[riffrafff]

I think that he is not unwise though

Indeed.

I still remember when the Internet was mostly civil, like back in the bang-path days.

 

[Pietro Cozzi Tinin]

I think that he is not unwise though

Indeed he's not other than even stating it comes over as a bit ....... controversial / feeding the sentiment.... like....
No offence here.

 

[Ripper2860]

It was just a friendly suggestion that has served me well and was directed at no one in particular.

No offense intended and I hope none taken.

 

[Pietro Cozzi Tinin]

OK, let's move on.

 

[Ripper2860]

OK. So I ran across a discrete R2R ladder DAC while tooling about the web. This was my 1st exposure to a discrete design and it got me to thinking if Mike might have some thoughts or comments regarding discrete sign-magnitude R2R ladder DACs in comparison to IC-based DACs. What would be the pros and cons -- sonic-wise, design-wise, and cost-wise. And what drove Schiit to IC-Based as opposed to discrete? Was it just a cost-based decision or were there other factors/benefits to doing so?

There's a bit of info floating around and it's difficult to sift through what real and what's not.

I'm hoping Mike ( @Baldr ) will see it here and enlighten us. Of course I'd like to hear from the group here, as well.

 

[RickB]

^I doubt very much if Mike will comment on a competitor's product.

 

[rlow]

To RickB's point, perhaps if you made it a more general question about discreet R2R implementations vs. ICs he would answer (removing any specific competitor mention - you may also want to tag him).

 

[Ripper2860]

I'm not asking that he comment on the product specifically, but rather on the practical differences between a discrete R2R design and a IC-based R2R design. There are several MFG's that make discrete R2R ladder DACs and I'm curious as to the pro's and cons of IC over discrete. I would assume IC-based R2R is more cost-effective, but what if any trade-offs are there to gain that cost savings?

I know it's a thin line, but I'm hoping it could be just an explanation as to why one might be better than the other and what drove Schiit to IC-based.

 

[Ripper2860]

Excellent suggestion. I'll revise my post!!

 

[bosiemoncrieff]

God, Boris Godunov was a snooze fest. I was there for three and a half hours, and heard maybe a half hour of interesting music.

 

[rlow]

I'm not asking that he comment on the product specifically, but rather on the practical differences between a discrete R2R design and a IC-based R2R design. There are several MFG's that make discrete R2R ladder DACs and I'm curious as to the pro's and cons of IC over discrete. I would assume IC-based R2R is more cost-effective, but what if any trade-offs are there to gain that cost savings?

I know it's a thin line, but I'm hoping it could be just an explanation as to why one might be better than the other and what drove Schiit to IC-based.

It's a good question, and something I'm curious about as well. Hopefully @Baldr will grace us with his thoughts on the matter. Or perhaps he's already written about it in the past, and others would have a link?

I have my doubts whether cost is the main reason however, but we'll see. My general understanding is that there really isn't anyone making/selling audio specific IC Multibit DACs anymore, which is why Mike ended up using a medical/military specific DAC chip. However it requires a crap ton of "care and feeding" to utilize it for audio purposes, as per the Yggy webpage FAQs:

So how come everyone isn't using the AD5791, if it's so great?
Because they require very, very special care and feeding. AD5791s aren't "bolt in and go" DACs, with pleasant little paint-by-numbers application notes for use with audio. They don't even accept normal digital audio formats. Managing their use with multiple input bit depths and sample rates is, well, challenging. And special care has to be taken with their output. Plus, people are more focused on silly claims like "32-bit" DACs and "Giga-Rate" DSD. "21 real bits," doesn't sound real sexy in that context.​

As far as discreet R2R implementations, one issue I've read is resistor quality and consistency/matching. A military/medical grade IC, I would assume, avoids that problem.

 

[madwolfa]

As far as discreet R2R implementations, one issue I've read is resistor quality and consistency/matching. A military/medical grade IC, I would assume, avoids that problem.

It doesn't avoid it completely, but definitely makes it more than good enough for audio applications, despite what some measurebators might claim.

 

[Ripper2860]

As far as discreet R2R implementations, one issue I've read is resistor quality and consistency/matching. A military/medical grade IC, I would assume, avoids that problem.

I thought so, as well. So why would anyone even venture out into discrete R2R? And as it relates to non-audio based DAC chips, if they are so difficult, why would one even bother with an IC-based R2R DAC. They should just stay in the paint-by-numbers DS DAC world. But that excludes the need of some to differentiate and market something not because it IS better, but because it's THEORETICALLY better.

I may have partially answered my own question -- it may not be that ONE APPROACH is better than the other -- It may be all about the EXECUTION of one over the other.

 

[madwolfa]

I thought so, as well. So why would anyone even venture out into discrete R2R? And as it relates to non-audio based DAC chips, if they are so difficult, why would one even bother with an IC-based R2R DAC. They should just stay in the paint-by-numbers DS DAC world. But that excludes the need of some to differentiate and market something not because it's better, but because it's "thoeretically" better.

I may have partially answered my own question -- it may not be that one is better than the other. It may be all about the execution of one over the other.

Jason's "Schiit Happened" book has most of your questions answered.

 

[artur9]

Hmm. I've not experienced this at all, with either Gungnir MB or with Yggdrasil.

I think one other person made a similar comment. One thought was that some sort of distortion became less perceptible so one's brain worked less and could pay more attention to the music.

Or maybe I just had too much scotch that night. That Aldi stuff sounds good.