Schiit Loki Mini+ Equalizer Review

[IAmAlwaysSerious]

I think I'd prefer them to effect just the frequency areas on which they're focused rather than skewing the whole frequency response

I'm fairly confident that Amirs graphs are making it seem... worse than it is, on that front. How do you feel about the version I posted on the first page of the thread?
Purple line in the middle is what you get with all knobs centered, the other lines are one knob maxed out one way or the other.

As for having all knobs maxed out at the same time... I can't actually think of a good reason that it should be flat at that point, but there's not much reason for it to be wavy either.
It makes some sense to me that the extremes would have more range since those are more likely to need bigger tweaking, I would think.

 

[mhardy6647]

Oh.
Like the ruby slippers* -- we've apparently had the answer all along. We just didn't know it.
From a very current tone control discussion @ AA.

Massive Passive - Bill Way - Amp/Preamp Asylum

Massive Passive - Bill Way - Amp/Preamp Asylum

www.audioasylum.com

from the link above:

And if you want an audio celebrity endorsement, Bob Ludwig has three of 'em.

https://www.manley.com/pro/msmp

well, twist me timbre, matey!

_______________________
* i.e., these ruby slippers...

 

[KR500]

The knobs on the Loki that control the lowest and highest frequencies 40 and 8 are 12 db cut/boost and the middle two knobs are 6 db.
I use minimal adjustments of 1 to 1.5db’s , the only exception is for my Beyerdynamic headphones to notch the spike out at 7 k where I cut 4 or 5db

 

[Robbo99999]

I'm fairly confident that Amirs graphs are making it seem... worse than it is, on that front. How do you feel about the version I posted on the first page of the thread?
Purple line in the middle is what you get with all knobs centered, the other lines are one knob maxed out one way or the other.

As for having all knobs maxed out at the same time... I can't actually think of a good reason that it should be flat at that point, but there's not much reason for it to be wavy either.
It makes some sense to me that the extremes would have more range since those are more likely to need bigger tweaking, I would think.

Yes, your graph shows more clearly what each of the knobs do. That makes the tone controls look better than I originally thought. For the bass tone control it could have been useful to have had one that mimicked a 105Hz Low Shelf (that Oratory uses as his bass filter tweak), that way you can manipulate the bass whilst staying away from the muddy zone - but we're praps talking more here about EQ rather than Tone Controls. Would be pretty cool as far a manual tone control would be, if you could have one that just tilted the whole frequency response perfectly linearly like a see-saw, but I expect there's reasons for why those might not exist in manual tone controls.

 

[PeteL]

effect just the frequency areas on which they're focused

It is called a brickwall filter, Impossible in the analog domain and would create breaks in the frequency response, that would not be pretty, now you can think ok then not brickwall but steeper slope, higher order? OK but then the chosen center frequency would be more enhanced than anything around it, but then which frequency exactly you want to enhance? Hey I am not saying that it's perfect. It does have limited applications. But I don't think, at all, it is about as you say as "Not well implemented". I do prefer a parametric EQ, even a semi parametric EQ. Or just a sweep able mid with bass and low shelf. Much more flexible, but I think what they implemented, they implemented it "well". Doing real EQ without measurements for the average music listener will create more harm than good. Not many have access to a measument microphone. Schiit customer base are not audio professionals.

 

[restorer-john]

??? That is not the point of that graph.

Well, what is the point?

You said the following:

"I set all the controls at max and min which should just set the unit up for flat response and act as extra or negative gain. Instead of that, we get quite bit of deviation. This will make it hard to interpret what the unit is doing by looking at the controls."

This is just plain wrong. When you set all the controls at max or min and sweep the unit, it will not, nor should not, be flat.

The controls should be set in the middle for such a sweep and that is merely to see the deviation from flat when in circuit, but not bypassed.

 

[Kc1234]

No it is not. There is no intelligence in the software. It is just a pure frequency sweep. That the graphs cross at 1 kHz is due to the way the filters are implemented in Loki. Not because I did anything special. Yes, I set reference signal at 1 kHz but that does nothing to automatically align anything.

The problem is exactly because you used reference signal at 1kHz.
Lets say if you use 20Hz as you reference, then you turn the bass knob (C1) up, then your graph will show that it does nothing to the 20Hz but attenuating the entire frequency range. Because all other frequencies is attenuated in relative to the boosted 20Hz signal.

 

[amirm]

The problem is exactly because you used reference signal at 1kHz.
Lets say if you use 20Hz as you reference, then you turn the bass knob (C1) up, then your graph will show that it does nothing to the 20Hz but attenuating the entire frequency range. Because all other frequencies is attenuated in relative to the boosted 20Hz signal.

What? If I set 20 Hz to reference when control is at 12:00 o'clock, the boosting bass will absolutely raise that level if the EQ works the way one expects. If it instead it kept 20 Hz the same and reduced the rest, it is a function of the device and totally non-intuitive.

 

[amirm]

Well, what is the point?

Just whether it is a flat line or not. It matters not how smooth the deviation was as you commented.

As to why it shouldn't, if the filters have the right slopes and are distanced correctly, then they should all sum to neutral at max in addition to center. On the other hand if each one has a different gain/slope then you get what I showed. Which is again, non-intuitive.

 

[Kc1234]

What? If I set 20 Hz to reference when control is at 12:00 o'clock, the boosting bass will absolutely raise that level if the EQ works the way one expects. If it instead it kept 20 Hz the same and reduced the rest, it is a function of the device and totally non-intuitive.

You might want to try to actually rerun your graph, using a different frequency as reference point and see what happens.
#17 graph already explained pretty well. It is really not that difficult to understand...

If you use 20Hz as reference, then 20Hz is always considered 0dB in the graph.
You turn up the C1 and boost the 20dB signal, then all other frequencies will show negative.. but 20Hz, being the reference point, will still be at 0dB..

 

[Mike-48]

Seems like a pretty pointless product nowadays (miniDSP and software parametric EQ possibilities to rival and outdo)

For EQ that always remain the same, I agree that miniDSP and parametric EQ can do that fine and with more flexibility, at least for digital signals. However, if you are trying to fix recordings that bother you, i.e., you want a more versatile tone control and something that can be adjusted in seconds without interrupting the music, the Loki models are better suited. The EQ on my miniDSP is nothing I'd want to adjust with music playing (I'm not sure it's even possible) to suit a particular recording.

 

[restorer-john]

As to why it shouldn't, if the filters have the right slopes and are distanced correctly, then they should all sum to neutral at max in addition to center.

The right slopes and distanced correctly? What are you talking about? It's not a graphic EQ with spacing at fractions of an octave across a 20-20kHz bandwidth and equal boost/cut for each band.

The device has 4 controls with specifically chosen centre frequencies. The maximum boost/cut is specified as +/-12dB at 20Hz and 8kHz and +/-6dB at 400Hz and 2kHz. Clearly it can never sum to 'neutral at max' and neither should it. That's not how tone controls work.

Shiit's own AP test report shows the controls and their boost characteristics are accurate.

 

[Kc1234]

The right slopes and distanced correctly? What are you talking about? It's not a graphic EQ with spacing at fractions of an octave across a 20-20kHz bandwidth and equal boost/cut for each band.

The device has 4 controls with specifically chosen centre frequencies. The maximum boost/cut is specified as +/-12dB at 20Hz and 8kHz and +/-6dB at 400Hz and 2kHz. Clearly it can never sum to 'neutral at max' and neither should it. That's not how tone controls work.

Shiit's own AP test report shows the controls and their boost characteristics are accurate.

Well, one can't wake a person who pretends to be asleep.

 

[mhardy6647]

The right slopes and distanced correctly? What are you talking about? It's not a graphic EQ with spacing at fractions of an octave across a 20-20kHz bandwidth and equal boost/cut for each band.

The device has 4 controls with specifically chosen centre frequencies. The maximum boost/cut is specified as +/-12dB at 20Hz and 8kHz and +/-6dB at 400Hz and 2kHz. Clearly it can never sum to 'neutral at max' and neither should it. That's not how tone controls work.

Shiit's own AP test report shows the controls and their boost characteristics are accurate.

Found this out in cyberspace & presume it's reflective of the performance of this beast. Looks like a four band EQ with very low "Q" for each band(?).

https://imgur.com/XGDm140

source:

https://www.reddit.com/r/headphones/comments/72r5x7

Looks like one could do quite a bit of damage (aesthetically, and -- with too much LF boost -- literally) with one of these little boxes.

Kind of reminds me -- umm... parametrically -- of this ...

source: https://www.radioshackcatalogs.com/flipbook/1977_radioshack_catalog.html page 25

 

[restorer-john]

Looks like one could do quite a bit of damage (aesthetically, and -- with too much LF boost -- literally) with one of these little boxes.

The issues may arise with vinyl.

I see no trailing off of boost, unlike something designed in the vinyl era, like the Realistic bass enhancer/subsonic filter I linked in a previous post/thread.

Topping D90LE - "Valve" and "Transistor" effects measurements

To add a basically but not utterly irrelevant aside, if I may: ;) Long, long ago, there was a flurry of "sub-bass synthesizer" products (all analog, AFAIK) that did the same thing, but in the other direction (i.e., added some signal at half the fundamental frequency). source...

www.audiosciencereview.com

 

[amirm]

If you use 20Hz as reference, then 20Hz is always considered 0dB in the graph.

Do you even know what you are writing? What on earth would force it to be zero if you boost that frequency? Magic?

 

[Kc1234]

Do you even know what you are writing? What on earth would force it to be zero if you boost that frequency? Magic?

No I don't. Actually you are right, you must be, I apologize.

With a second thought I find your graphs accurately represents how Loki works
Loki is designed this way that no matter how you turn the knob, 1kHz remains unchanged as your graph says. I take back whatever wrong comment and misinformation I posted. I am so sorry.

 

[IAmAlwaysSerious]

I see no trailing off of boost, unlike something designed in the vinyl era, like the Realistic bass enhancer/subsonic filter I linked in a previous post/thread.

It does drop off just a little as you get lower. Probably not as much as we'd like, for vinyl.

Sorry, all I have right now is a photo of the screen from back when I tested this.

 

[IAmAlwaysSerious]

Do you even know what you are writing? What on earth would force it to be zero if you boost that frequency? Magic?

I believe he's writing that -more or less- because that's precisely the definition of a reference frequency.

On your graph, anything higher in amplitude than the reference frequency will be above the zero dB line. Anything lower amplitude than the reference frequency will be below zero dB on the graph.
Since 1khz is the reference frequency, in your graph, it will always be equal in amplitude to the reference frequecy; the reference frequency is always equal to zero dB.

And for another way of looking at this... do you really think Schiit would have managed to keep a rocksteady 1khz when it's all implemented in analog? They couldn't keep the 20hz knob from affecting 20khz, after all.


Edit for clarity: what I said in the middle paragraph is true for any individual sweep you took. The reference amplitude is not the same for each sweep.

 

[pma]

Just whether it is a flat line or not. It matters not how smooth the deviation was as you commented.

As to why it shouldn't, if the filters have the right slopes and are distanced correctly, then they should all sum to neutral at max in addition to center. On the other hand if each one has a different gain/slope then you get what I showed. Which is again, non-intuitive.

In case of these simple analog tone controls the only setting where the frequency response should be flat is when you set them to the middle.