Revel Salon2 vs Genelec 8351B - Blind Test Preparations

[QMuse]

@ecxopraxia

As you are OP I will try to answer your question, although we stepped off-topic.

"What does that mean when your in-room response is so uneven that you have huge nulls and dips? Should you boost the sub until the minimum response matches the mains, or decrease them until the maximum response matches the mains SPL, or something in between? "

Say you are aiming for typical XO point at 80Hz. Most quality subs are quite linear in the 50-110Hz range so you should be able to draw a straight line in that range. Adjusting gain on the sub (or on the sub channel if dealing with passive sub) should be done so that it meets same that line from the mains response. That should give you a good starting point to start matching the phase. As @Duke pointed out, it is an iterative process, so patience and lots of measurements are required to make it right.

 

[patate91]

If I end up blind testing with subs (though I don’t know if we have a strong consensus either way yet on that), your advise and others would be helpful on how to correctly set it up.

Just as a sanity check, I would want to make sure I haven’t been doing it all wrong. When you mention adjusting sub level to match main, I’m a little unclear on what this means if there are large nulls and peaks — so what are you matching with what, for example?

I have usually followed a process of (1) adjust mains delay until REW measures they greatest response with the least nulls, (2) adjust sub level so that the lowest point is roughly around the SPL of the mains so there is EQ headroom, and (3) apply EQ corrections (Dirac Live via miniDSP SHD which as far as I know doesn’t seem to do the integration itself, just the EQ).

Maybe I’ve been doing it all wrong? Maybe I need to adjust sub level so the crossover slopes are more symmetric with each other rather than worrying about EQ headroom. But I also don’t know how it works if a huge boost is needed at some frequencies; does that risk clipping?

Maybe you can use REW's auto target level feature in the EQ menu.

 

[QMuse]

If I end up blind testing with subs (though I don’t know if we have a strong consensus either way yet on that), your advise and others would be helpful on how to correctly set it up.

Just as a sanity check, I would want to make sure I haven’t been doing it all wrong. When you mention adjusting sub level to match main, I’m a little unclear on what this means if there are large nulls and peaks — so what are you matching with what, for example?

I have usually followed a process of (1) adjust mains delay until REW measures they greatest response with the least nulls, (2) adjust sub level so that the lowest point is roughly around the SPL of the mains so there is EQ headroom, and (3) apply EQ corrections (Dirac Live via miniDSP SHD which as far as I know doesn’t seem to do the integration itself, just the EQ).

Maybe I’ve been doing it all wrong? Maybe I need to adjust sub level so the crossover slopes are more symmetric with each other rather than worrying about EQ headroom. But I also don’t know how it works if a huge boost is needed at some frequencies; does that risk clipping?

With one sub you will certainly excite some (if in corner probably all) of the room modes. Dealing with peaks is easy as you can simply attenaute them via DRC. Dealing with dips (nulls) is much more complex as it requires adding (at least) one more sub and finding a position for it which will NOT excite the same dip that first sub did. That second sub should however also be phase matched around XO point with mains if it is crossed with them or in phase with 1st sub if it is crossed with it.

 

[richard12511]

@ecxopraxia

As you are OP I will try to answer your question, although we stepped off-topic.

"What does that mean when your in-room response is so uneven that you have huge nulls and dips? Should you boost the sub until the minimum response matches the mains, or decrease them until the maximum response matches the mains SPL, or something in between? "

Say you are aiming for typical XO point at 80Hz. Most quality subs are quite linear in the 50-110Hz range so you should be able to draw a straight line in that range. Adjusting gain on the sub (or on the sub channel if dealing with passive sub) should be done so that it meets same that line from the mains response. That should give you a good starting point to start matching the phase. As @Duke pointed out, it is an iterative process, so patience and lots of measurements are required to make it right.

I don't see this as off topic. Integrating the subs well is an important aspect of the test, imo.

 

[QMuse]

Let me also point out to another problem that very often messes with the LF response. My current setup consists of 2 main towers that are linear and able to produce enough SPL with low distortion for the kind of music I'm listening to so I'm not using subs.

Here is how their phase response measured at LP (app 4 meters from speakers) looks:

Both mains are at the same distance from the LP so they are phase aligned to each other, but you can still see 3 problematic areas (circled red) where their phase is far from being equal to each other and that will cause dips in their joint response. That difference is coming from their different position in room and there is nothing much that can be done. I can of course hammer the phase to be linear but that will cause serious pre-ringing which will be audible. The only solution to that problem is to search for better position of speakers, which in my case is not the option. Here I'm showing phase response before correction - I have managed to make it better with some mild phase correction but the joined response is sitll showing some dips, luckilly for me not that deep to cause trouble.

With this example I wanted to show that it is not only the room modes that cause issues with sub summation, but also the phase mismatch that comes from problematic position. For that reason, I cannot agree with @Duke more when he emphasized the importance of finding a good position for the speakers, no matter if they are mains or subs. As if you don't do that be aware that no gain and/or phase adjustment, nor any DSP trick can make it right.

 

[echopraxia]

I don't see this as off topic. Integrating the subs well is an important aspect of the test, imo.

Yeah I‘m thinking the off topic reminder was probably referring to some arguments getting heated and a little too personal quite a few posts prior, so I’m guessing nothing to do with @QMuse post that just unluckily happened to be the last post at that point.

Anyways IMO the topic of positioning and integration is definitely relevant when we are talking about how to set up these tests.

 

[Duke]

Yeah I‘m thinking the off topic reminder was probably referring to some arguments getting heated and a little too personal quite a few posts prior...

I think that was aimed at me and March Audio, though starting with his post #348 and my reply #352 I thought we were both being civil. But we were off-topic.

 

[stevenswall]

If I end up blind testing with subs (though I don’t know if we have a strong consensus either way yet on that), your advise and others would be helpful on how to correctly set it up.

Just as a sanity check, I would want to make sure I haven’t been doing it all wrong. When you mention adjusting sub level to match main, I’m a little unclear on what this means if there are large nulls and peaks — so what are you matching with what, for example?

I have usually followed a process of (1) adjust mains delay until REW measures they greatest response with the least nulls, (2) adjust sub level so that the lowest point is roughly around the SPL of the mains so there is EQ headroom, and (3) apply EQ corrections (Dirac Live via miniDSP SHD which as far as I know doesn’t seem to do the integration itself, just the EQ).

Maybe I’ve been doing it all wrong? Maybe I need to adjust sub level so the crossover slopes are more symmetric with each other rather than worrying about EQ headroom. But I also don’t know how it works if a huge boost is needed at some frequencies; does that risk clipping?

For what it's worth, I flatten the sub and the mains (Auto EQ with REW is what I mean by flatten), then adjust the flattened sub line to match with the mains. (Crossover is set as steep as the MiniDSP will go at 80hz, then I just run a full frequency sweep and turn down the woofer volume nob until things are about even.) Seems as balanced as it gets considering it's a cheap sub and some DIY speakers. Doesn't sound like two separate units anymore.

 

[DonH56]

Just to be contrary, I have owned Maggies for decades, and have almost always preferred to damp the back wave.

On-topic re. subs: The easy answer is to say that the sub crossover should be set to 80 Hz for both (assuming the Genelecs play that low) and then dial in the subs normally. That can be done independently of the mains. The caveat is that settings (delay/phase) may need to be tweaked to optimize integration with each speaker (pair). That is one reason I originally suggested testing without subs, rolling off the LF to the speakers, so you could compare the range in which they would be used and assume the sub(s) would augment either pair in the same way.

There's no hard and fast answer. If you have the ability to load multiple setups in your processor/AVR/miniDSP/whatever, you could switch from one to another, so each pair of speakers is optimized with the sub. That would probably obviate instantaneous switching, natch. Certainly the easiest thing is to set them up and see if sub integration is close enough, or a reasonable compromise can be made, to allow fast switching without having to change the sub setup.

 

[echopraxia]

There's no hard and fast answer. If you have the ability to load multiple setups in your processor/AVR/miniDSP/whatever, you could switch from one to another, so each pair of speakers is optimized with the sub. That would probably obviate instantaneous switching, natch. Certainly the easiest thing is to set them up and see if sub integration is close enough, or a reasonable compromise can be made, to allow fast switching without having to change the sub setup.

The miniDSP SHD allows me to do this, and I would likely be doing switching via this system anyway. It’s the only reliable way I can easily switch and also ensure the right levels and bass EQ are being applied. Only downside maybe it is takes a few seconds to switch. But it is extremely easy: just one button on a remote.

The easy answer is to say that the sub crossover should be set to 80 Hz for both (assuming the Genelecs play that low) and then dial in the subs normally.

The Genelecs extend down to around 30hz so this shouldn’t be a problem. I’ve measured the Genelecs putting out 105db at 40hz, though the Salon2’s naturally do so at less distortion at extreme SPL (otherwise, the Genelecs have slightly lower distortion overall). In any case, I think these both could easily handle any normal crossover.

I do like the idea of just rolling off the bass response though once room modes start to enter the picture. This combined with mono listening might make it much easier to distinguish the favorite.

I am probably going to plan for some stereo + subs tests as well, to be as complete as possible. (And, because I want to try some new ideas like a phantom center image spatial quality test via dual mono playback.)

 

[BenB]

I do like the idea of just rolling off the bass response though once room modes start to enter the picture. This combined with mono listening might make it much easier to distinguish the favorite.

With the bass rolled off, even a flat speaker may be perceived as bright. Bass tends to smooth out the sound and diminish the perception of any harshness from the speaker. The speaker that sounds like it has a better spectral balance without bass may not be the one that has the better spectral balance with bass. If I were doing the test, I'd do my best to get the subs integrated in a way that's fair and balanced. I'd also do mono and stereo tests if time permitted.

 

[QMuse]

If I end up blind testing with subs (though I don’t know if we have a strong consensus either way yet on that), your advise and others would be helpful on how to correctly set it up.

Sure, I'll glady help.

Btw, IMO without subs you are risking that mighty Salons outweight the Genelecs in LF response in an audible way so listeners will be able to immediatley tell them apart. For that reason you may want to consider scenario where you introduce subs crossed at say 40Hz to put both speakerrs on equal ground considering lowest LF reach but to still allow for some difference in LF response above it which can contribute to realistic evaluation.

 

[echopraxia]

Sure, I'll glady help.

Btw, IMO without subs you are risking that mighty Salons outweight the Genelecs in LF response in an audible way so listeners will be able to immediatley tell them apart. For that reason you may want to consider scenario where you introduce subs crossed at say 40Hz to put both speakerrs on equal ground considering lowest LF reach but to still allow for some difference in LF response above it which can contribute to realistic evaluation.

Yeah, the Salon2's bass response is definitely 'unfair' and from the in-room frequency response, it certainly looks like it doesn't need a sub (SPL headroom concerns aside)

Salon2 Full Range In-Room Response:

And distortion:

BTW I am curious to get thoughts here about the below phase plots, and which crossover is best for this room: Which of the following crossovers are expected to be the best starting point, assuming that I will be EQ'ing the bass response down via Dirac once I set the crossover and delay parameters? Note: In all the below examples, I found a 7ms delay on the mains to give the best (loudest, smoothest) bass response, so that's what I'm using here.

Salon2 + 80hz crossover, 24db/octave (via Linkwitz-Riley filters):

Salon2 + 80hz crossover, 48db/octave (via Linkwitz-Riley filters):

Salon2 + 60hz crossover, 24db/octave (via Linkwitz-Riley filters):

Salon2 + 60hz crossover, 48db/octave (via Linkwitz-Riley filters):

Example of results after Dirac EQ is run for the Salon2 + 80hz 24db/octave LR configuration:

 

[echopraxia]

I pushed the Salon2's up to quite loud levels (very slowly and carefully increasing each time with distortion measurements). Here are the results where I stopped since I got a little nervous to the woofer excursion -- even though THD still looks surprisingly good (not really any worse than my dual 18" subwoofers)!

Salon2 ~110db bass test:

Same test with subs crossed at 80hz with 24db/octave LR filters:

Same test with subs crossed at 60hz with 24db/octave LR filters:

Same test with subs crossed at 60hz with 48db/octave LR filters:

 

[bigjacko]

I pushed the Salon2's up to quite loud levels (very slowly and carefully increasing each time with distortion measurements). Here are the results where I stopped since I got a little nervous to the woofer excursion -- even though THD still looks surprisingly good (not really any worse than my dual 18" subwoofers)!

Salon2 ~110db bass test:

View attachment 80109

Same test with subs crossed at 80hz with 24db/octave LR filters:

View attachment 80110

Same test with subs crossed at 60hz with 24db/octave LR filters:

View attachment 80111

Same test with subs crossed at 60hz with 48db/octave LR filters:

View attachment 80112

So subwoofer make even more distortion than those woofers on salon 2? That is pretty crazy, what are the woofers used on salon 2?

 

[echopraxia]

So subwoofer make even more distortion than those woofers on salon 2? That is pretty crazy, what are the woofers used on salon 2?

In a stereo pair, a total of 6x 8” titanium cone woofers:

 

[bigjacko]

In a stereo pair, a total of 6x 8” titanium cone woofers:

View attachment 80115

Nice speaker porn, I love it haha. Do all three woofers on salon 2 play the same content from sub bass to mid range? I thought they would of used 2.5 way to reduce IMD and Doppler effect, but anyway the performance is good enough on this configuration.

 

[echopraxia]

Nice speaker porn, I love it haha. Do all three woofers on salon 2 play the same content from sub bass to mid range? I thought they would of used 2.5 way to reduce IMD and Doppler effect, but anyway the performance is good enough on this configuration.

It's a 4-way speaker, with crossovers at 150hz, 575hz, and 2.3khz.

 

[bigjacko]

It's a 4-way speaker, with crossovers at 150hz, 575hz, and 2.3khz.

Oh I see, so three 8" crossover at 150Hz, that's why it can go that low with low distortion.

 

[echopraxia]

Oh I see, so three 8" crossover at 150Hz, that's why it can go that low with low distortion.

Yeah between the stereo pair, it’s about 30% more woofer surface area than two 12” subs, or would be roughly equivalent surface area to a single hypothetical 20” sub. Not sure about excursion limits though, but I’d anticipate much less than a large diameter subwoofer.

It’s nice to run with subs anyway though for extra headroom, if only because my subs are have active protection limiters, and it makes sense to send to them the frequencies otherwise most likely to clip the mains amp in extreme cases.

My subs together can exceed 120db long term continuous output. Even if the Salon2’s were physically capable of this, the amp power required to achieve it alone would cost more than the price of both my subs combined.

There are some high dynamic range orchestral tracks which if played loudly contain bass drum hits that will clip almost any sub less capable than the dual 18” subs I have. I don’t know if it would clip the amps I have or bottom out the Salon2’s but I’d rather not find that it does the hard way.