miketlse
Headphoneus Supremus
- Joined
- May 8, 2016
- Posts
- 5,963
- Likes
- 3,970
Didn't realise that you were a Genesis fan.I’ll give Duke a go
Didn't realise that you were a Genesis fan.I’ll give Duke a go
That's one of the reasons I subscribe here. The people and their musical picks.I really liked noticing over time that the M Scaler thread seems more active than most other high-end threads with music suggestions and reports of subjective enjoyment of certain pieces, new ones as well as rediscoveries in our library.
It appears that the HMS helps in reconciling the 'music lover' and 'gear lover' within us audiophiles quite nicely
Below are my hits for today
Regarding ferrite cores. Dipping my toes into the rabbit hole, I see a comprehensive product catalog of Laird ferrite cores, which are available from Mouser -- PDF link.
Their ferrite cores come in three categories:
You can see the impedance graphs for each on page 4 of that PDF.
- Low frequency
- Broadband
- High frequency
Which would be most appropriate for use with the M Scaler, I wonder? I'm leaning towards a mix of broadband and high frequency, but that's just a guess...
In fact I use a track which has a tendency to be on the harsh side of the spectrum anyway and with any significant RF noise it soon becomes unbearable in terms of fatigue and brittleness of sound.
I will try to give general guidance [...]
This is a very deep and important question. We know RF in the 2-2.5GHz is audible with mScaled signals (353/384kHz on each BNC line). Get a WiFi base station configured for 2.4Ghz and put it next to your DAC and you'll hear it. From the early ferrite experiments, ferrites targeted to the 2-2.5GHz range had a nice positive impact. I never got to point where it was too much (just kept adding ferrites, and eventually I couldn't hear any more difference).
But what happens as you filter more and more frequencies closer and closer to 353/384kHz? As you filter more and more high frequency components, wave forms get softer and softer (lower rise times, less sharp transitions, etc). At what point does the wave form of the digital signal start to impact SQ? With the phase lock loop on the receiver side, do the details of the waveform matter at all?
The reason this is curious for me is that Rob has pretty clearly demonstrated that jitter doesn't do anything material with his DACs. However, server changes are still audible, even via optical inputs. If jitter isn't the root cause, the only things left are waveform signal integrity changes (rise times, overshoot/undershoot, amplitude variations/modulation, etc) or some air transited RF, or some mechanism I haven't thought of yet. Some how whatever this factor is is inducing something on the receiver/DAC side. The alternative is some air transited something impacting the DAC, but I'm more bearish on that route (there would be too much room to room variability...reports of SQ changes with server changes seem to me to be too consistent for that).
If we figure this piece out, I'm convinced it will be the X factor that lead to some optimal digital source design/implementation. I'm very much looking forward to hearing what Dan hears and sees in his experiments, and maybe what Nick found out in his experiments with different ferrites and configurations (I trust Nick's ear for RF impact on Chord DACs to be amongst the most practiced and refined out there).
What's your favorite test track Nick? I also have a group of favorites and similar strategy for quick checks.
In particular, I lean heavy on Handel: Messiah - And the Glory oof the Lord by Dunedin Consort (Linn) for the spatial presence of the individuals in the choir (height, depth, width) and room, but also for borderline harshness in key sections. Taming RF and better power is very clearly audible to me.
For more nuanced evals, Arnesen: Magnificat - Fecit potentiam by TrondheimSolistene (2L) has been and continues to be go to to assess "emotional connection" (along with all the amazing spatial richness...as RF goes down, the sense of being in that church with that choir and ensemble on that day goes through the roof)
This is a very deep and important question. We know RF in the 2-2.5GHz is audible with mScaled signals (353/384kHz on each BNC line). Get a WiFi base station configured for 2.4Ghz and put it next to your DAC and you'll hear it. From the early ferrite experiments, ferrites targeted to the 2-2.5GHz range had a nice positive impact. I never got to point where it was too much (just kept adding ferrites, and eventually I couldn't hear any more difference).
But what happens as you filter more and more frequencies closer and closer to 353/384kHz? As you filter more and more high frequency components, wave forms get softer and softer (lower rise times, less sharp transitions, etc). At what point does the wave form of the digital signal start to impact SQ? With the phase lock loop on the receiver side, do the details of the waveform matter at all?
What goes down the cable will be lot higher than 384kHz.
384.4kHz x 32bits = 12.3Mbps per channel.
So you want a ferrite that targets the 2GHz range and stays away from 12MHz, the HF ones look best.
Edit: For the mscaler it runs at 768kHz per BNC cable so double that above.
Thanks for the thoughtful reply. The reason I asked is because I tried an experiment with an RF blocker I happened to have on hand; a pair of these: http://powerwraps.com/
My guess is these wraps filter at a lower frequency as they really mucked with that sound. They initially resulted in a darker sound - but it was a step backwards in transparency. I was surprised by how much the sound changed - for the better - when I removed them.
I’m now experimenting with a more moderate application of these in that I have my coax cables running through only about a quarter of the wrap. It seems to have stripped away a bit of hardness without doing any harm. Next step will be to obtain some HF ferrites to see what effect they have.
Thanks to you and @Triode User for your great contributions here. Much appreciated.
If something quotes a patent I always like to read it.