked difference using the KEF LS60 and Roon/ethernet connection vs any other input of the LS60?
I seem to remember that from watching a some reviews of this device when it came out, that it resampled all digital and analogue inputs to go through the KEF stack.
Maybe that is countering the work done by the MU1.
If it is going through directly from the Ethernet port via Roon then maybe it is not being reworked twice. Just a thought.
3 Likes
Bill_Janssen
(Wigwam wool socks now on asymmetrical isolation feet!)
3
I think that’s right. KEF, both the company and the speakers, know what they’re doing. Performing multiple transformations outboard could induce distortion.
Why should the Tambaqui sound better? One well-done DAC should sound pretty much like any other well-done DAC, to begin with. And the extra A/D conversion step required by the speakers can’t improve the sound.
As for the MU1, what are you comparing the LS60’s to, when you say “same benefit”. Something else sounds better when fed via S/PDIF? What is that?
Let’s say that I am streaming a symphony via Roon into the MU1 for conversion and upsampling and whatever else it does. Then, that result is output via the coax SPDIF into a TA HA200 headphone amp and listened to with the best headphones I own. What I hear is a marked difference. Musicality, texture and engagement is very apparent vs bypassing the MU1. The same is true if I send the Roon stream to the MU1 first, then via AES into the Tambaqui, then analog into HA200 amp or analog into a Warwick Acoustics Aperio headphone amp. There is always a marked improvement. If you’re interested, there is a lot out there on the synergies using the MU1 and the Tambaqui. Now, admittedly, these are high end headphones I’m ultimately listening to and not floor standing speakers. That might be part of the disconnect, but to my ears the LS60’s coax SPDIF and analog inputs don’t seem to benefit from the improvements using the above mentioned equipment and connections. I am just trying to understand why. I am not debating whether a $13,000 Tambaqui is worth it, I am just saying I have heard improvements on the headphone side of my setup. I have written to KEF to see what they say. I bought the LS60s without audition, based on the reviews and raves. I am just trying to get to know what they can and can’t do, especially with regard to ROON streaming. If, in fact, the MU1/coax and analog connections don’t improve things, so be it. I’ll disconnect the MU1 and Tamabqui and keep these as headphone only devices. Thanks for your comments!
Yes, but to experience the full effect of what they do, you need to play them a bit louder than I’m comfortable with in my apartment setting. They do come alive when they are cranked up a bit. I am still experimenting with the KEF EQ options along side the ROON EQ options. I am trying to achieve a natural texture for violins and other stringed instruments. I haven’t quite figured it out yet. I ordered a KEF KC62 subwoofer just to take some of the work off the midrange/treble driver so maybe that will help it bloom a little bit more.
If your using the analogue in on the kefs they will be resigitized by its ADC then it will go through its internal DSP as crossover and other stuff they do under the hood to its inbuilt dac and amplifiers for each driver. Using an outboard DAC is utterly pointless with active speakers like these as your doing too many conversions.
Going coax likely again not the best option again as the device is likely downsampling it only supports 192/24 and if using wireless link will downsample. Going Ethernet direct less processing is involved and the kefs are doing it the best way for their speakers. You’re over complicating the chain. These are designed as standalone convenient devices that negate the need for outboard DACs and streamers, and have the options to plug other stuff in but likely not using the best components in those bits. They are primarily designed as streaming speakers.
I am going from a router directly into the LS60 Ethernet jack. I am saying that, to my ear, the LS60 sounds much better being connected via ethernet than using any of the other inputs. I am beginning to think that the LS60 is by design meant to stream via ethernet (or wireless, but I haven’t tested) rather than any other digital or analog input. I’m not sure what this all means to me yet, I’m just trying to figure it out. These may not be the right speakers for classical music, early music in particular, at low to moderate volume levels. I’ll need more time to get used to them. My point in posting was to see what others are perceiving as they listen to the LS60 streaming from Roon.
The difference is that I can insert my MU1 FPGA upscaler and Roon streamer into the Coax signal path to play through the LS60 speakers. But as others have pointed out, the LS60 probably isn’t designed to work this way which explains why the direct ethernet path from Roon into the LS60 sounds better. I didn’t realize this before I bought these speakers.
Bill_Janssen
(Wigwam wool socks now on asymmetrical isolation feet!)
18
They’re not identical at all! S/PDIF provides a source-clocked interpretation of the audio data, said interpretation being provided by whatever is generating the S/PDIF stream. Direct connection via Ethernet to an in-speaker Roon or KEF SDK, however, allows player-clocked interpretations to be done by KEF software, apparently to the benefit of the sound.
Wireless: all sources resampled to 24bit/96kHz PCM
Wired: all sources resampled to 24bit/192kHz PCM
So, if the secondary speaker tops at 24/192, it makes sense for the primary speaker to do the same, even if you feed it 24/384 or DSD through Ethernet. Most probably, the internal DSP works at this rate. And, since they do have internal DSP, an external DAC will have absolutely zero benefits. I agree with @Bill_Janssen that this kind of speakers are not designed to work with an external DAC.
If you want to make the most of them, and if you think the internal conversions are not optimal, you should probably convert everything to 24/192 in Roon and use Ethernet.
You are not describing any difference between the two (the same clocking is done in different devices) but you are ascribing an “apparent benefit to one.”