Does the 'Buck' Start with the DAC?

Some members have expressed that the server / streaming front end is less important than the DAC.

I subscribe to the position that "Everything Matters"

As I consider front end choices, I'm trying to gauge if I should be leaning more heavily on a DAC versus a SOURCE (server / streamer).

Feedback and perspective from both camps, or any others, will be helpful. Thank you.

Please Note:  I'm not looking for validation of the position that "only the bits" matter... I accept that some believe this wholeheartedly... and that's okay by me. 
@skanda  Thanks for letting us know and the link. Looking forward to learning more from your journey and experience.
FYI for those interested in the Taiko Extreme....a 5 part review is online at Audiophile Style.
It depends on what you’re streaming. If it’s from a NAS or other home based server source then that’s your front end. If you’re streaming from internet sources then your modem is the front end. If you're in the camp everything matters how much importance is placed on your modem?
What a lovely thread David :-) what audiogon could and should be !

As a dark science devotee, and hard core math dude - most Economists are, I await a new law to replace diminishing returns...

but, while we wait, let’s listen to fantastic music !

yes, everything matters and not in equal weight, but innovation and improvements don’t necessarily flow to higher weighted components of what should be an engineered system. I agree, noise and signal quality are hyper important and frankly solutions are affordable- linear PSU and isolation of any switching supplies remaining on a dedicated circuit into a power conditioner with no analog devices on it is imo a must. Budget some $ for vibration management as well, I use HRS Nimbus and a Damper block on NAIM servers ( SSD and Spinning disc ) as well as isolators on Synology NAS. Teddy Pardo makes a variety of LPS.

David - I would be interested in your finding on listening to various DCS components... especially the Rossini.

finally I don’t think we can totally separate hardware and software innovation and a companies longevity of the firm and product upgrades over time. Clearly DCS, NAIM, etc ( many others ) have a track record. I value that my Aesthetix DAC is FPGA, somewhat future proof, hope Jim White engages with that future !!!!

best to all

Everyone is always shocked at how much improvement can come from a streamer, a DAC, an amp, etc. They also tend to think that with any major improvement they don’t have much left to improve. That is a poor conclusion.

This is typical; all systems have many levels of improvement available to them. Power cords, Ethernet to optical conversion, interconnects, preamps, tube changes, etc. Whatever touches the power and signal paths can be upgraded with efficacious results.

For some reason, likely inflated pride, the audiophile tends to think their rig as it sits is oh, so close to SOTA. Not really, as most are quite a ways from it. I know that my system is still quite a ways from what can be done ultimately. There are still many of levels of improvement available, and I will keep seeking them.

A comment was made in regard to the SONORE sonictransporter, "I admit to some curiousity on the SonicTransporter i9 Optical that has fiber optic output (through built-in Sonore Optical Module). " I wrote up both the SonicOrbiter and Signature Rendu SE for, and later added outboard the affordable systemOptique converter from Ethernet to optical. Sensational change, a gorgeous improvement. I have done a lot of changes to rigs, but the systemOptique conversion was efficacious beyond my expectations, a bonus in digital playback.

My guess is there will be some who scoff at such things. Good for them; that’s their problem. Let them sit with their mediocre sound, as they deserve it. The same group trots out misnomers such as the "jitter immunity" argument, that supposedly DACs immune to jitter render other such changes as server/streamers obsolete. Don’t you believe it. See my review of the Benchmark DAC3 and AHB3 Amps at, wherein I found that Benchmark products are every bit as sensitive to cable, playback changes as any other component. 

The closest I have found in regard to an agnostic result in regard to the outcome given the signal input is the Exogal Comet, but even that unit can be changed significantly through matching with different components, and surprisingly, by switching from use of the on board hardware attenuation to the media software attenuation! No, I'm not interested in the snarky comments of the ignorant/Inexperienced in that regard.  :) 

I developed an adage in regard to the hard-headedness and skepticism of some: The greatest impediment to advancing an audiophile system is the audiophile.
Worldwide, I have a NET on the way to use with my WD signature.
What i wanted to add here, from my previous experience with Aurender N10, is to elevate the SQ from stream vs internal SSD/HDD, you need to work considerably on your ethernet.
I have opted for:
Ubiquity edgerouter X Sfp (no wifi, have separate wifi hub from router, optically isolated), 
Directly by AQ rj45 vodka into etherregen switch, then AQ diamond rj45 into aurender, powering both ubiquity and etherregen with ifi ipower low noise bricks (with Paul hynes SR dual rail on the way to replace those).
With above arrangement, i could almost swear SQ from stream is equal if not in certain cases better than local files (depending on quality of local file which can vary). I look forward to see how that works for the NET, now that aurender is out of the picture.
Try to improve your ethernet, you might be shocked of the results. Big impact was from the cheapest component, the ubiquity edgerouter X sfp! 
Will update in the forums once i put my hands on the NET...probably another month of wait.
@azwill Thanks for the post. I do appreciate the simplicity of the X1 system. You only have 1 input but that is considered a feature and not a bug by Lumin. The X1 is supposed to be a little darker (or not bright) sounding and that may not work the best with my next speakers so I am looking at others.

Another DAC system that has me intrigued is the AudioByte VOX 3 unit DAC system for $6K (sister company of Rockna). This DAC system has an RJ45 Ethernet input in the HUB unit and that unit is not released yet (likely July 2020). They do some jitter and other cleanup in the HUB. The 3 unit configuration sounds complicated but you only need 1 power cable to power all 3 units.

I needed a new DAC ASAP for headphones (also to serve as a backup) so bought the Benchmark DAC3B a few days ago. I should get it tomorrow. I will use a microRendu to stream to it. So I am now building up a DAC stable too.


In terms of the topic of this thread - it's hard to say which contributes more to the sound (DAC/streamer).  It's sort of a chicken and egg scenario, a better DAC will help, but without a streamer to unlock it, you won't get the most out of the DAC.  Leads me to believe 50:50 is the appropriate ratio.  


It really is a phenomenal DAC.  What's interesting is I used to only use it in NOS mode, as the upsampling filters tended to thin out the sound.  Post streamer - that is no longer an issue, and it sounds absolutely sublime.  It's incredible how a streamer really ties everything together in the system.  I'm shocked how much music I was missing with a 20k system !

In terms of streamers that have caught my eyes - I like both the Rockna and the Musica Pristina units as they both support i2s, which would allow me to offload the SU1 KTE (USB -> i2s converter) and recoup more funds when I upgrade. 

I also like the Musica Pristina as it is a pure streamer, so I'm not paying for a CD drive, hard drive etc., which I will not use.  I'm purely Tidal via Roon.  That leads me to believe I'm getting more bang for my buck in terms of sound quality (or alternatively it is just overpriced !). 

Unfortunately not many reviews out outside of one agon user with the older A Capella II model (which is quite old).  

I recently have aquired a Lumin X1 and have connected via optical fiber converted by etherRegen from copper.  My server currently is a SonicTransporter i5 CDR.  From the Lumin X1 I have Triode Wire XLR's into a Vinnie Rossi L2 Statement DHT preamp.  I am also trialing Qobuz. I have Roon and have experimented with Lumin App with an iPad. Previously I had only been using my music stored on the SonicTransporter.   For DACs, I have the X1, the L2 DAC in my Vinnie Rossi, a PS Audio DirectStream, and a Luxman D-06.  Yep, I will be reducing the herd of DACS.  My previous preference has been the Vinnie Rossi L2 2.0 DAC, though have enjoyed the other two as well.  I am using a microRendu 2.8 prior to the VR L2 2.0 DAC from the etherRegen for comparison.  The DirectStream has been very easy to implement, and the periodic software updates have been simple and interesting to try.

The Lumin X1 is still early on playing.  I had read that it takes 400 hours to reach its potential.  I am glad I was aware, because out ot the box, sound quality was not up to the others.  I have maybe 150 hours on the X1 now, and sound quality has dramatically improved.  Last weekend I did some back and forth listening between my L2 VR DAC vs the Lumin X1.   I still preferred the VR from that listening comparison, but it is much closer than prior.  If even subtle improvements keep developing on the X1, I expect it to be my favorite.  

I really like the simplicity of connections with the Lumin X1 and Roon integration of Qobuz and my stored music.  My preference with the other DACs has been native sample rate to the DAC. The VR DAC 2.0 has an upsample minimal phase upsample ablity to 768 kHz but I have preferred the NOS (non-oversampling setting).  I have tried experimenting with Roon upsampling and a short test of HQPlayer upsampling, but the i5 SonicTransporter lacks capability to reliably process very high sample rates without buffering and stalling.  The Lumin X1 and my VR can accept and play up to 768 kHz and 512 DSD fed to them.  I have curiousity about this, but am reluctant to embark on whole new server just for this without better knowledge of whether  I would use after all.  Innuos has many fans and I have mused about the line often.  The lack of ability to upsample is something I wish were different, but then again, I have not done so in my limited experience to date.

I admit to some curiousity on the SonicTransporter i9 Optical that has fiber optic output (through built-in Sonore Optical Module).  The SGM Extreme has optical as well, but pricing on it is well over $20,000.  I suspect we will see other servers add optical fiber outputs and hope to consider those options when they come out.

The Lumin X1 has some upsample capability, but again have so far preferred native input resolution.  However, I am experimenting with upsampling to 24-bit over 16-bit through the Lumin and have so far liked doing it.

Regarding which is more important--source/server vs DAC, my opinion is that both matter and are overall system dependent for results. I think a far separation in quality of either limits the final result, but feel the source/server with connection chain to the DAC to potentially be the bigger influencer.  


@ag3__  the dac you have is great (i have the same one!). if you are in the market for a server, deff consider the wavedream net. it very well could be an endgame for your rig. it has been serving me well and if not for the taiko extreme on the way i'd have no plans to sell it. i will update this thread with my thoughts on the taiko and maybe some comparisons with the wavedream net as well.

if you look at my post history here, i detail in another thread my progressions from laptop to server. the wavedream net is worth the cost for the rockna and not to solicit, if the taiko lives up to it's reputation i'll be selling my net so if you're up for waiting you might be able to score a killer deal :)

the other reason i'd suggest the net is that it has a specialized clock for the wavedream. now, from what i understand, in taiko's testing they found the source clock to not cause a large difference/offer expected benefits so there is no clocking of the signal in the extreme. however, in my system just this morning i was playing with local clock vs. stream clock and stream clock sounded better to me. not huge difference, but better. i'm very curious to see if the taiko pairing is better than the synergy offered by the rockna stack
What you are describing @ag3 is exactly what people experience when switching from a general purpose computer (i.e. work or home laptop) to a dedicated, audio purpose built streamer. It's indeed eye popping! I was there a few years ago, and never looked back.

Unfortunately, there are still many people streaming from their laptop, simply because they refuse to try anything else. After all, zeroes and ones....if there is sounds, it works.... either it works, or it doesn't ....
Just plugged in my SOtM SMS 200 Neo (~ 600 CAD).  Suffice to phenomenal gains !

TLDR: 20k system will realize very strong gains from using a hifi streamer vs basic computer.  

For background - my system is worth about ~ 20k:
- Basic Computer
- SU-1 KTE
- Rockna Wavedream SE DAC
- LTA MicroZOTL pre
- Pass XA25
- Tannoy Eaton
- Audio Sensibility Statement cables throughout

I shied away from a streamer just because I use this computer for everything and as a TV, so I didn't want to have 3 remotes switching back and forth (basically 1st world problems).

I decided to start small on the streamer (replacing the basic computer) - hence the SOtM purchase.  

Unreal.  The refinement in sound stage, imaging, low level detail and bass is quite astounding.  Everything you would expect a better digital signal to produce. 

For reference - the gains realized are larger than going from a Yggy to the Rockna, and bigger than moving from the FW F7 to a Pass XA25 (from a detail perspective, although this is more subjective - I still really enjoy the F7).

As next steps - I will be looking at bringing in a Sonore OpticalModule, and linear power supplies.  Ultimately moving to an end game streamer in the 4-5k range over the next couple years.  
I’m enjoying this thread.  Interestingly, I posed a related question elsewhere about which might have more impact - server/streamer or DAC.  While I subscribe to the notion that everything makes a difference, I’ve also concluded that, In my experiences, the server/streamer is at least as important and likely the more important determinant of quality sound than the DAC.  

Several users suggest that, because their DACs are known to be relatively immune to jitter, the streamer isn’t particularly important.  I don’t find this to be true.  My DAC also rejects jitter very well, yet the differences in sound between a variety of servers is pretty easy to hear.  So, I might suggest they audition better servers to see if it improves their results - if they’re so included in that pursuit.

My $.02

Innuos is a great piece of hardware for sure. And PhoenixUSB does help with USB connection. A lot
Does the buck start with the dac ? 
 I’ve never experienced a bigger “all at once jump” in musically performance as I have with a Innuos Zenith 3 and it’s all in one box Phoenix USB regenerator , power supplies and master clock .

Having borrowed a few dacs some beyond my budget like MSB , Chord Dave and Total dacs the all at once leap in musically performance wasn’t near the equal to that of the server and Phoenix component.
Though it’s all important..
Been looking to finally add a dedicated streamer to my set up. I'm starting small with the SotM SMS 200 Neo. 

During my research - there's a massive thread on the new Sonore OpticalModule and quite a bit of feedback noting the improvement of feeding a streamer with fiber (or "cleaning" the ethernet signal). This device along with some power supply upgrades will be my next few purchases. 

Eventually the plan is to work up to 50:50 split between streamer and DAC. 
"rockna stack: wavedream signature se with wavedream net. it's a great combination in my system."

@skanda  Thanks for sharing your findings and results on your Rockna component stack. Hopefully your experience will help others making similar choices. And, of course, looking forward to your impressions once you have the Wavelight / Taiko Extreme in your system.

I'm in agreement on your "laws" perspective. Would love to hear more on that topic, wherever / whenever you choose to share.  : )
i am actually connecting the rockna stack using a wireworld gold aes cable. i've always been a fan of aes and yeah ive heard all the arguments for using one vs. another but in my system i'm getting fantastic/best-i've-ever-heard sound. will i keep upgrading? hell yea! this is my hobby, there is no end goal, just keep trying stuff and if it's not better move it on out. i have my eye on a lampi pacific and maybe even new speakers but i found myself randomly in this server upgrade so should be interesting.

i will have a chance to a/b when the taiko arrives as i can return it back to my buddy. if i like it, i'll pay for it otherwise i'll give it back to my friend who ordered 3. i'll be running the taiko on usb but i reckon it will be an easy resale, if needed, given the waitlist for it. or if i like it more, i can sell the rockna and make some of that cost back. if anyone is interested in learning more feel free to pm me.

Thanks for posting your feedback. I hope you were using i2s connection between the wavedream net and wavedream se DAC? I’ve heard through i2s the pair kicks into another level of audio bliss.

Personally, I would not sell wavedream net unless you get a chance to A/B with Taiko Extreme. I have no doubts that you will hear an uptick in performance with Taiko but only you can justify Taiko performance worth spending the extra $18K.

I am in everything matters camp, so I agree server (source) is just as important as your downline components....DAC, Preamp, Amp, cables and speakers.
hey guys, i'm reposting a response i wrote on a thread about rockna wavelights. @david_ten suggested it might be valuable here. also, i want to add that diminishing returns is a reality, but "law of diminishing returns" is extremely misunderstood. as a mathematician i have my fair share of issues with economics but calling things laws when they are strictly subjective points is something that frustrates me but i can save that for another thread :) :

i currently have a rockna stack: wavedream signature se with wavedream net. it's a great combination in my system. i was running the rockna with a metrum ambre to start - so i could utilize roon with my laptop serving as the roon core. then i bought the rockna wavedream net on a bit of a whim but with some reservations. this was - again, in my system - a huge step forward. i was frankly stunned as i'm a computer guy and bits are bits...

the first that thing that stood out to me was how much bigger the soundstage became in all dimensions. after that initial "wow" wore off over some days and weeks i noticed that everything was just much better over all. the tone, texture, and all the other words used to describe play back that differs from merely listening to being captivated.

my rockna experience and the experiences of others with other servers has lead me to believe the server is even more important than the dac itself in that changes in servers are resulting in bigger sound differences than changes in dacs. am i saying buy a cheap dac and an amazing server? no, not necessarily (though i know of one user using a taiko extreme with a dragonfly - while they wait for their new dac and enjoying the sound). but for me, i'd take any dac budget and try to spend half to 75% on the server and the rest on the dac. i mention this because the terminator from everything i read (and it was high on my list before pulling the trigger on the rockna) may really benefit from a nice server as a source.

okay, now that i've gotten through that i'll get to the bad news: i haven't heard the wavelight and can't comment directly. however if it carries any of the tonal qualities of the wavedream net + wavedream signature se, i reckon it will sound great. rockna has a density to the sound that is just so palpable and it's able to flow so beautifully from one note to another. but everyone has different tastes so all i can do is share my opinion. for what it's worth i'm probably selling the wavedream net soon as i'd like to purchase a taiko extreme and really take the server experiment to it's illogical/logical extreme :). not meant as solicitation but just wanted to be honest about it. if you have any other questions about rockna, feel free to pm and i'll do my best to answer
I am constantly optimizing @david_ten 😂😂

I am currently using a Triode Wire Labs Split Data & Power USB cable with surprisingly very good results. It beats some more expensive USB cables I have and have had. Eventually I will get a Shunyata Sigma USB. Also, a nice Audiogon member here recommended the Kimber Select HB cable that I would also like to try.
[Emphasis: "I have not heard...even two Music Servers that sound exactly the same."]

"So far, I have not heard a USB cable that sounds precisely the same as a direct Ethernet connection, or even two Music Servers that sound exactly the same. The trick is to identify what makes the music sound most convincing and involving, in other words, what will work synergistically in a given setup and choosing the cable accordingly."

- Christiaan Punter ( from his review of Pink Faun’s USB cable.
@thyname  Check out the review [from which I am copying some information] for perspective on USB cables, in case you are still looking to optimize USB out from your Innuos Statement.
Thanks @david_ten ! I very much enjoy the Statement.

It’s a little of long story: I had preordered the Innuos PhoenixUSB for months. When I finally received it, I was shocked how much it improved my. ZENith MK3. Then an opportunity presented itself for a used Statement 2TB in great condition at a good price, so I took it.

Perhaps this is a bit of oversimplification, but going from ZENith alone to ZENith + PhoenixUSB combo, is a bigger improvement than from that same combo to the Statement. Needless to say, this is my impression only, with my ears in my system, and considering using USB for streaming. Results may be different for those with no USB in the chain.

For me, Statement is end-game, and I am pretty sure I will not be looking for other servers AND streamers in the near future
@thyname  Congratulations on adding the Innuos Statement to your system. Would love to hear more about it's impact on sound quality and overall system performance...

FYI: I continue to be impressed (amazed) with the T+A SDV 3100 HV. I have not pulled the trigger on a server upgrade yet. 

That’s spot on. I totally agree. I think of the digital chain as ONE body comprised of four elements:

1) Server 

2) Streamer 

3) DAC

4) Peripherals (power supplies, Internet transmission, cables)

Often times, two or more of these combined in one unit (I.e. DAC with a network bridge built in).

IMHO, the most important decision is on the DAC, then building around it for best synergy, and without ignoring the other three elements in the chain.

BTW @david_ten ... I was able to get an Innuos Statement for myself a month or so ago
"Regardless, the comparison underscored that server preferences are ultimately DAC-dependent and may change as equipment (including cables and supports) changes."

[Emphasis: ...server preferences are ultimately DAC-dependent...]

The above quote is from Jason Victor Serinus’ very recent Stereophile review of the Wolf Audio Alpha 3 SX server.

I find his [this] statement valuable for the context of this thread.
Of course it all matters but practically one will choose the DAC for the sound and the streamer mostly for the features. 
The Lumin X1 was discussed a lot in the Best High End DAC thread but I have yet to see a mention of the X1 connected by the fibre optical input. It cost about $200 to buy a network switch with Optical and maybe another $100 to connect this to the X1 from the switch with fibre and the connectors.

It seems to me the goal of dedicated audio streaming servers (computers) is to lower RF noise, using an optimized computer for very low noise and maybe some horsepower for DSP/upsampling options. So I can understand why people praise them so much. However, if the fibre optical input of the X1 is supposed to be the best way to stream data because it eliminates RF noise since this noise cannot travel on fibre optical  wire. Then is a expensive dedicated audio server really needed for a DAC with fibre optical input? Only person that I have seen address this is the owner of the ComuterAudiophile web site on his review of the Lumin X1.

I am not saying that the Lumin X1 is the best sounding DAC but I am wondering if it the best streaming DAC. I am thinking of getting this DAC just for the optical streaming feature so I can eliminate boxes. I would use a cheapo computer, running ROON, somewhere on my network (not necessarily in the same room as the audio system).

The other option I was considering was a optical Rendu from Sonore but that adds another box and the data signal eventually goes back to USB when going into a non-Lumin X1 DAC. However, this option does open up the choice of DACs if the Lumin X1 is not your type of flavour. 

It also seems like most are saying that USB DAC streaming input sounds better in most DACs than RJ45 Ethernet DAC streaming input, for example Mola Mola Tambaqui DAC. I look forward to the day when all (or more) DACs have fibre optical input.
OK, these last few posts are a real aberration and at the highest 0.001% of this hobby. They should be ignored....;-)

I think the key is to match streamers and DACs intelligently.
As with everything, I target "92-95%" of what cost-no-object techos seek. The logic being the last 5-8% costs about $2,000 per 1%...;p-)

Some DACs are fine with a $40 streamer as they ignore jitter.

My $0.02...:

1. Get a $300-$500 streamer that can do both AirPlay and Bluetooth and UPnP.
2. Get a $750 - $1500 DAC that sonically matches your system.
3. Get a good digital cable.

If you want to go larger, then open your wallet wider.
Based on some of the responses here, it looks like you can fully isolate your DAC, whether Ethernet, USB, or otherwise for < $1,000. Given that, I am having a hard time understanding justification for a high end streamer?
I just did a major upgrade on my streamer - so I'm now about 50/50 in cost between DAC and streamer.  


  • DAC:  Mojo Mystique V3 - (single ended)
  • Digital front:  DCS Network Bridge + GigaFoilv4 ethernet/optical filter

If I add the cost of  the new LPS's for my cable modem and the new gigafoil, as well as ethernet cables, I'm more at 45/55.

I was talking with the owner of Mojo Audio about 6 months ago.  His opinion was that I was only getting maybe 30% of the sonic quality of my DAC with my current streamer (modded Sonos Connect).  He's a big proponent of source, power and cables in digital.   After this recent upgrade, I agree - just phenomenal improvements to my ears.

I just ordered a BlackCat Digit 75 coax.  Hoping for a final boost.

Innuos Statement would definitely be end-game for me on the server / streamer front. I just wish I had that much cash. Maybe one day. Probably at the next Statement MK2 😉
I’m in the camp of it all matters. But my digital front end is 36% streamer : 64% DAC

Thanks @biketony

@thyname  I have the i7 SGC server. I need the extra horsepower.

Regarding the Zenith Mk3 unit, it was only in my system for a few hours. At that time, I found it to be a toss up between my front end and the Innous...power cables/LPSs were a factor. I've made adjustments to the power side of my front end...and it is now performing above where it was when the comparison was made. Would the Mk3 + Phoenix surpass it, I cannot say, but I believe what I have vs that combo will be close. Which points to the Statement as the next step, should I go with Innuos.
I’m in the camp of it all matters. But my digital front end is
36% streamer
64% DAC 
Innuos zenith mkiii 
Aqua La Scala mk2 optologic.

But it’s what worked for me. Tried more than a few DACs and streamers. But that’s part of the fun of it for me. Looking and listening. I’m lucky to have a audio salon in my area that will let me take home just about any of the items they carry for weeks at a time to audition in my home. 
Just one more reason to support the local audio stores. 
BTW... I used the word salon instead of store. Kinda makes me feel like I’m “all that” using words like
Not sure if it is a good ratio to target. I definitely like how my system sounds. 
You no longer have the ZENith MK3? Why? I used to own the sonicTransporter i5 a while back, as Roon Core, and while it was more powerful in processing power compared to Innuos line, the ZENith sounded noticeably better to my ears, plus extra features, etc. I don’t do any heavy DSP, and my library is relatively small at slightly over 1,000 albums, so processing power not an issue for me
@thyname  Your current breakdown is roughly 3 : 1

Do you feel that allocation is a good one to target?

I have had the Zenith M3 in my system.

I chose to leave cabling out of the discussion, for obvious reasons. : )
@headphonedreams  Thanks for your suggested allocation of 2 : 1.

I have heard the dCS Bartok in a system, as well as the Rossini (twice) and a full Vivaldi stack.
Based on the responses so far, the DAC is a higher priority than the streaming source.

What percentage breakdown do members advise? 

DAC : Streaming Source

@david_ten : Is your sonicTransporter also your streamer?

Here is my current breakdown in MSRP prices (obviously, getting a nice deal or used helps 😉):

DAC: $13,500 (MSB Discrete with ProISL USB module and second optional discrete power supply)

Server / Streamer (USB out to DAC): $4,250 (Innuos ZENith MK3)

Then of course you have the USB cable, the power cords, the Ethernet cable, etc.
@jaytor - I use the Acoustic Zen MC-2 silver AES/EBU digital interconnect cable, because everyone recommends AES/EBU provides the best sound quality.  

I was surprised (and pleased) at how great the Yggy improved the sound of my system - would never have believed it had I not tried it.  No matter what all the reviews have raved about it, compared to all the $25,000-$50,000 Dac's.

@in_shore I currently have approximately 6K in my streaming source (components + LPS). And 4.5K in the DAC.

Based on what I am hearing within my own system as well as auditions, it is going to take a sizeable jump in spend to justify the upgrades on both fronts. Much, much more on the DAC front.

Unfortunate, but also a reality.
Wow , David you have a budget of 20 to 30 grand for a digital front end ? What fun that would be though some here suggested a Shiite dac and Bryson CD player can’t be beat and anything spent over that is getting on a fast train of disminishing returns .

Though I much prefer a server IE ripped CDs over streaming myself here’s one fairly easy example to highlight the critical importance of both source and dac .

This past summer we had a Bluesound vault and a Innous MK 3 with out board dacs for comparison and I can tell you the differences in musically performance was striking. 
Anyway have fun .

Some have hinted at rough allocations for DAC vs Streaming Source.

Assuming mid- to higher- level components, what allocations would you suggest for the DAC and Streaming Source?

- 20K combined budget

- 30K combined budget

Thanks. - David.
Thanks for the additional responses. So far it appears that the DAC has priority over the streaming source. And that the better the streaming source, the better the downstream system performance.