• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Placing speakers on top of subwoofers

DonH56

Major Contributor
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
2,805
Likes
3,099
Location
Monument, CO
The problem is usually that the best place for the mains is rarely the best place for the subs due to things like room modes and SBIR. and most full-range speakers distort badly when presented with large deep bass signals. I prefer to let subs do the job. Plus I get an extra octave-plus of bass beyond even what my Salon2's are rated.
 

andreasmaaan

Major Contributor
Patreon Donor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,701
I think the added complexity in this case comes from the fact that the crossover point will be 200-250Hz. Integrating subs in the 60-100Hz range is arguably easier than in the lower midrange, at which point you're doing something more akin to designing the bottom end of a 3-way loudspeaker.
 

stunta

Addicted to Fun and Learning
Joined
Jan 1, 2018
Messages
618
Likes
388
Location
Boston, MA
I think the added complexity in this case comes from the fact that the crossover point will be 200-250Hz. Integrating subs in the 60-100Hz range is arguably easier than in the lower midrange, at which point you're doing something more akin to designing the bottom end of a 3-way loudspeaker.
If I think of the FM8s as loudspeakers and integrate them with the REL sub, is it really that different? I can then repeat the process with my ATCs as mains and the FM8s+REL as the sub.

What is it specifically about the higher frequency range that makes things harder? I suppose I will find out on my own soon enough but I best be prepared mentally :)
 

andreasmaaan

Major Contributor
Patreon Donor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,701
If I think of the FM8s as loudspeakers and integrate them with the REL sub, is it really that different? I can then repeat the process with my ATCs as mains and the FM8s+REL as the sub.

What is it specifically about the higher frequency range that makes things harder? I suppose I will find out on my own soon enough but I best be prepared mentally :)
In lower frequency ranges the room dominates more and getting the drivers to integrate is as much about integrating with the room as with each other. Above around 200Hz (depending on the room) this is less the case. Also, the wavelengths are shorter as frequency increases, so things like phase and position in space become more important. And finally, the ears are more sensitive in this frequency range than in the bass.
 

graz_lag

Addicted to Fun and Learning
Joined
Oct 13, 2018
Messages
845
Likes
648
Location
Le Mans, France

stunta

Addicted to Fun and Learning
Joined
Jan 1, 2018
Messages
618
Likes
388
Location
Boston, MA
I got very lucky. My Integra processor allows for crossover frequency to be set as high as 200 Hz. The Classe preamp only allows up to 140 Hz.

I am still tweaking things but I went from 80 Hz xover with the RELs to 120 Hz with the dual Rythmiks and I am now at 200 Hz with the Rythmiks. I have disabled the LPF on the Integra since it maxes out at 120 Hz (which is weird given the HPF on the speakers goes till 200 Hz) and set the x-over on the subs to 200 Hz with a 24 db/octave slope. Next, I will try with 12db since from what I am reading most HT receivers/processors implement a 12 db/octave slope. Once I get this sounding right to my ears, I will take some measurements and fine tune from there.

So far, things seem to be heading in the right direction but the room is severely in need of some treatment; I can't listen for more than 15-20 minutes as its just too bright and reflective.

EDIT: Integra's documentation is quite lacking. It doesn't say much about the crossover implementation and does not offer any help with subwoofer integration. The EQ parameters are quite limited (no way to set Q). For now, I have let AccuEQ (Onkyo's proprietary version of Audyssey) calibrate things I don't see any EQ paramters set; it did calculate speaker distances and applied a filter for standing wave. Another area where documentation is sorely lacking. It is also not clear if analog sources will have the EQ applied.
 
Last edited:

DonH56

Major Contributor
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
2,805
Likes
3,099
Location
Monument, CO
The 120 Hz might be the LFE low-pass filter, not the main bass-management LPF? The LFE channel is limited to 120 Hz by design...

Edit: Here is a link to Dolby's explanations: https://www.dolby.com/uploadedFiles/Assets/US/Doc/Professional/38_LFE.pdf IIRC the LFE channel is sampled at 240 Hz and only 8 bits, with a 10 dB de-emphasis so you need to boost by 10 dB when summing back into the subwoofer or main (large only) channels. Some folk have told me the rate and resolution may have increased but it is not something I have researched (haven't cared enough to dig).
 

stunta

Addicted to Fun and Learning
Joined
Jan 1, 2018
Messages
618
Likes
388
Location
Boston, MA
Thanks Don. I am going to call Integra and will get this confirmed among other things.
Oh dear, Integra tech support couldn't answer most of my questions. I will have to email their engineers in Japan. But I did confirm that LFE setting is as Don pointed out - its for that special signal encoded into the stream.

Questions I couldn't get answered:
1. Slope on the crossover filters (I am pretty sure they use 12 db since that appears to be the standard)
2. AccuEQ did not apply any filters. Why?

A few more that were specific to my use case.
 

andreasmaaan

Major Contributor
Patreon Donor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,701
Oh dear, Integra tech support couldn't answer most of my questions. I will have to email their engineers in Japan. But I did confirm that LFE setting is as Don pointed out - its for that special signal encoded into the stream.

Questions I couldn't get answered:
1. Slope on the crossover filters (I am pretty sure they use 12 db since that appears to be the standard)
2. AccuEQ did not apply any filters. Why?

A few more that were specific to my use case.
You could always take a punt on a few settings, measure, then take it from there...
 

Sal1950

Major Contributor
The Chicago Crusher
Joined
Mar 1, 2016
Messages
5,375
Likes
2,699
Location
Central Fl

DonH56

Major Contributor
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
2,805
Likes
3,099
Location
Monument, CO
I read it as "take a pint" and went for a drink. :)

In my part of the USA, when we "punt" something we give up and move on with the (game, analysis, drinking, whatever).
 
Last edited:

stunta

Addicted to Fun and Learning
Joined
Jan 1, 2018
Messages
618
Likes
388
Location
Boston, MA
Here is the REW MDAT file for my latest round of measurements. I've applied the filters as shown below and there is a slight but noticeable improvement in the bottom end.

I looked at the waterfall plot and can't understand why some frequencies are missing. Any thoughts?

waterfall.JPG
 

andreasmaaan

Major Contributor
Patreon Donor
Joined
Jun 19, 2018
Messages
3,517
Likes
2,701
Getting there :)

You need to change the scale for the vertical axis. ATM you only have 15dB between the highest peak and the bottom of the graph. Try to get that to at least 40dB or so.

Also, how and where are you taking this measurement?
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
6,870
Likes
3,044
Location
Riverview, FL
I looked at the waterfall plot and can't understand why some frequencies are missing. Any thoughts?
Graph Axis Limits

1543893424307.png


Change to lower value

1543893496792.png
 

stunta

Addicted to Fun and Learning
Joined
Jan 1, 2018
Messages
618
Likes
388
Location
Boston, MA
Thanks. I find REW defaults confusing. I am not clear on how to interpret this waterfall graph. It looks like the decay times are lower for higher frequencies. What is bad here and what is good? Given the flutter and echo issues in my basement, I would have expected the mid to high frequencies to have a long decay time. Perhaps they do, but not as much relative to the lower frequencies?

I am getting some acoustic panels next week for the back wall and I am hoping to do a before/after measurement so it would be good to understand what I am measuring.

REW documentation is a bit too advanced for me at this point. I would prefer a simpler step-by-step for dummies that may not get me close to 100% but if it improves things noticeably, that would be a good step forward without spending hundreds of hours learning this stuff.

Also, how and where are you taking this measurement?
Using a UMIK-1 mic + REW on my laptop. Mic is placed at the listening position pointed up with the corresponding 90 deg calibration file loaded. The only reason I am pointing it up is because I haven't yet found a way to point it forward without it tipping over at the place i am setting it. I am hoping this won't make a big difference.

Meanwhile, the Integra processor has a bi-amp mode, so I connected my REL sub to one of amp/speaker outputs (set to full range) used for biamping, set the crossover to around 50 Hz and switched the Rythmiks to mid-bass (50 Hz - 250 Hz). With this, I am finally getting the dynamics I was hoping for!
 

Soniclife

Major Contributor
Joined
Apr 13, 2017
Messages
1,335
Likes
831
Location
UK
I think that rt60 might be more useful for your room decay problems, easier to understand, it's in rew.
 
Top Bottom