For music studio: should I EQ each monitor separately or not?

[NathanW]

Hey,

I'm in the process of EQing my monitors in my attic studio and have been using REW for measuring all sorts of things. I'm at a point where I want to EQ the monitors through Equalizer APO.

After looking online, I can't seem to figure out whether or not I should EQ both monitor outputs separately (so both monitors different EQ settings) or one EQ setting over both monitors. I see a lot of people with conflicting opinions/ideas/views on this.

On one hand it sounds logical to EQ them separately, but doesn't this only affect the final signal of that one speaker monitor in isolation? When 2 monitors playing, doesn't that then also create new interferences or effects in the room, compared to when one is playing the exact same signal? Both monitors could have a flat response after EQ (when played solo) but when combining the two signals, acoustics change. Sounds in the room change, sounds overlap and do things to the overall sound.

What would be the best thing to do?

I love to hear how you go about it in your music studio.

Thanks!

 

[whazzup]

Measure your L and R results through your monitors first. If they're matching with only very small variance then don't bother with the separate eq. If there are also environmental factors, like a closer wall or more equipment on one side, then definitely.

 

[staticV3]

I think the most sensible approach is to follow what your tracks are doing.

In the bass where your tracks are most likely Mono (no panning), look at the combined response of L+R and correct that.

In the mids and highs where elements are frequently panned, look at the individual response of each speaker.

Of course if the room is setup well and the left and right speaker responses essentially overlap, then you can apply one set of corrections to both channels.

As for the exact crossover frequency, so where you switch from Mono to Stereo, try analyzing a bunch of your tracks and see if you can find a common trend.

 

[Sokel]

As you're at music production along with listening, I would suggest EQ'ing them separately and then do a sanity check with both.

Taking this post from someone that really knows what he's talking about, the choice is easy:

How to find what music has stereo bass (with EasyEffects)

It is very interesting indeed, glad ya took the time to read it. And agree, 5 full range can be pricey (I know, paying for such a system now!), but worth it! Will in fact have 5x full range cardioid towers ;) Soundfield does full range cardioid, as does I believe GGNTKT, Gethain, and...

www.audiosciencereview.com

 

[NathanW]

In the bass where your tracks are most likely Mono (no panning), look at the combined response of L+R and correct that.

Do you mean playing one of my tracks and then measuring that sound response in the RTA panel for example?

Of course if the room is setup well and the left and right speaker responses essentially overlap, then you can apply one set of corrections to both channels.

Interesting.

When I look at my latest measurement, and put the smoothing to something like psychoacoustic or VAR, both curves for L and R monitor in isolation seem quite similar. The only big difference seems to be that 150hz dip in the right channel. I included the .mdat to this measurement. I'd love to hear your perspective on the measurements and how you think I should go about this - whether to EQ both monitors separately or EQ them both together with 1 set of corrections. Just if you have the time!

As for the exact crossover frequency, so where you switch from Mono to Stereo, try analyzing a bunch of your tracks and see if you can find a common trend.

With this, do you also mean playing some of my music and measuring the sound coming from the speakers using the RTA panel?

Thanks!

 

[staticV3]

Do you mean playing one of my tracks and then measuring that sound response in the RTA panel for example?

No, measure the room response using a measurement sweep or Pink noise in REW.

Then let the type of panning found in your tracks decide how you work with the data.

So as a baseline, I'd measure L, R, and L+R using the Moving Microphone Method.

With this, do you also mean playing some of my music and measuring the sound coming from the speakers using the RTA panel?

No, Stereo effects you can best analyze fully digitally, like in Audacity or your DAW.

For example, subtract the left channel from the right and analyze the residual.

 

[NathanW]

No, measure the room response using a measurement sweep or Pink noise in REW.

Then let the type of panning found in your tracks decide how you work with the data.

So as a baseline, I'd measure L, R, and L+R using the Moving Microphone Method.

Cool, thanks, I did exactly that in the file I sent in my previous reply.

Now I'm just wondering whether or not I should EQ them separately, or have them both operate under 1 EQ curve.

My biggest "problem" is the 150hz dip on my right speaker. I can't figure out where it's coming from. And perhaps some 250hz to 400hz on the left speaker - but this one I feel I can live with. I'd love to hear your thoughts.

 

[thcdru2k]

A dip in only one speaker usually points to a channel specific placement or reflection issue. A dip in both speakers individually suggests a shared room effect such as SBIR, floor bounce, ceiling bounce, or a room mode. A dip that appears only when both speakers play together indicates interference between the speakers from timing, phase, or geometry. If a dip appears individually but improves when combined, the speakers are filling in each other's weaknesses. I find lower frequency dips are more often caused by room modes and boundaries, while higher frequency dips are more often caused by reflections and interference.

 

[NathanW]

A dip in only one speaker usually points to a channel specific placement or reflection issue. A dip in both speakers individually suggests a shared room effect such as SBIR, floor bounce, ceiling bounce, or a room mode. A dip that appears only when both speakers play together indicates interference between the speakers from timing, phase, or geometry. If a dip appears individually but improves when combined, the speakers are filling in each other's weaknesses. I find lower frequency dips are more often caused by room modes and boundaries, while higher frequency dips are more often caused by reflections and interference.

Thanks!

Yes, the dip and overall graph very much improves when doing R+L combined sweeps and MMM measurements.

So would you say it's best to EQ both monitors with one EQ curve based on a L+R combined measurement? Instead of them eq'ing both separately?

 

[thcdru2k]

Thanks!

Yes, the dip and overall graph very much improves when doing R+L combined sweeps and MMM measurements.

So would you say it's best to EQ both monitors with one EQ curve based on a L+R combined measurement? Instead of them eq'ing both separately?

As staticV3 noted, the combined response matters for bass where much of the content is centered, while Sokel is also right that individual speaker response matters for panned content in a studio. Since your L and R responses are already fairly similar and the 150 Hz dip improves when combined, I would use mostly shared EQ based on the overall MMM trend and only apply separate EQ for broad, obvious channel-specific issues rather than trying to correct every dip seen in an individual measurement.

The individual measurements help identify what is happening but the combined measurements help determine what actually needs correction.

 

[NathanW]

As staticV3 noted, the combined response matters for bass where much of the content is centered, while Sokel is also right that individual speaker response matters for panned content in a studio. Since your L and R responses are already fairly similar and the 150 Hz dip improves when combined, I would use mostly shared EQ based on the overall MMM trend and only apply separate EQ for broad, obvious channel-specific issues rather than trying to correct every dip seen in an individual measurement.

The individual measurements help identify what is happening but the combined measurements help determine what actually needs correction.

Cool, understood. Thanks!

Will try to EQ as you and Sokel recommended and see how it plays out.

I appreciate it!

 

[thcdru2k]

Cool, understood. Thanks!

Will try to EQ as you and Sokel recommended and see how it plays out.

I appreciate it!

I personally do one eq change at a time, remeasure each channel separately and combined so you get a better understanding of what that filter did

 

[NathanW]

I personally do one eq change at a time, remeasure each channel separately and combined so you get a better understanding of what that filter did

Hmm.

So I did a EQ based on the L+R (combined) MMM measurement, and it gave me this prediction:

So I uploaded the EQ text document into Equalizer APO, and this were the results afterwards:

Which almost look the same, if not worse than the before EQ curve here under:

I'd love to hear your thoughts, and whether or not I might be doing something wrong.

Thanks so much.

 

[Timcognito]

Much on topic is discussion and videos in this thread

The ultimate low cost solution for room correction?

Hello to all of you from ASR, I've been following you for about three years and this is my first post...I'm using Google translator, so excuse my English... I wanted to let you know that IK Multimedia has just introduced a new dedicated product to environmental correction, called: "ARC STUDIO"...

www.audiosciencereview.com

 

[thcdru2k]

Hmm.

So I did a EQ based on the L+R (combined) MMM measurement, and it gave me this prediction:
View attachment 538953

So I uploaded the EQ text document into Equalizer APO, and this were the results afterwards:

View attachment 538954
Which almost look the same, if not worse than the before EQ curve here under:

View attachment 538955
I'd love to hear your thoughts, and whether or not I might be doing something wrong.

Thanks so much.

This is exactly why I usually do one filter at a time and remeasure. Sometimes a filter that looks great in the prediction doesn't produce the expected change in the actual measurement. Working incrementally makes it much easier to see whether the filter is helping, doing nothing, or causing a new problem.

 

[thcdru2k]

One thing I personally do is build the EQ manually instead of importing a large set of filters all at once. I'll usually look for the biggest broad peak, estimate the center frequency by finding the middle of the elevated region, then choose a Q based on how wide that feature is. Wide humps get lower Q values, narrower features get higher Q values. For gain, I usually start conservatively and cut only part of the excess rather than trying to flatten it in one shot. For example, if I saw a broad hump from roughly 120–180 Hz peaking around +5 dB, I might start with something like 150 Hz / -3 dB / Q 1.5–2 and then remeasure L, R, and L+R to see what actually happened. I've found that working this way helps me understand what each filter is doing, and I often end up using fewer filters than the optimizer originally suggested.

 

[NathanW]

One thing I personally do is build the EQ manually instead of importing a large set of filters all at once. I'll usually look for the biggest broad peak, estimate the center frequency by finding the middle of the elevated region, then choose a Q based on how wide that feature is. Wide humps get lower Q values, narrower features get higher Q values. For gain, I usually start conservatively and cut only part of the excess rather than trying to flatten it in one shot. For example, if I saw a broad hump from roughly 120–180 Hz peaking around +5 dB, I might start with something like 150 Hz / -3 dB / Q 1.5–2 and then remeasure L, R, and L+R to see what actually happened. I've found that working this way helps me understand what each filter is doing, and I often end up using fewer filters than the optimizer originally suggested.

Thanks so much, this really helped!


So yesterday I sent you the results of me having EQ'ed both L and R seperately and seemingly having not too impressive results in L, R and L+R combined. (might be my overthinking).

Today, I went and made an EQ curve based on the L+R combined response (the second pic I sent yesterday) from my measurements yesterday after I applied separate EQ to both. monitors.

It gave me a nice prediction once again:

So after applying this one EQ curve on top of both monitors (on the all channel in Peace GUI) this is what my new real measurement looks like:

I'm quite happy with the results for now. The only thing that really stands out is the 350hz region with a dip. But beyond that, I feel like I'm getting somewhere now.

Even the left and right channel when measured solo look a bit flatter now as well.

Hope this image isn't too confusing, lol.

But I'm glad to see that the L and R channel in isolation also showed some positive improvements by adding another EQ onto the main "all" channel tab in Peace GUI.

Absolute before & after:

If there is anything you feel like I might be overlooking, I'd love to hear it!

Thanks so much for your time.

Note: everything is a MMM measurement.

 

[thewas]

If your loudspeakers have decent pair matching and their placement isn't extremely asymmetrical I would recommend common equalisation above 300-500 Hz based mainly on anechoic responses and separate left and right room related filters below that.

It was also hard and took a long time for me to accept that not the nicest looking responses at the listening position are the best sounding ones, like Dr. Toole correctly writes, our hearing is very different than a non-gated omnidirectional microphone.

 

[Cbdb2]

Im with thewas. The most important part of the sound is the direct sound. It gives you most of your imaging and impulses. A vocal in the middle will stay in the middle even in a bad room. EQing the speakers differently above that 300 hz can skew that imaging. Why would Genelec, Nuemann etc match there speakers F-response as tight as possible just to have you screw it up with EQ.

 

[Ype]

Hey,

I'm in the process of EQing my monitors in my attic studio and have been using REW for measuring all sorts of things. I'm at a point where I want to EQ the monitors through Equalizer APO.

After looking online, I can't seem to figure out whether or not I should EQ both monitor outputs separately (so both monitors different EQ settings) or one EQ setting over both monitors. I see a lot of people with conflicting opinions/ideas/views on this.

On one hand it sounds logical to EQ them separately, but doesn't this only affect the final signal of that one speaker monitor in isolation? When 2 monitors playing, doesn't that then also create new interferences or effects in the room, compared to when one is playing the exact same signal? Both monitors could have a flat response after EQ (when played solo) but when combining the two signals, acoustics change. Sounds in the room change, sounds overlap and do things to the overall sound.

What would be the best thing to do?

I love to hear how you go about it in your music studio.

Thanks!

As others already mentioned, EQ bass in mono, which means playing both speakers at the same time.
Above 200Hz-300Hz I would ABX filters individually, as we are now in the territory where neither the room or speaker are dominant, this is the the transition zone.
I usually decide to just not touch anything above 300Hz.

EQing higher frequencies (500Hz or so) can (will) add dips and peaks (resonances) to speakers on axis response if you use in room response measurements.
You need to EQ from a spinorama for that and if the off axis response isn't smooth, a better speaker.

As for some unsolicited advice
Mixing in nearfield will give you a more direct sound field, off axis response will be less critical beyond listening window, but the problems of stereo will also be very audible.
Mixing in mid and far field will have the benefit of having reflections filling in that stereo dip, I would use both to check mixes.

Once you get the bass EQ and speakers right your brain will take care of the rest.
You could always add in some room treatment at that point but get the basics right first.

Hope that makes sense, not much else I can add to it, must smarter people then me have written books on this subject.
Sound reproduction 4th edition would be where I'd start if you want to go that route.

Edit: you want to use a 60dB scale for your REW measurements and I'd zoom in on the frequency axis as well to have a beter idea what is going on at bass frequencies.