Read this again:
“It is a mistake to think that measurements are all that matter.”
I read it the first time, and for the record, I didn’t say that measurements are all that matter. I just said that in my view, they’re important.
I appreciate that there are some things you just can’t measure, like rolling tubes. I’ve just replaced the ageing Harma 6922 cryos in my preamp stage with a pair of Russian S6N23P-EV tubes. The difference spectacularly unsubtle.
My observation is based on CPU utilization, if the Nucleus handling DSP and Upsampling, the CPU usage will be high and the power rail noise may arise, will affect the USB bus power supply to DAC connected to it, if the USB out have a isolator/reclocker with LPS Supply maybe less affected, if the Nucleus only do the light work just a straight playback, connected directly to Nucleus should be fine, otherwise connected with Bridge or Pi based endpoint, the endpoint only receiving and playback processed audio data from Nucleus, should be way quieter than Connected to Nucleus directly without handling computation intensive jobs, and the cost of providing LPS for Bridge and Raspberry Pi is way cheaper than LPS for Nucleus.
3 Likes
You want spectacular? Get some Western Electric WE396 or JW 2C51 tubes along with 5670 to 6922 adapters and try them.
Graeme, I think that for some pieces, the measurements are almost all that matter. Not all the pieces, but some.
This is no longer 1970, where the sources were stereo FM or tape or vinyl. We are using digital sources and running Roon, a sophisticated music manipulation system with built-in DSP, which can add any tweak you may want, any euphonic distortion you may desire, to your music. We no longer need to swap tubes or op amps to get “the right sound”; we just need to learn to use the tools we’ve already paid for, and already have in the audio chain.
That is, provided the rest of the chain provides accurate reproduction of the sound Roon works so hard to build.
This means getting a DAC which can accurately reproduce the analog waveform captured in the digital samples, and an amp which can amplify that signal without distortion, and has enough power (voltage) to drive your speakers to a desired volume, and enough muscle (current) to throw the speaker cones or horns around as needed, that’s what counts. And those capabilities can be measured. Quite accurately. So, for DACs and amps, measurements really are all that matter.
Almost. If you go by other people’s measurements, like Amir’s or Archimago’s, well, they are measuring a different unit from the one you have, even if it’s the same model. So you need to either obtain the test equipment and do your own verification of their measurements, or only choose equipment from companies with enviable records of quality assurance and consistency of manufacture, to ensure that the equipment other people measured is as much like your units as possible. And that can be hard to judge, for most people. I think that’s really where professional audio reviewers offer the most help, in their view of the companies and how well they’re run, rather than their subjective assessments of the performance of specific pieces of equipment.
Now, the speakers interface with the physical environment, and are a different matter. They actually generate the sound we listen to, but only in conjunction with the setting they’re placed in. Ideally their coupling with the environment could also be measured, and that’s what the automatic room correction software built into modern amplifiers attempts to do. But it’s a complicated process, and the mic’ing on the automatic systems is usually insufficient. We’ve seen many satisfied reports on this forum of measurements done by hand, then fed back into the Roon DSP, so a manual process can work, but what if your speakers so mis-match the room you can’t ccrrect for it? Always listen in your room before buying speakers.
Anyway, my two cents.
5 Likes
Might have to try those. They’re much cheaper per valve than the S6N23Ps too!
I wonder if connecting a jitterbug thingy between a Nucleus and Dac would resolve many issues vs a dedicated streamer?
Who knows? I mean, who knows if there actually are any issues to resolve? And if there are, who knows if the Jitterbug would do any good? I have to say, it seems unlikely, if your DAC is any good. Take a look at this evaluation, for instance:
4 Likes
I don’t perceive any issues with connecting a DAC directly to a Nucleus. DAC or streamer depends on how close your hi-fi is to your Nucleus. If it’s close enough, you can plug a DAC straight into it. If it’s too far away, then a streamer is the way to go.
The jitterbug is pure snake-oil audiophoolery. Save your money. If you need a streamer, the cost of the jitterbug would go a long way towards a Raspberry Pi 4.
5 Likes
Is the nucleus (no internal HD) quiet enough to place on my actual audio rack? I may go that route vs buying a streamer. I only play anything within the Roon framework.
Roon’s Nucleus is passively cooled, so if it’s anything other than totally silent in operation, I’ll be very surprised.
1 Like
Having owned quite a few hi-end streaming systems, and recently reverted back to an NUC (roonserver) > (USB out) > Vitus RD-100 DAC (USB in) … I’d have to say NO, and a streamer is completely pointless IMHO.
The NUC (10th gen i7) is used as RAAT rooncore. The library is on a QNAP NAS, which has 4TB’s of SSD and 8GB RAM. As its all totally silent, it sits in the (rear vented) cupboard less than 1m away from the system.
6 Likes
I think I will add a Nucleus and go direct into my DAC. I’m going from an old imac as my core with a long USB cable. I’m probably going to have an improvement just by moving the Nucleus to my audio rack. I can always add a streamer later if I feel the need.
3 Likes
@Dale_Clark That’s the best formation IMHO; rooncore next to USB DAC > 1m cable.
Music on a NAS and the Nuc-leus streams from it over Ethernet.
1 Like
Any competently designed DAC should deal with upstream issues within USB spec, though, so there’s a reason you can’t hear it.- it doesn’t generally come out the other end.
Sonore, for instance, puts a lot of real effort into addressing streaming artifacts, but to little audible, or even measurable, effect. I asked them, they won’t supply measurements, which is a tell for anyone claiming a lower noise floor (measurable).
1 Like
I used to be in the streamer between Roon Nucleus+ and DAC crowd, but have since pivoted and have not looked back. I used to use an Auralic Aries G1 streamer USB to Denafrips DDD to convert USB to I2S then into my Terminator+ DAC, but now I simply go NUC+ USB to Denafrips Gaia DDD for conversion to I2S, then to DAC. I can now also upsample to 768k, which my DAC handles with ease. The Aries G1 streamer can’t process above 384k, so that was alwasy a bottleneck as well. Next stop is a nice LPS for the Nucleus+.
2 Likes
A very simple answer: do a test.
For test purposes, put the DAC close enough so USB connection is possible.
Set up both a USB direct, and an internet-through-MicroRendu-USB-to-DAC connection (or if not MicroRendu, whatever streamer you are thinking about).
Compare the sound.
If there is no difference, return the streamer.
If the distance is to great for USB, yeah, you need a streamer.
2 Likes
I recently made some changes to my config. Was using a dedicated computer in main listening room, connected to BDA-3 DAC USB and on other USB port connected directly to my Lumin U1-Mini. to me the Lumin always sounded better than the DAC connected directly to PC. So when I moved Roon-Core to my main workstation in another room, was surprised the sound improved on the Lumin U1 Mini. I assume better Galvanic isolation but don’t have a way of measuring. On switching back just using the directly connected (USB) BDA-3 to the workstation as an endpoint I was able to confirm the sound was slightly diminished, less polish and detail. Now both workstations have upgraded power supplies that measure well for low voltage ripple so not quite that simple. Then the hydro bill came in and was 10 less for month and we have a win win situation…
3 Likes
I also have a Bryston BDA-3 DAC and a Lumin U1 Mini and can confirm that the Lumin makes a quite surprising improvement in the sound quality coming from the DAC. I can’t explain why, but it’s definitely an improvement.
4 Likes
I have a JCAT USB Isolator between the Lumin U1 and my DAC. A definite improvement IMHO in SQ.
1 Like