That was for Lumin models older than X1 T2.
If your Max Volume is 75%, but the main volume (in the Lumin app volume circle or the volume number seen in Remote Remote app) is 60, then the actual volume = 60 * 75% = 45 when Max Volume is 100%.
Thx. That the current volume and max volume setting are multiplied is clear. The Lumin settings support page explains this.
My question is, what does a total volume of say 50% mean in terms of output voltage?
Say my T2 is in output mode normal, using RCA. So the max output voltage is 3V. How does 50% volume affect it? Is is a simple 0.5*3V or is there more behind it? Could you shed some light.
Regarding the output setting âlowâ lowering the signal by 9.5db - thx for confirming. Turns out I was doing all math wrong to figure out if my new amp has enough power to drive my speakers loud enough⌠it would be great if the settings page could be corrected to explain the differences per Lumin device.
As always, I am impressed by the way you/Lumin is taking care of customers questions. Highly appreciated!
1 Like
Exclude the lowest part of the volume scale, and exclude Lumin M1, each volume step is roughly -0.5dB.
50 steps decrease from 100 = -0.5dB * 50 = -25dB
Feed this into the bottom calculator of this web page:
You get a voltage gain of 0.0562341, multiply this by 3V results in 0.1687V
Although people will disagree with me on an objectivist forum, do not just use math as a basis to determine whether your amp has enough power. You should google for real world experience with your speakers. Simple math based on superficial specs do not give you the whole picture with dynamically changing speaker impedance when playing real music.
Sure Peter! I was thinking the T2 output âlowâ was around 2v (-4db) but as you confirmed it is actually around 1v (-9.5) db.
The former would mean I am driving the Pass XA25 already past its officially rated spec - at around 50W@4Ohm. The later means I am not even close to the max rated spec, only using around 12W@4Ohm. So this is actually really good news to know there is quite some headroom left. I would not be able to do any further math 
T2 X1 Analog Output Level Normal for XLR is 6V rms, RCA is 3V rms.
T2 X1 Analog Output Level Low for XLR is 2V rms, RCA is 1V rms.
Peter,
You comment that it is optimal not to mix the Leedh volume control with another digital volume control. This implies that the optimal gain staging for me is to set the volume control in my active D&D 8Cs to 0 dB and set my Benchmark DAC2 HGC in bypass mode and then regulate the volume with Leedh activated in my Lumin U1 Mini.
This works very well when listening at a bit louder volume, but I often listen at (very) low level, which for many modern recordings means that the volume control in my U1 Mini is between 20-30. I guess that this large attenuation takes Leedh out of the range in which it is lossless. Am I correct here?
Actually, while I definitely hear an advantage using Leedh when the volume is above at least 40, I am not so sure when the volume is at 20-30. In this case I think I prefer attenuation 10 dB in the D&D 8Cs or in the Benchmark DAC so that the volume in the U1 Mini is between 40-50 instead of 20-30.
I read in the white paper at the Leedh web site that the researchers suggested that adding dither to the Leedh control would be an improvement in some cases and I wonder if my case, i.e. very large attenuation, is such a case?
D&D 8C is a DSP speaker, instead of a pure analog pre-amp volume control Leedh Processing Volume is usually compared to. With more bits DSP can work better. In this case, youâll need to experiment to find the best configuration most suitable for your needs.
Ok, thanks for the reply.
In case you donât need the other inputs from the DAC, have you compared the DAC output vs Lumin AES output to the D&D 8C (note: need to use Roon to convert DSD to PCM)?
That is a very interesting question. I have not done that comparison yet simply because I do not have any AES/EBU cables.
I currently feed the D&D 8Cs with a balanced analog signal from my Benchmark DAC that I am using as pre-amp since I have more sources than the U1 Mini. The D&D 8C only has two inputs: (1) XLR input that that can be toggled between AES/EBU and analog XLR and (2) Ethernet, the 8Cs will soon become Roon Endpoints and when that happens I can connect them directly to my network with one Ethernet cable to each speaker if I want to use Roon.
In any case, since I have other sources than the U1 Mini, I have postponed the investment in AES/EBU cables until I find the optimal solution for a pre-amp with very good digital volume control.
I am a bit amazed that the market for quality digital pre-amps with a full set of digital inputs and outputs is almost non-existent. So far I have only found one such unit that fulfills my demand for I/O and that is the MiniDSP Studio. Hopefully, though, (and I reiterate) Lumin will see this market potential and develop the U1 Mini to something similar to the MiniDSP studio, but with the better Leedh volume controlâŚ
I have now received my AES/EBU cables (Mogami 3080) and compared going the from U1 Mini directly into my speakers over AES/EBU and going over my Benchmark DAC2 and analog XLR to the speakers.
My initial observations from comparing attenuation in the DAC and attenuation in in the U1 Mini with the Leedh processing remain the same when going only digital. That is, unless attenuation is large, attenuation through the Leedh volume control (no attenuation in the DAC) seems superior compared to attenuation through the digital volume control in the DAC (no attenuation in the U1 Mini). Similarly, with not too much attenuation the Leedh volume control works very well when going directly to the speakers over AES/EBU.
In short, using the Leedh volume increases resolution, in particular in the bass and midrange (perhaps I am too old to hear the same in the treble). The sound is also a bit âleanerâ, but that may of course be the effect of loss of distortion. I bit like comparing Audirvana and Amarra, with Audirvana more precise and âlight-footedâ and Amarra thicker and more euphonic (but the differences are bigger when comparing the volume controls).
However, when applying large attenuation, subjectively things start to change. When going below 40 or so (depending on recording) using the Leedh volume control in the U1 Mini, the sound starts turning thinner and sharper/harder. Using the same digital attenuation in the DAC (no attenuation in the U1 Mini) or the speakers produces a smoother sound, perhaps less resolution but a more relaxed sound.
This is a bit of a problem for me since I normally listen at rather low levels. In particular, when going directly to the speakers over AES/EBU the signal is very strong (compared to the Benchmark DAC in which I have the internal passive output pads in the -20dB position). In this case, normal listening levels for me are with the Leedh control between 20-30 and for some recordings below 20. This is when having the volume in the speakers at -10db (it goes from zero and downwards). To put this in perspective, when having the volume at 100 in the U1 Mini and using max attenuation in the speakers (- 80 dB) I easily hear music from the other end of the room about eight meters from the speakers. To obtain a bit less attenuation using the Leedh control I have now reduced the volume in the speakers to -30dB.
I have a question for Peter after reading the âLeedh processing documents for engineersâ, document â2-LPV-How it works.pdfâ at:
If I understand this document correctly, with 24-bit processing loss-less attenuation is possible down to -30.10dB. However, with 32-bit processing attenuation is loss-less down to -78.27dB (and more if one accepts larger attenuation steps than about 0.5dB).
My own observations are in line with 24-bit processing. Peter you wrote âExclude the lowest part of the volume scale, and exclude Lumin M1, each volume step is roughly -0.5dB.â That is, 30dB attenuation corresponds to a volume around 40 in the U1 Mini. This is also the point where I (subjectively) start to notice a reduction in SQ when using the Leedh control.
So my question to Peter is: is the U1 Mini operating with 24-bit or 32-bit processing? Or is this not a relevant question with respect to loss-less attenuation in the U1 Mini. Perhaps I have misunderstood what they are saying in the document that I refer to above. They are talking about what happens in a 24-bit or 32-bit DAC, which perhaps is a different question than what happens in the U1 Mini where no D/A conversion occurs?
AES can output 24-bit, while USB can output 32-bit. I think this difference makes less impact with Leedh Processing Volume than traditional volume calculation, especially for low levels of attenuation.
1 Like
Thanks for the answer. Interesting. For me this boils down to a true blind test of the effect of lossy digital volume attenuation. It was only today that I read the âLeedh processing documents for engineersâ that I referred to in my previous post, but I have all the time perceived that when going below a volume of 40 (which seems to correspond to the level when the Leedh control starts to become lossy for 24-bit processing) things start to get more and more unpleasant. From the white paper on the Leedh Processing web page I understand that the Leedh control is not dithered as the researchers suggest adding a dithering stage as an improvement of the Leedh processing. The fact that the Leedh control is undithered may explain why I perceive a significant detoriation of SQ at very low volume levels.
I think Peterâs answer to my question also gives me the information I needed to get the best possible gain staging. I simply attenuate the volume in my speakers (which is dithered) until I donât have to go below 40 in the U1 Mini volume control (which is about 25dB attenuation in the speakers).
This creates one more question to Peter or someone else here that understands this better than I do. Since I send a lossless signal to the speakers if I donât attenuate more than about 30dB in the U1 Mini, will the remaining 25dB attenuation in the speakers also be lossless given that the volume control in the speakers is a 24-bit control? Or am I trying to have the cake and eat itâŚ
If this sounds good to you, then itâs a good setup for you.
Please donât worry about the digital volume inside a DSP speaker, as itâs only a very small part of everything it needs to do.
1 Like
Yes, you are right. I donât worry that much, it is more that I am curious and lacking in knowledge regarding the technical aspects about digital attenuation. Which brings me to a new question. I mainly use the U1 Mini to stream Tidal, i.e. 16-bit CD or MQA converted to 24/96 PCM. What is outputted to the SHD Studio (that I am now waiting delivery of) over the AES/EBU and the USB output respectively? is the Tidal stream up-sampled to 24-bit or 32-bit?
Simplistically speaking, yes.
Thanks. I have asked so many questions now so I donât want to ask what âSimplistically speakingâ means⌠
I have now extensively tested the difference between AES/EBU and USB out from my Lumin U1 Mini and I am a bit surprised that it is indeed an audible difference.
My signal path is Lumin U1 Mini â AES/EBU or USB â MiniDSP SHD Studio â AES/EBU â Dutch & Dutch 8C.
All AES/EBU cables are Mogami 3080 (I guess the most common AES/EBU cable in recording studios) and the USB cable is Audioquest Diamond.
AES/EBU is more dynamic, bass is slightly louder and everything is more staccato compared to USB. I also think the resolution is slightly higher and bass is more delineated.
USB is smoother, and to use a standard phrase, a tad more organic and more relaxed. The better the recording, however, the better AES/EBU fares and the harsh sound (in comparison to USB) when listening to poorer recordings translates into high resolution and perceived neutrality with good recordings. But still, the USB is very relaxed and niceâŚ
I am, as I said, surprised that there is a clearly audible difference and I wonder if the difference is due to cables, the Lumin U1 Mini or the MiniDSP SHD Studio. Any MiniDSP SHD Studio users with experience?
1 Like
I often ask users to try both AES and USB and determine which sounds better to them. What you reported is interesting, and much more detailed than the typical answers I got.
I have one more exercise for you. 
With 44.1kHz music, try resampling to 48kHz in Roon vs bitperfect output. See which sounds better in AES. (The 48kHz figure is due to your DSP speaker.)
Good idea. I will try that. I, however, have to confess (this is a Roon forum after all) that I currently do not use Roon even though I have a license (my daughter has expropriated the computer with the Roon coreâŚ). Furthermore, I only use Tidal and the few music files I have is on a USB memory directly attached to my U1 Mini. For now I simply donât see much advantage by going over a Roon core instead of going directly from the Tidal server over Ethernet into the U1 Mini. Is there any advantage?
Perhaps the answer is the possibility to resample 44.1kHz to 48kHz. I will connect the computer with the Roon core in order to test your suggestion.
By the way, I am curious if the users you ask in general prefer AES/EBU or USB out from the U1 Mini? And are their impressions similar to my impressions?