Genelec GLM Review (Room EQ & Setup)

[DJBonoBobo]

Hello in the round, do I understand correctly that the Neumann MA 1 System uses FIR Filter and the GLM not? Thx in advance

I don´t know for sure, but Neumann writes: "As a unique benefit, the software not only optimizes the amplitude response but also optimizes the phase." (https://en-de.neumann.com/ma-1)

 

[Sancus]

This seems not correct, but I may be missing something.

GLM could always integrate multiple subs, including time of flight and level to main listening position.
Even in a mixed system where some monitors have their own dedicated sub or adaptive W371A SAM™ Woofer.

For multiple subs outputting the same signal, just calibrate as usual and edit the 'Acoustic settings' of each subwoofer to reduce the ‘Level Compensation’ value. Section 10.4 in the current GLM4 System Operating Manual even has a table if you don't know by how many db.

Depends on how you define "100% properly". You're correct that GLM does those things, but that's an inferior solution to DLBC or Multi-Sub Optimizer. GLM does not correct the combined response of the subwoofers at all, and its multi sub but one signal solution is basically just a manual hack as you noted.

This isn't really a big deal for the intended mixing/mastering studio use, as those basically never use bass management as far as I know. But it's definitely suboptimal for domestic use.

 

[DigiPete]

Depends on how you define "100% properly". You're correct that GLM does those things, but that's an inferior solution to DLBC or Multi-Sub Optimizer. GLM does not correct the combined response of the subwoofers at all, and its multi sub but one signal solution is basically just a manual hack as you noted.

This isn't really a big deal for the intended mixing/mastering studio use, as those basically never use bass management as far as I know. But it's definitely suboptimal for domestic use.

Could you be more specific, as I see absolutely nothing that DLBC adds over GLM?
And how is the pinnacle of sound needed for mixing/mastering studios suddenly not applicable in home audio?

 

[Sancus]

Could you be more specific, as I see absolutely nothing that DLBC adds over GLM?
And how are the pinnacle of sound needed for mixing/mastering studios suddenly not applicable in home audio?

Sorry but.. did you read the thread? GLM is not a SOTA room correction system. It's a fast, easy-to-use basic one that gets the core things right without adding any latency.

All GLM does when you add a subwoofer is match the phase to a specific monitor, and add IIR PEQ filters to flatten the response in isolation. If you have multiple subwoofers and are using bass management(one bass signal), this is not very useful. What you want is for the system to correct the response at the listening position(s) to be as flat as possible when all subwoofers are playing simultaneously, and this is a very complex mathematical optimization problem.

DLBC, Multi-Sub Optimizer, and presumably other advanced solutions(Trinnov Optimizer, etc) calculate different filters for each subwoofer so that the simultaneous response is flattened. Also unlike GLM, Dirac additionally uses a combination of FIR/IIR filters and corrects phase/impulse response.

Whether or not studios should use bass management is off-topic, but they normally don't. Among other things, it conflicts with the creator's goal of making sure that the signal recorded on each channel is exactly as intended, since it sums signals, and there's no real standard for it either.

Genelec hasn't put a lot of work into a bass management system because their main customers don't want it.

 

[testp]

There is this AutoPhase in GLM which allows you to integrate subwoofer with one speaker. So with two subs in stereo setup, one sub for left and one for right. But if you would like the subwoofers to playback the same signal (i.e. mono) that could cause them being out of phase between each other.

(out of phase) but that would be mainly the point as i saw from videos, it would deal with dips below 100hz..

 

[Sean Olive]

This is a review and detailed measurements of the Genelec GLM "SAM Loudspeaker Management System." Its main use is automatic and manual room equalization for select Genelec speakers with such support. It is on kind loan from a member and costs US $299.

I was pleased to see it come with a nice looking microphone and an audio interface:

View attachment 151754

At first I thought the communication was through Ethernet so plugged in the speaker into my home network but nothing happened. Then I read the manual. I realized that the RJ-45 is a custom connection and protocol for the above adapter to talk to speakers in a daisy chain manner. Once I plugged the speaker into it, the GLM software recognized it and I was able to configure it. FYI the version I tested is the 4.1.

The interface is a bit tricky in that you have to drag and drop the speaker onto this unusual grid to then group speakers and manage them:

View attachment 151756

Post calibration to see how the filters are setup, you need to click on the Genelec speaker icon in that grid which was not easily discoverable. I found it by accident.

For testing, I used the Genelec 8330A which I had recently reviewed. The process could not be simpler. You tell it to start calibrating and next thing you know, this super loud chirp signal plays and that is that. No multiple measurements. No averaging. I would have wanted some control over volume on this but I did not realize the role of the vertical slider until later. The good news is that the whole affair was done in just a few seconds. Compare this to automated Room EQ on AV products that can be very time consuming with multiple measurements and such.

Before I show you the measurements, I really appreciated a couple of other cool features. First, you can calibrate the SPL level where you sit with the microphone. Once there, you can then leave the microphone there and watch the real time SPL values shown at the bottom level ("77" in this snapshot). I was able to go into low 90s before the speaker clipped with the speaker icon turning red above that. With all other systems, once you are done calibrating, the mic and software collect dust. Nice to see that is not the case here especially if the displayed SPL values are correct.

You can also preprogram a couple of SPL values as you see on the left. Once again, I could not discover how to change them. There are presets for the system/speaker but I could not figure out how to edit them.

The whole interface is very responsive which makes the included volume control slider useful.

Gelenec GLM Room EQ
The main event here is how well the automated GLM calibration performs. Per above, post correction you can examine what it did and thereby also understand its underlying technology:
View attachment 151757

I truncated the above display as it goes to 16 filters. The system is not different than automated filter creation in Room EQ. The system is simply using a combination of shelving and parametric EQ to build an inverse of the response with some care. For example, the sharp dip around 80 Hz which is caused by cancellation, is left alone sans a constant boost provided by the shelving filter. On the other hand, the peak at 200 Hz is attacked strongly with a complex curve created out of a nearly a dozen filters to invert it. The filter response is in blue and the measurement is in red. Filter applied to measurements is the green.

From what I can tell, it didn't attempt to make any correction above 300 Hz which is fine but then there is a lonely filter at 818 Hz with just a -0.3 dB gain. Such a filter will not have an audible effect. I wonder if it is using the knowledge of the speaker to make a tiny correction there?

Sadly as with all of its competitors sans JBL ARCOS, the corrected response is a simulation. No post calibration measurement is performed to see if that is the results that is generated. That is left to us to mess around with another microphone and measurements software to use. Problem with this type of verification is that the microphones will be different as will the absolute positions so we can't do a precise determination. But we can get close.

I used my Earthworks measurement mic with my RME Babyface Pro FS in combination with REW software to make a before and after measurement. I used 1/12 octave smoothing to keep detail there but soften the results some so we can make sense out of it:

View attachment 151759

Please ignore the levels. They are not calibrated.

The graph in red is the measurement prior to calibration. We see the standard impact of the room causing similar dual peaks around 55 and 180 Hz as GLM software showed. There is also the same dip at 100 Hz or so. We have a peak between 400 and 500 Hz that was not in the GLM measurement but this may be due to me not matching the mic location.

It is interesting that the correction for the first peak around 55 Hz is almost not there and much less so than predicted by GLM software. I can't explain this other than the filter implementation not having the resolution it thinks it has. Maybe the Q should have been lower.

On the other hand, the second peak around 180 Hz is corrected well with a response that actually matches the peak for the 55 Hz one.

Genelec GLM Room EQ Listening Tests
Turning the calibration button on and off is very fast allowing for quick AB comparison and boy is this a stark comparison (as it usually is with Room EQ). Turning on the EQ instantly removed some amount of low frequency boominess but importantly, it brought the vocals forward which I really, really liked. Once you listened to the calibrated sound, you just did not want to go back to not having it.

One of the key benefits of the GML is that you have full visibility into all the filters and you can add your own and turn them on and off and see which version you like better. So I added a filter at 449 Hz to fix the third peak with a Q of 5 and gain of -3 dB. This was a subtle change but it brought even more clarity to vocals. What I was hearing then was stunning! The vocals in one of my reference and favorite tracks, Biscuits from the live album by Fink had a realism and fidelity that was just a joy to listen to:

There, I did the obligatory thing of putting in some music in a review! But really, it was just wonderful and showed how good these Genelec speakers are when you take out the impact of the room.

Note that the overall signature was somewhat bright as is typically the case when you take out the excess bass. Genelec provides dual shelving filter overrides to boost the lows and reduces the highs. This is limited 3 dB max correction however. I found the effect subtle even at maximum correction and wanted to have more room for adjustment.

Conclusions
The Genelec GML Room EQ is a straightforward automatic generation and execution of filters. This makes the system easy to understand and modify but perhaps takes out the mystique of something magical going on. I personally may take a shot at just programming the filters manually or measuring and then modifying. While I have accepted the fact that consumer EQ products don't want to provide transparency on what they have done, I wish a Pro product like GLM would make a post EQ measurement and show that rather than simulated, feel good but made up response. It would take just a few seconds. Heck, that measurement could be used iteratively to optimize the filters more.

As is though, the system provides 90% of what an expert could do on his own in almost an instant. The improvement is dramatic and you would be silly for not using it if you have a Genelec SAM speaker. Indeed any system used without an EQ is producing incorrect and far less than satisfying sound in your room. You must have an EQ strategy and if you can't provide it upstream, having it this easily programmed into each speaker is a great help for very little money.

Overall, I am going to recommend the Genelec GLM. It should be mandatory for anyone buying a SAM speaker that works with it.

----------
As always, questions, comments, recommendations, etc. are welcome.

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

A single measurement? No multiple microphone measurements with spatial averaging?That doesn't seem like a good idea.

You never mentioned what their target curve is and whether it can be adjusted? Is it a flat in-room target curve or something different? Thanks!

 

[Trell]

A single measurement? No multiple microphone measurements with spatial averaging?That doesn't seem like a good idea.

You never mentioned what their target curve is and whether it can be adjusted? Is it a flat in-room target curve or something different? Thanks!

GLM can do multiple microphone measurements for each monitor/subwoofer as many as the user wants. After the calibration is done the user can set shelving filters to taste and change it without any new calibration needed.

 

[Sancus]

A single measurement? No multiple microphone measurements with spatial averaging?That doesn't seem like a good idea.

You never mentioned what their target curve is and whether it can be adjusted? Is it a flat in-room target curve or something different? Thanks!

There's definitely an option to take multiple measuremenrs if you wish. As for target curve, elsewhere they've stated the target is "flat perceived-direct sound". My experience with GLM is it generally does not modify the natural monitor response above the transition frequency.

You can modify the overall response post-calibration with their Sound Character Profiler or by setting up to 4 shelf filters but there isn't a way to modify the target curve prior to filter calculation.

 

[amirm]

You never mentioned what their target curve is and whether it can be adjusted? Is it a flat in-room target curve or something different? Thanks!

There is no concept of target curve that I can find. It shows a relative flat line when it does its thing and then you can add shelving. I am not a fan of this as I like to have a sloped response and this is not easy to do with just shelving filters.

 

[thorvat]

This is a misunderstanding of what DRC/DSP FIR filtering products actually does... There is indeed filter insertion loss as there is no boosting, but it is typically 5 to 7 dB of loss. But can easily be recovered by adding some digital gain. It is a completely different paradigm than PEQ.

I also think there is a misunderstanding about eq'ing just to the transition frequency. Sounds waves don't just turn into rays above the transition frequency. It is a gradual transition as illustrated in this chart:

View attachment 151774

The art in applying DRC is to use a frequency dependent window so that normal modes are getting the full correction, but above the transition frequency, the correction becomes less and less so by the time you get through the diffusion zone (4fc) no more correction is being applied (or if any, it is to the direct sound only and not including any room reflections). Typically that is around 600 Hz in most rooms. I offer my clients partial and full range correction and they can decide which they like better. The results of which on what most people prefer are overwhelmingly in one direction.

Sorry, I did no mean to barge into your review. Just wanted to make you and your readers aware that there are DRC/DSP products that can offer a much higher resolution room eq than just PEQ. My offer still stands on producing some high resolution (partial correction) FIR filters for you to listen, measure and compare.

I fully agree with everything you said except for the "can easily be recovered by adding some digital gain" part. I assume this was a typo and that you wanted to say analog and not digital. As you had to apply attenuation to compensate for the max filter gain to avoid clipping adding digital gain would get you back to clipping, but applying analog (post DA conversion) gain would indeed compensate for the attenuation. But I'm sure you're fully aware of this hence I believe it was a typo.

 

[thorvat]

A single measurement? No multiple microphone measurements with spatial averaging?That doesn't seem like a good idea.

Judging by this post @amirm believes that spatial averaging job is to "filter the response" and not to provide more realistic picture of the response at LP compared to a single point sweep.

P.S. I believe most of the folks who are into room EQ already know that MMM and averaging of multiple measurements done from a single points around LP provides practically the same spatially averaged response curve hence I used the term "spatial averaging" instead of MMM as it covers both techniques.

 

[Robbo99999]

There is no concept of target curve that I can find. It shows a relative flat line when it does its thing and then you can add shelving. I am not a fan of this as I like to have a sloped response and this is not easy to do with just shelving filters.

Folks using this GLM then therefore might find it useful to use 3 High Shelf Filters in conjunction to act as a virtually linear tone control throughout the whole frequency range. You can do this by using the following three High Shelf Filters 63Hz Q0.5 / 632Hz Q0.5 / 6324Hz Q0.5 , you set the Gain to the same dB value for each one - only very small dB changes required to have big effects on tonality as affecting the whole frequency range in a linear fashion. (You can use the same 3 filters as Tone Control to solve Circle of Confusion issues on a per music album basis, and I use these filters permanently baked into my miniDSP (&EqualiserAPO for a different chain) to subtly tweak my Anechoic EQ of my JBL 308p speakers.)

 

[Costas EAR]

Dear Amir,

As Dr Sean Olive told you very gently:

<<A single measurement? No multiple microphone measurements with spatial averaging?That doesn't seem like a good idea.>>

I suspect Dr Floyd Toole would have said exactly the same.

And all other Dr's and experts, like the guys in trinnov audio.

A single point measurement from you, is really of no use.
There is no meaning at all for me, and for all of us with enough experience at room eq software using and evaluation, to this single point measurement practice of yours.

You have to use multiple point measurements and averaging, or (even better and easier and quicker) the moving microphone method (but you have to practice for that).

Everything else, is simply, unacceptable.

You are using a state of the art measuring equipment for speakers, which is averaging thousands of measurements, and when you want to measure by yourself a speaker you are using... just one measurement, a magic one! Yeap, it's kind of funny (and it's a kind of magic).

Anyway..
I really enjoyed your listening evaluation of the glm auto equalization software, but there is no meaning to watch the single measurement diagrams of yours and all the rest.. I, a man of measurements, can find usefulness only in your listening evaluation. Oh, life sucks.


PS. Room curve, home curve or target curve, is always the holy grail, when we use eq. Without setting the right for the room home curve, we never listen to music..

 

[GDK]

I fully agree with everything you said except for the "can easily be recovered by adding some digital gain" part. I assume this was a typo and that you wanted to say analog and not digital. As you had to apply attenuation to compensate for the max filter gain to avoid clipping adding digital gain would get you back to clipping, but applying analog (post DA conversion) gain would indeed compensate for the attenuation. But I'm sure you're fully aware of this hence I believe it was a typo.

In most cases, the large filters will all be negative, as to fill a null you have to pull down the surrounding peaks. Therefore, there should be plenty of headroom remaining to apply digital gain. That is what I do in my set-up and I have never encountered digital clipping.

 

[q3cpma]

PS. Room curve, home curve or target curve, is always the holy grail, when we use eq. Without setting the right for the room home curve, we never listen to music..

That's only an opinion. I'm not the only one thinking that room curves should never be targets and only describe what an anechoically flat speakers gives in a "normal" room. Because, remember, your ears aren't a mic and you would compromise the much more important direct sound by trying to reach that target whatever the conditions.

Basically, I would only use such a curve as target if I don't know that the speakers are actually flat; otherwise, not EQing above the transition frequencies makes much sense.

I don´t know for sure, but Neumann writes: "As a unique benefit, the software not only optimizes the amplitude response but also optimizes the phase." (https://en-de.neumann.com/ma-1)

Note that GLM can also implement phase correcting filters with 4.1.

 

[Sancus]

Note that GLM can also implement phase correcting filters with 4.1.

It does? As far as I know, this is a specific feature that extends phase linearity of the Ones only down to 100hz, I didn't see any indication that it affects any other speaker models or has anything to do with the room correction filters.

 

[q3cpma]

It does? As far as I know, this is a specific feature that extends phase linearity of the Ones only down to 100hz, I didn't see any indication that it affects any other speaker models or has anything to do with the room correction filters.

Well, it may be restricted to the Ones for some reasons, but it means that the capability is present. Personally, I've alway thought that time domain perfection and coaxial speakers were meant for each other; after all, being time aligned only on a narrow horizontal plane is less impressive than in all directions and I wonder how it interacts with ground/ceiling reflections.

S&R's measurements of the 8350A and S360A also show some builtin phase correction (down to 100 Hz for the S360A and 200 Hz for the 8350A).

 

[amirm]

Dear Amir,

As Dr Sean Olive told you very gently:

<<A single measurement? No multiple microphone measurements with spatial averaging?That doesn't seem like a good idea.>>

I suspect Dr Floyd Toole would have said exactly the same.

The context is not what it seems. Correction here is only in bass. There is no need for spatial averaging if that is all you are doing. Yes, you have to worry about non-minimum-phase nulls but you can do that without any spatial measurements.

If you want to do correction in higher frequencies, then yes, you have to have a strategy for smoothing as slight changes in microphone location can make a big impact on measured results. Comb filtering for example is a big problem in higher frequencies which smoothing helps to resolve.

In bass frequencies, modal peaks can have resolution of less than one Hertz. Smoothing can produce incorrect results for filter programming for example.

As I noted, the problem with multipoint measurement is repeatability. You absolutely want to listen to experiment here. To the extent you can't repeat a multi-point measurement because you can't identically move a microphone to different spot in 3-D space, then I rather use single point measurements.

Now, if you have multiple seating locations, then that is an entirely different game and you do need to measure more than one seat to build a model of what to correct. Even here, spreading the measurements across multiple seats equally can reduce the preference in the outcome. From Sean and crew's own AES paper:

The Subjective and Objective Evaluation of Room Correction Products
Sean E. Olive1, John Jackson2, Allan Devantier3, David Hunt4 and Sean M. Hess5

Any kind of averaging gives you suboptimal results for the corner cases. Here, if you all you care about is a single location, then averaging just reduces the resolution there, not increase it.

Bottom line: keep bass optimization separate from rest of the frequency response. They are completely different animals as Sean and Dr. Toole would tell you.

 

[dominikz]

As I noted, the problem with multipoint measurement is repeatability. You absolutely want to listen to experiment here. To the extent you can't repeat a multi-point measurement because you can't identically move a microphone to different spot in 3-D space, then I rather use single point measurements.

Actually, in my experience one gets much better repeatability of measurements with MMM vs single position sweep. The random movement of the mic does a great job of averaging out the exact mic position differences between takes (I usually stop around 30 averages, to be on the safe side). Here I'm assuming MMM within a diameter of about 0,5-1m, i.e. covering approx. one seat.

As you said, in the LF (below transition frequency) such MMM response typically matches well to single point sweep, though - as also shown by @thewas here.
In the HF MMM provides much more consistent results between takes, making comparisons between individual measurement takes easier IMHO.

 

[amirm]

Actually, in my experience one gets much better repeatability of measurements with MMM vs single position sweep. The random movement of the mic does a great job of averaging out the exact mic position differences between takes (I usually stop around 30 averages, to be on the safe side). Here I'm assuming MMM within a diameter of about 0,5-1m, i.e. covering approx. one seat.

??? If you put the mic on a tripod, how are you getting all that variability?