• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Genelec 8330A Review (Studio Monitor)

milosz

Addicted to Fun and Learning
Forum Donor
Joined
Mar 27, 2019
Messages
584
Likes
1,643
Location
Chicago
Looking at the waterfall plot, the 2 kHz "resonance" seems more like ringing than just a peak in amplitude response. You can correct an amplitude response peak with EQ, but ringing / stored energy is a dynamic, time-related behavior and cannot be corrected with EQ.
 

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,335
Likes
6,700
Do they, though? The on-axis appears worse, however, these were measured with the higher number of points so one would expect it to be worse. The 8330A's PIR though ends up marginally better, perhaps. The slight broad dip from 100-500hz is better corrected, so is the sharp drop above 15khz, and the broad crossover dip is instead a dip and then peak. It's possible that this is intentional and ends up sounding better than the broad dip, I don't know.

In any case even looking at Genelec's own measurements and specs I don't think they were ever a sound upgrade, it's purely for GLM. Unlike some other 8000-series, the 8030C already has the newer Class D architecture.

To me, they do look slightly worse than the non DSP version.

Side by side:

Screen Shot 2021-08-12 at 6.53.30 PM-min.png


The only real flaw with the 8030c is that slight 1-2dB trough from 2-4kHz. To me it looks like they tried to correct that but messed up the DSP. It looks like they made the filter, 1. Too low in frequency(or two low Q), and 2. Too great in magnitude. They traded a 1-2dB trough from 2-4kHz for a 2-4dB peak from 1.5-3kHz. We know that peaking bothers us more than the opposite. Directivity also seems a little less smooth.

I know there is even some dispute as to whether that depression in the 8030c is even real. I assume it is, as it shows up in all of my in room moving mic measurements(more so than in the anechoic data).

It's not much worse, but I suppose I'm just disappointed that it's not better. I expected it to be at least as good as the same speaker with no DSP, but the DSP seems to have hurt more than it helped. @Maiky76 's filters for the 8030c seem do a better job of fixing it.

The new problem is even above the frequency at which GLM auto calibrates, so users will have to correct it manually.
 

Spocko

Major Contributor
Forum Donor
Joined
Sep 27, 2019
Messages
1,621
Likes
2,999
Location
Southern California
Great! Here are my GLM measurements of 8330 which I use as surrounds. It shows a peak at 1,8 KHZ as well (more prominent on my right speaker which is in a corner => probably refections?). Also here is my room set up.

First image is left surround and the second image is right surround.
View attachment 147117
I assume you can correct these peaks with GLM, right?
 

Sancus

Major Contributor
Forum Donor
Joined
Nov 30, 2018
Messages
2,923
Likes
7,616
Location
Canada
To me, they do look slightly worse than the non DSP version

Yeah but you're focusing on the on-axis. I was looking at the PIR specifically because I think the 8030C's dip is worse than it looks on-axis due to dispersion narrowing. So my argument was effectively that Genelec tried to fix it at the minor expense of the on-axis measurement. Considering Amir tried to EQ the issues that created, and did not find it conclusively better, I think it may be more complicated than just flattening the on-axis graph.

The alternative viewpoint, of course, is that they tried to make it sound better a little too hard and failed, but I don't think that's an easy conclusion to make without a real blind test.

It's not much worse, but I suppose I'm just disappointed that it's not better. I expected it to be at least as good as the same speaker with no DSP, but the DSP seems to have hurt more than it helped. @Maiky76 's filters for the 8030c seem do a better job of fixing it.

It's impossible to know this without applying that EQ to an actual speaker and then measuring with the Klippel though. Something about the crossover interaction or design of the speaker seems to result in small resonances in that range, which is not something that a simplistic EQ simulation is going to account for.

I do think that the idea that you can EQ the on-axis to perfection with a simulation, especially in sensitive ranges that have nonlinear behaviour, is inherently wrong, and we do already know that sometimes it does make sense to sacrifice the on-axis to improve the overall sound of a speaker.

But I am just speculating here. Perhaps if @Ilkka Rissanen is permitted and willing to talk about implementation details, he could shed some light on the crossover and EQ differences between the 8330A and 8030C and why Genelec thinks the 8330A is an improvement. Assuming they believe the measurements are accurate, representative of the model, and there isn't some separate issue here or something.

I assume you can correct these peaks with GLM, right?

Sure, but the question is whether that actually makes the sound better. The review states that it was inconclusive.
 

markb

Member
Joined
Apr 29, 2021
Messages
55
Likes
101
i'd mostly agree, but there is a nagging thing about GLM kit approach for me atleast, no-PC setup with Genelecs, i need their volume control from their remote and its akward, small cheap remote, unless i can use a custom remote, this is not what i want to use, ever...
Yes - although the GLM adapter can be used in standalone, with either a wired volume control (a potentiometer) or a wireless, RF remote for volume control, I agree it remains a bit of a clunky solution that doesn't always meet all needs.

I am actually working on some open source software now (and perhaps some hardware later) that may be able to help with this in some use cases. I hope to post more details within a few days.
 

markb

Member
Joined
Apr 29, 2021
Messages
55
Likes
101
This is part of the smart speakers from Genelec so naturally sports Ethernet connectivity:

Although they are RJ45 sockets, it's actually not Ethernet, but a proprietary serial protocol, needing to be daisy chained to the GLM adapter and other SAM monitors/subwoofers. People shouldn't connect these to Ethernet switches.
 
Last edited:

Tangband

Major Contributor
Joined
Sep 3, 2019
Messages
2,994
Likes
2,789
Location
Sweden
To me, they do look slightly worse than the non DSP version.

Side by side:

View attachment 147228

The only real flaw with the 8030c is that slight 1-2dB trough from 2-4kHz. To me it looks like they tried to correct that but messed up the DSP. It looks like they made the filter, 1. Too low in frequency(or two low Q), and 2. Too great in magnitude. They traded a 1-2dB trough from 2-4kHz for a 2-4dB peak from 1.5-3kHz. We know that peaking bothers us more than the opposite. Directivity also seems a little less smooth.

I know there is even some dispute as to whether that depression in the 8030c is even real. I assume it is, as it shows up in all of my in room moving mic measurements(more so than in the anechoic data).

It's not much worse, but I suppose I'm just disappointed that it's not better. I expected it to be at least as good as the same speaker with no DSP, but the DSP seems to have hurt more than it helped. @Maiky76 's filters for the 8030c seem do a better job of fixing it.

The new problem is even above the frequency at which GLM auto calibrates, so users will have to correct it manually.
Yes - and that can be done in GLM without trouble . And saved in the loudspeaker. The only trouble is that you have to buy the GLM kit .
 

Tangband

Major Contributor
Joined
Sep 3, 2019
Messages
2,994
Likes
2,789
Location
Sweden
Regarding 8330 - the sound is slightly improved with digital connection . This is another advantage with the 83xx serie vs 80xx .
Feeding the dsp crossover with a digital signal makes the sound a little more clear and the dynamics are slightly better.

I use my own Genelec 8340 with an iPhone/Ipad and USB bridge using Apple Music lossless.
I can then change the volume digitaly with the iPhone . The xu208 in the USB bridge is converting the signal to 32 bit floating point then puts out 24 bit with dithering at the digital output.
https://www.audiosciencereview.com/forum/index.php?threads/bits-are-not-bits-…-or.25101/#post-854084

The connection from USB bridge to Genelec 8340 is rca - AES xlr. I have modified the output of the USB bridge with a resistor , thus having the same impedance output/input.

————————
Regarding the high gain in 8330 and 8340. It can be lowered digitaly with the GLM kit and then saved in the loudspeaker . After that, you can put away the GLM. I use -20 dB digital gain in my 8340 in digital mode . It works internally after the dsp crossover .

So , if you do that you can use the 8330, 8340 digitaly with a small sound benefit with an external digital source that can change the volume like Yamaha wxc 50 or USB bridge running digital out . Theres no need for using and buying the variable volume control with GLM .
 

Attachments

  • CADA6DCB-4180-4902-9C07-6A93BC5B503C.jpeg
    CADA6DCB-4180-4902-9C07-6A93BC5B503C.jpeg
    378.8 KB · Views: 316
Last edited:

Trell

Major Contributor
Joined
May 13, 2021
Messages
2,752
Likes
3,285
The only real flaw with the 8030c is that slight 1-2dB trough from 2-4kHz. To me it looks like they tried to correct that but messed up the DSP. It looks like they made the filter, 1. Too low in frequency(or two low Q), and 2. Too great in magnitude. They traded a 1-2dB trough from 2-4kHz for a 2-4dB peak from 1.5-3kHz. We know that peaking bothers us more than the opposite. Directivity also seems a little less smooth.
...
The new problem is even above the frequency at which GLM auto calibrates, so users will have to correct it manually.

The new GLM 4.1 did add a notch filter at 2K for both left and right channel on my 8330A.
 

Ismapics

Active Member
Joined
Nov 21, 2020
Messages
213
Likes
285
They usually have a switch though or a knob. With neither here, I am surprised the default is consumer levels.
Hello @amirm : do you have any lower priced or affordable speakers in the pipeline. It seems that under $500 a pair segment either is not making the grade or they are not making the way to you. I dont mean DYI, meaning more I want to enjoy my music , but dont want to spend a lot.
 

zym1010

Active Member
Forum Donor
Joined
May 31, 2019
Messages
168
Likes
95
Location
San Francisco Bay Area
The new GLM 4.1 did add a notch filter at 2K for both left and right channel on my 8330A.

With GLM 4.1, I subjectively feel the bass part of my 7350A + 8330A combo improves a lot. Kudos to Genelec
 

dorirod

Active Member
Joined
Oct 9, 2019
Messages
249
Likes
249
Hello @amirm : do you have any lower priced or affordable speakers in the pipeline. It seems that under $500 a pair segment either is not making the grade or they are not making the way to you. I dont mean DYI, meaning more I want to enjoy my music , but dont want to spend a lot.

I count 28 reviews with speakers under $500/pr that are recommended:
https://www.audiosciencereview.com/forum/index.php?pages/SpeakerTestData/

Sorted by score 5.0 or greater, the 5 non-DYI options available are:

1628870149249.png
 
Last edited:

Spocko

Major Contributor
Forum Donor
Joined
Sep 27, 2019
Messages
1,621
Likes
2,999
Location
Southern California
I assume you can correct these peaks with GLM, right?
Sure, but the question is whether that actually makes the sound better. The review states that it was inconclusive.
I don't know if it would have made a difference, but I don't believe @amirm used GLM when he added EQ the 8330a in his review. Now, for those wondering, Amir does have my GLM kit (updated to the latest firmware) and will be testing it on the 8330a so let's see if it's any better than his normal EQ software. Hopefully, it offers some insight into the limits of what Genelec's advanaced DSP can do to improve a speaker's fundamental design flaws.
 

Sancus

Major Contributor
Forum Donor
Joined
Nov 30, 2018
Messages
2,923
Likes
7,616
Location
Canada
The new GLM 4.1 did add a notch filter at 2K for both left and right channel on my 8330A.

That's very interesting. First time I've heard of GLM being patched to correct the response of a speaker itself and not just the room.

I wonder if these speakers were on the newest firmware when measured?
 

Trell

Major Contributor
Joined
May 13, 2021
Messages
2,752
Likes
3,285
That's very interesting. First time I've heard of GLM being patched to correct the response of a speaker itself and not just the room.

I wonder if these speakers were on the newest firmware when measured?

It's part of the GLM 4.1 upgrade instructions to check that all monitors/subwoofers as well as the GLM Kit has the latest firmware, so I did that. Same instructions for GLM 4.0 as well.

I've no opinion if the 2K notch is in order to correct the response of the monitor as such or just my room.
 

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,335
Likes
6,700
The new GLM 4.1 did add a notch filter at 2K for both left and right channel on my 8330A.

That's really cool. With all the new features of 4.1(positive shelf filters, positive notch filters, speaker correction, phase alignment to 100Hz), it almost seems like 4.0 -> 4.1 is a bigger jump than 3.0 -> 4.0.
 

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,335
Likes
6,700
It's part of the GLM 4.1 upgrade instructions to check that all monitors/subwoofers as well as the GLM Kit has the latest firmware, so I did that. Same instructions for GLM 4.0 as well.

I've no opinion if the 2K notch is in order to correct the response of the monitor as such or just my room.

2kHz is well into speaker correction territory, so it's not a room correction. It seems they've chosen to correct the 2kHz resonance via GLM on not internal DSP. It makes sense that they fix it with DSP, but it is interesting that they do it via GLM and not internally. If that's the case, then it seems that all of these 83XX anechoic measurements may be somewhat meaningless now(since they don't have proper DSP applied). To get true measurements, Amir likely needs to :

1. Measure once anechoically
2. Run GLM
3. Compare GLM filters to anechoic measurements to determine which are room filters and which are speaker filters
4. Store only the speaker filters to the monitors memory
5. Re run the NFS measurements

Really complex, but that's the only real way I see to get accurate measurements now if they've started relying on GLM to do both speaker ad room DSP. That, or maybe Genelec could just give Amir a list of speaker filters that he needs to apply.
 

Sancus

Major Contributor
Forum Donor
Joined
Nov 30, 2018
Messages
2,923
Likes
7,616
Location
Canada
2kHz is well into speaker correction territory, so it's not a room correction. It seems they've chosen to correct the 2kHz resonance via GLM on not internal DSP.

Well, maybe. The GLM filters can be stored in the speaker after all. It is POSSIBLE that the firmware update applies the filter and then GLM just shows it because they chose to use a GLM filter slot and not one of the "hidden" ones because those were already filed with important filters.
 

richard12511

Major Contributor
Forum Donor
Joined
Jan 23, 2020
Messages
4,335
Likes
6,700
Well, maybe. The GLM filters can be stored in the speaker after all. It is POSSIBLE that the firmware update applies the filter and then GLM just shows it because they chose to use a GLM filter slot and not one of the "hidden" ones because those were already filed with important filters.

So, do you think the speaker Amir measured already had the 2kHz filter in place? Maybe he just needs to apply the 4.1 update. Honestly, it makes perfect sense to do both speaker correction and room correction using GLM's filters. It just makes it a PITA to measure, if that's what they're doing.
 
Top Bottom