Not sure I agree...
There is no clear proof of a root cause or causes here - just some interesting hypotheses that remain untested.
Certainly not conclusive for me but am willing to acknowledge we may never know and am ok with it.
-
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!
Buchardt S400 Speaker Review
- Thread starter amirm
- Start date
Not sure I agree...
There is no clear proof of a root cause or causes here - just some interesting hypotheses that remain untested.
Certainly not conclusive for me but am willing to acknowledge we may never know and am ok with it.
Not sure I agree...
There is no clear proof of a root cause or causes here - just some interesting hypotheses that remain untested.
Certainly not conclusive for me but am willing to acknowledge we may never know and am ok with it.
Resonance of passive radiator at the frequency of that notch has been measured and, as @ctrl, @BYRTT and @Juhazi explained it directly points to that notch.
What exactly is your theory, do you think it is SBIR related so the position would change something? Are you aware that Klippel takes out SBIR effects?
This "phenomenom" is acoustic interference as we measure it by spl variations, and that is exactly why it makes different spl peaks/dips at different angles horizontally and vertically. Basically change of speaker position (in relation to front and side walls) will have more changes (reflection summ/negat) at listening spot in many other frequencies that will mask this single narrow band deviation.
This means that also eq in signal is useless as "cure". Radiation pattern in 3D and room reflections don't change at all that way. But the most important thing is the narrow band and small amplitude of deviations - most likely not audible at all with music and very difficult to hear even with a slow sinusoid signal sweep!
Last edited:
While I agree with everything you said I still think that changing speaker position will not cause radical change at 500-550Hz frequency. IME it simply doesn't happen so high at freq range. You may want to try to move any of your speaker 1m to any side and compare the response at LP and you will see that at 500-500Hz change is minimal.
Btw, as I demonstrated here IME you would have more chance to catch that notch at LP using MMM with pink noise with 60+ samples than with any kind of in-room sweep.
Many posts ago, I expressed agreement that there is additional output from the passive radiator. How narrow or wide is up for interpretation. Narrow Q resonances generally show up in impedance plots and that is not the case for the S400. As I posted earlier as well, I generally align with @Juhaszi modeling of an acoustic event due to a change in directivity. Without a speaker in hand, I have no way to independently test or try any remediation that might prove or disprove his model.
Am not a Klippel expert (Amir's job), but would think that their SBIR comp was meant to take the existing measurement room and compensate it to be quasi-anechoic? This would not say that taking a speaker in a normal room and moving it closer to the back wall or further would have minimal effect on its response at 550 Hz. I have not tried this with a single speaker, but seems counterintuitive to have no effect if I start to get within a wavelength or less. In this case, it would not be unusual for someone to place a speaker less than 24" from the back wall. Let's say , I moved it to within a foot.
Would you still contend there is minimal change in the 500 Hz wave propagation from the back of the speaker than at more than 24" from the back wall?
Last edited:
Yes, I strongly believe that notch would stay there as it is. I have tried moving my speaker more than 24" and measured reponse at LP above 450Hz barely changed at all. I would be happy to see your measurements proving the opposite.
Btw, I trust Klippel's measurement to be precise enough for speaker's EQ like this one, especially as inter-specimen consistency has been confirmed by @amirm , so we have all reasons to believe that adding room EQ in the 20-400Hz region would result in a very decent SQ in average room.
As I stated, I do not have the S400 or any comparable speaker that I might try to prove or disprove your statement.
For that matter, no Klippel either!
Naturally am interested, but will have to leave it to someone else.Finally, unless it is more audible than anyone has suggested, was already concerned that the panther was dead and still getting kicked.
As I stated, I do not have the S400 or any comparable speaker that I might try to prove or disprove your statement.
For that matter, no Klippel either!
Finally, unless it is more audible than anyone has suggested, was already concerned that the panther was dead and still getting kicked.
I wasn't expecting you to have S400 nor Klippel at your home - what I suggested was to measure your speaker with REW positioned at different distances from front wall. I'm pretty sure differences from 500Hz above would be minor, if any.
GelbeMusik
Senior Member
- Joined
- Apr 19, 2020
- Messages
- 445
- Likes
- 290
A few notes on that speaker. I have the naked bass driver at my hands, and evaluated it recently (SB Acoustics, IndONESia, sell to hobbyists too). It is well designed, only that it is a bit too soft to be used in a bass reflex enclosure. The tube has asymmetries in its air flow, which tend to pump-up the enclosure. A soft, high compliance driver follows the pressurization too easily, losing the center position.
So they opted for a passive radiator. But in a box that small it has to be quite heavy. When it moves there are counter-forces, which eventually shake the box more than expected. When it comes to thin walls--they are as good or even better than thicker ones. The resonances of the enclosed air make them vibrate, hence these shall be avoided in the first place. The 500Hz resonance (half of a wavelenght is about 33,4 cm) might find it's root cause there.
Some other design decisions are less agreeable. The oversized waveguide for instance, together with the high crossover frequency. The subsequent dip can possibly, gradually be mitigated by using the box top down
.
Regarding the distortion measurement I would recommend to keep the levels in tighter bounds. This size of a speaker is just not capable to deliver 96dB@1m in the deep or even upper bass region. To ask for it shows the too obvious: utter fail! If You get 86dB You're lucky. Any more of bass would at least distort the associated midrange beyond recognition (by intermodulation).
So they opted for a passive radiator. But in a box that small it has to be quite heavy. When it moves there are counter-forces, which eventually shake the box more than expected. When it comes to thin walls--they are as good or even better than thicker ones. The resonances of the enclosed air make them vibrate, hence these shall be avoided in the first place. The 500Hz resonance (half of a wavelenght is about 33,4 cm) might find it's root cause there.
Some other design decisions are less agreeable. The oversized waveguide for instance, together with the high crossover frequency. The subsequent dip can possibly, gradually be mitigated by using the box top down
.Regarding the distortion measurement I would recommend to keep the levels in tighter bounds. This size of a speaker is just not capable to deliver 96dB@1m in the deep or even upper bass region. To ask for it shows the too obvious: utter fail! If You get 86dB You're lucky. Any more of bass would at least distort the associated midrange beyond recognition (by intermodulation).
Last edited:
I finally got a good outdoor measurement. I only had time to perform a single on-axis measurement. The mic was 2-meters away (since this was a ground plane measurement). This was taken outdoors with a window time of approximately 50ms wide. Half-Hanning. Naturally, my results differ a bit from both Buchardt's and Amir's but overall it's a really close match.
The low-end bump appears on my data and is similar to Buchardt's. I have a steeper top-end dip in these measurements. It's interesting to me that the bump in question is different in all 3 sets of data. In Amir's data (blue) it's steeper and at about 520hz. In Buchardt's it is more mild in amplitude and at about 460hz and in my data (orange) it's at about 500hz and not quite as pronounced.
When I have time I will perform a full suite of tests. But for now, here's a comparison of all 3:
The low-end bump appears on my data and is similar to Buchardt's. I have a steeper top-end dip in these measurements. It's interesting to me that the bump in question is different in all 3 sets of data. In Amir's data (blue) it's steeper and at about 520hz. In Buchardt's it is more mild in amplitude and at about 460hz and in my data (orange) it's at about 500hz and not quite as pronounced.
When I have time I will perform a full suite of tests. But for now, here's a comparison of all 3:
Also, here's the distortion measurement. Again, 2.83v @ 1m equivalent. Below 1% to about 200hz. (gating means no data below this for HD)
No sign of distortion at the infamous 500hz spot.
No sign of distortion at the infamous 500hz spot.
Thanks man share research and a nice better coherence to Amir's curve, one variable is changing ground plane verse NFS that is there is no free space below the bottom of enclosure for ground plane and imagine that variable can influence a bit on frequency for bump.
- Joined
- Mar 17, 2020
- Messages
- 1,071
- Likes
- 4,548
I certainly have found this discussion interesting from an academic standpoint and far above the quality found on other forums. I wouldn't say the back and forth suffers from a forest vs. trees problem, but I can't help thinking that we're focusing on the wrong tree. I doubt very much that my unenthusiastic reaction the the S400's was due to a very narrow spike in the lower midrange. What I heard seems to correlate much more closely with the directivity mismatch in the critical region between 2k and 3k, which is very evident when you compare directivity plots of the S400 with those of the Revel and KEF speakers. I'm pretty sure that's what is robbing the S400 of an open and detailed sound. Of course, the only way I could convince all of you would be to transport everyone into my listening room (wearing masks, of course) and let you hear the S400's along with speakers that have more even directivity in this frequency band.
OP
- Thread Starter
- #416
96 dB @ 1 meter target is at 1 kHz. The rest, i.e. bass, will be whatever they will be (much lower levels).
OP
- Thread Starter
- #417
I hardly see it at 2.83 volts as well. You have to go up in drive voltage to see it. In my case, I had to drive it at 11 volts to get the measurements I showed (although I measure at 0.33 meters).
I agree, that notch is most probably not audible. The EQ I did based on Klippel won't change DI but that wide dip at app 2400Hz may influence your perception. Can you try listening with that filter?
I assume you corrected for your room effects?
- Joined
- Mar 17, 2020
- Messages
- 1,071
- Likes
- 4,548
I don't have EQ capability in my listening room. I'm not sure what you mean by correcting for room effects. I auditioned and compared all of the speakers in the same room using a volume-compensated switching preamp, so I'm not seeing where room effects enter the equation.
Your (and every other) room heavilly modifies frequency response in the range below 400-450Hz. So auditioning a speaker without proper room correction is simply not making justice to a speaker.
As an example, this is frequency response of Revel performa3 F208 loudspeaker I measured at dealer premises. Not very smooth below 450Hz, right?
Similar threads
- Poll
