MDAC first listen (part IV)

[ADL]

mods

can anything related to audiolab MDac or johnW be put in "trade"

this is getting a very tedious advert

JMO

Martin

Very annoying statement. I'm interested in making the mdac works at its best and this means coupling it with others JohnW product.

And the main thing is that someone asked for the price and JohnW simply answered...

So please, don't read this thread...

 

[amershamman]

This is also very useful news for me, as I plan my set-up for the future. I've never felt that John is a money grabbing hard sell man...

 

[Plutox]

in my setup and room, volume is about -30db for nighttime listening to -20db for daytime listening

which is preferable:

(1) -30db on the MDAC's digital volume control
(2) use a -10db in-line attenuator
(3) wait for the MPAX and use its built-in -10db attenuator?

(4) do nothing and enjoy it. This sounds like audiophile paranoia in extremis.

You will not detect 20dB (of MDAC attenuation). You will certainly not detect -30dB at low (night) levels.

The point to be concerned about this is when -30dB is uncomfortably loud

 

[JohnW]

I face the same concern: in my setup and room, volume is about -30db for nighttime listening to -20db for daytime listening. Numerous posts (by John and others) suggest putting a -10db in-line attenuator at the power amp's input, but I also read (http://www.enjoythemusic.com/magazine/equipment/0803/rothwell.htm) that attenuators do have a detrimental effect on sound quality.

So question to John, which is preferable:

(1) -30db on the MDAC's digital volume control
(2) use a -10db in-line attenuator
(3) wait for the MPAX and use its built-in -10db attenuator?

Doe you do much listening at -30dB or is it just background music levels?

If its your normal listening level, you could try the rothwell's - I suggest inserting the attenuators at the Amplifier end of the cable.

John

 

[liszt]

(4) do nothing and enjoy it. This sounds like audiophile paranoia in extremis.

You will not detect 20dB (of MDAC attenuation). You will certainly not detect -30dB at low (night) levels.

The point to be concerned about this is when -30dB is uncomfortably loud

I humbly disagree. -30 dB is too high and will rob the music of its dynamics, killing off the bass resulting in a thin and sterile sound. It is even more of a problem, at low levels, because that's when we really need more bass to compensate for our low frequency 'deafness' (fletcher munson curve)

 

[sq225917]

I'd say you are just guessing and actually have no idea how much info is lost at any attenuation setting. It would be nice if John had a table to show us exactly what the loss in resolution was with level. I'm just guessing but I'd assume that -20db is still 20bit or better.

 

[liszt]

Ah, that makes perfect sense. Yes i do have some resistors lying around. I'm using both RCA and balanced. RCA for the active sub, balanced for the power amp.

About 20-30 dB, depending on the material. How should i hook up the resistors, and what values should i use? Thank you, John

John, could you kindly advice, please?

Is there any formula i can use to calculate the resistor value and network i should use, for a particular attenuation level?

I've been doing some more reading about source direct -> amp and it seems that impedance mismatch is another possible problem to watch out for. What are your thoughts on this, and when will this be an issue if i go with fixed attenuation? What is the output impedance of the MDAC's rca and xlr outputs? Thanks

 

[JohnW]

John, could you kindly advice, please?

Is there any formula i can use to calculate the resistor value and network i should use, for a particular attenuation level?

I've been doing some more reading about source direct -> amp and it seems that impedance mismatch is another possible problem to watch out for. What are your thoughts on this, and when will this be an issue if i go with fixed attenuation? What is the output impedance of the MDAC's rca and xlr outputs? Thanks

I'm lazy and just use my Spice simulator

But you can download any of the online attenuator programs. such as:

http://www.madsencircuits.com/acalc.html

Select the Pi Attenuator - and "Nomilize" to say a 330 to 470 Ohms (this is both the Load impedance as seen by the MDAC - and source impedance to the Amplifier - keeping the load impedance low (but not too low) this will help driving the front-end of the amplifier cleanly - place the attenuator at the Amplifier end of the cable.

If your designing for a Balanced connection then design as two separate single ended attenuators (as two Pi Attenuators - one for each "Arm"), referenced to center Ground connection.

John

 

[janerik]

Hello John,

After a stunning three days of music, my M-DAC now refuses to start. When I press the power button, I can hear relays clicking, but the front panel doesn't light up. I've unplugged and replugged the power cables, USB cable and S/PDIF cable, but to no avail.

Is there anything I can do to reset the unit or is it defective ?

Many thanks,

Jan

 

[travelfotografe]

I read this document (http://www.esstech.com/PDF/digital-vs-analog-volume-control.pdf) from ESS about digital volume control in the SABRE 32bit DACs. Slide 15 says even at -35db, there is no loss in resolution (no numerical loss of accuracy in the digital signal). Whereas slide 16 says the S/N ratio when using digital volume control will behave just like using an analogue volume control.

John: In an earlier post, you mentioned that the "noise" you are more concerned about when using a high attenuation in the digital volume control is the system noise from the digital circuitry in the MDAC. Is this system noise the reason why you suggest using a -10db Rothwell attenuator? And back to my question about -30db using the MDAC digital volume control. It seems ESS thinks dialing in -30db on a SABRE 32bit chip is no problem. Does the MDAC and/or your experience with the SABRE chips suggest otherwise, that lower than -20db will hurt (however slightly) sound quality?

I'd say you are just guessing and actually have no idea how much info is lost at any attenuation setting. It would be nice if John had a table to show us exactly what the loss in resolution was with level. I'm just guessing but I'd assume that -20db is still 20bit or better.

 

[elnero]

I read this document (http://www.esstech.com/PDF/digital-vs-analog-volume-control.pdf) from ESS about digital volume control in the SABRE 32bit DACs. Slide 15 says even at -35db, there is no loss in resolution (no numerical loss of accuracy in the digital signal). Whereas slide 16 says the S/N ratio when using digital volume control will behave just like using an analogue volume control.

John: In an earlier post, you mentioned that the "noise" you are more concerned about when using a high attenuation in the digital volume control is the system noise from the digital circuitry in the MDAC. Is this system noise the reason why you suggest using a -10db Rothwell attenuator? And back to my question about -30db using the MDAC digital volume control. It seems ESS thinks dialing in -30db on a SABRE 32bit chip is no problem. Does the MDAC and/or your experience with the SABRE chips suggest otherwise, that lower than -20db will hurt (however slightly) sound quality?

This is something I'm very curious about as well, I have tried my Westone 4 IEM's with the headphone out and I was in the -30db range. I'm really curious just how much loss, if any, I would be incurring in this scenario.

 

[Plutox]

I humbly disagree. -30 dB is too high and will rob the music of its dynamics, killing off the bass resulting in a thin and sterile sound. It is even more of a problem, at low levels, because that's when we really need more bass to compensate for our low frequency 'deafness' (fletcher munson curve)

You appear to be living in audiophile cloud-cuckoo land.

If -30dB of digital attenuation is causing "a thin and sterile sound" or, indeed, any other substantive and repeatable "problem" with the sound quality, you are experiencing the effect of other issues.

Are you aware of just how far down in the dirt the data below bit 20 actually represents? When you are listening quietly enough to feel that things are bass-light, it's quite probable that 8 bit audio would be quite sufficient*.

And if you are listening at low enough a level to be worried by a lack of bass, you need some bass boost from somewhere - it has nothing whatsoever to do with attenuation, digital or analogue.

* The range between threshold of audibility and instant hearing damage is about 90dB. 20 bits allow for a dynamic range at least 10dB more than that, 8 bits give you 50dB, give or take.

 

[JohnW]

Hello John,

After a stunning three days of music, my M-DAC now refuses to start. When I press the power button, I can hear relays clicking, but the front panel doesn't light up. I've unplugged and replugged the power cables, USB cable and S/PDIF cable, but to no avail.

Is there anything I can do to reset the unit or is it defective ?

Many thanks,

Jan

Hi Jan,

Try leaving your unit powered up for say 1 minute - dos the display eventually come up with "Comms Error"?

You could also try re-powering the unit in quick succession.

Where are you located?

John

 

[JohnW]

I read this document (http://www.esstech.com/PDF/digital-vs-analog-volume-control.pdf) from ESS about digital volume control in the SABRE 32bit DACs. Slide 15 says even at -35db, there is no loss in resolution (no numerical loss of accuracy in the digital signal). Whereas slide 16 says the S/N ratio when using digital volume control will behave just like using an analogue volume control.

John: In an earlier post, you mentioned that the "noise" you are more concerned about when using a high attenuation in the digital volume control is the system noise from the digital circuitry in the MDAC. Is this system noise the reason why you suggest using a -10db Rothwell attenuator? And back to my question about -30db using the MDAC digital volume control. It seems ESS thinks dialing in -30db on a SABRE 32bit chip is no problem. Does the MDAC and/or your experience with the SABRE chips suggest otherwise, that lower than -20db will hurt (however slightly) sound quality?

The issue with any Digital product is RF leakage levels.

At say -30dB to -40dB the ratio between "Audio signal" and RF becomes smaller - i.e Higher RF levels in relation to the Audio signal.

Most amplifiers don't like high levels of RF on there input - they became hard and bright...

If you use an analogue attenuator then you also attenuate the RF level.

When using Headphones RF is not an issue.

 

[mark.king]

John Dom

When the new firmware comes out can you start a new thread please as I don't want to lose it within this one.

I am waiting so I can plug my set top box into the Mdac. Quick question. Will I be able to set up the inputs independantly? I ask as I want the usb to as good as it can be but I know some setting will have to be changed for say my optical from my set top box so it doesn't stutter/drop out.

Cheers

 

[JohnW]

John Dom

When the new firmware comes out can you start a new thread please as I don't want to lose it within this one.

I am waiting so I can plug my set top box into the Mdac. Quick question. Will I be able to set up the inputs independantly? I ask as I want the usb to as good as it can be but I know some setting will have to be changed for say my optical from my set top box so it doesn't stutter/drop out.

Cheers

Hi Mark,

The DPLL settings will not effect the USB input, and they will be configurable for each input.

We should have a very early Beta version (One optical input set for High DPLL BW to confirm it solves the issue, then we will build the full version) - hopefully by the end of this week.

John

 

[Leprish]

Hi Mark,

The DPLL settings will not effect the USB input, and they will be configurable for each input.

We should have a very early Beta version (One optical input set for High DPLL BW to confirm it solves the issue, then we will build the full version) - hopefully by the end of this week.

John

John,

Surely this approach will result in all of the "sample rates" being set for a high DPLL bandwidth. In the case I've quoted previously the only sample rate where a problem exists is 88.2KHz. The others (44.1, 48 and 96KHz) all work absolutely fine. With all sample rates set to high I will loose the benefits of your super low jitter system.

 

[arthur]

John,

Surely this approach will result in all of the "sample rates" being set for a high DPLL bandwidth. In the case I've quoted previously the only sample rate where a problem exists is 88.2KHz. The others (44.1, 48 and 96KHz) all work absolutely fine. With all sample rates set to high I will loose the benefits of your super low jitter system.

''they will be configurable for each input''

 

[Leprish]

''they will be configurable for each input''

Input as in Optical 1/2. I have the Mac mini connected via optical and music is a mix of 44.1/48/88.2/96KHz. I'm not going to flaff around changing from one optical port to the other.

 

[ChrisPa]

Input as in Optical 1/2. I have the Mac mini connected via optical and music is a mix of 44.1/48/88.2/96KHz. I'm not going to flaff around changing from one optical port to the other.

The fix is nothing to do with sample rate. You will leave it connected to the same input for all music from the mac mini whatever its sample rate

The fix is to do with how easily/rapidly/readily the dac will relock/track variations within a single clock rate - the problems people have been reporting with tvs, digiboxes etc

In any case you'd probably be best using the usb input with the mac mini