GRIMM Audio LS1c & SB1 DSP Speaker Review

[tktran303]

It's a pity that NFS can not have a more realistic in room prediction using more information about the placement of the speaker. Now we get quite some misleading results..

That's Treble Technologies, or other 3D simulators

 

[amirm]

It's a pity that NFS can not have a more realistic in room prediction using more information about the placement of the speaker.

It is not part of NFS to predict anything. The graphs you see are computed per CEA-2034 standard.

 

[Vuki]

I decline from further comment re: full space, half space or quarter space. But one doesn't have to take my word for it. Just look at the data.

Exhibit A:
88dB is the red line.
20Hz is 72-75dB depending on which line you take (16-13dB down)

View attachment 452382


Exhibit B: at 1m- average response is 88dB between 500Hz and 8KHz.
20Hz is 54dB (34dB down)


View attachment 452383


Exhibit C: Estimated in-room response (bold: my emphasis)-

View attachment 452384

Exhibit D:
Bass response at the actual listening position? Depends on the actual room and actual listening position.


The whole bass response below 100Hz is different, depending on how/where you measure.

But there is no room in the world where bass would be so much down as in "estimated in-room response". So this estimation is not really correct.

 

[tktran303]

Then you have done more research than I have

When listening doesn't correlate well with distortion measurements, I connected the dots when I also noticed that harmonic distortion measurements from various sources are different eg. Voice Coil Test Bench vs others

2. And this is why it's an Estimated or Predicted in-room-response.

estimated: adj
roughly calculated; approximate:

 

[SorenTyson]

Thanks for this review, both to Amir, and to Grimm for sending. I often look at the super expensive speakers and wonder how they measure. Normally only subjective reviews of them, here we get the data

 

[Pearljam5000]

Thanks for this review, both to Amir, and to Grimm for sending. I often look at the super expensive speakers and wonder how they measure. Normally only subjective reviews of them, here we get the data

The next expensive speaker @amirm should measure

 

[Endibol]

I find Darko insufferably boring and smug, although I acknowledge he’s a good presenter and his production standards are very high.

There are a lot of people who hear differences between DACs. Expensive and esoteric DACs are for these people.

But these esoteric DACS must be "broken" then, otherwise you wouldn't here any difference at all.. Transparent = Transparent!

 

[Karmacoma]

Ugly, overpriced and bass shy are too big of negatives.
This radiation pattern is the star of the show, makes me very curious to listen to them in a large room, I don't expect them to be very at ease in a studio configuration, or is there something I'm missing ?
Big thanx to the company for sending these, and of course thanx to our master reviewer.

 

[mar1o]

Thank you Amir and Grimm Audio. Long waiting for a Klippel measurement of the LS1. I don‘t know why, but i like these speaker so much…

 

[fredoamigo]

Ugly, overpriced and bass shy are too big of negatives.
This radiation pattern is the star of the show, makes me very curious to listen to them in a large room, I don't expect them to be very at ease in a studio configuration, or is there something I'm missing ?
Big thanx to the company for sending these, and of course thanx to our master reviewer.

They are not built for studios, however, in the near field (nearfield) I do not expect huge differences with the typical studio monitors. On the other hand, in more distant fields, the imaging and the soundstage are certainly much better with the GRIMM.

 

[Matt_Holland]

But there is no room in the world where bass would be so much down as in "estimated in-room response". So this estimation is not really correct.

But the estimated response above Schroeder frequency is very accurate. Estimating anything below Schroeder frequency in any way other than anechoic is pointless. We can use the anechoic bass response to decide where we think the speaker may work optimally in a room. The KEF extended bass shelf is clear in NFS measurements and we know this is optimised for near wall placement.

With the LS1c/SB1 we see a typical closed box shallow roll off in the LF. If we didn’t know what the product looked like we might make a fair assumption that the speaker (sub) would work well close to a boundary. And guess what, the SB1 is designed to go on the floor.

 

[charlielaub]

I have heard this speaker a couple of times at shows and it indeed does sound very neutral and accurate. It's a pleasure to listen to.

I was a bit surprised to see the low frequency (below 100Hz, or even between 100-300Hz) distortion level getting a bit high in the 96dB measurements. Grimm's SB1 (subwoofer) has a sophisticated feedback system (there is a tech note about it) that should keep distortion very low. I had very high expectations of it but honestly I don't consider circa 10% distortion at 50 Hz,96dB@1m to be all that good from the SB1 and its 12" driver, all things considered. Grimm makes some strong claims about their feedback system such as "has 30dB less distortion" and that "all problems that are related to low frequency resonances and distortion" are "solved". It seems that this is not the case, unless perhaps the feedback loop was defeated during the measurements?

Adding to the above and regarding the subwoofer, I am now wondering if the level for the sub was turned all the way down or the subwoofer was defeated entirely during the Klippel measurements. Here's why I suspect that might be the case.

First, look at the plot of the nearfield responses for the drivers in the system:

Now look at the frequency response in the HD measurements of the "system":

What I see is that the fundamental (the upper most green line) in the HD plots is rolling off too early for the subwoofer to be playing at the proper level relative the the rest of the loudspeaker system. It is not difficult to imagine the nearfield measurement of the subwoofer in the upper plot being brought up in level and added to the "fundamental" in the lower HD plots. This should fill in down to and below 50Hz, not roll off starting at or above 100Hz. The "fundamental" line looks more like the "woofer" line in the nearfield plot.

If the subwoofer level is very low or it is defeated entirely, it is only the woofer that is contributing to the output and of course it will have much more distortion than if a 12" subwoofer with motional feedback is reproducing those frequencies. I would expect that, if the subwoofer level was set properly, distortion would show a peak around 80-100Hz and then HD would begin to decrease again at lower frequencies as the purportedly very low distortion subwoofer system begins to dominate the system output.

I think that @amirm should check into this, if the intent was to include the subwoofer in the HD and other system measurements.

 

[Matt_Holland]

Adding to the above and regarding the subwoofer, I am now wondering if the level for the sub was turned all the way down or the subwoofer was defeated entirely during the Klippel measurements. Here's why I suspect that might be the case.

First, look at the plot of the nearfield responses for the drivers in the system:
View attachment 452408


Now look at the frequency response in the HD measurements of the "system":
View attachment 452409


What I see is that the fundamental (the upper most green line) in the HD plots is rolling off too early for the subwoofer to be playing at the proper level relative the the rest of the loudspeaker system. It is not difficult to imagine the nearfield measurement of the subwoofer in the upper plot being brought up in level and added to the "fundamental" in the lower HD plots. This should fill in down to and below 50Hz, not roll off starting at or above 100Hz. The "fundamental" line looks more like the "woofer" line in the nearfield plot.

If the subwoofer level is very low or it is defeated entirely, it is only the woofer that is contributing to the output and of course it will have much more distortion than if a 12" subwoofer with motional feedback is reproducing those frequencies. I would expect that, if the subwoofer level was set properly, distortion would show a peak around 80-100Hz and then HD would begin to decrease again at lower frequencies as the purportedly very low distortion subwoofer system begins to dominate the system output.

I think that @amirm should check into this, if the intent was to include the subwoofer in the HD and other system measurements.

Assuming that Amir placed the sub between the speaker stands when measuring the system, could it be that the NFS mic when aligned to the tweeter axis at 1m is then further than 1m away from the sub resulting in a relative drop in SPL? I guess we’re only talking about 20log(1/1.4) though.

 

[charlielaub]

If you compare the frequency response of the woofer measured in the nearfield and the "fundamental of the HD plots, below 100Hz they are essentially identical. To me this seems to say that there is no contribution from the subwoofer, e.g. it seems to be "off". I don't think that can be explained by a reduction in level by a few to up to 6dB due to a longer pathlength to the sub from the mic.

 

[Sigberg Audio]

So, I have been measuring subwoofers extensively both in-room, nearfield, ground plane (outside) and in an anechoic chamber (@Seas), which have been verified by @AscendDF to measure very closely to the NFS below 100hz.

The result in the anechoic chamber is way leaner than even a nearfield or ground plane measurement, and obviously nothing close to what you'd typically measure in-room. While what we see below 50-70hz in an anechoic or NFS measurement may technically be "correct" given what we are measuring, it is not a good way to understand real world performance.

Based on my experience I think the result we see here is misleading for the real world bass performance of the Grimm, and everyone saying they are disappointed in the bass performance are likely jumping to conclusions based on measurements you do not properly understand.

TL;DR: If there is anything wrong with this measurement I don't know, but even it were correct, it may not properly indicate real world performance below 100hz.

 

[Sigberg Audio]

An additional interesting fact is that the bass is not omnidirectional in the same way in an anechoic chamber.

Here is an anechoic measurement (20-50hz) of a ported speaker. Drivers + port pointed at the microphone is in red, pointing away is in green.
Additionally, this speaker is typically flat to at least 25-30hz in-room, while in the anechoic measurement we see 10dB drop from 50hz to 20hz even when pointed at the microphone.

 

[Sokel]

So, I have been measuring subwoofers extensively both in-room, nearfield, ground plane (outside) and in an anechoic chamber (@Seas), which have been verified by @AscendDF to measure very closely to the NFS below 100hz.

The result in the anechoic chamber is way leaner than even a nearfield or ground plane measurement, and obviously nothing close to what you'd typically measure in-room. While what we see below 50-70hz in an anechoic or NFS measurement may technically be "correct" given what we are measuring, it is not a good way to understand real world performance.

Based on my experience I think the result we see here is misleading for the real world bass performance of the Grimm, and everyone saying they are disappointed in the bass performance are likely jumping to conclusions based on measurements you do not properly understand.

TL;DR: If there is anything wrong with this measurement I don't know, but even it were correct, it may not properly indicate real world performance below 100hz.

Correct.
It's strange though that both SP and audio.de got to get flat response down to 20Hz:

I insist, something is amiss at its configuration.

 

[Sigberg Audio]

Correct.
It's strange though that both SP and audio.de got to get flat response down to 20Hz:

View attachment 452415

I insist, something is amiss at its configuration.

Yes, that may be. How do audio.de measure?

 

[Sokel]

Yes, that may be. How do audio.de measure?

Don't know their methodology but the results of this one (it's the previous model) looks a lot similar with one of SP modes (without the usual bump)

 

[Matt_Holland]

Are we saying that whilst absolute LF accuracy might not be correct with NFS, that relative accuracy isn’t either? Or can we confidently compare NFS data to assess performance differences between two speakers?