Years ago I read an online article that tested just this on L-C-R speakers. They used different quality speakers, both matched and mismatched. The matched set won, even tho they were of lesser quality. (Radio Shack, IIRC).
I cannot find that anywhere, now - of course. But it's certainly interesting and should be considered.
Just yesterday this thread came up on the RePhase thread. Instead of taking that one way off topic I'll share what I got from this thread in the long run.
I've done many fun experiment, all of them pretty much based on the theory we came up with in this thread.
I've done:
- Mid/side EQ
- Play with the phase shuffler extensively, none of them bettered Pano's RePhase 2 shuffler.
- Tried to mimic the early XTC algorithms from Prof. Choueiri, as presented as convolution files on ambiophonics.org long ago.
- Altered my ambient setup after seeing Jim1961's measurements.
First, I had a lot of fun experimenting with all of the above. It was spread out over a long period as usually a little change can stand out, without being necessarily better in the long run.
At first I was struck by the shuffler, it did something I could not get with all my regular tricks. After living with it for a longer period I did start to hear some tonal imbalances between the left and right side. I documented the probable cause somewhere in this thread.
The XTC was fun too, it could do pretty much all of what the shuffler did. It kept me happy for a long time. Though there was a slightly less defined top end. A bit more diffused. I only noticed it after hearing some pro monitor speakers up close in a near field setup.
In the end, the mid/side EQ stayed. It is an excellent solution to be able to control the tonal balance differences between the phantom centre and the sides. It gives more body to vocals panned to the sides while solving part of the phantom problems.
I also kept the ambient speakers with a changed way of processing. It always consisted of (L-R) and (R-L) with a little of phantom centre (L+R) mixed in and band passed and delayed 20 ms after the mains. I've also got a hint of reverb on that mix, and aimed the ambient speakers to the sides to diffuse them of nearby borders. Played by themselves it gives the impression of a far sound coming from behind, hitting the listening spot from a lateral angle.
The change is that I routed the phantom mix to go around the reverb algorithm before mixing it in with the (L-R) and (R-L). This has made just enough difference for me compared to the other solutions I tried. It's a pleasant and open sound, wide sweet spot and good tonal balance.
In my thread there are a couple of reviews, written by visitors. I linked them in the first post of my Two Towers thread.
The first review by Jan Fekkes was without ambient channels, but with mid/side EQ.
The next one by member Boden was in the middle of my XTC experiments. It still included ambient channels and mid/side EQ.
The last two were with the altered ambient recipe. Mid/side EQ being part of this setup. As far as I can tell, the best compromise I've found to date.
Clear open sound, good tonal balance throughout and a large sweet spot.
So, how many are still running something extra, originating from or discussed in this thread?
I've done many fun experiment, all of them pretty much based on the theory we came up with in this thread.
I've done:
- Mid/side EQ
- Play with the phase shuffler extensively, none of them bettered Pano's RePhase 2 shuffler.
- Tried to mimic the early XTC algorithms from Prof. Choueiri, as presented as convolution files on ambiophonics.org long ago.
- Altered my ambient setup after seeing Jim1961's measurements.
First, I had a lot of fun experimenting with all of the above. It was spread out over a long period as usually a little change can stand out, without being necessarily better in the long run.
At first I was struck by the shuffler, it did something I could not get with all my regular tricks. After living with it for a longer period I did start to hear some tonal imbalances between the left and right side. I documented the probable cause somewhere in this thread.
The XTC was fun too, it could do pretty much all of what the shuffler did. It kept me happy for a long time. Though there was a slightly less defined top end. A bit more diffused. I only noticed it after hearing some pro monitor speakers up close in a near field setup.
In the end, the mid/side EQ stayed. It is an excellent solution to be able to control the tonal balance differences between the phantom centre and the sides. It gives more body to vocals panned to the sides while solving part of the phantom problems.
I also kept the ambient speakers with a changed way of processing. It always consisted of (L-R) and (R-L) with a little of phantom centre (L+R) mixed in and band passed and delayed 20 ms after the mains. I've also got a hint of reverb on that mix, and aimed the ambient speakers to the sides to diffuse them of nearby borders. Played by themselves it gives the impression of a far sound coming from behind, hitting the listening spot from a lateral angle.
The change is that I routed the phantom mix to go around the reverb algorithm before mixing it in with the (L-R) and (R-L). This has made just enough difference for me compared to the other solutions I tried. It's a pleasant and open sound, wide sweet spot and good tonal balance.
In my thread there are a couple of reviews, written by visitors. I linked them in the first post of my Two Towers thread.
The first review by Jan Fekkes was without ambient channels, but with mid/side EQ.
The next one by member Boden was in the middle of my XTC experiments. It still included ambient channels and mid/side EQ.
The last two were with the altered ambient recipe. Mid/side EQ being part of this setup. As far as I can tell, the best compromise I've found to date.
Clear open sound, good tonal balance throughout and a large sweet spot.
So, how many are still running something extra, originating from or discussed in this thread?
Last edited:
Phase (and time anyway, if we are in the digital domain) should be matched. I tried to get this into HobbyHifi, but Mister Timmermanns does not want to load his boat that full, after he had sealed the holes.
Dear Wesayso, M/S EQ is sure possible, but not for actual notches, unless you have your head fixed in space. Once you bend your shoulders and pitch your head to the left, notch frequency for left ear has lowered round, and notches have disappeared for reit ear, which is at center now. So all one should do is raise and lower presence level broadly, say +-3dB with Q=0.5. If you are already there, congratulations for gold!
Last edited:
Based my visit wesayso's living room couple weeks ago which included latest DSP correction steps think he is there and have gold sound in room, well on par the insane technical and costly BeoLab 90
. That said its audio and always stuff to try investigate and try out. To back up my opinion there is three more reviews than mine on his speaker build thread.First: http://www.diyaudio.com/forums/full...er-full-range-line-array-169.html#post4589772
Second: http://www.diyaudio.com/forums/full...er-full-range-line-array-243.html#post4816031
Third: http://www.diyaudio.com/forums/full...er-full-range-line-array-275.html#post4905160
Fourth: http://www.diyaudio.com/forums/full...er-full-range-line-array-277.html#post4906905
Hello Grasso,
My mid/side EQ scheme does not use high Q cuts or boosts. It is a gradual change where there's more mid energy (around 400 Hz) in the left + right and gradually less above ~ 1700 Hz compared to the phantom centre. It's not a huge deviation (around that 3 dB number). But it does help retain a more balanced tonality between the phantom centre and left and right channels over a wide area.
BYRTT, thanks for mentioning my system in the same sentence as the Beolab 90, let alone comparing it favourably.
As I haven't heard the Beolab 90 myself I have no clue how that one sounds. I do think that some of the things I do will be very familiar. I bet they do more than just beam steering with all that DSP power.
A few months ago I asked their Tonmeister Geoff to post an IR and STEP. He politely responded he would try to post it at a later date. Sadly, I haven't seen it yet.
My mid/side EQ scheme does not use high Q cuts or boosts. It is a gradual change where there's more mid energy (around 400 Hz) in the left + right and gradually less above ~ 1700 Hz compared to the phantom centre. It's not a huge deviation (around that 3 dB number). But it does help retain a more balanced tonality between the phantom centre and left and right channels over a wide area.
BYRTT, thanks for mentioning my system in the same sentence as the Beolab 90, let alone comparing it favourably.
As I haven't heard the Beolab 90 myself I have no clue how that one sounds. I do think that some of the things I do will be very familiar. I bet they do more than just beam steering with all that DSP power.
A few months ago I asked their Tonmeister Geoff to post an IR and STEP. He politely responded he would try to post it at a later date. Sadly, I haven't seen it yet.
Today, I have tested M/S EQ with 2 speakers. I also tested 3 ch, sending M signal to center speaker and S to the side speakers without EQ.
2ch M/S EQ made some improvement for the old MONO recordings, but for stereo recordings, there was a problem. It seems like some mastering engineers today has already applied the phantom center treatment with their M/S EQ sometimes, NOT ALWAYS. So additional M/S actually made things worse sometimes. All depends on the song or mastering engineer.
3 ch experiment didn't go well either. I think it is because I didn't use the same center speaker as side one. I didn't test a lot, but I guess it should have the same issue as M/S EQ 2ch treatment for stereo recordings.
To me, the winner was norman bates' post#8. A barrier. I thought it is a crazy idea, but it is easiest by far, best sounding and easily on/off. Try this with your chair cushion, you'll be surprised how effective it is.
http://www.diyaudio.com/forums/multi-way/277519-fixing-stereo-phantom-center.html#post4399039
2ch Stereo was a great idea, but it was far from a perfect idea. We should have started with 3-ch from the beginning, sigh...
2ch M/S EQ made some improvement for the old MONO recordings, but for stereo recordings, there was a problem. It seems like some mastering engineers today has already applied the phantom center treatment with their M/S EQ sometimes, NOT ALWAYS. So additional M/S actually made things worse sometimes. All depends on the song or mastering engineer.
3 ch experiment didn't go well either. I think it is because I didn't use the same center speaker as side one. I didn't test a lot, but I guess it should have the same issue as M/S EQ 2ch treatment for stereo recordings.
To me, the winner was norman bates' post#8. A barrier. I thought it is a crazy idea, but it is easiest by far, best sounding and easily on/off. Try this with your chair cushion, you'll be surprised how effective it is.
http://www.diyaudio.com/forums/multi-way/277519-fixing-stereo-phantom-center.html#post4399039
2ch Stereo was a great idea, but it was far from a perfect idea. We should have started with 3-ch from the beginning, sigh...
Last edited:
One more finding. I always feel that sitting center of the room (the same distance to the speakers), facing front center is somewhat tiring. Now I know it is due to the comb filtered phantom center illusion that the stereo system insists. Sitting a little off, facing side is much more enjoyable for music listening.
I guess this comb filter symptom is only apparent with excellent room / speakers that can project very small and clear phantom center image. I have no problem with my second system, which is far from perfect and its center image is huge. If this symptom starts annoy you, it means you're closer to the end, I mean dead end.
I guess this comb filter symptom is only apparent with excellent room / speakers that can project very small and clear phantom center image. I have no problem with my second system, which is far from perfect and its center image is huge. If this symptom starts annoy you, it means you're closer to the end, I mean dead end.
Last edited:
Hi,
I do not think so. 3-ch array stereo (by array stereo I mean at least two playback sources spanning a distance) has the same issues as 2-ch, say loss of sync causing destructive interference, just starting one octave higher, as source-to-source distance has halved.
The barrier idea is an emulation of headphones playback.
I have come to the conclusion that 2-ch stereo is a fine thing, if done rite. Rite means co-incident on both recording and playback, see my blog for details. This needs a new kind of loudspeaker, I call it the beam stereo loudspeaker, just one box for stereo reproduction.
Uli
I do not think so. 3-ch array stereo (by array stereo I mean at least two playback sources spanning a distance) has the same issues as 2-ch, say loss of sync causing destructive interference, just starting one octave higher, as source-to-source distance has halved.
The barrier idea is an emulation of headphones playback.
I have come to the conclusion that 2-ch stereo is a fine thing, if done rite. Rite means co-incident on both recording and playback, see my blog for details. This needs a new kind of loudspeaker, I call it the beam stereo loudspeaker, just one box for stereo reproduction.
Uli
hmm, this should be reasonable. I'm not sure if I understood correctly.
The similarity between the barrier and headphone is lack of comb filtering. Everything else is different.
I read this approach before. Beam stereo speaker setup was popular at a time in 70' or 80' in Japan, but it quickly disappeared. Interesting idea, though.
Do you want to listen to music with a cushion or panel rite up your nose? While dancing, socializing, xxx? Even if you do, you do not get good stereophonics that way.
With a proper beam stereo loudspeaker, you have a) mono compatibility with a clean response of first wavefront anywhere within the room, and b) decent stereo effect anywhere within the room but most balanced and strong, when the listener is centered between side walls and backed off a good deal from the loudspeaker. The stereo effect, what you call "image", does not collapse when moving around. Just the contrary, sonic impression is more stable.
Those Japanese beam stereo loudspeakers are most imperfect. My prototype is better yet far from perfect.
With a proper beam stereo loudspeaker, you have a) mono compatibility with a clean response of first wavefront anywhere within the room, and b) decent stereo effect anywhere within the room but most balanced and strong, when the listener is centered between side walls and backed off a good deal from the loudspeaker. The stereo effect, what you call "image", does not collapse when moving around. Just the contrary, sonic impression is more stable.
Those Japanese beam stereo loudspeakers are most imperfect. My prototype is better yet far from perfect.