[dougms3]

Having double the arrays is a good thing. So is I2S (hopefully without te clocking so better external clocks on DDCs are actually meaningful?)

Still, I can’t get excited from these measurements: https://alpha-audio.net/review/multitest-the-best-dac-for-1000-euros/2. Distortion is mainly on uneven harmonics, dynamic range is around 24 bit and the roll-off on the NOS doesn’t look right. There is aliasing but with some very steep dips and not how NOS should roll off both in theory and practice.

You may not care if this is a great sounding DAC but without having heard it this is what raises a bit of an eyebrow with me.

It seems like its becoming a trend. People think they can predict if something is good based on biased measurements instead of listening.

Everyone wants to be an expert and all it takes is a measurement rig apparently.

 

[roderickvd]

I remember reading something by Ti Kan of AMB fame. Not sure if it’s his quote but it’s where I got I first. Paraphrasing: if it measures poorly then it is likely to sound bad, if it measures well then that is not yet a guarantee that it will sound great.

I like that.

 

[sajunky]

Having double the arrays is a good thing. So is I2S (hopefully without reclocking so better external clocks on DDCs are actually meaningful?)

Still, I can’t get excited from these measurements: https://alpha-audio.net/review/multitest-the-best-dac-for-1000-euros/2. Distortion is mainly on uneven harmonics, dynamic range is around 14 bit and the roll-off on the NOS doesn’t look right. There is aliasing but with some very steep dips and not how NOS should roll off both in theory and practice.

You may not care if this is a great sounding DAC but without having heard it this is what raises a bit of an eyebrow with me.

I can't comment on these measurements, having trouble reading parameters from these screenshots and a missing description. I think I saw the same article 4 years ago. Audio GD definitely has NOS roll-off conforming to the theoretical sample-hold figure, I have seen it on other tests. In practice you can't hear this drop, as high-frequency images (do not mistake images with aliasing) stimulate our ear receptors. It is my own explanation. My ears are old, with my 12kHz limit I don't notice any drop on my R2R-11, but I should.

Other than that, comparing measurements of Delta-Sigma DAC with R2R type is pointless, as spectrum analysers take averaged results, our ears pickup things that happen during transients. The author says about blackness present in other DACs, I agree with. This is effect of having a dominant tone exposed on the expense of everything what happen in the background, but he mention it in connection to the 'precision' which I think he use a wrong term. In fact a timing precision is much better in R2R DAC. Our ears are more sensitive to the timing than amplitude errors. It translate to the coherent harmonics across all frequencies, together with unsuppressed timbre and texture it makes a music. Audio GD website compares R2R sound to the oil painting, I think there is no better comparison.

In other words, comparing R2R sound to the DS type is like comparing apples to oranges. There are more subtle differences when comparing Denafrips sound to Audio GD. Both sound very good, but I think the author would prefer a laid-back character of Denafrips sound. Especially in NOS mode, as Denafrips use a different decoding method, not true NOS. In OS mode Audio GD sound is more similar, but still R-1 is more dynamic and putting into a chair.

You must hear R2R DAC first, listen to the music at least for two weeks, then connect DS type and you will immediately notice you are missing a lot.

Regarding I2S, you are right, it doesn't require reclocking, but the best results can be achieved when both DDC and DAC are referenced from the same external clock. This is a TOTL solution. R-1 do not have external clock port, don't look for such port in Pontus and Venus. The cheapest model with external clock support is R-8Mk2.

 

[roderickvd]

No doubts there I am a R2R guy myself. I compare with measurements from the Ares II, not analytical sounding DS ones. Even then these graphs don’t give the full picture, I agree.

The NOS roll off as always pictures only starts beyond 10 kHz but in practice colors two octaves down and that is what you and I probably hear and love. Still, that NOS graph on the linked site is definitely off. May be a wrong measurement or visualization too, I don’t know.

 

[sajunky]

@roderickvd. Which figure you are refering to? I will look at later on PC, currently on mobile.

 

[gimmeheadroom]

If you want better connectivity and true NOS + OS modes, spend the extra money for the R8 Mk2. It's shockingly good.

 

[roderickvd]

This one:

 

[gimmeheadroom]

The vertical scale seems uselessly gigantic in the audible range.

 

[sajunky]

This one:

Oh, it looks completely different to any other tests. It is because a frequency scale (X-axis) is linear, not logarytmic as usual. On this graph it would be also difficult to get exact numbers, as a vertical scale is to large as @gimmeheadroom said.

 

[George Hincapie]

It seems like its becoming a trend. People think they can predict if something is good based on biased measurements instead of listening.

Everyone wants to be an expert and all it takes is a measurement rig apparently.

How are measurements biased? They are what they are.

 

[roderickvd]

Oh, it looks completely different to any other tests. It is because a frequency scale (X-axis) is linear, not logarytmic as usual.

Good catch.

 

[sajunky]

How are measurements biased? They are what they are.

These are static tests, averaged and subject to DSP filtering. They are insensitive to dynamic changes. A different type of ears I would say in a human language.

 

[szore]

USB can sound good or very bad, a reason is pointed out by @godmax. Coax is less sensitive to the specific configuration assuming there are transformer coupled as it should. Transformer coupling cannot be used on USB lines, so isolation of the USB port is a tricky job. Ares has no USB isolation, it is why i.e. for some users Bifrost 2 sounds better, but it is only because BF2 has USB isolator. It will perform more consistent in various systems. For Ares you should try ground loop redirectors like ifi Defender, converters to S/PDIF, then at a higher price level network streamers that are designed to generate less noise than ordinary PC/laptop.

Try the Matrix H usb card.

https://www.moon-audio.com/matrix-audio-element-h.html

 

[dougms3]

How are measurements biased? They are what they are.

Even though they are all dacs, they don't measure the same. Its not a one way to measure all devices.

Might as well do something stupid like judge the sound quality by looking at the measurements and never listen it.

 

[gimmeheadroom]

Might as well do something stupid like judge the sound quality by looking at the measurements and never listen it.

This is exactly what they do. I remember in the 1970s when they did that with cars it was called "Armchair racing."

So, time moves on, technology improves, and nothing has changed