WiiM Ultra Streamer Preamp Review

[SoySalsaPicante]

…And the funny thing is that people constantly ask for higher sample rates where things can go really bad as computing demands skyrocket.

Is this something that they can optimize a hardware DSP implementation for (with the aim of better efficiency/accuracy compared to general-purpose computation devices)?

 

[Sokel]

That's either inaudible, or close (in other words inaudible in real-world listening).

Agreed, what you see at this chart is way inaudible.
The thing is that what you see has to be translated in analog at some stage, adding all the analog "sins" to it.

Think of it as if it didn't have the penalty, take the glorious -144dB THD+N shown at the chart and just pass through the best DAC out there, you'll get -125dB THD+N something?
Now do the same for the degraded. I would say that it would still be inaudible but you won't like the resulted figure visually. Net, net,when you know that there's other, proper way. Of course for a cost, people have to get payed accordingly when they do a good job.

 

[Sokel]

Is this something that they can optimize a hardware DSP implementation for (with the aim of better efficiency/accuracy compared to general-purpose computation devices)?

It needs both hardware (a dedicated DSP chip probably with enough computational power) and software (this includes some deep knowledge and education, you saw the TI's engineer paper and how important filter structure is) .

 

[James240]

Room Correction is something i havent been able to get right with the Wiim i would definitely put it down to user error using Umik mic with REW first then tried the RC on Wiim

Limited the RC frequency to 54-400HZ with a Gain of 3 and Max Q of 4 it just sounds tinny and skinny but the midrange sounds clearer (analog out to a pair of G3) problem is if i increase the volume seems to be distortion happening and ive read to much gain with powered speakers can be damaging?

 

[pogo]

Limited the RC frequency to 54-400HZ with a Gain of 3 and Max Q of 4 it just sounds tinny and skinny but the midrange sounds clearer (analog out to a pair of G3) problem is if i increase the volume seems to be distortion happening

Is it possible that a Wiim does not automatically reserve headroom for a correction, as is the case with Dirac Live, for example? If so, some users will probably unconsciously listen with slight distortion!

 

[Mort]

Room Correction is something i havent been able to get right with the Wiim i would definitely put it down to user error using Umik mic with REW first then tried the RC on Wiim

Limited the RC frequency to 54-400HZ with a Gain of 3 and Max Q of 4 it just sounds tinny and skinny but the midrange sounds clearer (analog out to a pair of G3) problem is if i increase the volume seems to be distortion happening and ive read to much gain with powered speakers can be damaging?

I've used DIracLive, Audyssey, GLM. I don't like my results with Wiim mostly. That said, I put a ton of time into the others and they are costly. Wiim is ridiculously fast and cheap.

 

[pogo]

I don't see that the Wiim RC is any faster than a single-point autotarget curve calibration from Dirac Live via the smartphone app. Also, there is no need to make presets such as 'gain of 3 and maximum Q of 4', which also need to be selected correctly for a good result.

 

[morillon]

you just have to understand that wiim-linkplay is a new player in audio and that it will take time for this function, most difficult, to be more mature...
mmm on the way, management of positive-negative gains, management of the sub etc.... but already the presence of a high-performance 2*10 peq (but not its graphic display, it's stupid) allows many things, see by "third-party" software, for very affordable consumer machines... (mini too now!)

you just have to encourage them (and they don't get lost with accessory-gadget functions like all those attached to the ultra's too small screen... which only concerns it itself....)
;-)

 

[pogo]

The Wiim RC no longer looks like 'one touch' to me. You have to know exactly what you are doing to avoid distortions.

 

[morillon]

The Wiim RC no longer looks like 'one touch' to me. You have to know exactly what you are doing to avoid distortions.

It will certainly progress... "in devellopement"

in terms of the relevance of measurements/corrections, the priority, and we can imagine towards simplified-integrated modes (perhaps one day an "expert" mode in parallel?)
but hey, the mere presence in the whole range, mini too, of a high-performance peq mode is a commendable thing, and the development of RC mode...a difficult path..
;-)

 

[Heat11]

I have an Ultra and use its sub crossover/output without issue. It's actually very flexible as you can high-pass, or not, the main speakers, low-pass or send the full signal to the sub, control phase and sub volume, as well as time-sync the sub & mains. I can't imagine what problem the Amazon review user was having other than maybe misconfiguring the settings. However, there is a lot of recorded material that doesn't have much bass output so a sub doesn't do much for those songs.

awesome, right now i am trying to figure out which amp to go with, the pro or the vibelink. any thoughts?

 

[Yorkshire Mouth]

it will take time for this function, most difficult, to be more mature...

Depending on how you take it, I find that statement to be either wrong or blindingly obvious.

If you're saying it will take time to be useable, you're wrong.

If you're saying that everything can improve given more time (whether Dirac, WiiM, or whoever), that's blindingly obvious.

 

[Mort]

The Wiim RC no longer looks like 'one touch' to me. You have to know exactly what you are doing to avoid distortions.

It's still the fastest and is the closest to 'one touch'. I can't even get through the setup screens in DiracLive in the time it takes Wiim to be done. But I don't like the results as much as DiracLive, Audyssey, REW or GLM.

 

[morillon]

Depending on how you take ir, I find that statement to be either wrong or blindingly obvious.

If you're saying it will take time to be useable, you're wrong.

If you're saying that everything can improve given more time (whether Dirac, WiiM, or whoever), that's blindingly obvious.

Either... I don't have too much time to lose with you... ;-)

 

[Yorkshire Mouth]

The Wiim RC no longer looks like 'one touch' to me. You have to know exactly what you are doing to avoid distortions.

I'm not sure that's particularly true in a real-world situation.

The measurements, both from Amir, and those Sokel has produced, suggest any issues are going to be inaudible or close.

And there's a very easy way to tell if they're inaudible or not, and that's to listen to the results.

Now, if you find that, after running RC, you hear distortion, then yes you can look at what RC has done, see where the biggest PEQ increases are, and see if those frequencies coincide with what you're hearing. You can then dial those boosts back a little until you can no longer hear an issue. Then A/B the two to see which you prefer. Now, yes that's not one touch if you hear issues. But it is one touch if you don't.

And importantly, if you do hear issues and need to try this solution, then whilst no longer being one touch, it's still substantially than most of the alternatives.

Don't make the perfect the enemy of the good.

 

[Yorkshire Mouth]

Agreed, what you see at this chart is way inaudible.
The thing is that what you see has to be translated in analog at some stage, adding all the analog "sins" to it.

Think of it as if it didn't have the penalty, take the glorious -144dB THD+N shown at the chart and just pass through the best DAC out there, you'll get -125dB THD+N something?
Now do the same for the degraded. I would say that it would still be inaudible but you won't like the resulted figure visually. Net, net,when you know that there's other, proper way. Of course for a cost, people have to get payed accordingly when they do a good job.

Just one more thing. Any noise in hi-fi is, by definition, audible. At least if it's going to matter. Asking if inaudible noise matters in hi-fi is a bit like asking about witnessless tree fellings.

If the PEQ from RC creates noise, then the noise should be there, whether music is being played or not, as long as the PEQ is running.

So, run RC, allow the PEQ filters recommended, and turn it on. Without music. Then turn up your amp to a level just above that which you listen to at maximum, and stick your ear to your speakers' drivers.

If you can hear noise, it's there. If you can't, it isn't, or is at least inaudible.

 

[Sokel]

Just one more thing. Any noise in hi-fi is, by definition, audible. At least if it's going to matter. Asking if inaudible noise matters in hi-fi is a bit like asking about witnessless tree fellings.

If the PEQ from RC creates noise, then the noise should be there, whether music is being played or not, as long as the PEQ is running.

So, run RC, allow the PEQ filters recommended, and turn it on. Without music. Then turn up your amp to a level just above that which you listen to at maximum, and stick your ear to your speakers' drivers.

If you can hear noise, it's there. If you can't, it isn't, or is at least inaudible.

No, that's not how it works, it's not like some kind of low hiss.
This noise rises with the signal (and also lowers, maybe modulate, etc), it's an error.

Read the paper I posted and the thread about it.
That's the elephant in the room these days and it's not confined only to Wimm.

It took miniDSP a decade to apply a partial fix after a gazillion of complains about audible degradation, go to their forums to see.
Same with bluesound where things go really south.

 

[Yorkshire Mouth]

No, that's not how it works, it's not like some kind of low hiss.
This noise rises with the signal (and also lowers, maybe modulate, etc), it's an error.

Read the paper I posted and the thread about it.
That's the elephant in the room these days and it's not confined only to Wimm.

It took miniDSP a decade to apply a partial fix after a gazillion of complains about audible degradation, go to their forums to see.
Same with bluesound where things go really south.

Yes, noise rises with the signal, but at the same time the signal covers the noise. Noise is measured, not as an amount in itself, but as a proportion of total; a signal to noise ratio.

Any noise which you can't hear because the signal is loud enough to make it inaudible is like that tree.

But I suspect we both agree. The levels of noise we're discussing are irrelevant in most real-world use situations. I appreciate that some people will want the best measurements there are, but at the same time many of us don't care about inaudible problems. Issues caused by your speakers' interaction with your room can be very audible indeed. If room correction (on any device) can fix (or at least part-fix) these, and the only penalty is noise you can't hear, then I think many people will just shrug and take that inaudibility on the chin.

But both approaches are valid, and I'm not criticising anyone.

 

[Soniclife]

It took miniDSP a decade to apply a partial fix after a gazillion of complains about audible degradation, go to their forums to see.
Same with bluesound where things go really south.

Are there measurements that show Wiim also needs a fix? (Expects to be told the measurements are in this thread)

 

[Sokel]

Are there measurements that show Wiim also needs a fix? (Expects to be told the measurements are in this thread)

Not on this thread, no.
But there are some early tells here:

WiiM Pro - Review & Measurements (Streamer)

WiiM Pro - Review & Measurements Hey folks, It is time for my first review of 2023... This one is about a complete objective analysis of the new WiiM Pro streamer. The aforementioned has been recently announced, back in November 2022, and I can see quite a strong hype around it since. That...

www.audiosciencereview.com

I mean look what happens at fixed volume and EQ just been ON, without filters (digital measurement, you see that the reduced level fixes it completely) :

I believe (?) they fixed this horror show though, or at least reduced it.
This resulted by only putting EQ on, so clearly a half-baked one (at the time?)

But imagine been an early adopter.