LUMIN - Audiophile Network Music Player

[barrows]

On 6/3/2020 at 8:09 AM, pl_svn said:

almost for sure better than Roon's vc

There is no actual reason to believe that....

[barrows]

1 hour ago, ogs said:

Maybe you are right, maybe not, but what is your explanation..?

Roon volume control is 64 bit.  It is lossless unless one is using enormous amounts of attenuation.  Even LeedH themselves appear to compare their volume control to ones operating at only 24 bits when they claim superiority.  I am unaware of any aspect of LeedH, or of any possible aspect of LeedH, which could make it perform any better than what Roon does.  They are both likely transparent.

[barrows]

30 minutes ago, pl_svn said:

what doesn't make sense, at the very least in terms of "politeness", is Sonore guys regularly popping-out in threads (even in sponsored areas) about their competitors products "depreciating" them

Umm?  Please point out any comments which I have made which are critical of Lumin products, and I will ask Chris to remove them.  I try to make a point of never criticizing any other companies' products (with the very rare exception of products which might cause actual damage or poor performance is used together with a Sonore product).

The fact is I like and appreciate Lumin products, and have spoken positively of them many times on these forums.

[barrows]

Just now, BlueSkyy said:

This comes to mind:

 

"There is no actual reason to believe that.... "

 

While not a slam against any particular vendor, it caused me to take note and wonder if there was some agenda, hidden or otherwise, at work.  Maybe it's just a sign that some of us, me included, are all a little too sensitive these days................

In my follow up comment I made the statement that both the LeedH volume control and the the volume control in Roon are likely transparent.  i do not find anything about these statements to be critical, but I am sorry if you do.

[barrows]

3 minutes ago, pl_svn said:

it actually is the very first time on a Lumin product

can't say the same about Allo's or UpTone's

 

also... it's not you the one who usually does what I said and this wasn't directed at you, I reacted here 'cause... enough is enough 😶

so forgive me for calling out *you*

 

my point, though, doesn't change

I can only speak for myself here.  And again: I like and appreciate Lumin products, and have said many positive things about them on these forums.

[barrows]

1 minute ago, BlueSkyy said:

Thank you for taking the time to explain.  As I mentioned, I did not find the comment critical.  I simply found it to be to be "impolite" possibly because of the abruptness of the original comment.

 

My sincere apologies for overthinking things and please enjoy the rest of your day.

No worries, I am just glad we have it straightened out.  

[barrows]

5 hours ago, matthias said:

 

I think gain matching of the complete set-up is most important as well because the Leedh Processing is not lossless with greater attenuation than 30dB.

 

Matt

It should be noted that this holds true for analog volume controls as well.  An analog volume control is far from "perfect", and analog volume controls add their own noise and distortion to the signal.  The more attenuation one applies with an analog volume control, the more noise it will add (this is resistor noise, which increases with the amount of resistance).  So proper system gain matching is always advised for best performance.

An exception would be the analog volume controls which adjust the gain of the circuit directly, rather than just being an attenuation stage (Ayre, Mola Mola, PS Audio, maybe some others).  But it is important to note that these types of volume controls are not perfect either, generally speaking, every circuit is going to have a certain range of gain in which it performs its best, and large deviations away from that gain level will come with increased distortion and noise.

 

All this being said, in my experience, one can get an increase in transparency and overall fidelity, for sure, no matter how you achieve it, by getting rid of the preamp (assuming the DAC has a robust enough output stage to drive the amplifier's input stage properly).  And I feel, that given good implementation, and gain matching, a digital volume control, or an analog one, is not going to result in audible sonic problems if critical listening happens at -20 dB or above.  Just avoid pairing an amp with 30 dB of gain and 500 watts of power with 93 dB efficient speakers!

 

[barrows]

2 minutes ago, pl_svn said:

... and what about a VC that works by, as its designer says, "variating DAC's voltage reference plan"? (Metrum Adagio)

I have no experience with this approach, and I have seen no measurements of a DAC which uses this approach to confirm or deny how well it works.  I can think of many potential problems with such an approach, such as varying impedance on the Vref supply rails which is contrary to my experience of how to get performance out of a DAC, but without measuring and seeing the results there is no way to know for sure.

[barrows]

1 hour ago, wklie said:

The combination of upsampling/resampling everything via Roon's DSP engine to DSD 512, Leedh Volume processing by the Lumin X1, and the elimination of the DSD to PCM filtration in Roon has enabled my system to push past some threshold into a realism I've never before achieved.

Huh? What?  My understanding is that in order to apply the LeedH volume control the X-1 has to convert back to PCM, Peter is my understanding incorrect?

[barrows]

10 hours ago, wklie said:

At least the Lumin CPU / FPGA is not converting DSD to PCM in this case.  I do not claim to know what really happens inside the ESS blackbox, but I believe no decimation occurs.  Many experts have tried to guess what exactly ESS does, in this forum and others.

The generally accepted understanding is that the ESS chips take in DSD, and do not change it to a lower sample rate, they just re-modulate it to a multibit single, but keep the sample rate high.  So, unlike DACs like Chord (excepting DAVE) which turn DSD into 352.8 or 705.6 PCM, ESS keeps the original sample rate and just turns it into a form of "DSD wide", so it sounds like Peter is saying that this is how the Lumin X-1 handles LeedH volume control in their FPGA, then sending that "DSD wide" signal to the ESS chips.  Peter, please correct me if my understanding is incorrect.

[barrows]

11 minutes ago, Ropet said:

Thanks, @One and a half. I will contact Bosch referring to your tip.

 

At the same time I did not have this issue with AES. I guess it is less sensitive (higher voltage?) than USB. But for USB I lost contacts many times per day during my first months with Lumin. But since approx 2 weeks using USB again it has worked with no hick ups. The only changes made is new version 13/13a. Could not USB connections be more or less sensitive against HF noise?

As this happens when you something turn on, or off, the sensitivity is most likely to a DC offset on the AC line.  When motors turn on, and off, there can be a momentary DC offset.  This may also happen because of home wiring problem related to your ground connections.  I would not expect this to happen with a product from Bosch, for example, so perhaps you might investigate whether the home wiring has any ground issues.  Proper grounding for AC power is always a good idea, just for general safety, so I would consider having an electrician inspect your home wiring.

 

[barrows]

2 hours ago, BlueSkyy said:

I received a new X1 power cable a few days ago from Angus Leung of WestminsterLab in H.K.  Upon opening the box I received a bit of a surprise because I had read somewhere, or maybe imagined, that the Lumin X1 power supply had used a 9-pin connector.  To my surprise, the new cable had 10-pin Hirose connectors which sent a chill up my spine and I thought it might not work.  No worries as the OEM cable was also a 10-pin configuration.

 

I borrowed a friend's Keithley 2182A nano-voltmeter to do some, probably useless, measurements of the noise at the output of the X1.  I connected the meter to the X1's left channel XLR output terminated in a "dummy load" of 20kohms and played a track ripped from a "SilentCD".  The tracks on this disk are described as absolute silence and, with the OEM cable, I measured an RMS voltage level of about .0063741mv which I assume is the noise level of the output.  I then swapped in the new 0.6m Westminster cable with the optional carbon fiber wrap, played the same track and took the measurement.  The RMS value of the noise I measured dropped from .0063741mv to .0042476mv.  If my calculations are correct, that appears to be about a -3.5db drop in "noise" at the output just by changing from the long OEM to this shorter custom cable.  I am also not saying that the drop in output noise level would ever be audible, but it is just a data point, but it still caught my attention that something changed measurably.

I have listened to the system now for several days with the new cable and my initial impressions are that the music emerges from a darker, more velvety background.  Before the cable swap, I used to have to almost strain to enjoy the opening bars of Leonard Slatkin and the Orchestre National de Lyon’s rendition of Ravel’s Bolero in 192khz 24bit.  Either due to the drop in noise floor and/or increase in my system’s ability to reveal more of the low-level information, i.e. “micro dynamics” (?), brought by the WestminsterLab cable, my system now allows me to become fully involved in the music without having to ride the gain control early on in the piece.  Also, on Shelby Lynne’s “Just A Little Lovin’, also at 192khz 24bit, the cymbals shimmer a bit more to my liking and the drum rim shots seem to have quite a bit more “attack” or sharper transients in the opening sequence.   Finally, in Leopold Ludwig and the LSO’s version of Dvorak’s New World Symphony also in 192khz 24bit, the English Horn solo in the Largo, or second movement, is the best I’ve ever heard it.  Although I didn’t do any A/B comparisons to the cable I do intend to revisit that idea at some point.

 

My initial observations tell me that my overall system performance has been improved with the replacement of the OEM power cable.  As the cable runs in and if any other changes, pro or con, are noticeable I will come back with my findings.

I think something is wrong with your measurements.  The noise floor of the DAC while playing with a 16/44.1 bit file (you said the silence track is ripped from CD) should not be able to rise above the inherent noise floor of the quantization noise of the CD source-that is, I would expect the noise floor of a DAC at the level of the X-1 to be lower than that of a Redbook CD.  I think you would need a 24 bit "silence" source file to see any kind of a difference here.  With most of the good DACs I have heard here, the higher noise floor of CD vs 24 bit material is quite audible....

 

@EvilTed: If this aftermarket cable is sleeved in carbon fiber, it is indeed shielded.  Carbon fiber can be a very effective shield, it is highly conductive, and it is also capable of absorbing and dispersing RF energy (hence stealth technology).

[barrows]

28 minutes ago, BlueSkyy said:

Hi Barrows.  I am not savvy enough to understand fully what you are saying and I will defer to you.  I did confirm the files were 44.1khz 16bit and not 24bit.  I wonder if JRiver MC could have manipulated the files it was streaming to the X1?  I'll check that later.  Inn any event, the readings are what I got and will probably go back to confirm the measured change at some point further along with the cable.

 

Maybe not playing anything would have been a better idea?

What I am saying is that the noise floor of CD based playback is higher than the noise floor of most good DACs.  So the noise floor should be the same with CD based playback for both of your measurements, as the noise floor of the DAC itself (what could/might change with the new cable) is not going to rise above the noise floor of the CD playback. 

 

I would certainly recommend repeating a measurement like this a number of times to confirm it.

 

Most DACs mute their output with no signal present, so measuring that way will not give any meaningful results.  This is why your approach of measuring with a file of "digital silence" is a good idea.  But even then, the ESS chip used in the X-1 has an auto-mute feature, which (if activated) will mute the chip's output when playing digital silence.  It would be helpful to ask Lumin if they have the ESS chip registers set with the auto-mute feature "on".

 

The best way to measure the noise floor of a DAC is using a 24 bit file, which has some low level content in it.  Like a couple of tones at perhaps 1 kHz -90 dB, and 500 Hz at -70 dB and then looking at the spectrum.  Otherwise, it is very likely that the DAC is muting its outputs.  By having some signal present, and the lower inherent noise floor of a 24 bit file, one can actually see on the spectrum what the floor really looks like.  The way you are measuring you should see no differences, and if you do see a difference, something is wrong with the measuring technique.  take a look at John Atkinson's DAC measurements at stereophile.com where he shows the difference between noise floors with a 16 bit and then 24 bit file, like here, where the magenta spectrum is the noise with a 16 bit file, and the red spectrum is the noise with a 24 bit file:

 

 

 

 

[barrows]

54 minutes ago, EvilTed said:

It is a carbon fiber sleeve, so it is NOT electrically shielded (in that it is not connected to ground at either end).

Yes, it may provide some benefit, but not as much as a traditional screened cable which uses a metal braid that is connected electrically to ground.

How do you know it is not connected to ground?  For example, i have made AC cables with carbon fiber shields and connected them to ground, it is not difficult to do so, as carbon fiber is highly electrical conductive.  Additionally, carbon fiber shields (or damping is a better term in this use) are still effective if left floating, at least to some frequencies, as they absorb and damp RF fields (they couple with RF energy and dissipate it by turning it into heat).  Consider stealth tech in a non grounded airplane, for example.

How much benefit a carbon shield gives is frequency specific, so one cannot just say it is "not as good", it depends what frequencies one is trying to deal with.  Also, typical metal braid shields are also not all that effective, they help a little, but their level of attenuation is not really that high or that broadband.

[barrows]

8 minutes ago, BlueSkyy said:

Thanks Barrow.  What you say now is more understandable to me and I can see where you are going.

 

Another thought that I had after reading your response is that, even though the DAC's output may be muted with the system not playing a file, it is possible that the analog circuitry post DAC driving the output transformers is contributing to the overall noise floor and would it be possible for EMI introduced by the interconnect between the X1 and its power supply somehow contaminate the output and be detectable whether or not the DAC was muted or not?

Yes, that sounds possible.  One would have to have pretty good knowledge of the design of the X-1 to really investigate that with any certainty.  Some DACs, for example, have muting relays directly at their outputs.  Are you measuring the single ended or balanced outputs?  

[barrows]

2 minutes ago, BlueSkyy said:

I was measuring the balanced output and, although I know relays can be nearly silent and any solid state switching arrangement would be completely silent, I have heard nary a "peep" out of the X1 or, for that matter, the A1 and S1 that came before.

Output muting can also be handled by silent solid state switches at the output, so it is impossible to know for sure, unless one has the schematic of the DAC, or confirmation from the design team.

Suffice it to say though, measuring for noise while playing a 24 bit file with some, single tone, content, and then looking at the spectrum of the noise floor is the only way to truly have an idea of what is going on, as it is the only way to assure that the DAC is not muting.  This method is also much more relevant to how the DAC performs while playing music, as some DACs produce some weird noise floor modulations which are signal related.  In fact, using a signal with a bunch of unrelated low level tones, which might reveal noise floor modulations, is likely the best way to do this-this is probably one of the reasons why single tone measurements (like the -90 dB 1 kHz tone shown in the stereophile measurement above) do not tell us  everything about how a DAC actually sounds in real life, with complex music signals.

[barrows]

Guys, I am not a Lumin user, but I can sympathize with @wklie's point of view here.  It sounds like the DACs which are having problems while using the LeedH VC in a Lumin streamer are not properly applying a mute signal when the DAC needs to re-establish a lock when a file stops playing, and then a new file starts.  A DAC should mute its signal anytime a reliable lock is lost from the incoming signal.  This mute function must be very quick, and it must be reliable.  Some of us who build DIY DACs are well aware of how important a properly functioning mute capability is.  With many DACs problems will occur especially when changing sample rates from file to file, but it can also happen when playback stops for one file, and then begins with another file, and the DAC tries to re-establish a lock.

If one uses Roon as playback software, there is a feature in Roon which may help some.  In the device set up there is a function called Resync Delay, this allows one to set a time where Roon will send silence to the DAC to allow the DAC a certain amount of time to re-establish lock with the  signal.  I typically set this to 500 mS.  I do not know if Lumin's own software has such a feature, but it is worth looking into.  

I would also suggest that anytime one wants to experiment with using a software based VC, one first do so with an analog VC control in place, set to a low level, just to make sure everything works with your DAC without any problems, before fully relying on the software VC.  Especially test what happens when a playlist includes different sample rates, as often the problems can be especially bad when the DAC is re-establishing a lock with a new sample rate, or DSD.

No company can test their streaming product and verify "perfect" operation with every DAC. 

[barrows]

2 hours ago, Liam said:

@barrows Thank you for that very clear information as to why some of us have this issue with Leedh VC.  I noticed the previous poster with the issue was using a Denafrips Terminator and in my own case it is a MHDT Orchid DAC, both R2R design.  Is this relevant at all in your opinion.  Finally is it possible that the appropriate dac delay could be implemented by a firmware update?

I cannot speak specifically to those DACs as I have no direct experience with them.  As to R2R, that specifically would pose no problem, as long as the manufacturer incorporated a mute circuit correctly, but in these cases it appears they have not.  But, to look at it a different way, for example, a DAC made with an ESS chip can usually just rely on the mute circuit built into the chip itself, making development easier, whereas something like the Denafripps which uses a discrete conversion would require the manufacturer to develop a good muting circuit on their own.  Often these mute circuits can be incorporated into the USB interface, but then they would not work for the other inputs.  In my DIY DSC-2 for example, the mute circuit is on the (discrete) converts board, but is controlled by a mute signal coming from the USB interface, and it took the Amanero folks a long time to get their software just right to have the mute work correctly (in order to avoid loud pops at start up, etc).

It is also worth noting that mixed sample rate playlists can be a challenge for any DAC.

 

There is a tendency for streamer makers to blame the DAC, and for DAC makers to blame the streamer, as there are no strict standards of how to connect these things together, and exactly what "should" be required, it is not surprising that both sides often need to make software changes in order to accommodate each other's needs.  At Sonore a lot of software development time is taken up by trying to guarantee good operation with many different DACs and their different softwares-it is a difficult and complex process, and throwing digital volume control into the mix magnifies all of the potential problems.  On the customer side, it would be nice if all components always worked perfectly together all the time, but given all the potential combinations that is just not possible.  The best approach for the audiophile, is always to err on the side of caution with new combinations until one is sure things are working together well.  

[barrows]

5 minutes ago, EvilTed said:

I'll take BADA as an example.

The BADA DACs have a built in volume control. 

They recommend using it at 54 out of 60 (max).

What happens when there is a power cut?

Does it start at 60?

No, it comes back up at 24.

I have no disagreement with the above example, even better would be user stable parameters for what happens, with a default position of  -40 dB or so.  But the above is not the only problem being described here.  My Bricasti DAC just remembers the last VC setting and starts up there.

I would add, that in the case of the X-1, Lumin should have everything dialed as this is a controlled and known situation (at least when using the X-1 as a DAC from the analog outputs).  But separate Renderers and DACs pose a much more difficult set of problems to provide solutions for, and there is no universal answer possible, as there is no way any company can test every possible combination before releasing a product.

[barrows]

1 minute ago, BlueSkyy said:

Regarding the X1's fiber interface, I am currently using single mode transceivers and cables to connect the X1 to the switch and I believe it was based upon a recommendation here but I can't find it.. To set the record straight, Is single mode connectivity "better" than "multi-mode" and, if so, why?

No one "knows" definitively.  But there have been reports of SMF sounding "better" than MMF, I believe some have reported this with the X-1. I use SMF with my Sonore Signature Rendu SEoptical, and it sounds great.  Also, be advised that some are finding sonic differences with different SFP modules.  These are things to experiment with if you are so inclined, only you can really decide if you want to experiment.  I would suggest that if SMF is working for you and sounding good, there would be no reason to then try MMF.  Theory is that because SMF has to be more precise in transmission for longer distance, it may have precision advantages at shorter distances as well-but this theory is unproven, there are only anecdotal reports of "better" sound quality.

[barrows]

1 hour ago, wklie said:

If I'm not mistaken, the SFP module photo on the Sonore web shop is a multi-mode TL-SM311LM and the fiber is also multi-mode, so I'm a little surprised that you use a single-mode setup.

I am running Cisco SMF SFPs with Corning Clear Curve SMF, these things are easy and affordable to try, so why not. The official Sonore recommendation is for customers to use OM-1 MMF with the provided SFPs, but I am interested in trying different things.  I would not mix SMF with the OE SFPs though...

Also going to try some Finisar SMF SFPs soon, so SFP rolling is alive and well, at least for some.  I cannot say for sure I heard any difference with the SMF set up, but it certainly did not sound worse with my Sonore gear.

[barrows]

24 minutes ago, dark said:

I tried a lot, different hardware fuses, in the end I liked Synergistic Research BLACK Fuse the most.

Did you try Orange?  I find Orange easily trumps Black, Blue, and everything else in most components.  Orange seems to have the rare combination of higher resolution AND greater ease in presentation and natural sound.

[barrows]

3 minutes ago, BlueSkyy said:

Quote:  "Did you try Orange?  I find Orange easily trumps Black, Blue, and everything else in most components.  Orange seems to have the rare combination of higher resolution AND greater ease in presentation and natural sound."

 

 

Orange "trumps" Black, Blue and everything else?  (I really have to stop watching so much politics on the telly).

Haha, I know Chris frowns on political comment here, but just so there is no mistake, I believe in globalism, unity, and equality.

[barrows]

2 hours ago, chungjh said:

Has anyone replaced the stock fuse with a high end fuse like SR orange on X1? Does it REALLY make a difference?

Only you can decide these things in your system, I would recommend giving the orange fuse a shot though, as they are returnable fo you do not like what they do.

[barrows]

5 hours ago, FIndingit said:

They are returnable if one lives in the US. Let’s not forget the international audience of this forum. 

You do not have access to a dealer who will allow for a trial?