-
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!
Kef Q150 Bookshelf Speaker (review by Erin)
- Thread starter sweetchaos
- Start date
-
- Tags
- kef
Steve Dallas
Major Contributor
That is... Not terrible for the size and price, and its flaws are easily EQ-able.
This could be a great speaker with a few tweaks to the cabinet and crossover.
I would LOVE to see the Q series towers measured.
This could be a great speaker with a few tweaks to the cabinet and crossover.
I would LOVE to see the Q series towers measured.
You get 3 EQs:
v1: optimise for PIR -> too many PEQs around the crossover that may not work
v2a: optimise for PIR but smooth target -> better but still too many small and peaky PEQs
v2b: remove the last PEQs and keep only the first 3 from v2a -> recommended EQ.
edit V2a
Preference score 4.8 and 6.9 with a sub
with an EQ 6.1 and 8.2 with both
I added some smoothing to the eq designed in V1. This yield a lower score but more realistic as it is less sharp around the crossover and both LW and PIR looks good.
V2b
If you look at the EQ, you see that all the last one have low gain and high Q. You can remove all of them and get most of the benefit from the EQ. Here are the results with 3 PEQs and that's my recommended one:
Preference score 4.8 and 6.9 with a sub
with an EQ 5.7 and 7.8 with both
V1
Preference score 4.8 and 6.9 with a sub
with an EQ 6.9 and 8.2 with both
I optimised for flat LW and smooth PIR. The automated algorithm is optimising around 1kHz and it should not. The huge variations around the crossover are likely to drive a non minimal phase behaviour that IIR eq cannot correct. So the scores post EQ are likely to be optimistic.
v1: optimise for PIR -> too many PEQs around the crossover that may not work
v2a: optimise for PIR but smooth target -> better but still too many small and peaky PEQs
v2b: remove the last PEQs and keep only the first 3 from v2a -> recommended EQ.
edit V2a
Preference score 4.8 and 6.9 with a sub
with an EQ 6.1 and 8.2 with both
I added some smoothing to the eq designed in V1. This yield a lower score but more realistic as it is less sharp around the crossover and both LW and PIR looks good.
Code:
EQ for KEF Q150 computed from ErinsAudioCorner data
Preference Score 4.8 with EQ 6.2
Generated from http://github.com/pierreaubert/spinorama/generate_peqs.py v0.14
Dated: 2022-03-01-14:04:03
Preamp: -0.5 dB
Filter 1: ON PK Fc 13062 Hz Gain -4.34 dB Q 0.50
Filter 2: ON PK Fc 164 Hz Gain -1.36 dB Q 1.50
Filter 3: ON PK Fc 951 Hz Gain -1.63 dB Q 2.28
Filter 4: ON PK Fc 2528 Hz Gain +0.63 dB Q 3.99
Filter 5: ON PK Fc 1022 Hz Gain +1.29 dB Q 4.00
Filter 6: ON PK Fc 1177 Hz Gain -1.43 dB Q 3.98
Filter 7: ON PK Fc 792 Hz Gain -0.73 dB Q 3.97
Filter 8: ON PK Fc 1067 Hz Gain +1.01 dB Q 3.95
Filter 9: ON PK Fc 1568 Hz Gain +0.70 dB Q 3.98V2b
If you look at the EQ, you see that all the last one have low gain and high Q. You can remove all of them and get most of the benefit from the EQ. Here are the results with 3 PEQs and that's my recommended one:
Preference score 4.8 and 6.9 with a sub
with an EQ 5.7 and 7.8 with both
Code:
EQ for KEF Q150 computed from ErinsAudioCorner data
Preference Score 4.8 with EQ 5.7
Generated from http://github.com/pierreaubert/spinorama/generate_peqs.py v0.14
Dated: 2022-03-01-14:06:26
Preamp: -0.2 dB
Filter 1: ON PK Fc 13243 Hz Gain -4.37 dB Q 0.49
Filter 2: ON PK Fc 165 Hz Gain -1.30 dB Q 1.40
Filter 3: ON PK Fc 951 Hz Gain -1.63 dB Q 2.29V1
Preference score 4.8 and 6.9 with a sub
with an EQ 6.9 and 8.2 with both
I optimised for flat LW and smooth PIR. The automated algorithm is optimising around 1kHz and it should not. The huge variations around the crossover are likely to drive a non minimal phase behaviour that IIR eq cannot correct. So the scores post EQ are likely to be optimistic.
Code:
EQ for KEF Q150 computed from ErinsAudioCorner data
Preference Score 4.8 with EQ 6.1
Generated from http://github.com/pierreaubert/spinorama/generate_peqs.py v0.14
Dated: 2022-02-27-18:03:10
Preamp: -0.7 dB
Filter 1: ON PK Fc 12987 Hz Gain -3.99 dB Q 0.60
Filter 2: ON PK Fc 159 Hz Gain -1.36 dB Q 1.67
Filter 3: ON PK Fc 951 Hz Gain -1.73 dB Q 3.67
Filter 4: ON PK Fc 1012 Hz Gain +1.44 dB Q 3.99
Filter 5: ON PK Fc 1188 Hz Gain -1.80 dB Q 3.99
Filter 6: ON PK Fc 1339 Hz Gain +1.35 dB Q 3.37
Filter 7: ON PK Fc 805 Hz Gain -1.09 dB Q 3.99
Filter 8: ON PK Fc 1279 Hz Gain -0.93 dB Q 3.99
Filter 9: ON PK Fc 1054 Hz Gain +1.40 dB Q 3.98
Last edited:
I’ve had these for a few months now and I agree with Erin that the treble is a bit hot when listening directly on axis, but that’s also why I listen to them with less toe in. The soundstage seems wider to be that way too.
Not sure I necessarily agree with needing a subwoofer though. Compression at 96db looks fine and I would never come close to listening to that volume.
The DI mismatch and cabinet resonances are a bit disappointing unfortunately.
Could someone explain to me why the enhancement at 96db is higher than 102db? Is this what’s called a linearity problem?
Not sure I necessarily agree with needing a subwoofer though. Compression at 96db looks fine and I would never come close to listening to that volume.
The DI mismatch and cabinet resonances are a bit disappointing unfortunately.
Could someone explain to me why the enhancement at 96db is higher than 102db? Is this what’s called a linearity problem?
Last edited:
Tried this speaker a few weeks ago and didn't care for the tweeter. My high-passed and EQ'd Sony SS-CS5 bookshelves sounded better overall to me than the KEFs. The KEFs do image well though. I tried to like these but just couldn't vibe with them and ended up returning them.
Resonances of about 1khz I also experienced with the SP1587 driver in Q100. Resonance is a consequence of suspension. If a dumping mass is added, it resolves a lot. In the pictures, I marked where some flexible material should be applied.I also took off the HF and changed the dumping, both the amount and the type and then measured the impedance to see what the differences are.
The biggest concern is in the meager crossover.
The picture is not mine but a member from another forum.
Picture source:https://www.diyaudio.com/community/...ks-like-inside-a-real-kef-q100-driver.368717/
The biggest concern is in the meager crossover.
The picture is not mine but a member from another forum.
Picture source:https://www.diyaudio.com/community/...ks-like-inside-a-real-kef-q100-driver.368717/
Attachments
Larry B. Larabee
Senior Member
- Joined
- Nov 24, 2021
- Messages
- 347
- Likes
- 194
The time-alignment can't be perfect. Can anyone tell me where the transition is between drivers on the step response graph? Is it just a fraction of a millisecond? If it matters what maximum differential in a conventional design would be cause for concern. Do we know? Is the tweeter wavefront leading or lagging the woofer, does that matter?
Can someone tell me what's going on with the Q150?
Track: Frozen 2 - Into the Unknown
The "oh ohhh ohhh" voice cracks and sounds terrible. Not sure if the full tragedy comes through in the video.
I bought the Metas a couple weeks ago but when the Q150s went on sale I bought them hoping to save $1300+ and have been a/b'ing them with different music. Q150 was winning me over until "ohhh ohhhhh"
Track: Frozen 2 - Into the Unknown
The "oh ohhh ohhh" voice cracks and sounds terrible. Not sure if the full tragedy comes through in the video.
I bought the Metas a couple weeks ago but when the Q150s went on sale I bought them hoping to save $1300+ and have been a/b'ing them with different music. Q150 was winning me over until "ohhh ohhhhh"
What is the time stamp of this crackling voice?
The first 10 seconds of the song. I just fixed the video link.
The crackling and fuzz is in the recording. It's a defect.
It has to be more than that. It's like a terrible resonance that crackles the whole speaker. The metas don't come close to making the sound. Like the Q150 takes the defect and multiples it by 20x
OK. Then the next step is to frequency sweep the speakers to check for resonances and stuff.
Online Tone Generator - generate pure tones of any frequency
www.szynalski.com
I'm familiar with that song. That is definitely a speaker resonance, not inherent to the source. None of my speakers have such a strong resonance playing that part.
Not sure if it's a flaw with that particular Q150 or if it's the resonances around the 1khz mark being triggered. Oftentimes with cheaper speakers that have multiple significant(enough to show up in measurements) resonances, even if they're not too bad on a sweep, if you get the exact right frequency it's very noticeable.
That happened to Amir on the Elac UB52.
I suspect there is something defective about the speakers from your confirmation of the fault/complaint. The next time this happens frequency sweep the speakers and you'll pick up on any issues very accurately.
If it is what I think I am hearing, it is exactly the resonance issue I heard with Elac speaker that created a lot of fireworks. You need a clip like this that is borderline having that resonance. Then the speaker hits on the precise note and it gets amplified. Your track sounds very similar to the one and only track I have that brings this to surface.
maty
Major Contributor
With my modded KEF Q100 "no problema". PEQ (frequency and delay) + Convolution (filter made with rePhase). Nearfield, out off axis. And other tweaks.
The driver is unmodified. Box with two layers of Tecsound SY 70 inside (viscoelastic material).
Viscoelasticity - Wikipedia
Google: speakers+viscoelastic+material
Last edited:
Similar threads
- Locked
