Which MAAT EQ to choose?

[Ploki]

Red I haven't used much because how the mid side is implemented…

Hey damayor,

A future release of thEQred will have a re–worked M/S vs stereo workflow. We think you’ll like it…Basically, now it’s modal; either an instance is M/S or it’s stereo. You can’t have both in the same instance.

In future, you’ll be able to intermix M/S and stereo in the same instance, as with thEQblue and thEQorange. We had to rebuild the engine to achieve what we wanted but, so far, our testers are liking the improvement.

revolutionary. something Melda does for free and FabFilter ProQ3 does better for less

[Halonmusic]

Just took a look at the website... Take your money and run You can get more and better stuff everywhere, even for free.

indeed!

[OMas@MAAT]

1. thEQorange looks like it has (along with red) the most & most advanced features and would be the ‘if you could only choose 1’ choice. But why is it geared towards mastering and not also mixing? Imo it’s toolset would be very welcome in mixing too. Wouldn’t it be great to fiddle with that great sound on individual channels too?

Hey 128kbpsMP3,

Sorry, you started this thread and I never got back to you! As to #1; basically, high latency and CPU utilization. You have to have a MERGING+MASSCORE rig or equivalent to support more than a few instances of either thEQred or thEQorange. So…primarily mastering except for those few who have a mondo setup.

3. …For daily mixing tasks or usage while composing even, which one would you choose?

Both thEQred and orange are linear phase EQs (LP EQ) so, they don’t “sound like” analog EQ (no such thing, strictly speaking, as linear phase analog). So, I’d recommend you try thEQblue12 first.

thEQorange is very much a mastering tool. It’s best used to “fix” unfortunate mixes or to correct issues like voicing monitors. thEQred is for artistic EQ where you don’t want to add extra color. Notice that thEQred can also dial up super shallow slopes…

4. I’m not the most technical person so what is the difference in noob terms between the way orange & red work compared to blue…so is blue ‘not so special’ as it’s a normal EQ which has those ‘disadvantages’ like ‘regular’ eq’s? Why is it offered when they market so much on that phase stuff and how the way those other eq’s work is ‘superior’?

Oooph, lot’s to unpack…First, thEQred works in the frequency domain, thEQorange in the time domain. Basically, who cares! The take–away is that they sound fundamentally different because of those two different approaches in the underlying engines. thEQorange has exceptional LF response, which a frequency–domain LP EQ like thEQred or most all of the LP EQs out there, just can’t do as well.

On the other hand, thEQblue is a minimum phase EQ, as are all analog EQs. So, that aspect isn’t anything special. What is special is the no–compromise implementation. In thEQblue, we took the concept of an analog circuit, and created an idealized version. Not a model, not an emulation, but an 80 bit DSP version of the underlying physics that makes an equalizer. We did that twelve times, for twelve different circuits. Those are the twelve Architectures.

Being minimum phase, all Architectures have group delay (temporal distortion) and ripple (amplitude distortion, cab thought of as mis–tracking between what the gain numeric value says and what you actually get at any particular frequency). They are supposed to! That’s what gives them their “colors,” along with their individual curve shapes. After dialing up some EQ, try switching Architectures without changing anything else. You’ll see the shape changes which, in turn, effects subjective sound quality.

5. …I wish they made it stand-alone…

We hear ya, bud. Please stand by…

For what it’s worth I’m downloaded the demos an hour ago and intent to intensively try it out next couple of days, but am also very interested in other people’s expertise and experiences.

Thanks for actually…using your ears! to form an opinion. What a concept.

Whew, that’s a big post

Amen. Thank you, bruh!

[DPhil]

On the other hand, thEQblue is a minimum phase EQ, as are all analog EQs. So, that aspect isn’t anything special. What is special is the no–compromise implementation. In thEQblue, we took the concept of an analog circuit, and created an idealized version. Not a model, not an emulation, but an 80 bit DSP version of the underlying physics that makes an equalizer. We did that twelve times, for twelve different circuits. Those are the twelve Architectures.

Please, could you explain why those curves can 100% nullify with a couple other EQ plugins, even the free "cookbook EQ" implementations, without them rebuilding those architectures? I initially thought this could be just another marketing claim, but for the benefit of the doubt there might be a good explanation.

[Halonmusic]

most likely just a marketing claim. just download luftikus and be happy .. for free
https://lkjbdsp.wordpress.com/luftikus/

[OMas@MAAT]

…could you explain why those curves can 100% nullify with a couple other EQ plugins, even the free "cookbook EQ" implementations, without them rebuilding those architectures?

Sure! As we know, in audio [as well as rocket science and brain surgery ] it’s the details that make the difference. Just as Plugindoctor is a great tool if you know what you’re looking at — its implications and limitations — so too with “curves.” EQ curves are static representations of large signal, amplitude–only response. They don’t tell you about low amplitude response and only minimally about distortion in the time domain, frequency domain, or interactions with other sections that may be in circuit.

Curves are a simplistic tool, a snapshot of a very dynamic process. For 75 years, engineers made some fabulous sounding records without having curves to look at, except those in service manuals…

[martinjuenke]

1. thEQorange looks like it has (along with red) the most & most advanced features and would be the ‘if you could only choose 1’ choice. But why is it geared towards mastering and not also mixing? Imo it’s toolset would be very welcome in mixing too. Wouldn’t it be great to fiddle with that great sound on individual channels too?

Hey 128kbpsMP3,

Sorry, you started this thread and I never got back to you! As to #1; basically, high latency and CPU utilization. You have to have a MERGING+MASSCORE rig or equivalent to support more than a few instances of either thEQred or thEQorange. So…primarily mastering except for those few who have a mondo setup.

3. …For daily mixing tasks or usage while composing even, which one would you choose?

Both thEQred and orange are linear phase EQs (LP EQ) so, they don’t “sound like” analog EQ (no such thing, strictly speaking, as linear phase analog). So, I’d recommend you try thEQblue12 first.

thEQorange is very much a mastering tool. It’s best used to “fix” unfortunate mixes or to correct issues like voicing monitors. thEQred is for artistic EQ where you don’t want to add extra color. Notice that thEQred can also dial up super shallow slopes…

4. I’m not the most technical person so what is the difference in noob terms between the way orange & red work compared to blue…so is blue ‘not so special’ as it’s a normal EQ which has those ‘disadvantages’ like ‘regular’ eq’s? Why is it offered when they market so much on that phase stuff and how the way those other eq’s work is ‘superior’?

Oooph, lot’s to unpack…First, thEQred works in the frequency domain, thEQorange in the time domain. Basically, who cares! The take–away is that they sound fundamentally different because of those two different approaches in the underlying engines. thEQorange has exceptional LF response, which a frequency–domain LP EQ like thEQred or most all of the LP EQs out there, just can’t do as well.

On the other hand, thEQblue is a minimum phase EQ, as are all analog EQs. So, that aspect isn’t anything special. What is special is the no–compromise implementation. In thEQblue, we took the concept of an analog circuit, and created an idealized version. Not a model, not an emulation, but an 80 bit DSP version of the underlying physics that makes an equalizer. We did that twelve times, for twelve different circuits. Those are the twelve Architectures.

Being minimum phase, all Architectures have group delay (temporal distortion) and ripple (amplitude distortion, cab thought of as mis–tracking between what the gain numeric value says and what you actually get at any particular frequency). They are supposed to! That’s what gives them their “colors,” along with their individual curve shapes. After dialing up some EQ, try switching Architectures without changing anything else. You’ll see the shape changes which, in turn, effects subjective sound quality.

5. …I wish they made it stand-alone…

We hear ya, bud. Please stand by…

For what it’s worth I’m downloaded the demos an hour ago and intent to intensively try it out next couple of days, but am also very interested in other people’s expertise and experiences.

Thanks for actually…using your ears! to form an opinion. What a concept.

Whew, that’s a big post

Amen. Thank you, bruh!

You may not have noticed it but that guy is banned for KVR.
This happens when someone violates the forum rules. That may be a serious misbehaviour (or something not so dramatic), however, banned is banned.

[DPhil]

…could you explain why those curves can 100% nullify with a couple other EQ plugins, even the free "cookbook EQ" implementations, without them rebuilding those architectures?

Sure! As we know, in audio [as well as rocket science and brain surgery ] it’s the details that make the difference. Just as Plugindoctor is a great tool if you know what you’re looking at — its implications and limitations — so too with “curves.” EQ curves are static representations of large signal, amplitude–only response. They don’t tell you about low amplitude response and only minimally about distortion in the time domain, frequency domain, or interactions with other sections that may be in circuit.

Curves are a simplistic tool, a snapshot of a very dynamic process. For 75 years, engineers made some fabulous sounding records without having curves to look at, except those in service manuals…

Thank you so far. I probably have used words that can be misinterpreted. Of course I did not mean just literally looking at the EQ curves Indeed we're in audio industry, our ears are the key. And I can get complete silence with a null-test between like FabFilter ProQ3/Toneboosters/HOFA EQ/TDR and theEQBlue architectures. So my question is how is that possible? Either a) you are not doing anything special or b) you have done all that circuit rebuilding work with the result of recreating very basic cookbook filters, which then would only confirm it's old technology that has long been overhauled by modern methods. That would also explain why theEQBlue consumed at best 2x the CPU.

It's just hard for me to see the reasons behind the pricing or those marketing buzzwords. I know Algorithmix and I know they once have been some good stuff (even if they were already overpriced back then), but I also saw Friedemann Tischmeyer stumbling about certain claims when other experienced guys took a closer look at them. I saw/experienced the audio tests where those EQs were not outstanding or the results easy nullable with free existing plugins.

[Obsolete462444]

C'mon MAAT, not this BS again! You already lost 95% of your credibility back then on this Mastering Lab livestream, when Tischmeyer failed to validate his hillarious claims when discussing his products with other renown developers.

Now you come up with a new technology, FiDef, which is based on a trivial noise audio file added to the signal path? Are you serious?

I recommend ikjb Orange EQ to everyone thinking about spending their hard earned cash on imho overpriced MAAT EQs. Ikjb's free EQ has actually performed superior in all domains in a test last year on another forum, including latency, CPU and RAM utilization (MAAT thEQrange is barely usable on modern computers), not to mention that ikjb and multiple free EQs all perfectly nulled with your so-called "architecture".

Oh, did I mention your absurd claims about neuroscientists, who supposedly developed your whole "FiDef" technology? I did expect more from neuroscientists than simply adding static noise files to the signal path, not to mention that this kind of dithering is decades old and nothing new under the sun!

If you MAAT guys want to convince potential users to invest in your products, you need to offer value, not BS!

Why would anyone invest in your EQs, when they can be perfectly nulled with multiple free EQs and also perform inferior in all domains? In the blind test your EQ was actually the least favoured, while at the same time the most expensive!

Hence I will repeat my previous statement:

Here is my advice to MAAT: stop trying to fool your potential customers by utilizing bad marketing and psychoacoustic tricks and introduce a more realistic, market-oriented price policy. Your products are not priced competitively, when superior (in sound, function and design) well renown EQs cost a fraction.

[Burillo]

i thought IK Multimedia reps were being evasive in answering direct technical questions, but wow, MAAT guys are on another level

[OMas@MAAT]

…I can get complete silence with a null-test between like FabFilter ProQ3/Toneboosters/HOFA EQ/TDR and theEQBlue architectures. So my question is how is that possible? Either a) you are not doing anything special or b) you have done all that circuit rebuilding work with the result of recreating very basic cookbook filters…

Hey DPhil,
All I can tell you is: I, my testers & my dev lead are not able to achieve a “complete silence” null test. After first asking our CTO, his opinion being “It's almost impossible, the libraries would somehow have to be EXACTLY the same for all EQs,” I tried it in a 88.2 session (with VST3 versions for better parity between products). Using TDR’s SlickEQ as an example: -3dB at 400 Hz [“Normal for SlickEQ, Q of ≈1.66 for thEQblue], either of our parallel Architectures produced a good null, around -70 dBFS, but nowhere near -144 dBFS. That was after considerable time dicking w/settings.

In fact, with either or both EQs in circuit and with zero EQ gain (or gain of any kind) in both, I don’t even get a complete null. That doesn’t surprise me as, again, somehow our implementation and anyone else's would have to magically be the same. If you tried this with four EQs, then you are more patient and skilled than I. Actually, I would very much like to get your null test session so we can see where we’re making an error in setup. Please ping us at <support @ maatinc . zohodesk . com>.

BTW, I also tried a null test with thEQorange and QRange…Please do try that yourself. Again, no complete null and the residual is, shall we say, interesting.

[Ploki]

if there's no gain and the EQ's don't cancel each other, something is wrong in one of the EQs.

[Burillo]

After first asking our CTO, his opinion being “It's almost impossible, the libraries would somehow have to be EXACTLY the same for all EQs,”

it's perfectly possible, especially if you're comparing linear phase EQ's (so that differences in phase shift do not affect the result of the experiment). an EQ is a linear process, so by definition any well-made transparent EQ (i.e. one that doesn't intentionally color the sound with saturation or noise etc.) will to a large extent null with any other well-made transparent EQ, given the same curve.

non-linear phase EQ's producing similar phase shifts will also null with each other, but it's harder to do so because everyone has different definitions of what "analog", "digital", "minimal", "natural" etc. phase mean, which is why it's easier to compare linear phase (or linear phase-capable) EQ's. now, there are also caveats regarding oversampling filters etc., but i'm assuming you're not comparing boosts at Nyquist.

anyway, point is: EQ is math. if you did your math right, any other math done right will come to the same result, even if you used a different method to get there. there's no magic there.

Using TDR’s SlickEQ as an example

did you try to null MAAT EQ with TDR, or did you try nulling TDR EQ with MAAT? if you're using MAAT as a source, the comparison is not exactly fair because TDR is a comparatively limited EQ compared to a full parametric EQ like MAAT.

either of our parallel Architectures produced a good null, around -70 dBFS

-70dBFS is barely above audible threshold, sooo.... essentially, it does null to a very large extent

In fact, with either or both EQs in circuit and with zero EQ gain (or gain of any kind) in both, I don’t even get a complete null. That doesn’t surprise me as, again, somehow our implementation and anyone else's would have to magically be the same.

no, that just means you're adding either phase shift, or distortion, or noise when no one asked your EQ to do that considering previous claims about "FiDef" being pink noise added to the signal, the lack of null will be expected, in which case your definition of "null" should be modified as "higher than noise floor of MAAT EQ". or maybe you're using one of the colored modes of TDR, which would give you a similar result - extraneous factors affecting the results of the test.

[Ah_Dziz]

Damn. This is lots of eq talk. I use my saw eq and have no problems getting professional work done. When I need something more (dynamic eq or weird filter shapes) I use Melda stuff. I've never heard an eq that did much special outside of distorting in a pleasant way in the case of analog stuff. I have somewhere in the neighborhood of 70 3rd party was and I will just as happily use any of them for any given task (I don't as I said above because I don't want to deal with the ui differences or possible incompatibility down the road). If you want something nifty that sounds noticeably different, build a passive stereo pair of analog boxes or a couple tilt filters. In the box just pick one (that has all the options) and move on.

[damayor]

So what is the verdict do they null or do they null? Lol