advertisement


MDAC First Listen (part 00101100)

Status
Not open for further replies.
AFAIK the faulty Apple USB implementation is only causing problems with USB 1.1 devices. You could try using an USB 2.0 hub with your MDAC. FDAC should be fine.
 
As soon as I have tried shortening the modified RS cable, and heard what that does, I will try to open up the Calex and look at changing capacitors. Even with the long cable, there is an improvement, perhaps at the "1%" level, but a definite plus.

Yes - that follows my own experiences here with the MDAC and external HQ PSU... And to be honest it offers too little improvement for the cost (lets say GBP350).... Hence the FDAC Project to achieve that "Leap" in performance.
 
Ah, I was under the impression that the M-DAC's non-native filters were implemented in DSP. Anyway, what I mean is the filters that aren't built into the SABRE DAC.

On a neutral system, I really feel the treble roll-off of the optimal transient filters just sounds too dull. Minimum phase, on the other hand, sounds a little too glary and strident. I was surprised to find that the most listenable DSP filter was optimal spectrum, and I was even more surprised when I found the built-in SABRE filters sounded better still! These filters appear to 0.2dB quieter, which will probably make them sound worse to most people. I initially preferred the fuller sounding midrange of the non-bulit-in filters, but this also made them more prone to chestiness, and the lighter midrange balance of the built-in filters also makes it sound more open and transparent. Plus the bass definitely sounds more even with the built-in filters, although this could be very room and system dependent.

I know you've said you think time domain accuracy is more important than frequency domain accuracy, but I guess for me the opposite is true. And with the slow roll-off filter, you can have your cake and eat it too anyway, as it is excellent in both the frequency domain and the time domain.

The MDAC HAS no DSP, the Digital filters are internal to the ESS Saber.

I cannot believe ANYONE can hear a level difference of 0.2dB!! 1dB is recognised as the lowest level difference perceivable by humans...

Its Optimal Transient filters that I believe are the most popular (Not Optimal Spectrum) - these have a "Perfect" time Domain performance.
 
I also have a question which is probably very dim, but it has been bugging me for a long time .... so here goes:

Some people here are in for more than one FDAC with a master - slave configuration in mind. I don't understand how one would use this. I sort of understand that the master unit will provide the clocking for the data stream before the DAC conversion, but I can't see why you would want to send that to another slave-FDAC (perhaps that's in another room/system or somehow connected to another CD/Bluray source?)... I'm wondering if I'm missing anything by just going for a single master unit.

Anybody wish to explain? especially the practical use of Master-Slave FDAC set-ups.

Cheers:)

Basically for any system where you need more then two channels - eg. Surround Sound systems and Active crossover systems (using the internal DSP) where you perform the Crossover in DSP and send the "filtered" Audio to different DAC Channels / amplifiers to drive the Bass / Mid Range / Tweeter directly.
 
Is the FDAC Lite going to be made available now? Last thing I remember reading it was only going to be something for JohnW to develop as a test bed on the way to the FDAC.

There will be early engineering units of the FDACLite and a very select few released "into the Wild" :) for 'early' evaluation of the FDAC audio quality - these will help us define the "Fusion Bulk foil resistors" etc. for the flagship FDAC test the USB drivers etc.
 
It's for use as an active digital crossover with separate D/A conversion for each drive unit.

How interesting.....So I guess you take the digital file separate this in frequency digitally (in a PC or DSP crossover) and then feed each frequency range to 3 separate FDACs......the analogue signal from these then goes to 3 separate power amps (VFETS)....into 3-way speakers.

I guess this gives you complete control over the crossover frequencies and filters etc. etc.

And I guess you need a master FDAC to control the others for volume and data clocking.....very ingenious.

Wow...seems like a good solution. I have a 3-way active system but the crossover is analogue. I'm guessing that digital crossovers can be made 'perfect' whereas analogue crossovers have some inherent compromises.

Thanks!
 
Basically for any system where you need more then two channels - eg. Surround Sound systems and Active crossover systems (using the internal DSP) where you perform the Crossover in DSP and send the "filtered" Audio to different DAC Channels / amplifiers to drive the Bass / Mid Range / Tweeter directly.

Thanks John, I think I understand now, very interesting......For a 3-way active set-up for example one would need a mindsp 2x8 kit to do the processing and provide the 6 digital outputs to be fed into 3xFDACs and then to 3xVFETS (VFETS may be monoblocks and then you would need 6 of them!!).....Oh the penny is dropping! :D

I wonder which speakers Adamdea has?
 
The MDAC HAS no DSP, the Digital filters are internal to the ESS Saber.

I cannot believe ANYONE can hear a level difference of 0.2dB!! 1dB is recognised as the lowest level difference perceivable by humans...
I was under the impression that 0.1dB was the strict requirement for blind testing and that that was regarded as being just under the minimum perceptible difference.
This source quotes 0.2-0.4dB
http://www.indiana.edu/~emusic/etext/acoustics/chapter1_loudness.shtml
(mind you 0.2 dB is still 0.2dB more than the difference caused by an asynch usb reclocking device)
 
How interesting.....So I guess you take the digital file separate this in frequency digitally (in a PC or DSP crossover) and then feed each frequency range to 3 separate FDACs......the analogue signal from these then goes to 3 separate power amps (VFETS)....into 3-way speakers.

I guess this gives you complete control over the crossover frequencies and filters etc. etc.

And I guess you need a master FDAC to control the others for volume and data clocking.....very ingenious.

Wow...seems like a good solution. I have a 3-way active system but the crossover is analogue. I'm guessing that digital crossovers can be made 'perfect' whereas analogue crossovers have some inherent compromises.

Thanks!
Yes, I think some active speaker manufacturers do this all in one box (Genelec for example). The problem if you are doing it bespoke you need to make sure the dacs are exactly synchronised, hence the master/slave thing.

Apart from the fact that the digital crossovers should be more accurate, there is the attraction that you can apply room correction at the same time.
 
There will be early engineering units of the FDACLite and a very select few released "into the Wild" :) for 'early' evaluation of the FDAC audio quality - these will help us define the "Fusion Bulk foil resistors" etc. for the flagship FDAC test the USB drivers etc.

Sheffield is renowned for being "Wild" ;)
 
I was under the impression that 0.1dB was the strict requirement for blind testing and that that was regarded as being just under the minimum perceptible difference.
This source quotes 0.2-0.4dB
http://www.indiana.edu/~emusic/etext/acoustics/chapter1_loudness.shtml
(mind you 0.2 dB is still 0.2dB more than the difference caused by an asynch usb reclocking device)


A very interesting acrticle...Cheers!
The balance on the MDAC goes up in 0.5dB steps .... I need at least 1dB to notice a change.....but I'm glad designers out there can hear 0.2 - 0.4 dB differences...hence why John wants a panel of good testers/listeners for the FDAC....maybe he's getting too old (like me):D
 
A very interesting acrticle...Cheers!
The balance on the MDAC goes up in 0.5dB steps .... I need at least 1dB to notice a change.....but I'm glad designers out there can hear 0.2 - 0.4 dB differences...hence why John wants a panel of good testers/listeners for the FDAC....maybe he's getting too old (like me):D
On the other hand this article suggests 0.75dB to 1db with music, which seems to match your experience
http://www.audioholics.com/room-acoustics/human-hearing-amplitude-sensitivity-part-1
 
Thanks John, I think I understand now, very interesting......For a 3-way active set-up for example one would need a mindsp 2x8 kit to do the processing and provide the 6 digital outputs to be fed into 3xFDACs and then to 3xVFETS (VFETS may be monoblocks and then you would need 6 of them!!).....Oh the penny is dropping! :D

I wonder which speakers Adamdea has?

Except you won't need the minidsp 2x8 since the FDACs will do all the processing.
 
Status
Not open for further replies.


advertisement


Back
Top