"ancient" DAC chipset; (?!)
being sarcastic- 'yes', cause designed for mass market that favours 'esoteric sampling rates' and 'long battery life' as features
greater than tuning (top priority) being absolute SOUND QUALITY, flavours the NEW MARKET CHIPS towards measuring well.
I seriously consider my DAC chip (that will be implemented in a circuit design, for me- ideally without wireless radios nearby), and as someone who listens to 99% of my audio source files (CD/'redbook' being 44khz @ 16bit), as needing to do this one simple task well.
Back before large corporation buyouts when Wolfson and BurrBrown were 'the stuff', they built designs that did the task
'very well'.
The job required by the redbook format (established in the seventies?) was 'pretty confident' and a task that was now refined and affordable/doable with 'simpler' REFINED chips.
The market had kept what worked- the desire being ULTIMATE
TWO CHANNEL audio...
But then market direction wanted surround formats, and licensed (royalties expiring) for Philips/Sony on redbook; needed a new replacement format. Minidisc never got the traction needed for a market dominance, and hi res format wars (DVD audio and SACD) never were widely adopted, so all the hybrid DACs and '
crap' that was churned out whilst the 'small' (by todays' standards) audio chipset companies were easy buyouts to larger corps (like Texas Instruments).
Cirrus Logic liked being 'cutting edge' (were the PC soundcard chipset of choice mostly) and targetted spec sheets numbers more than the companies building 'audio' first designs.
The modern company equivalent was the ESS designs (pretty sure a whole slew of sound cards in the early nineties may have been ESS, but then I am not going to do research for a 'coffee rant post', and don't care about the origins of a lot of the companies that have seldom delivered a chipset that I care for.
I know that some late eighties Philips used parts were phenominal, and those players held there own very well OVER A DECADE LATER, vs the 'better spec sheet' Wolfson and BurrBrown designs of the late nineties/early 2000s.
The market wants and eight channel DAC chips that sips power, and two channel DACs (ideally that can do other functions than just two channel Digital Audio conversion), and filters and all sorts of things that a great design may have used OFFBOARD to the DAC chip (and required DAC chip pinouts for flexibility in implementation.)
The designs that were floating around in the market circa 2000 (give or take four years) were very refined, supported methods that allowed a clever circuit builder to 'take them MUCH FURTHER' (meaning total BOM for any given product being highly flexible), and didn't waste a tonne of their silicon space with multiple pathways for various formats and unproven sampling rates in case a user wanted to dabble with some 'soon to be forgotten about by the market' piece of tech.
(and I have DVD audio and SACD sources and discs)
@Bob - yes I know you like DSD conversion of everything you play
@majority of worlds consumers.. mp3s are great and subscription internet audio (or chewyyoube) as audio sources is acceptable for a wee bluetooth speaker (including 'in ear' variants)..
Given that most of the world listens to audio contained in formats that are not as good as 16bit CD (eg MQA on 'non MQA kit', and well
anything compressed from a CD 'master'-=ahem=-), a great DAC chip from a time period when audio actually mattered FOR THE SOUND QUALITY is all fine and dandy.
Most wouldn't care and majority of DAC chips existing or used in majority of the worlds' tech would LOVE an upgrade to the 'lowly' BurrBrown 'sound'.
Like all 'future formats'; technically brilliant and deliver excellence when implemented in perfect circuits and review well, they are then implemented in lesser kit and sold to the mainstream and generally considered as 'great' based on the white DOC spec sheet of the chips used in the circuit (and not the measured output of the ACTUAL IMPLEMENTATION), things like HDR and 4K start to equal what the elite owned in their reference kit of the 'old generation tech'.
As an example - flagship parts (TVs or surround receivers) of the old tech is generally matched in performance by the 'newer' replacement tech after it starts to become 'mass produced'... (and the quality output slips back to what the mass market can 'get away with').
That 4K TV that has 600 lines of motion resolution is no better (for motion resolution) than many a flagship 720p ('high def') TV sets.
The rich get to own the nice stuff, and then the mass market gets stuff compatible with it.
Some companies have streamlined this process, and is why companies like Panasonic leave markets cause Samsung are so great at 'fudging numbers' and representing products as better (when they are inferior) and consumers being 'yellow belt consumers' all buy based on spec sheets.
Well- DAC chips are products than need to market themselves and so target 'the numbers'. This wasn't necessarily 'always the case', or the number/spec measured points didn't need to scale/vary by large margins to be 'all things to all people', and so the designs still were largely capable of being tuned for musical quality output.
Any circuit builder working with a part that they know how to squeeze the most from (DAC chip white DOCs often give a few suggestions for a range of ways to implement the chip); I'd take a part that can be maxed out or pushed to perfection reliably, over a part that may never get the support (software/firmware) to make it shine, or 'just does a few formats to achieve sales sheet "bullet points"!'.
True I have only played with DSD output from DAPs and PC 'a little'.
I find that the DSD sound can improve things, sure,..
But feeding redbook into a GTO filtering iFi product gets me to the same place, quality wise, give or take a smidge (I prefer it), and is a lot easier to implement (even my Nintendo Switch gets to enjoy the benefit).
The majority of the time I use high quality transport and Non OverSampling (NOS) purity.
This has often been a hard 'high watermark' to
get right.
Most consumers don't notice
dithered 44khz (say, playing at 48khz sampling rate), so semantics is super easy to argue here.
True,"most consumers" break the notion of any chance at audio being a subjectively agreeable thing.
An mp3 with an overall louder presentation is going to sound 'better' to most people. Enough so to skew any results as to be meaningless when testing kit and using blind testing or whatever some fanbase decides are needed to 'prove themselves right' *the phenomenological universe is filled with stimulus and observable exchanges that will support findings towards just about any outcome- 'the world is a large place' and so "we (can) see what we want to see"/"can find evidence to support just about any notion"..
No matter how we feel towards the numbers written on the pages of the accompanying spec sheets of modern world 'refinement' DAC chip parts- whether a chip built with its first goal being to 'use less power' (an attractive spec sheet measurement as majority of consumers WANT THIS more than quality), and lost a quarter of its total die space to supporting formats an end user
may never use (or risk being passed over for a product that allows "more sales sheet bullet points"), I have no qualms using 'an older design' built in a time period ignorant to future format wars (that amounted to *nothing*)..
And given how cheap most mainstream stuff needs be made in order to compete in the market, off board high quality filter chips etc is generally passed up as a function more cheaply handled by using a general purpose CPU reprogrammable to do whatever is needed to handshake and provide compatibility for anything a consumer may desire, if only for bragging rights or 'feeling safe' that their new purchase can DO EVERYTHING and is THE BEST.
In my experience, 'the best' is often performed by parts built to do one task and do one task 'well'.
I have DVD players from the nineties with red colour output that actually nearly matches typical 'last generation' flagship laserdisc players, and even flagship settop boxes with red colour output that is very impressive.
But everyone wanted cheaper products and their set top boxes built into the TV, which for 'cost savings'/'competitive' reasons, couldn't afford the space and implementation cost (or heat) and so were 'compatible' with the standards, but didn't really deliver what the reference units were establishing that the products
were capable of.
In hundreds of 'fast switching' moments comparing the Diablo vs a Topping D90, the Diablo sounds like a super high quality DAC box...
The edge of human voices and drum tonality and texture are HUGE giveaways that these products are simply not in the same league.
From an engagement perspective, one piece of tech is simply mass market junk by comparison (or 'very obviously' not doing the same outstanding job of turning zeros and ones/'digital' into analogue wavelengths... )
From a marketing perspective, being close to the same pricepoint, and the D90 doing a plethora of functions and having the joy of using 'unlimited power' to do so, whilst containing 'fifteen years+' newer DAC chip part(s), it has a very obvious 'digital grain' or 'thinness' that the Diablo simply does not- by comparison the Diablo has a warmth to every instrument (if called for), and sounds like large expensive DACs built at the height of two channel audio..
The Diablo isn't equal to the best, sure,.. but given most consumers can't spot sampling rate mismatches, or discern volume discrepancies as 'better', the shortfallings in the Diablo sound (vs esoterically priced stuff, and GOOD R2R), is 'good enough' and of no real difference, especially when played back into 'matching price point parts'.
@Bob Parish - like you mate, I use a vastly better sound setup (by price point) than the DAC cost I spend, on 'flavour of the moment' technology, that DACs seemed to have become.
Like replacing Valves to tune an amplifier sound, the DAC
could be considered/implemented this way, sure..
I actually feed my Diablo through a Grace Design m903, that uses a near identical BurrBrown DAC chip with a great circuit implementation too; but the magic that iFi once bought, when having an experienced top tier engineer, a veritable legend, come up with a
total design that squeezes WELL BEYOND the rated spec sheet performance that these 'ancient' BB DAC chips were listed as giving, is proof/testimony that knowing a part and implementing it well
might just yield better than being flavour of the moment and simply targetting ALL consumer wants.
We know the old saying; "jack of all trades (master of none)".
The Diablo is a master of 'some', and it just so happens that 'some' is the EXACT SECRET SAUCE I want my DAC to perform.
The Diablo blows away the Grace m903 (but not in terms of features) and if sound quality was my single goal, the Diablo is hands down the best DAC box I have come across for sane money.
The Diablo might fudge its brilliance in ways that I am not aware, but that is fine- it is called
suspension of disbelief and it does its sole task admirably.
If I really want the sound that all these 'future formats' that have come along since the old BurrBrown designs, then I can always employ the iFi "GTO" filter.
With a massive price miss-match in my system design (and all sorts of silly spendings on power conditioners and cables and 'stuff'), the lowly Diablo holds its' prestige well in a system that might simply use a Diablo as a doorstop (by price point)..
I'd not say the DAC part of the Diablo lets down the amp part (I use it the 'other way around' personally) (improving the DAC output by using better amps downstream!).
YMMV
(new isn't necessarily better)