advertisement


MDAC first listen (part VII)

Status
Not open for further replies.
If this is so, then that's great as the "ESS Left Channel Locking" counter-measures are turned off during Bit Perfect testing due to MCU overhead limitations...

Are the "counter-measures" correctly reinstated following a bit perfect test?
 
No I'd hope not, but second order effects (differences in Data processing patterns) might have changed the sound. The Phase Inverts feature was requested by a PFM owner... so we added it, even though I do not believe there will be a difference in SQ due to the system "phase" - although second order effects could be possible.
John

It was me and some others who suggested the option to invert phase. Thanks for fulfilling the wishes of us with ears sensitive to absolute phase ;) . Remains to be proven though.

I only have MAC computers so currently go on with the factory version of firmware.
 
It really sound different on 0.99 compare to 0.90. But you need to do factory reset after update the firmware.

it sound thicker, more body, more defination in the lower base as well. Vocal are not as forward, airier and sweater high. I like it.

John, I am ready and side aside some saving for your mamp and mpax and bdac and ... what else... lolz?

Mav

The problem is, you said you thought it sounded better, so now it maybe a mass hysteria that we have all convinced ourselves it sounds better. When it comes to audio you can never downplay too much the relevance of psychoacoustics including the influence of others and simply things like your own mood at the time. Unfortunately there is no easy A/B comparison possible unless we had two MDACs with different firmware.
 
My M-dac now crackling madly from the left channel. What a pisser!

Second time this week, both stopped when re-booted to factory defaults.

I found that mine stopped when I switched off and then (later) on again. I didn't need to restore the defaults.

Edit: And, BTW, I've never done the bit perfect test - using Linux.
 
It really sound different on 0.99 compare to 0.90. But you need to do factory reset after update the firmware.

it sound thicker, more body, more defination in the lower base as well. Vocal are not as forward, airier and sweater high. I like it.

John, I am ready and side aside some saving for your mamp and mpax and bdac and ... what else... lolz?

Mav

Mav - Did you update straight from 0.90 to 0.99?

If so that might explain why your MDAC did not automatically reset - I recall an early version of software did not handle the Reset - although it might have been even earlier then 0.90...

Also, don't forget that the selected Digital filters might not have loaded correctly with 0.90.... :rolleyes: sorry about that one...
 
Are the "counter-measures" correctly reinstated following a bit perfect test?

Could be one of two issues.

1. We have not re-enabled the "counter-measures" after the BPT.

2. During the BPT, the "counter-measures" are disabled and the ESS has misslocked - and once its got its knickers in a twist it will stay that why until we force it to re-lock...

I suspect the second option - so we basically have to force a re-lock event after the BPT - the software is full of stupid workarounds trying to make the ESS behave in a semi-orderly fashion.

ESS should really sort there house out... and should have RESPUN THE SILICON a long time ago!

John
 
I found that mine stopped when I switched off and then (later) on again. I didn't need to restore the defaults.

Edit: And, BTW, I've never done the bit perfect test - using Linux.

Hi Richard,

The BPT is just a WAV file - so you can still use it under Linux (I guess once you have created it from the Java applet...)

John
 
ESS should really sort there house out... and should have RESPUN THE SILICON a long time ago!

John

Quarter of a million dollars just to fix a bug that most designers are willing to workaround. How do you rate the chances?
 
Quarter of a million dollars just to fix a bug that most designers are willing to workaround. How do you rate the chances?

ESS make about US$50K from our projects alone every month (just based on our Qtys of the 9018).

I can understand that - then at least they should not promote features that don't work correctly.

The second point is that some bugs cannot be 100% worked around... We have to perform off board processing to "Hopefully" resolve the left channel crackling... we are lucky to have this off board processing power, other designs would be stuck...

John
 
Well I could wait no longer for an MPAX so bought a Behringer UCA 222.

Project X pack>TR. Dino phono amp>Behringer UCA>Audacity(WinXP).
(The Behringer has SPdif optical out)

>MDAC>Yamaha AS2000>CelestionA3.

So Vinyl into the Digital domain :) sounds surprisingly good.

(Just waiting for the organic caps in the Behringer to burn in :rolleyes:)

£ 20 on fleabay!, If you want to try it buy the UCA 202 and download Audacity.

Fairly plug and playish too (lucky for me).
 
The second point is that some bugs cannot be 100% worked around... We have to perform off board processing to "Hopefully" resolve the left channel crackling... we are lucky to have this off board processing power, other designs would be stuck...

John

They seem to be ready to live with the risk. It is quite widely used chip, two of them inside same popular Oppo players for instance. Quite a risk if issues begin to escalate.
 
John, yes it's straight update from 0.90 and I only use USB so the filter thing don't affect me.

Now I like even more optimum spectrum filter as the full bodied sound compliment the dynamic sounding of it. I also have to move away my speaker little more from rear wall to clearer some "air" from the woofer.

Mav


Mav - Did you update straight from 0.90 to 0.99?

If so that might explain why your MDAC did not automatically reset - I recall an early version of software did not handle the Reset - although it might have been even earlier then 0.90...

Also, don't forget that the selected Digital filters might not have loaded correctly with 0.90.... :rolleyes: sorry about that one...
 
I just got left channel crackle with the latest 0.99 unfortunately.

I was doing a bitperfect test with a Sonos on COAX1 (which worked and passed), and then after I dropped out of the test mode, the crackle was there.

Thanks for the feedback - indeed, when in Bitperfect Test, the DSP runs a completely different workload and the ESS crackle countermeasures were not applied there. As always, once the crackle sets in, it will only disappear upon re-lock ie. change of input etc. No need for power cycle, no need for factory defaults. It has nothing to do with the M-DAC itself, it's just the ESS input decoder missing a beat. I've put the countermeasures into the Bitperfect Test code already, expect it fixed starting from the next build.
 
ESS make about US$50K from our projects alone every month (just based on our Qtys of the 9018).

John

I'd guess that until everyone stops buying them that they won't choose to fix it. You are tacitly accepting the fault as being ok by continuing to pay for the chips. Whatever else designers and engineers like yourself might communicate to them, your decision to continue to use the chip is what counts.
 
Thanks for the feedback - indeed, when in Bitperfect Test, the DSP runs a completely different workload and the ESS crackle countermeasures were not applied there. As always, once the crackle sets in, it will only disappear upon re-lock ie. change of input etc. No need for power cycle, no need for factory defaults. It has nothing to do with the M-DAC itself, it's just the ESS input decoder missing a beat. I've put the countermeasures into the Bitperfect Test code already, expect it fixed starting from the next build.

Great. Thanks Dominik. I suspected I had found an "edge case" here, but its still good to tidy them if you can :)
 
However, I do have an old Cisco NSLU2 (probably one of the first hackerble NAS devices!) which I've re-flashed an got a light Ubuntu distro on it. I used to use media tomb media server on that, although it didn't perform too well. The Cisco is quite underpowered though, I think it's processor is only 233MHz. I think is's power-pc, rather than x86 - will LMS run on that do you think?

I've owned an NSLU2 for about 2 days, many years ago. I was so disappointed by the performance that I sold it on eBay. I have no idea how it performs with LMS on it but there is some documentation available in the Slimdevices Wiki on how to install it on a NSLU2. You might give it a try: http://wiki.slimdevices.com/index.php/Category:Installation_Guide_Pages
 
I am considering to use an external DAC. My first candidate was a Buffalo3, until I saw the MDAC. Although I spent several evenings on reading the threads here on this DAC, I am still missing a piece of the puzzle: the direct comparison of the internal DAC with the MDAC (it is quite certain that I did not read all messages on this subject).

I am one of the owners of an SBT / M-DAC combo but I have never compared the SBT internal DAC with the M-DAC. I’ve been using an external DAC since I bought the SB3 (back in 2005). The SBT has replaced the SB3 in my main listening room about 6 months ago and since I already was using an external DAC I never took the trouble to compare the internal DAC with the external DAC. For sure the M-DAC will sound miles apart from the internal SBT DAC due to the difference in technology (DAC chip), the analogue stage and the power supply designs of both systems.
The Squeezebox is a great streamer; the M-DAC is a great DAC. The combination is best of both worlds I would say.
 
Status
Not open for further replies.


advertisement


Back
Top