This isn't a quest for sufficiency, he stated the mission parameter as "best in the world".
That's why it went up from 24-bit to 28-bit, as soon as someone invents a 29-bit the sky-is-limit trophy is lost.
In engineering the best is approached only by narrowing down the focus. For example no-one expects an F1 car to be the best at carrying home the shopping or pulling a trailer. Similarly in DACs, the more sample rates one needs to cover the poorer the performance at any single one of them.
Last edited:
Hi,
Said another way round, the DAC may perform better with the lower clock.
If You're into high clock rate numbers then choose a sigma-delta
jauu
Calvin
Best here doesn't necessarily mean higher clock rate, as for example the issues with parasitic capacitances rise with increasing frequency.
Said another way round, the DAC may perform better with the lower clock.
If You're into high clock rate numbers then choose a sigma-delta
jauu
Calvin
It's already not the best in the world as the ES9018 already uses a 32 bit volume control.
Perhaps in engineering circles it is not the "best", but can the difference be detected by the human ear?
If you're trying to make the best DAC of its kind in the world, things like 'can it be detected by the human ear' fall by the wayside, technical perfection is as important as the way it sounds is.
Hi Soekris,
What about this question please ?
What ate the electrical outputs data measured at J7
What can be found between the last passive filter cap on the Resistors bench and J7 ?
I try to understand if the oaps can be totally bypassed and/or if the circuit need I/V conversion or gnd after the R2R bench ?
As not a power dac and if the embeded digital attenuation is not totally loss free (ES9018 volume control hint above?), I would like to understand how can be maid the impedance adaptation before an amp if needed (mine e.g. is 100K ohms input)
WHat is the best solution : digital attenuation with a streamer à la SqueezeBox : but here we have the same theorical loss (than O can't here myself with any dac after My squeezeBox when volume control is used in it !) or SOTA ladder resistor pot in an analog pre after the DAC ? If adaptation impedance is not as important with voltage than current, it seems a good ratio for good sound is around 1/10 to 1/20 ... at least this is my simple understanding.
Sorry for the two cents question, I'm not at technician !
Last edited:
The R-2R DAC resistor network is output directly at J7, no parts in between. The output voltage is 1.4V RMS, the output impedance is 1.2 Kohm. The output resistance is also purely resistive, so if heavily loaded the output voltage will just be lower.
So no I/V stage is needed and it's powerful enough to connect directly to a power amplifier....
The DAC volume control is better than any external one, is done precisely in the digital domain with high resolution, and the sign magnitude DAC works the same as the level goes down. T.ex. when volume is at - 72 db it still has 16 bit resolution, just look like there is a -72 db resistive divider after it as the 12 MSB bit resistors's are connected to GND.
Don't know about the rest of yous, but I'm excited about this DAC design. The IV and output stages has always involved a bunch of compromises of some sort in previous DACs I've played with. I'd be regularly thinking of possible tweaks to this or that to get 'better' sound. I never felt it was 'perfect'. Nothing's perfect I guess, but maybe this design could help relieve me from some of the stresses
Then again, how this dac will sound like is still a complete unknown and that can be bit stressful too! Lol
Then again, how this dac will sound like is still a complete unknown and that can be bit stressful too! Lol
Last edited:
Well, as from the listening experiences (high resolution equipment and using acoustic music): You should drive the DAC with ALL possible bits to get the highest listening resolutions...
and afterwords attenuate while as Soren above declared, you will deal with lower used bits and you will sooner or later
Hp