Lower than Spring 3!?
1 Like
I’m just looking what is there on the specs section for DSD128 input and XLR output:
Cyan2 : 2.5 Vrms , THD+N -114dB
Spring3 : 2.9 Vrms, THD+N -112 dB (Holo May is same)
These are Holo’s numbers
Just consider an extreme example, a really low output voltage , feeding same preamp and power amp and speakers as a higher output voltage DAC
You need to amplify that noise and distortion more that with the higher output voltage DAC, to reach same SPL out of the speakers
And you may not even have enough gain to reach that SPL in extreme example
Just an example
2.5 vs 2.9 of course is not an example of “extreme” difference obviously.
AFAIK 2.5V is the industry standard.
I.e. more on par with other sources one might have.
A26 is 2.5 for RCA and… 5v for XLR with 127db SNR.
Ok, it’s not a R2R… but DSD Direct ! 
but this Holo interests me…
1 Like
Old fashioned standard I agree. And not applicable to pro audio sources.
And i also have consumer headphones amplifiers now comfortable with 10Vrms input
Having said all that 2.5Vrms is not bad. 2.9Vrms better for me
My incoming SMSL DO300EX is 3.7Vrms in DSD Direct mode
1 Like
Nowadays the problem is more vice versa. Preamp should be called “preatt” these days. Since it doesn’t amplify anything, instead it attenuates. And then it gets amplified again by power amp. So if DAC has lower output voltage, preamp needs to attenuate less, so there is less back and forth attenuation / amplification and you get less noise as result.
So in practice, the voltage coming out of preamp to power amp is typically less than 1/10th of the DAC output voltage… And the voltage going to loudspeakers is close to the DAC output voltage. (typical floor stander having sensitivity of for example 89 dB at 2.83V input)
1.3 dB…
2 Likes
I know my DSC-1 inspired DAC had something like 1.2Vrms output and was a problem with my headphone amp
I literally ran out of volume for some tracks (combination of low output voltage and low headamp gain)
Hi folks,
Just to share some update. I got ASDM7ECv3 working with PCM 192 by changing the following:
change multicore = “auto” to multicore =“1” on line “engine” in /etc/hqplayer/hqplayerd.xml
Hopefully there is no other side effect.
I have the following setup:
Qobuz → Roon ROCK (NUC10 i3 with local music) → HQPlayer embedded (i7 gen 6, no GPU)-> RopieeeXL → Topping E30 (DAC mode) → Topping L30 → Hifiman Sundara
HQPlayer setting:
crossfeed and matrix pipeline enabled
filter: 1x: poly-sinc-gauss-long, nx: poly-sinc-gauss-hires-lp
shaper: ASDM7ECv3
Output: 11289600 (DSD256/44.1)
1 Like
I am interested to purchase dac that is capable of dsd direct, to use with HQP dsd upsampler. Currently on Chord Qutest doing pcm.
The problem is, my amplifier is unbalanced/rca only.
If i were to acquire a gustard a26/smsl d400
ex, is there anything i need to note regarding their analogue output (unbalanced vs balanced) performance?
That’s a different thing. But it is partially due to headphones having much wider impedance and sensitivity variety than loudspeakers. For this reason, many headphone amplifiers like Ferrum Oor have multiple gain settings. I don’t have such issues with the Ferrum Oor, or with Schiit Jotunheim. iFi headphone amps have also multiple gain settings. For example typical IEMs need very different gain than for example my 600 ohm AKG K601’s.
SMSL won’t do for now. Gustard A26 is an option given that the fixed DAC controller firmware is installed.
Nothing special about the analog outputs. Only note that you cannot use balanced and unbalanced simultaneously, as they don’t seem to be independently buffered.
1 Like
But if you use the same headamp and same headphones and same gain settings - and only compare 1.2V vs 3V output DAC , and you run out of gain/volume with 1.2V out
Difference can only be due to voltage outputs right?
All other things being equal and 47kohm input impedance headamp
Yes, but if you run out of gain with such difference, your combination of headphone amp and headphones is not optimal. That is still just a bit over 6 dB difference.
With high voltage DAC output it is much more common to have preamp having volume set to something between -40 and -30 dB and then power amp having gain of about 36 dB. So the net result is near 0 dB and there is totally unnecessary back and forth attenuation and gain of 36 dB which will just make SNR suffer.
If you run DAC direct to power amp, 1.2V is enough for most power amps to run way into clipping.
For example with Ferrum Wandla, I need to set DAC output trim to -12 dB to avoid amplifier input circuitry from clipping.
2 Likes
All noted. In my example I compare 2 different DACs with headamp in low gain setting
In reality, for low voltage DAC i simply switch to high gain headamp setting and problem disappears for any headphones
Headamp has < 0.5 ohm output impedance , good match for all headphones
1 Like
Not that, but the gain… Something like 16 ohm IEM’s need completely different voltage gain than the big over the ear 600 ohm AKG K601’s. This is why for example iFi DAC/headphone amps have the IEMatch and gain settings. Because when the voltage gain would match K601’s, it would be unbearably narrow adjustment range with highly sensitive low impedance IEMs and you would just have danger of breaking your ears…
I have wide variety of headphones and just need to use different headphone amp gain settings with different headphones. Since their sensitivity in terms of dB / V is totally different. One could give 100 dB at 100 mV while other one could require 3 V for the same 100 dB.
For the ones that give 100 dB already at 100 mV input, it would have pretty big negative SNR impact if DAC output is for example 4 V.
1 Like
I have high anticipation for this Cyan2 performance!
I’ve been wanting a discrete design (non chip based) DSD to analogue converter for < USD1000 for a long time.
If it performs well this could be an instant classic !
6 Likes
Hi All,
Singxer seem to be releasing an update to the SDA-6 PRO using dual 4499EX’s. SDA-6 PRO2 now listed on Audiophonics.
“its digital-to-analog conversion stage is based on two AKM AK4499EX DAC chips coupled with the AK4191 digital filter. These support high-resolution streams up to PCM 32Bit 768kHz and native DSD up to DSD512. The DSD stream operates in NOS (Non Over Sampling) mode, without passing through the AK4191 filter, in order to respect the original signal as closely as possible.”
Sounds like it is always running in DSD Direct mode
These really need to be checked for DSD Direct by measurements. Since failing on this front with this chip pairing seems to be be more norm than exception…
That one is not so cheap, so I have to see when I’d have budget to get one for testing.
3 Likes