Rockna Wavedream DAC: Awesomesauce

I do something similar with a Saga OG (that has a remote) and a DNA Starlett. With these SET designs, the tubes are heated to the same point no matter what the potentiometer is set to, so unless someone corrects me they are being run just as "hard" either way. If instead of analog I was using a digital volume control in front of the Starlett (e.g. the Wavedream), I would use the pot on the amp such that I limiting how much digital attenuation I was regularly using for the reasons @GoldenOne stated.

 

This is crazy talk, if you already have a pot in the signal path then using digital attentuation is just the worst of all worlds. Even small amounts of digital gain reduction (eg ReplayGain) sound really bad imo.

I just tried once again dialling 20dB of reduction into the Wavedream and turning the pot up on the EC Aficionado. It sounds totally busted. Nicolae is an amazing digital wizard but says some funny things (up to 40dB digital attentuation is fine, my DACs don't need more than 30 mins warm up time).

 

Honestly, I wonder if this experiment says something more about your chain then anything generalizable about "small amounts" of digital gain reduction, although I don't think 20dB really is a "small amount", though < 5 db and maybe even less then < 10 db probably can be usefully termed "small reduction".

Looking at Goldenone's measurement I see the Wavedream output a hot 7Vrms signal. Could it be you like the sound of your over-saturated input transformers on your Aficionado and a weaker signal sounds "busted"?

 

I take your point that 20dB is a lot. That was because Nicolae says <40dB is fine which is definitely optimistic. But even shaving 4-5dB off to me loses a lot of immediacy and microdynamic appeal.

Mine is the less hot Wavedream and using SE connection. So only 2Vrms and no transformers to hit. This is also consistent with various other configs I tried (Gungnir and Yggdrasil -> powered monitor input trafos, Yggdrasil -> Freya, Yggdrasil -> Af). Digital gain is just very disappointing.

 

Do you have Roon? If so perhaps another experiment - try turning on DSP and using 'headroom management' or para EQ (it's the same sox implementation either way if I'm not mistaken) and shave off -1db and run the same 16/44 PCM files as before. Can you hear a difference (beyond a -1db volume reduction)?

No offense, but I can't tell the differnce between statements like yours I quoted and "I find tubes very disappointing" or "I find SS very disappointing" or "I find black painted amps disappointing"

 

I also assumed this would be the case and initially had a goldpoint after my Wavedream to do the heavy lifting but ended up realizing I preferred it out of the chain and going DAC straight to amp. This isn’t ReplayGain (which I agree sucks), the signal’s being processed in the DAC at 63 or 68 bits anyway. Can’t remember which of the two.

 

Volume leveling in Roon is done at 64bit. I use it and, to be honest, I never felt any sound degradation.
Only lower volume, as intended.

 

I liked the digital attenuation up to about -20db with my adam a77x. When i moved to atc150aslt i needed to use about -30-40db to bring it down to comfortable levels and i felt like i was losing some detail and dynamics going beyond -20db.

I compared it with the atc sca-2 preamp which for me was a step up from the digital attenuation with rockna.

Some people seem to prefer or atleast believe the rockna digital volume beats every preamp, but in my experience a good preamp beats rockna especially if you have to attenuate beyond -20db.

 

Thanks for sharing, will keep this in mind.

 

Thanks for the hands on, helpful post. Hoping I hear no audible downgrades using up to -20db. Simply for times I'd rather be lazy and use the remote with the Studio. Not interested in an added preamp in the chain, nor the likely cost of one that would do the WD and SB justice...

 

As another viewpoint. I bought a Goldpoint to use with my Wavedream, intending to do bulk volume adjustment with it, and use the digital control and remote from my seat, assuming that would be ideal.

But when I did comparisons of using the chain with and without the gold point, I greatly preferred the system without the goldpoint inserted, doing all attenuation in the digital domain, usually between 30-40db depending on the album. staging depth improved, but more importantly immediacy and detail took a step up.

As with all things, ymmv, and you should try for yourself, but don't count out the digital volume control of the wave dream based on dogma.

 

There's a bit more to it than that.
Dynamic range is always a concern, and generally it'd be advised to go with the highest dynamic range option to do vol control. If you have a preamp with higher dynamic range than your DAC then in theory that will be better for preserving DR of the whole system.

BUT, with digital volume control there's also the distortion profile vs level to take into account.
Most DACs have a THD+N vs level graph that looks something like this:



Basically a straight line. Distortion profile does not change depending on level, it just rises/lowers proportionally to the signal.
The only thing that's different is that on many DACs sometimes distortion will go up a bit toward the upper end as you near full output. In the case of the one above, not by much, in the case of the one below, a bit more drastic and you'd probably actually WANT to have some DSP vol control to avoid higher level signals being un-necessarily distorted:



But the Rockna Wavedream shows different behaviour:



Between -20dB and -50dB, the distortion level/profile changes quite a lot.
This means that as you apply digital volume control, you're shifting where this area of changing distortion is in relation to the content of the music itself, and so you will change the sound.
There's not necessarily a 'right' answer as to what you should do. Leave it at no digital attenuation and that area will be firmly within almost all music's DR. But apply enough DSP to avoid it and you'll lose almost 50dB of dynamic range.

The point though is that in this instance, digital attenuation will CHANGE the sound to some degree more than on other DACs. And so isn't realistically a 'transparent' option.


(Side Note: Personally I have a hunch that this change in distortion vs level is what helps the Wavedream to sound so dynamic. Other dynamic DACs like the Yggdrasil also show something similar:



Would be cool to have a way of simulating this behaviour on other DACs to try it out!)

 

I've been trying to figure out why I much prefer the Goldpoint out of my chain. I thought it was because the Goldpoint might have been subtly doing something odd to the signal, but maybe there's a little bit of something in the changed distortion profile that is nice.

Anyway, I just ordered a slagleformer AVC and will have a few others over to hear it in and out of the chain, for science. Will report back.

 

Good stuff, please let us updated once you have that DAC.

 

Calm.

That's the first word that came to mind within 3 seconds of hearing the WD. Calm, not to be confused with laid back or relaxed. Very dynamic. Amp of course Studio B w/ EML mesh, Walthen driver and rectifier. Ether 2 and AD2000s. Microdetails are great, and my headphones are officially my weakest link now for once. Microdynamics are spectacular, and being able to hear exactly how hard each piano key is being hit, all within an overarching dynamic crescendo is really incredible...dynamics within dynamics. Piano in general just sounds so damn real and epic, and all percussion for that matter.

Only minor subjective nitpick so far is that I wish the mids were a titch more forward like my Eximus dp1, but I certainly wouldn't call them recessed. I also feel the Eximus is slightly more transparent than the WD, but even then it could be each recording. The dp1 tends to make anything sound kinda sweet, whereas the WD takes on every aspect of what it's fed.

Those are my still early impressions, but really feeling this WD is better served with some higher end headphones, or great speakers...

 

It needs about a week of on-time to reach equilibrium. With a good source, clock mode stream rather than local I think puts a bit more texture on the mids.

 

I believe it...I usually put my thoughts on the backburner in general for the first 200 hours for all of my audio stuffs. Sounds great though, and again piano/percussion is addicting in terms of realistic timbre and dynamics, it's most outstanding attribute to my ears thus far.

 

According to Nicolae the clock mode should be set always on "local" with good streamers.

I just paste here below his answer:

"local clock is used when input clock is good (should be used most of the times).
Stream clock is when input source is poorer quality."

 

This is counter intuitive. To me "local clock" would mean the DAC reclocks the incoming data, while "Stream" would keep the timing of the incoming signal intact, hence preferred for better quality sources.
Although it might be a case where "Stream" uses a looser VCO reclocking scheme and "Local" uses a tight VCXO instead (kinda like Schiit's Yggdrasil and original Gungnir Multibit implementation).

 

Another explanation I received from a Rockna reseller:

"If source is bad... and your Rockna unit cannot clock correctly.. you use stream as solution...
With good source always local"