Genelec 8330A Measurements. (Is the issue resolved?)

[Nuyes]

I measured a pair of 8330A's a couple days ago.








(The owner of these speakers bought them a month ago).

Frequency Response
Acoustic axis : outer rim of the woofer as instructed in the manual.

Directivity

Polar plot

THD

Multitone test

Each SPL.

Compression test (with Multitone)

Compression test (with Sine sweep)

Since I use nearfield measurement, the grilles on this speaker prevented me from getting close enough to the woofer.
Therefore, LF is different from the NFS data measured by @amirm

But... See?


We can see the same thing at 2.2 kHz.





I thought this issue was resolved with GLM two years ago.


So I took one of the two speakers the owner brought in and measured it without resetting the GLM settings.

However, the response at that frequency was the same.





Can this really be explained as a coincidence due to product variation?





I use a different microphone than Amirm.
(Earthworks M30)

 

[Sancus]

I don't recall Genelec confirming there was an issue. The idea that GLM was patched for this was suggested by a user in the review thread, and it also was never confirmed by Genelec.

You'd have to get a statement from Genelec. The rises in response at 500-600hz and 1-3khz do exist in their own measurements, if with different magnitudes, so I doubt they're unaware.

 

[Nuyes]

I don't recall Genelec confirming there was an issue. The idea that GLM was patched for this was suggested by a user in the review thread, and it also was never confirmed by Genelec.

You'd have to get a statement from Genelec. The rises in response at 500-600hz and 1-3khz do exist in their own measurements, if with different magnitudes, so I doubt they're unaware.

Below is the response that a Korean audiophile received from Genelec about the issue.

However, the two additional samples I measured were at the same level.

 

[YSC]

Below is the response that a Korean audiophile received from Genelec about the issue.



View attachment 283161


However, the two additional samples I measured were at the same level.

Nice, unfortunate for the systematic deviation of a batch, but great that they are prompt to include a fix and face the issue and unlike some manufacturers just blame for your measurement being unprofessional

 

[Sancus]

Below is the response that a Korean audiophile received from Genelec about the issue.
However, the two additional samples I measured were at the same level.

Interesting, if you're sure they're up to date in terms of firmware(it's possible for inventory to sit around for years and be on an old firmware version) then I'd say Genelec has a support issue they need to resolve.

 

[Nuyes]

Nice, unfortunate for the systematic deviation of a batch, but great that they are prompt to include a fix and face the issue and unlike some manufacturers just blame for your measurement being unprofessional

This is from a Korean user when the 8330A was measured by Amir here at ASR two years ago.

 

[Nuyes]

Interesting, if you're sure they're up to date in terms of firmware(it's possible for inventory to sit around for years and be on an old firmware version) then I'd say Genelec has a support issue they need to resolve.

The owner of these samples purchased the product a month ago, and GLM is the latest version.

 

[Sancus]

The owner of these samples purchased the product a month ago, and GLM is the latest version.

Last I did this on my Genelecs, being on the newest version of GLM isn't enough. You also have to go into the group view and manually select each speaker and then confirm the firmware update.

I did that a while ago, it's possible they've cleaned that process up because it was annoying.

E: Genelec speakers also have a manufacturing date on the serial number sticker, so if you confirm that's months or years after they said the bug was fixed, then that would also confirm that something has gone wrong over there.

 

[Nuyes]

Last I did this on my Genelecs, being on the newest version of GLM isn't enough. You also have to go into the group view and manually select each speaker and then confirm the firmware update.

I did that a while ago, it's possible they've cleaned that process up because it was annoying.

E: Genelec speakers also have a manufacturing date on the serial number sticker, so if you confirm that's months or years after they said the bug was fixed, then that would also confirm that something has gone wrong over there.

Thank you from the bottom of my heart.

I wasn't aware of this at all.
I have reached out to the owner of the speaker in question about the firmware version and have asked them not to proceed with the update if possible.

If the owner allows, I will bring this samples back to my studio to compare before and after the firmware update.

 

[YSC]

Thank you from the bottom of my heart.

I wasn't aware of this at all.
I have reached out to the owner of the speaker in question about the firmware version and have asked them not to proceed with the update if possible.

If the owner allows, I will bring this samples back to my studio to compare before and after the firmware update.

actually the original measurement from you is good enough for comparison even if he'd already update it, the broad shelf is easily noticeable to be present or not. It would be interesting to see if Genelec did really fixed the issue or they just kept as is and fix when user find out

 

[Nuyes]

Last I did this on my Genelecs, being on the newest version of GLM isn't enough. You also have to go into the group view and manually select each speaker and then confirm the firmware update.

I did that a while ago, it's possible they've cleaned that process up because it was annoying.

E: Genelec speakers also have a manufacturing date on the serial number sticker, so if you confirm that's months or years after they said the bug was fixed, then that would also confirm that something has gone wrong over there.

I was just contacted by the owner of these products.
The samples had the latest version of firmware installed at the time of measurement.

 

[goat76]

Nice, unfortunate for the systematic deviation of a batch, but great that they are prompt to include a fix and face the issue and unlike some manufacturers just blame for your measurement being unprofessional

If it's just a certain batch that has this fault, how can it be fixed by a firmware update? Wouldn't that update affect all existing units the same way or can Genelec make firmware updates for a certain batch or a specific unit?

 

[Sancus]

I was just contacted by the owner of these products.
The samples had the latest version of firmware installed at the time of measurement.

Guess you should contact Genelec directly and ask what's going on here It's possible the adjustments made by the firmware update are targeted by serial number or another unique identifier and they missed a few.

It doesn't seem like they could adjust all speakers equally because they already measure and correct them individually at the factory, and they differ enough that a "one size fits all" filter probably wouldn't work.

 

[YSC]

If it's just a certain batch that has this fault, how can it be fixed by a firmware update? Wouldn't that update affect all existing units the same way or can Genelec make firmware updates for a certain batch or a specific unit?

well, all we know is that AT LEAST it is affecting a batch, and Genelec said themselves it is a systematic error where they are fixing in who knows starting from when produced 8330A, and they did issue a firmware update to acknoledge the issue, that is the nice part, the systematic error is the bad part but fixing any minor defect is much better than dodging no?

 

[YSC]

Guess you should contact Genelec directly and ask what's going on here It's possible the adjustments made by the firmware update are targeted by serial number or another unique identifier and they missed a few.

It doesn't seem like they could adjust all speakers equally because they already measure and correct them individually at the factory, and they differ enough that a "one size fits all" filter probably wouldn't work.

actually I've just emailed Genelec support for the findings in this post, see what they will do to the issue of affected units/all. it's unfortunate that the higher end DSP 8330 on axis is worse than the Analog 8030, and with same directivity performance. will be a shame if they can't fix that, what I would guess is they could add a gated measurement option in GLM to see if the speaker is affected and apply the firmware to have a flattened FR.

 

[goat76]

well, all we know is that AT LEAST it is affecting a batch, and Genelec said themselves it is a systematic error where they are fixing in who knows starting from when produced 8330A, and they did issue a firmware update to acknoledge the issue, that is the nice part, the systematic error is the bad part but fixing any minor defect is much better than dodging no?

I think it's great that they want to fix the problem, I'm just asking how they can address the problem for a certain batch, or for single units with larger deviations than other units of the same model.

 

[YSC]

I think it's great that they want to fix the problem, I'm just asking how they can address the problem for a certain batch, or for single units with larger deviations than other units of the same model.

I am asking them essentially the same question, a wild guess would be since they factory calibrate every unit to match their tolerance, it's likely the equipment at the QC station or so was faulty at who knows how long a period of time, coz Amirm and Nuyes measured FR is very consistent I believe it should be the in house calibration problem, so basically they could do the same PEQ and implement over the original calibrated speaker and that should work for anechoic FR, for identifying there can be multiple ways like tracing the troubled equipment usage date before they found out it was faulty (I believe that could be as late as they are aware of Amirm's data) and fix all unit serial nos. before that fix using that equipment, and maybe for un-related serial units, add an option of gated measurement with GLM to see if that unit is affected. it's their QC chain so we can only guess until they reply on what they will do or have done

 

[Nuyes]

Sorry about that. There was a mistake in labeling when comparing samples.

It was not compared as Sample A and Sample B (with GLM), but as Sample A and Sample A (with GLM).

When we looked at the Klippel project where the measurements were taken, we found that the problem was significantly improved in Sample B.

(The 400 Hz in Sample B is the effect of the owner's room correction).

 

[SDC]

So GLM does something to the issue.

But it is not fixing it correctly. And now there is sample to sample difference in all 8330. Correct?

 

[YSC]

Sorry about that. There was a mistake in labeling when comparing samples.

It was not compared as Sample A and Sample B (with GLM), but as Sample A and Sample A (with GLM).

When we looked at the Klippel project where the measurements were taken, we found that the problem was significantly improved in Sample B.

View attachment 283438

(The 400 Hz in Sample B is the effect of the owner's room correction).

Sorry, I can't see where sample B comes from, so both A and B are from the owner as a pair? wondering are the serial no. differs significantly.

Since both sample A and B are with GLM I bet it was not GLM corrected but sample B is what I would consider on axis anechoic flat response with only +1db shelf, while Amirm's review sample and sample have a peak ~3db (still within +/-3db but the peak is what I would consider defect in Genelec 8x3x).

it seems like either sample B is actually with the latest firmware correctly identifying the problematic unit and fixed it, or it used a correct facility to do the factory calibration.