You are not logged in. Please login or register.
RME User Forum → ADI-2 & 2/4 Pro series, ADI-2 DAC series → Does EQ lowers quality ?
I would like to buy RME's ADI-2 DAC, but this guy on youtube is telling that activating EQ makes the sound "crappy" :
Does the EQ really lowers dac quality ?
No, in no way whatsoever.
I would like to buy RME's ADI-2 DAC, but this guy on youtube is telling that activating EQ makes the sound "crappy" :
Does the EQ really lowers dac quality ?
Don't be worried, this is an excellent product, IMHO you will be very satisfied with this product.
And if not - for whatever reason - there should always be the possibility to give it back
if you should be frightenend to spend so much money, simply check that upfront.
BTW .. I owned the 1st versions of ADI-2 Pro and DAC ~6y ago and even those did already an excellent job in front of
- very good / proven active studio monitors
- planar phones and
- a high-end HiFi
No issues with the PEQ detected so far.
elapouya,
We would be delighted to welcome you to the RME group of users.
Does EQ hurt Sound Quality? No it does NOT! (Good thing too! Else we would be watching Silent-Movies, etc as without EQ, Sound Recording Studios would be in an awful situation)
Truthfully, it's the Non-Linearities in Speakers, Headphones, Rooms, and even our own Ears that are what hurt Sound Quality. The expertly implemented RME Digital PEQ serves as a Tool to help us correct those deficiencies, and it does so brilliantly.
Best Regards,
Curt
I would like to buy RME's ADI-2 DAC, but this guy on youtube is telling that activating EQ makes the sound "crappy" :
Does the EQ really lowers dac quality ?
I watched the video, then had a look into the Monolith THX-AAA 788.
The main criticism, besides that the poster had no use of ADI-2 features like the remote control and comprehensive menus and display (!):
He did not like the sound of ADI-2’s EQ.
Following his description, EQ the way he uses it seems to be limited to bass boost.
Here’s the point: ADI-2 and THX do use different shelve EQ shapes, no way to match the curves and sound:
ADI-2 uses 2nd order (12 dB/Octave) shelve filters, this enables the possibility to adjust Q-factor.
THX-AAA 788 uses 1st order (6 dB/Octave) shelve filters.
1st order filters in general do not offer a Q-factor option, and it’s not possible to match 1st and 2nd order filters for the same curve.
This means, his comparison is apples to oranges, he just likes 1st order filters better.
From my recording studio experience I know that the two types basically do sound quite different:
1st order filter do spread a lot more into the midrange, this makes them less “surgical” and precise, limiting their usability, but can prevent a muddying when used as simple bass boost.
Interestingly: on cuts the sound effect is just the other way round, 1st order - thin and punch-less, 2nd order - keep the impact and warmth and just remove excessive bass.
I do use both depending on the task.
@MC: I suggest to implement a 1st order shelve filter option into ADI-2, shouldn’t be a big deal because 1st order filter architecture is more simple.
Could be just an additional step in the Q-factor setting, if you dial one step lower than 0.5, the filter switches to “6dB”, or “1st” (order), or Q:“Off”.
Another purely BS Anecdotal "Review"
Aside from being seemingly unable to Read (ie: The User Manual), the reviewer then found himself befuddled by the RME's vast array of functionalities which "Drove him to the Forum of Elitists.. " Perhaps as a reviewer, he perhaps might have availed himself to that which IS explained in detail in the User Manual. Better still...He also didn't care much for the RME's lack of Coloration. What? Isn't lack of coloration the Whole Idea behind a premium device such as the RME?
In the end, that Video was little more to me than "Video Compost" (read:BS)
While quite misleading, the reviewer IS entitled to his views just as I'm under no obligation to agree with any of them.
Simply 20minutes of my Life that were wasted forever.
Curt
I would like to buy RME's ADI-2 DAC, but this guy on youtube is telling that activating EQ makes the sound "crappy" :
Does the EQ really lowers dac quality ?
I have to admit, this guy almost prevented me from buying the ADI-2 DAC, I'm glad to have ignored him.
Another purely BS Anecdotal "Review"
Let’s look at the positive, the inspiration that can be taken out of it:
A 1st order shelving EQ option could be a true enhancement for ADI-2, at almost no cost.
This does make a difference, much greater than most other factors.
KaiS,
Our Reviewer Friend offered no such engineering suggestions. There was nothing inspirational in his commentary, just more of the same "Mouthwash" that we could have just as easily gotten from an Amazon Product Review.
Regards,
Curt
I'm not sure what to say about the idea now. I'm not that familiar with EQs.
However, I am surprised at the proposal of KaiS, that the actually better / more precise EQ by design (the PEQ as it works now) should now get an operating mode, which one has not implemented on the part of RME probably for good reasons.
One should perhaps also consider that many users are already overwhelmed with the variety of functions.
When I think back how concerned HiFi enthusiasts are and are always anxious to do everything 110% right, even if it already sounds great: sinfully expensive cables, save a few dB SNR, balanced phones, the "best" AD/DA filter choice, the best setting for level, reflevel and so on and so forth.
That's when we come up with a new EQ operating mode.
Either you decide then with arising questions, which should be better a priori or you simply tell the people that they should open their ears and minds to make their own decisions and come to own conclusions.
Have fun with new great discussions 
But maybe this is just a KaiS's specific feature to explain EQs to people for the next years until retirement :-) :-) :-)
Joking aside, what's the bottom line ... interesting colorations of the sound ... a sun rising all at once ? Have I missed anything fundamental in the past years ? What should we call the new operating mode ? Maybe affectionately "Mojo" or rather "Dirty Mode" ? Well, I don't know tbh if that really helps to increase "quality" or whether this is just a new "confusion mode" for some ? Don't get me wrong .. if it makes sense .. but at the moment I have some doubts.
EDIT: Then I am also of the opinion that a more precisely working PEQ is more needed in order to carry out precise sound shaping, for example. Adjustments of headphones and frequency corrections for rooms.
Ramses,
I have been a Native English Speaker/Reader/Writer a very long time, and have yet to encounter such an amazingly well-composed post such as yours! Brilliant mein Freund!
I'm not choosing Winners, or Losers, but such a well composed comment deserves my Applause.
On EQ? (The original topic) When used wisely, it is indeed beneficial.
With that said, I continue to Marvel at some European Users ability to use the English Language to such impressively Good Effect.
Best to All,
Curt
One should perhaps also consider that many users are already overwhelmed with the variety of functions.
Where are your statistics and research behind your statements
As I wrote in the totalmix section, rme eq needs a major overhaul IMO. These devices are for Pro\s engineers etc. not some random nobody youtube reviewer or the average hifi enthusiast. I\m literally having to stack DSP eq on top of DSP eqs to get the results I want currently, which is obviously not ideal. I dont agree for a second any normal RME customer would feel overwhelmed by having more options like this. Look at FabFilter plugins / you see anyone complaining about their new plugins_ I sure dont.
1st order shelving EQs are those with the least amount of coloration.
This is the perfect fit to a device with ADI-2‘s philosophy.
The 1st order (6dB/Octave) curves, both amplitude and phase response, run smoother, less abrupt than any other shape.
This is what the youtube reviewer heard and I can agree with.
These type of decisions, selection and adjusting EQ, are my daily bread, and EQ is the No.1 bread and butter tool in the recording studio.
In fact I’m not satisfied with ADI-2‘s EQ sound, more specific from day one I have a problem with ADI-2‘s 2nd order treble shelving EQ, for the same reason.
In the studio I mostly end up with a 1st order treble shelving in case I need it.
The way I imagine implementation is very natural from the user‘s point of view:
You dial down Q-factor, and the last step is the logical completion in this dial direction, as 1st order EQ further opens up the curve into the midrange.
You don‘t even need to know the technical background to use it.
@KaiS: your post leaves me scratching my head. The ADI PEQ and B/T can be set to Q = 0.5, which IS first order (or 6 dB/oct). It 100% replicates the simple RC filters of analog devices, including phase. So what are you talking about?
@KaiS: your post leaves me scratching my head. The ADI PEQ and B/T can be set to Q = 0.5, which IS first order (or 6 dB/oct). It 100% replicates the simple RC filters of analog devices, including phase. So what are you talking about?
If that‘s the case my postings are pointless and RME was faster than me 
I don‘t know any other EQ working this way, 2nd order setting always stays 2nd order no matter what Q-factor is chosen, on the ones I usually work with.
The resulting response is different than switching them to 1st order.
Is there a dedicated internal switch function present, or is the 2nd stage nulling out?
I have to check and will come back to you.
The PEQ peaking and bandstop filters (Low Cut, High Cut) are 2nd order. The only way to see this is by the latter (the steepness after the cut-off frequency). The shelve filters are calculated differently, the 'order' does not have any relevance. As said setting them to Q = 0.5 gives the same curve as on my old analog units (and mathematically it is that, Q 0.5 equals first order behaviour).
There is a slight difference to some 'analog' designs (not all), but it has nothing to do with order. Raising the treble (as you mentioned, but same for bass) the amount of amplification and attenuation does not follow the frequency like a straight line, but is reduced again at the exterme borders (0 Hz and sf/2). So a shelve of a treble increase will look like a rising line that at some point is limited and turns into a horizontal line. This effect is both welcome (limits amplification of unwanted DC and critical, inaudible frequencies near Nyquist), and is not dependent on the filter order, but simply a different filter type. And not your point, you were talking about 'more mid range affected'. For that one would need a Q below 0.5.
@KaiS, it is not intuitive that Q=0.5 for 2nd order shelves is really equivalent to a first order shelf. We're so used to highpass/lowpass function where this is not the case, Q=0.5 still is a 2nd order function (two identical real poles). To get to first order there, we would need to use a Q approaching zero ("infinite" pole splitting, so that the second pole is outside of the usable bandwidth).
The shelve filters are calculated differently, the 'order' does not have any relevance.
This applies to the B/T shelves? The ones in the main EQ do react to their 2nd order Q setting.
MC wrote:The shelve filters are calculated differently, the 'order' does not have any relevance.
This applies to the B/T shelves? The ones in the main EQ do react to their 2nd order Q setting.
B/T implementation in adi-2 must be based on 2nd order. otherwise Q value in the setting page does not make sense anymore.
KSTR wrote:MC wrote:The shelve filters are calculated differently, the 'order' does not have any relevance.
This applies to the B/T shelves? The ones in the main EQ do react to their 2nd order Q setting.
B/T implementation in adi-2 must be based on 2nd order. otherwise Q value in the setting page does not make sense anymore.
Looks like you are correct. I never enable or use B/T as it is so easy to bump it by accident.
I wanted to wait until I gathered some test facts, but the topic is running on, based on assumptions, so I chime in again.
I have no doubt in MC’s explanation that the shelving EQ at Q=0.5 equals a 1st order filter (I’m NOT talking about HPF/LPF).
Only the developers do know which of the numerous filter architectures they chose.
A 2-stage design, eg, could neutralize the 2nd stage at Q=0.5, leaving a true 1st order filter.
I’d suggest to leave it for the moment, until me or someone else comes up with evaluation results.
My personal fact is: I don’t like the treble shelving EQ, but have to admit I don’t remember to ever have used it at Q=0.5.
This will be my first, subjective test ASAP.
Only the developers do know which of the numerous filter architectures they chose.
A 2-stage design, eg, could neutralize the 2nd stage at Q=0.5, leaving a true 1st order filter.
don't quite understand this. there's no '2-stage design'. an IIR filter's transfer function is always transformed into a time domain equation. all peq implementations are the same. as long as q/g/f are set, the six parameters are fixed.
A cheatsheet of this topic is here: https://webaudio.github.io/Audio-EQ-Coo … kbook.html
ADI-2's implementation should be highly similar to TI's official sample code: https://www.ti.com/lit/an/spra867/spra867.pdf
(The equations are listed in appendix and is equivalent to the cookbook above)
@MC. I really want to look at my Q on the B/T function. That said, it usually doen't matter. Typische? It's 0/0.
But, on occasion, I'll give it a nudge one way or the other. Just the bit of extra "something" Wife, and I needed
My Job had other needs, and now I'm 2000miles from my RME.
When I can, I want to do it MC's way!
I'm printing your Post, and inserting it into my User Manual.
Viele Dank!
Curt
KaiS wrote:Only the developers do know which of the numerous filter architectures they chose.
A 2-stage design, eg, could neutralize the 2nd stage at Q=0.5, leaving a true 1st order filter.don't quite understand this. there's no '2-stage design'. an IIR filter's transfer function is always transformed into a time domain equation. all peq implementations are the same. as long as q/g/f are set, the six parameters are fixed.
A cheatsheet of this topic is here: https://webaudio.github.io/Audio-EQ-Coo … kbook.html
ADI-2's implementation should be highly similar to TI's official sample code: https://www.ti.com/lit/an/spra867/spra867.pdf
(The equations are listed in appendix and is equivalent to the cookbook above)
Sorry, I didn‘t mean this literally.
I‘m not into digital development.
In the analog domain RC shelve filters are realized in various ways, passive, active buffered, bootstrapped in a feedback loop...
Most Hifi designs are a bit compromised as they have bass and treble shelving in one passive stage (“Baxandall-EQ).
A pure, unloaded (active buffered) RC-circuit gives a broad and smooth transition with least signal coloration.
Think I will do some comparative measurement based on this as soon as I find the time.
@KaiS,
A 2nd order shelf is basically a standard PEQ with "fudge factors" added.
As bass-boosting shelf, for example looks like this in Laplace notation:
a + b*s + s²
F(s) = --------------
1 + c*s + s²a is the gain at low frequencies, and b and c determine the PEQ part (gain and Q)
With the right values of b and c this can actually reduce to the 1st-order transfer function:
a + s
F(s) = -------
1 + s@KaiS,
A 2nd order shelf is basically a standard PEQ with "fudge factors" added.
As bass-boosting shelf, for example looks like this in Laplace notation:a + b*s + s² F(s) = -------------- 1 + c*s + s²a is the gain at low frequencies, and b and c determine the PEQ part (gain and Q)
With the right values of b and c this can actually reduce to the 1st-order transfer function:a + s F(s) = ------- 1 + s
Thanx, this is what I meant
take the low shelf as instance,
H(s) = A * (s^2 + (sqrt(A)/Q)*s + A)/(A*s^2 + (sqrt(A)/Q)*s + 1)
when Q = 0.5, H(s) = ((sqrt(A) * s + A)/(sqrt(A) * s + 1))^2.
setting H(s) = 0 and you'll find the two poles and two zeros overlap.
the system works exactly the same way as two 1st order filter combined together.
In this condition, it is possible to emulate the slope of a 1st order filter using different frequency.
One can very easily try the following: in any PEQ software (such as REW) add the following filters:
1. low shelf with quality factor, f=1000, gain = 12, q =0.5
2. low shelf 6db/octave , f = 500, gain = -12
you'll see the two filters cancel each other.
for 12db / octave, choose q=0.707. the following two filters cancel each other:
1. low shelf with quality factor, f=1000, gain = 12, q =0.707
2. low shelf 12db/octave , f = 1000, gain = -12
RME User Forum → ADI-2 & 2/4 Pro series, ADI-2 DAC series → Does EQ lowers quality ?
Powered by PunBB, supported by Informer Technologies, Inc.