Upgraded DAC makes no difference?

[SL1200]

Whether you're able to hear differences is debatable, but it is definitely possible to measure differences.

 

In this example I've used my sound card to measure noise / distortion (fft) on 3 different dac's. To maintain a degree of consistency of measurements, an off-ramp USB converter is used to send 2kHz @ pcm 24/96 0dBfs from the pc to each dac - via AES/EBU, and unbalanced R channel analogue is connected from each dac to the sound card input with 10dB attenuation. (to avoid sound card clipping).

 

Note that the measurements are affected by the sound card's analogue input - and potentially also by the noisy pc environment. 

 

3 dac's used are:

 

1. Wyred4Sound DAC2DSDse

 

2. Berkely Alpha DAC2 

 

3. Bricasti M1 DAC

 

Very interesting.  It seems that in all cases the noise floor is below what can be heard.  

 

@Addicted. the SE version of the W4S DAC does have a Femto clock etc

 

@@o2so.  The Recovery is a modest improvement.  I think it would be a better in systems that have a bit more jitter.

[GFuNK]

Very interesting. It seems that in all cases the noise floor is below what can be heard.

@Addicted. the SE version of the W4S DAC does have a Femto clock etc

@@o2so. The Recovery is a modest improvement. I think it would be a better in systems that have a bit more jitter.

Also, it depends what your looking at, that bricasti dac has a, comparatively speaking, large dip in its frequency response at high frequencies:

http://www.stereophile.com/content/bricasti-design-m1-da-converter-measurements#ph5JFKJgbleWFXi4.97

Sent from my iPad using Tapatalk

[BradC]

It's possible that the digital receiver is the cause of the measured differences.

Did you try any of the other inputs?

[Slartibartfast]

Very interesting.  It seems that in all cases the noise floor is below what can be heard.  

 

There are audible differences between all three - although to what extent it's related to above measurements is debatable, but the Bricasti is to my ears the the best sounding dac.

 

Also, it depends what your looking at, that bricasti dac has a, comparatively speaking, large dip in its frequency response at high frequencies:

http://www.stereophile.com/content/bricasti-design-m1-da-converter-measurements#ph5JFKJgbleWFXi4.97

Sent from my iPad using Tapatalk

 

Large dip?

[Slartibartfast]

It's possible that the digital receiver is the cause of the measured differences.

Did you try any of the other inputs?

 

Possible, yes, but I'm not convinced it's the cause. 

 

I used AES/EBU from the off-ramp - to avoid ground loop issues. 

[GFuNK]

There are audible differences between all three - although to what extent it's related to above measurements is debatable, but the Bricasti is to my ears the the best sounding dac.

Large dip?

Looks to be -0.25 dB at 20 kHz... Large compared with other DACs.

Sent from my iPad using Tapatalk

[Slartibartfast]

Looks to be -0.25 dB at 20 kHz... Large compared with other DACs.

Sent from my iPad using Tapatalk

 

Good luck if you can hear that. 

[GFuNK]

Possibly not, but if you're going to bring up measurable differences it's worth considering a range of metrics.

Also, for 14k I'd expect it to be measurably better in all respects.

Sent from my iPad using Tapatalk

[Slartibartfast]

Possibly not, but if you're going to bring up measurable differences it's worth considering a range of metrics.

Also, for 14k I'd expect it to be measurably better in all respects.

Sent from my iPad using Tapatalk

 

 

You pay for a nice case and plenty of desirable features and connectivity, but if you know of a cheaper dac that measures better then I'd be interested to know what it is.  

 

Fwiw, Steve Holding's parting summary after evaluating the technical performance for the Aus HiFi review:

 

"Whereas most DACs have at least one weak area of performance, the Bricasti M1 DAC exhibited outstanding performance right across the board. Brilliant design and implementation". Steve Holding

 

That's not to say that the Bricasti is perfect. Having compared it directly against some high end DAC's, I preferred the dCS Rossini with master clock and MSB Diamond, but both are more than 5 times more expensive. 

 

As is the case with all things hi-fi, the law of diminishing returns applies.   

[GFuNK]

You pay for a nice case and plenty of desirable features and connectivity, but if you know of a cheaper dac that measures better then I'd be interested to know what it is.

Fwiw, Steve Holding's parting summary after evaluating the technical performance for the Aus HiFi review:

"Whereas most DACs have at least one weak area of performance, the Bricasti M1 DAC exhibited outstanding performance right across the board. Brilliant design and implementation". Steve Holding

That's not to say that the Bricasti is perfect. Having compared it directly against some high end DAC's, I preferred the dCS Rossini with master clock and MSB Diamond, but both are more than 5 times more expensive.

As is the case with all things hi-fi, the law of diminishing returns applies.

Well I already pointed out one area of weakness so I'm unsure how old mate can claim outstanding performance across the board.

According to stereophile measurements, it also has higher levels of IM distortion than say a squeezebox touch or dacmagic.

Are these audible? Probably not, but it does beg the question: is the extra 13.5 k worth it?

Also, how can someone who, I'm assuming is reputable, make a claim which is clearly not the case?

Sent from my iPad using Tapatalk

Edited March 25, 2016 by GFuNK

[Slartibartfast]

Also, how can someone who, I'm assuming is reputable, make a claim which is clearly not the case?

Sent from my iPad using Tapatalk

 

 

Have you asked yourself why none of the technical reviewers have commented on your so-called areas of weakness? 

 

Perhaps it's because the areas of weakness to which you refer are not actually areas of weakness. Third order IMD <0.001% is a strength, not a weakness. 

 

Note also JA's closing remarks in Stereophile: "Bricasti Design's M1 has state-of-the-art measured performance".—John Atkinson

 

Before you can classify areas of strength or weakness you first need to establish a correlation between specific metrics and subjective performance. Fortunately, this has been established over (considerable) time, so it's expected that any experienced technical reviewer will know what to measure and how to interpret the results. 

 

I would suggest that this comment from JA goes some way to explaining the Bricasti's subjective performance: 

 

"The Bricasti M1's rejection of jitter was one of the best I have measured; any jitter-related spuriae lay below the resolution limit of the Miller Analyzer."

 

Sure, $14k isn't small bucks, and I'm almost certain that the Bricasti isn't the dac for you, but I know quite a few M1 owners and they're all very happy with their choice of dac, - even at the elevated Au rrp. 

Edited March 25, 2016 by Slartibartfast

[o2so]

Well I already pointed out one area of weakness so I'm unsure how old mate can claim outstanding performance across the board.

According to stereophile measurements, it also has higher levels of IM distortion than say a squeezebox touch or dacmagic.

Are these audible? Probably not, but it does beg the question: is the extra 13.5 k worth it?

Also, how can someone who, I'm assuming is reputable, make a claim which is clearly not the case?

Sent from my iPad using Tapatalk

Paying 13.5k extra for an improvement that is measurable but not audible, in a world where kids are starving, is against my very few moral principles, sorry.

Sent from my iPhone using Tapatalk

[GFuNK]

Have you asked yourself why none of the technical reviewers have commented on your so-called areas of weakness?

Perhaps it's because the areas of weakness to which you refer are not actually areas of weakness. Third order IMD <0.001% is a strength, not a weakness.

Note also JA's closing remarks in Stereophile: "Bricasti Design's M1 has state-of-the-art measured performance".—John Atkinson

Before you can classify areas of strength or weakness you first need to establish a correlation between specific metrics and subjective performance. Fortunately, this has been established over (considerable) time, so it's expected that any experienced technical reviewer will know what to measure and how to interpret the results.

I would suggest that this comment from JA goes some way to explaining the Bricasti's subjective performance:

"The Bricasti M1's rejection of jitter was one of the best I have measured; any jitter-related spuriae lay below the resolution limit of the Miller Analyzer."

Sure, $14k isn't small bucks, and I'm almost certain that the Bricasti isn't the dac for you, but I know quite a few M1 owners and they're all very happy with their choice of dac, - even at the elevated Au rrp.

Perhaps I'm missing something, but if we compare the jitter measurements of the M1 with a dacmagic they look almost identical... Am I missing something? In the words of Manuel, "keh?".

It's something that has confused me for some time, the words they write aren't always consistent with the piccies, the words usually correlate with price though. .

Sent from my iPad using Tapatalk

Edited March 25, 2016 by GFuNK

[joz]

Paying 13.5k extra for an improvement that is measurable but not audible, in a world where kids are starving, is against my very few moral principles, sorry.

Sent from my iPhone using Tapatalk

 

Also against my tight Ar$e.

I couldn't see any reason to spend BUX on any inaudible  change.

I still question spending BUX on anything that's not a significant improvement audibly.

Hence my system changes are very slow.

Though I must say just because an item may not make an improvement my system others may experience some benifit from the change.

As they say in the classics, YMMV.

[Slartibartfast]

Perhaps I'm missing something, but if we compare the jitter measurements of the M1 with a dacmagic they look almost identical... Am I missing something? In the words of Manuel, "keh?".

 

 

Yes, it's likely.

 

If you check the jitter performance of the USB interface it's very ordinary, but I would also suggest that it's false to assume that the SPDIF performance will be replicated in a typical system.

 

It's one thing to measure well in a lab environment with a low jitter signal generator, but quite another in the real world.

 

My advice to you would be to audition a Bricasti and compare it directly against a Dac Magic or any number of other good budget dacs. 

[GFuNK]

Yes, it's likely.

If you check the jitter performance of the USB interface it's very ordinary, but I would also suggest that it's false to assume that the SPDIF performance will be replicated in a typical system.

It's one thing to measure well in a lab environment with a low jitter signal generator, but quite another in the real world.

My advice to you would be to audition a Bricasti and compare it directly against a Dac Magic or any number of other good budget dacs.

Those measurements were using toslink not USB, they mention the SPDIF is "just as impressive".

We were talking about measurements, and the measurements I have spoken about were chosen by the folks at stereophile, they think they are relevant.

My point was, if you are handing over 14k for a dac, You would expect that it is measurably better in all regards.

In this case, stereophile present plots which demonstrate that the M1 has poorer IM distortion results and comparable jitter measurements to a $500 dacmagic.

Sent from my iPad using Tapatalk

[Slartibartfast]

Those measurements were using toslink not USB, they mention the SPDIF is "just as impressive".

We were talking about measurements, and the measurements I have spoken about were chosen by the folks at stereophile, they think they are relevant.

My point was, if you are handing over 14k for a dac, You would expect that it is measurably better in all regards.

In this case, stereophile present plots which demonstrate that the M1 has poorer IM distortion results and comparable jitter measurements to a $500 dacmagic.

Sent from my iPad using Tapatalk

 

Personally, I don't know anyone who pays more money for better measurements. The criteria is always subjective. 

 

Fwiw, the MSB Diamond DAC measures worse for THD and IMD than most DAC's. Applying your methods of assessment, it's a s**t dac, worse than a Dac Magic.

Edited March 25, 2016 by Slartibartfast

[GFuNK]

Personally, I don't know anyone who pays more money for better measurements. The criteria is always subjective.

Fwiw, the MSB Diamond DAC measures worse for THD and IMD than most DAC's. Applying your methods of assessment, it's a s**t dac, worse than a Dac Magic.

No comment [emoji4].

Sent from my iPad using Tapatalk

[joz]

Have been playing with a few Dacs of late comparing the subtle and not so subtle differences.

Has been very interesting considering the age and wild price disparity between them all.

[WhoRu?]

Have been playing with a few Dacs of late comparing the subtle and not so subtle differences.

Has been very interesting considering the age and wild price disparity between them all.

I have found that new and expensive does not always mean better...

After going on the DAC Merry-go-round after my long term DAC died... I ended up with a DAC from 1995 that I prefer to many FOTM modern dacs. Recently went to listen to the schiit Yggdrasil which was seen as the second coming on another forum and brought my DAC along and was somewhat underwhelmed.

I think we all listen for different things... Thank goodness there are plenty of options out there.

[joz]

Now I don't usually do this but yesterday I asked my wife to have a listen to the three I have on hand and asked her for her opinions.

Firstly she was worried that if she picked the most expensive unit it would give me Carte Blanche to go out and buy it.

But nonetheless she picked them in order of price blind.

She only knew them as 1,2,and 3.

But she also admitted that they were more similar than different.

 

Will do it again with her with a different track next time as I have found from my experience tracks can also make a large variation between what makes one component shine and others flounder.

[BradC]

@slartiblartfast What software did you use for the posted measurements?

Using rightmark audio analyser I get a noise floor of -140dB on my dac.

[Guest Sime]

wrong thread

Edited March 26, 2016 by Sime

[Slartibartfast]

@slartiblartfast What software did you use for the posted measurements?

Using rightmark audio analyser I get a noise floor of -140dB on my dac.

 

Noise floor will depend on FFT bin size and sample rate, so it depends on how you configure the software.

[GFuNK]

Noise floor will depend on FFT bin size and sample rate, so it depends on how you configure the software.

That should not be the case.

The bin size controls the frequency resolution, ie delta f, and the sample rate controls the highest frequency you can view.

Hopefully your analyser is implementing anti aliasing and windowing properly too.

It's worth doing some averaging to smooth out the signal this will require longer time records.

The only way to drop the noise floor in processing is to do coherant averaging. This is used to extract deterministic signals from noise, not to get a measure of the noise.

Sent from my SM-G920I using Tapatalk

Edited March 26, 2016 by GFuNK