What makes a DAC so expensive?


You can buy a Cambridge Audio AXA25 25 Watt 2-Channel Integrated Stereo Amplifier | 3.5mm Input, USB Input for $225, and most DACs seem more costly. 

I'm wondering what it is that makes a Bifrost 2 almost as expensive as an Aegir and 3x's as expensive as the Cambridge product, above. I would have thought an Aegir would out-expense a Bifrost by a factor of two or three. What are the parts that make the difference? 

I'm wondering if the isolated DAC concept is one that comes with a "luxury" tax affixed. Can anyone explain what I'm getting in a Bifrost 2, or other similar product that justifies the expense...?

Thank you.
listening99
Jitter isn't a problem for well designed DACs.  Take the little Schiit Modius at $199 on the USB input, even for all the talk about problems with USB, Jitter comes in at -150 dB much better than it's toslink and coax inputs  where it hovers around -120dB. Of course none of this is audible. It just shows how far engineers have come if a company can produce a DAC in the US for $199 with the measurements this little DAC has. 
@djones51 Have you paid attention to the new Denafrips GAIA DDC unit that recently came out. From a connectivity stand point it would be very convenient for me. A single Sonore microRendu connected to a GAIA which is then connected to a multiple DACs which as connected to separate preamp inputs. That is the convenience part for someone that is looking to get a few DACs into the system. I wonder about the claims of sonic improvements.

I have the Benchmark DAC3B. Looking to add the Audio Mirror Tubadour SE, and maybe something else if I win a lottery. Any opinions on the GAIA?
There are many assumptions or assertions buried in the original question and several replies. First, the DAC chip is a small part of the cost of a DAC, and in the case of the R2R Schiits, know that they had to do a bunch of custom work to use those AD DACs since they are designed for instrumentation and therefore glitch. Engineering is expensive.

next, for those who think most high end is just marked up a lot - reality check. This industry sucks financially. yep, manufacturers can charge what the market will bear, but the market won't bear much. Why? Lots of people want to be in the business and try their hand at design. When supply rises, prices and markups fall. That’s the invisible hand at work. Want to make money? get into something like cement. Nobody finds it romantic.

The cost of almost any piece of electronic gear (speakers, aka furniture are similar but different at the same time) is driven by, in rough order:

  • The chassis
  • The transformer(s)
  • The jacks and other switchgear/hardware
  • The box you ship it in
  • Warranty costs, made worse if one wishes to accommodate user error, which is common. Sometimes user error seems mandatory.
  • The custom PCBs, especially if large
  • And then a few bucks for the various electronic components, even when quality stuff is being used
Engineering costs must be amortize over small numbers.

R&D is a slower and more prototype-intensive process than most since measurements only take you so far in high end.

An aside, I’m beginning my DAC journey. I have three very early prototypes running. All use very different DAC chipsets. And yet, i have managed to make them sound fairly similar with the characteristics i always strive for. Lesson: the DAC is not the determining factor. Not saying I’m happy with any of these designs - they are are early days. I'm only saying that different chips wind up sounding the same when similar engineering recipes are applies to them.

In DACs you will find (or ought to!) multiple power supplies, and lots of money spend to contain and reduce noise - ground noise, radiated noise, reconstruction noise, blah, blah. Costly clocks and multiple series timing circuits. Isolation efforts between stages. A few dollars her and there add up.
G
Don't really know anything about it. Looks like it could be handy for 1 input to multiple DACs. 
Those discrete R2R DACs are expensive even back in the day when they ruled. Yes and there’s circuit board layout, earthing , filtering, output stages it all adds up. IMO though it’s not relevant to how it sounds unless it’s been purposely designed to not be accurate. The only connection DACs have with the "outside world" is the output terminal. If we measure the magnitude frequency and phase well beyond anything a human can hear does it really matter how it was achieved? DS, R2R , which brand of chip, exotic components? If the output is accurate what difference does it make ? I don’t even know what DAC I have, what chip it uses no idea. I know I have class D amps but no idea what brand. They’re inside my active speakers and all I know they are the most accurate most detailed, smoothest speakers I ever heard. That’s where the rubber hits the road , the speaker, it’s the least accurate component with the most distortion in the chain, along with the room.