Single Ended DAC vs Dual Differential XLR DAC

@gkg2k .. apologies for veering off course. If you decide to take a chance on the balanced version I hope you'll share your impressions (warts and all). 

A couple of years ago there was a long and heated discussion in DIYaudio about this topic (maybe it's still going on). Generally speaking higher SNR  means in the digital domain more bits which would translate into more resolution. The position from the developer (Soekris) was that the Soekris DAC would deliver 26bits of resolution (calculating it from the number of used resistors per channel in the R2R) while the other party insisted that the SNR numbers simply wouldn't verify this claim.

So whatever is true we can say at least that XLR has higher SNR compared to SE DAC.

My personal experience is that it also depends a lot on how one connects the DAC to the rest of the system and what source files (bit depth) are being used. Especially considering digital vs. analog volume control or a mix of both (like what I am doing). Because if one uses a 16bit signal and digital volume control one will listen in the end to  maybe 12bits of resolution only. Can still sound satisfying with the rest of the system and room acoustics and one wouldn't even notice. With 24 bit signal depth and a 26 bit DAC (like the Soekris claims) it works as long as the input sensitivity of the power amp and the sensitivity of the speakers are in range with the necessary digital volume settings for comfortable listening.

A combination of digital and analog volume control  make it easier to fine tune without losing resolution and allows better adjustment to the rest of the system in my experience.

The biggest flaw of analog volume control  is that it takes some part of the signal and connects it to ground (voltage divider principle) and a lot of transients and subtle elements of the signal get lost first. The lower the settings the more is dumped to ground. A combo of digital and analog allows the best compromise between bit depth and transients. It's clearly audible, at least in my system.

So XLR or SE considering  SNR, bit depth, and volume control  is the chain that would need to be looked at in order to find the correct answer.

@melvinjames ,

Measurements, done correctly, do not lie. Perhaps if manufacturers took their own measurements and published them, as opposed to others doing it, then sites like ASR would not even exist. As a consumer, I feel like manufacturers are taking advantage of us. When I started into audio in my teens in the 80's, ever speaker vendors with few exceptions posted response curves. Today that would be the exception. There is not desire on the part of manufacturers to educate their customers in general or even on their own products. We are expected to believe everything in their marketing blurbs. While I tend to find Amir's tone often arrogant, bordering on offensive, I do appreciate the work and other similar efforts as it levels the playing field between consumers and manufacturers.

I believe that most audio purchases are made with relatively little prior listening, and are based on reading reviews. Many of those reviews are highly suspect and involve little in the way of critical listening or critical comparison. Based on those reviews perhaps you buy this product, and you like it, but the noise background seems a bit high, and it sounds "good" but perhaps not quite right. So you come to a site like this, and people tell you, well you need to:

• buy this or that expensive interconnect
• buy a $500 power cord to make it really shine
• buy a $1000 reclocker if you want better performance

Now, another $2,000 in, not knowing any better, you still think it sounds "off".

It sounds "off" because it IS off. And none of those glowing review sites told you that, and no one here told you that, they told you to spend more painting the pig.

Measurements, done correctly, do not lie.

Nor do they the whole story.

I believe that most audio purchases are made with relatively little prior listening, and are based on reading reviews. Many of those reviews are highly suspect and involve little in the way of critical listening or critical comparison.

Agree with this big time and is a HUGE pet peeve of mine. TAS is a main offender of not doing any relevant comparisons or even disclosing what equipment is in the review system and I’d never rely on one of their “reviews” to form an opinion on a product. However, there are some good, rigorous reviewers out there who I find very helpful, so it’s important to find the more reputable reviewers out there and just ignore the trash.

You will get a better s/n ratio and a cleaner, clearer signal with a fully balanced unit.  That being said, if you do not have a fully balanced system, it is unlikely that you will get the full benefit of this.  This is largely due to the fact that at some point, the balanced signal will be converted to a single ended signal when it hits the unbalanced component.  Common mode noise rejection is lost and the quality of that transformer will have a BIG impact on sound.  

That transformer is often why, single ended devices sound better when using RCAs rather than XLRs with a transformed signal.  It is also why balanced devices XLRs almost always sound better than the RCAs that typically accompany the units.

If your system is fully balanced, you should get a fully balanced DAC.  SE will never sound as good.  If your system is single ended, a fully balanced unit will sound as good as a single ended unit at best.  Pay a lot of attention to the quality of the output of the RCAs if you get a fully balanced DAC.  Brands like Bricasti and Weiss have virtually no difference in performance between XLRs and and RCAs in practical terms but those are more the exception than the rule.  Many others see signal degradation in that XLR to RCA transformer.  

The reason the circuit is duplicated, (4 chips vs. 2 chips, etc...) is driven by the need to produce the inverse for common mode noise rejection.  That extra chip does not increase resolution, etc...