It's an L-shaped room, with the longest side measuring 8 meters. The opposite side is 6 meters. The adjacent sides are 3.30 meters and 2.22 meters respectively. The following diagrams are to scale: 1 large square = 1 square meter.
The speakers, viewed from above, measure 76cm x 47cm.
The first layout is called Tank. I kept this setup for about 18 months. It was good, but I was never satisfied because the deep bass was unbalanced. I attribute this to the stairwell behind the left speaker.
The second layout, called Sophie, aimed to widen the stereo image. Against all logic, I placed the speakers on the longest wall. This was already a significant improvement. The image was much wider, but the deep bass, while filling the entire left channel, remained absent on the right. I still attribute this aberration to the stairwell. No toe-in yet. Measurements with REW show a uniform frequency response from 20 to 20,000 Hz. But curiously, L+R shows two nulls around 55 and 75 Hz of about ten dB each, with a smoothing effect of 1/48 or Var that doesn't appear on either L or R when measured separately.
The third trace, called Fleuve, aimed to create symmetrical areas to the left and right in front of the listening position to eliminate the effects of the first two setups. It seems to work well. The deep bass is well-balanced and fills the entire space. The nulls on the L+R measurement shifted to 70 and 85 Hz.
I stopped using a house correction curve. I used one for 18 months, and all things considered, I've been actively listening to music for 55 years, and until I started experimenting with REW, I never needed it. I didn't even know it existed. And the music sounded perfectly fine as it was. When I changed my setup by 90 degrees (Sophie), I deactivated the room correction, and lo and behold, the treble was more present and more natural.
The speakers are each positioned at 25% of the room's width, and the main listening position is at 33% of the length for the last two setups. The speakers were positioned foot by foot, initially placed at the latitude of the listening position and then moved back foot by foot until the stereo image became fully clear and the speakers disappeared as an apparent source. At its final position, the back side of the speakers are at 30cm from the front wall, therefore the diaphragm is 47cms closer to the listener.
It's a stereo system (2 chan) tri-amped. The woofers are 15 inches in bass reflex enclosures. Whish is probably the cause of the nulls on the L+R trace. The arithmetic summation of L and R gives a trace that perfectly overlaps with the L+R measurement. This looks like destructive interference. That said, I'm wondering how to correct these two nulls when the two bass speakers don't have their own channel in REW. I don't have a 6-channel setup from REW's point of view, but a 2-channel one.
Any idea / comment appreciated.
The speakers, viewed from above, measure 76cm x 47cm.
The first layout is called Tank. I kept this setup for about 18 months. It was good, but I was never satisfied because the deep bass was unbalanced. I attribute this to the stairwell behind the left speaker.
The second layout, called Sophie, aimed to widen the stereo image. Against all logic, I placed the speakers on the longest wall. This was already a significant improvement. The image was much wider, but the deep bass, while filling the entire left channel, remained absent on the right. I still attribute this aberration to the stairwell. No toe-in yet. Measurements with REW show a uniform frequency response from 20 to 20,000 Hz. But curiously, L+R shows two nulls around 55 and 75 Hz of about ten dB each, with a smoothing effect of 1/48 or Var that doesn't appear on either L or R when measured separately.
The third trace, called Fleuve, aimed to create symmetrical areas to the left and right in front of the listening position to eliminate the effects of the first two setups. It seems to work well. The deep bass is well-balanced and fills the entire space. The nulls on the L+R measurement shifted to 70 and 85 Hz.
I stopped using a house correction curve. I used one for 18 months, and all things considered, I've been actively listening to music for 55 years, and until I started experimenting with REW, I never needed it. I didn't even know it existed. And the music sounded perfectly fine as it was. When I changed my setup by 90 degrees (Sophie), I deactivated the room correction, and lo and behold, the treble was more present and more natural.
The speakers are each positioned at 25% of the room's width, and the main listening position is at 33% of the length for the last two setups. The speakers were positioned foot by foot, initially placed at the latitude of the listening position and then moved back foot by foot until the stereo image became fully clear and the speakers disappeared as an apparent source. At its final position, the back side of the speakers are at 30cm from the front wall, therefore the diaphragm is 47cms closer to the listener.
It's a stereo system (2 chan) tri-amped. The woofers are 15 inches in bass reflex enclosures. Whish is probably the cause of the nulls on the L+R trace. The arithmetic summation of L and R gives a trace that perfectly overlaps with the L+R measurement. This looks like destructive interference. That said, I'm wondering how to correct these two nulls when the two bass speakers don't have their own channel in REW. I don't have a 6-channel setup from REW's point of view, but a 2-channel one.
Any idea / comment appreciated.
