VERY interesting design from Genelec, who don't give in to frivolous waffle. Nearly-flat coaxial loaded in large waveguides, with 2 oval woofers slot-loaded behind. The woofers are themselves contained in a reflex box. Any DIY precedents? I like the idea.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Just from appearance, I would have more faith in great sound if a coaxial speaker like the KEF was used on the front baffle. That would provide a smoother directivity function from midrange up to tweet frequencies, and allow easy woofer blending.
The BMS 6CN160 and 6C150 6.5" coaxials also have some good engineering. Nice computer speakers.
The BMS 6CN160 and 6C150 6.5" coaxials also have some good engineering. Nice computer speakers.
Attachments
From my experience, I think KEF's coaxials are state-of-the-art. The LS50s and old Reference 205/2 were a revelation for me. From Genelec's measurements, these seem to do well too. Their unique driver complement differs from KEF's in some significant areas:
KEF uses a ribbed surround to reduce surround reflections. Genelec uses a 3-way design, thereby reducing the need to consider excursion in surround design. The midrange is apparently some sort of foam (polyurethane?) with a hidden surround (some sort of rear suspension system?) that is then waveguide loaded. Diffraction where the horn meets the slots seems likely though. I also don't like the idea of oval woofers. More Sd but a horrible chore to fix. Nevertheless, seems intriguing.
KEF uses a ribbed surround to reduce surround reflections. Genelec uses a 3-way design, thereby reducing the need to consider excursion in surround design. The midrange is apparently some sort of foam (polyurethane?) with a hidden surround (some sort of rear suspension system?) that is then waveguide loaded. Diffraction where the horn meets the slots seems likely though. I also don't like the idea of oval woofers. More Sd but a horrible chore to fix. Nevertheless, seems intriguing.
An externally hosted image should be here but it was not working when we last tested it.
is 260hz THD peak from bass desing?
Wow! Where did you get these graphs? Genelec only provides polar graphs (though that's already more than can be said for most companies!)
Seems to be some issue with the oval bass driver. Too low to realistically be some surround/cone edge resonance. Only 2nd harmonic spikes badly. Some sort of motor linearity/symmetry issue? I'd imagine oval drivers to fall flat under Klippel.
http://www.diyaudio.com/forums/multi-way/272346-manninen-presents-steel-real-6.html#post4367902
measured myself. it is nearfield in a room, did it few other spots, similar peaks on same places
measured myself. it is nearfield in a room, did it few other spots, similar peaks on same places
Last edited:
Thise Gene does have a coaxial mid-treble driver! It is so well integrated tha tit's hard to see!
Or do you mean the diffusor that KEF uses? Manninen's on-axis is very smooth, no typical wg dip around 8-10kHz.
Thanks for posting this. I was quite enamored with these speakers until now that I see they have terrible distortion in the bass area. My own diy dual woofer slot loaded sub integrated with a full range FAST is way better from HD perspective. Look at the climbing distortion below 150Hz. And the peak at 250Hz is quite frankly unacceptable for a pro reference monitor. That frequency may be cabinet rattle or resonance, or turbulence due to the narrow channels used for the bass slots.
I honestly think a better idea would be a slightly larger coaxial driver mounted similarly crossed to dual opposed conventional woofers instead of the fanciful slot loading design. But that'd give us the KEF Blade
BTW, have you ever measured the intermodulation performance of your speakers? The Vifas are great, but surely running them for such a wide frequency range introduces significantly more IMD, especially in multitone performance, would it not?
https://dl.dropboxusercontent.com/u/33080121/hifi/steelisreal/8351.mdat
here is the rew file for 8351.
for being finnish im not patriotic about genelec, 8351 is way too bright for me. -2,5db 5000hz shelf on GLM
i have roommode in 30hz, i think somehow room affects the graphs
here is the rew file for 8351.
for being finnish im not patriotic about genelec, 8351 is way too bright for me. -2,5db 5000hz shelf on GLM
i have roommode in 30hz, i think somehow room affects the graphs
Last edited:
I honestly think a better idea would be a slightly larger coaxial driver mounted similarly crossed to dual opposed conventional woofers instead of the fanciful slot loading design. But that'd give us the KEF Blade
BTW, have you ever measured the intermodulation performance of your speakers? The Vifas are great, but surely running them for such a wide frequency range introduces significantly more IMD, especially in multitone performance, would it not?
I have not measured IMD, but nowadays I typically use the Vifa or other full range driver in a 2-way FAST with a XO between 225Hz and 400Hz so I don't expect IMD to be a problem now that it is relieved of bass duties.
I would have used it as a wideband midrange driver with a low-order XO, crossed to a good 3/4" tweeter like the SEAS 22 TAF/G or the new 19mm SB. Not that the Vifa can't do treble well.
Have you tried the Scanspeak 8414? Seems to be a more affordable version of the SS you tried in your subjective comparison part 2 and it has been measured to have very clean decay and HD (As Troels states, almost like a dome tweeter!)
Back to topic, certainly disappointing. Interesting coax design though. There are not that many good coaxs that can be used as a 2-way system with decent extension. Gradient has a custom one from SEAS that way outperforms the trash SEAS has for DIYers.
I applaud Genelec for its new more conventional designs. The M040/M030 entry-level series are extremely flat with good polars, but the amp clips fairly easily. I wanted to love the 8351s too, as it seemed radical but ultimately has been shown to be half-baked is certain important regards.
I honestly think a better idea would be a slightly larger coaxial driver mounted similarly crossed to dual opposed conventional woofers instead of the fanciful slot loading design. But that'd give us the KEF Blade
Opposed mounted woofers produce a completely different radiation pattern. The Genelec is optimized to produce a nearly constant directivity down to 200 Hz. Without the special woofer alignment this would not be possible. I simulated both with ABEC.
Last edited:
No, I have not tried the SS 10F/8414 - it looks pretty good too. For a FAST or would have worked just fine. I certainly enjoy the sound from the 8424.
Yes, but is less directivity control or lesser distortion preferable? As xrk posited, there is a possibility that the distortion problems are caused by their prioritization of constant directivity, causing phenomena like turbulence at the slot. I suspect that huge 2nd-order peak at 260Hz is due to improper woofer design. Probably does badly in Klippel. Another possibility is XO being too low for the midrange to handle, but the overriding priority was to control directivity at all cost.
Any graphs/screenshots? Would be interesting to compare.
While distortion graphs are nice to look at, they are mostly irrelevant when it comes to the quality of sound. Unless the distortion is gross, it is not be audible. Geddes has done some research in this area, look it up.
Also, I would not believe distortion measured in the home environment. I'm not sure how reflections add/detract from the distortion measurement. At the very least, one could do a nearfield measurement for the bass drivers, or for frequencies below 500 Hz. That should make for a reflection free measurement.
You guys should look at the link below. How much distortion can you hear?
Listening Test
Genelec have correctly focused on the most important thing in these monitors: linearity of the frequency response and off-axis response. 99.99% of what you hear is defined by these parameters. Besides, nearfield monitors are not expected to go that loud, which further reduces the importance of distortion measurements.
Also, I would not believe distortion measured in the home environment. I'm not sure how reflections add/detract from the distortion measurement. At the very least, one could do a nearfield measurement for the bass drivers, or for frequencies below 500 Hz. That should make for a reflection free measurement.
You guys should look at the link below. How much distortion can you hear?
Listening Test
Genelec have correctly focused on the most important thing in these monitors: linearity of the frequency response and off-axis response. 99.99% of what you hear is defined by these parameters. Besides, nearfield monitors are not expected to go that loud, which further reduces the importance of distortion measurements.
I have heard this Geddes study before and think it refers to when you collapse distortion to a single THD value that it is not clear. In this case, the distortion is shown for various harmonic orders over the audible range, and it is quite large, 10% at 50Hz and approaching unity at 30Hz. I am more concerned about the rather significant 3rd harmonic peak at 250Hz which would be quite audible. 10% distortion is perceptible and is about where we think it sounds bad.
An externally hosted image should be here but it was not working when we last tested it.
That would very unlikely be caused by the surrounding environment.
Last edited:
2nd and 3rd order distortion is pretty benign. The 250 Hz peak is about 30db down, which is about 3% distortion. Also, if you look closely, both the 2nd and 3rd order peak at the same point. Not sure what mechanism in the loudspeaker can cause such a distortion. Could it be a room reflection?
A much better way to measure low frequency distortion is in the nearfield, where the mic is placed a few mm away from the center of the cone. That significantly improves the signal to noise ratio and the reflections are significantly below the main signal level.
The other question is drive level. At what SPL is this measured at? How far from the speaker? All these things matter in a distortion measurement. You cannot simply look at a graph and say that the distortion is bad without knowing the specifics of the measurement.
A much better way to measure low frequency distortion is in the nearfield, where the mic is placed a few mm away from the center of the cone. That significantly improves the signal to noise ratio and the reflections are significantly below the main signal level.
The other question is drive level. At what SPL is this measured at? How far from the speaker? All these things matter in a distortion measurement. You cannot simply look at a graph and say that the distortion is bad without knowing the specifics of the measurement.
Any object in the room with a self resonance at 250 hz will vibrate at 250hz when you hit it with 125hz, adding 2nd harmonic distortion to the measurement. If not done anechoicly, the way Genelic does them, they have to be taken with a grain of salt. The room will definitely affect the measurements.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.