For me, it is a lot of fun my son is getting more interested in the results I got. Every time the two of us have the house to ourselves we play a movie after which we play a couple of songs, loudly (but not too loud 
, set at ~87 dB average). He's really into the Stereo effect. The fact that you can't 'hear' where the speakers are. The choices the artists have made in placement or as effects.
He's become a huge Queen fan after the "Bohemian" movie and we play the old and new, like "The March of the Black Queen" which is real fun as a Stereo composition.
So he's opening up to bands like "Pink Floyd", he gets into the 80's rock (even without me pushing that genre, he looks up everything on Youtube)... He definitely likes "Infected Mushroom". He's really discovering music, it makes me remember my own trip down that lane. I don't push anything, he needs to discover all this for himself. Ever since he was about 6 years old he was into "ACDC". While I do like some of their songs, that's never been "my thing". He just learned about it trough Youtube and games.
He's 12 years old now and I love to see his reaction listening to his personal favourites and discovering new songs. A new world opens up for him. Not a bad thing to have for us, to share this passion for music, even if we don't always like the same stuff...
, set at ~87 dB average). He's really into the Stereo effect. The fact that you can't 'hear' where the speakers are. The choices the artists have made in placement or as effects.He's become a huge Queen fan after the "Bohemian" movie and we play the old and new, like "The March of the Black Queen" which is real fun as a Stereo composition.
So he's opening up to bands like "Pink Floyd", he gets into the 80's rock (even without me pushing that genre, he looks up everything on Youtube)... He definitely likes "Infected Mushroom". He's really discovering music, it makes me remember my own trip down that lane. I don't push anything, he needs to discover all this for himself. Ever since he was about 6 years old he was into "ACDC". While I do like some of their songs, that's never been "my thing". He just learned about it trough Youtube and games.
He's 12 years old now and I love to see his reaction listening to his personal favourites and discovering new songs. A new world opens up for him. Not a bad thing to have for us, to share this passion for music, even if we don't always like the same stuff...
May I suggest Genesis, Peter Gabriel and Rush for your son?
I grew up on Genesis.... Rush was an acquired taste later... but now I can't get enough.
On that subject, Rush - Beyond the Lighted Stage is probably one of the best Rock documentary of all time. That along with Sound City. Both amazing rock documentaries.
And Peter Gabriel's Secret World DVD a great show to watch.
I grew up on Genesis.... Rush was an acquired taste later... but now I can't get enough.
On that subject, Rush - Beyond the Lighted Stage is probably one of the best Rock documentary of all time. That along with Sound City. Both amazing rock documentaries.
And Peter Gabriel's Secret World DVD a great show to watch.
Another week of listening tests... and a different conclusion!
Somehow I don't seem to like the x-talk compensation in the long run. I keep adjusting it downwards until there's hardly anything left to hear. While it really brightens up the upper mids, to be more inline with the perceived sides, something bugs me.
So if I want to compensate for this cross talk problem of Stereo speakers, I'll have to find another way.
It really is impressive to listen to, at first. A subjective increase in detail. However in the long run it seems to 'hurt'. It's also subjectively louder sounding. So much that I wanted to turn down the volume. That should have been a good indicator something is "off". As I used to be able to play louder and louder and not really perceive this loudness as being awkward until trying to speak.
A fallback to my reference sound, a saved setting that does not tense me up, even after weeks/months of listening, proved successful.
On to the next idea...
Somehow I don't seem to like the x-talk compensation in the long run. I keep adjusting it downwards until there's hardly anything left to hear. While it really brightens up the upper mids, to be more inline with the perceived sides, something bugs me.
So if I want to compensate for this cross talk problem of Stereo speakers, I'll have to find another way.
It really is impressive to listen to, at first. A subjective increase in detail. However in the long run it seems to 'hurt'. It's also subjectively louder sounding. So much that I wanted to turn down the volume. That should have been a good indicator something is "off". As I used to be able to play louder and louder and not really perceive this loudness as being awkward until trying to speak.
A fallback to my reference sound, a saved setting that does not tense me up, even after weeks/months of listening, proved successful.
On to the next idea...
Last edited:
Maybe I spoke too soon... in this previous test I spoke of, I was using pointers learned from the phase shuffler out of Pano's thread. This shuffler tries to shake up the response so each ear gets a slightly different result, while the grand total of both ears is what one would need to hear.
Sorry to expres myself so badly, I just mean the sum of both ears fills up the holes that the cross talk from both speakers creates. As the cross talk which happens in the exact sweet spot is the same for both ears, this "problem" reveals itself more with a listening position that has an absence of early reflections. By shaking up the sound each ear gets, we hear the sum of both ears, similar as it would sound at positions slightly beside the sweet spot.
After typing my previous reply, I changed the x-talk signal back to just one EQ-ed negative spike for both channels timed at 0.27 ms. (applied only to phantom sounds)
I have played with this before, quite a while back (years fly by) and decided to revisit that part again, even though I had already done it quickly in the HT test I spoke about a few weeks ago.
What I heard yesterday and today gives me hope there's something to be had after all. Not quite there yet but it made me positive enough to work on it for another few days/weeks.
It's kind of hard, as I gamble here and listen, without having a clear graph that tells me I'm moving in the right direction. But the signs are positive.
Some songs clearly show me an increase in depth, while bringing the central singer forward. The shuffler and it's variations I tried almost always push the center backwards, subjectively speaking.
I can move the sides wherever I want them, just by changing a couple of variables. The phantom center is much harder to get "just right". This cross talk cancellation does seem to give me that power.
And without changing the perceived balance, like adding early and late reflections do. It's hard to put into words to describe what happens. If I hear the sides I'm completely content with what I hear. The phantom part is great etc. but lacks just that little bit of that true sense of realism that I do find in side panned sounds. Well, I hope that came out more clearly, this cross talk cancellation seems to make me able to adjust the phantom part to suit the sides. I can throw any mid/side EQ at it to try and balance both similar or the same. But it only gets me half way. Adding the cross talk might just cover the other "missing" half.
It isn't easy, but I did find me a nice foam football that might let me get a picture of it with the microphone. It will still be hard as .... to interpret the measurement, but with a bit of luck it will help me fine tune what I hear.
I will never forget my first crude cross talk cancellation experiment. I was playing a Bowie song and the crude setting made Bowie pop out and sound believable. At that time, the sides were (sounding like) a mess... It was a very weird experience that made me play with this concept for quite a while.
Just this morning I've listened to about 4 songs, 3 of them improved a lot, subjectively. One failed to impress. Tonal balance wasn't quite where I want it yet. But I am positive...
Sorry to expres myself so badly, I just mean the sum of both ears fills up the holes that the cross talk from both speakers creates. As the cross talk which happens in the exact sweet spot is the same for both ears, this "problem" reveals itself more with a listening position that has an absence of early reflections. By shaking up the sound each ear gets, we hear the sum of both ears, similar as it would sound at positions slightly beside the sweet spot.
After typing my previous reply, I changed the x-talk signal back to just one EQ-ed negative spike for both channels timed at 0.27 ms. (applied only to phantom sounds)
I have played with this before, quite a while back (years fly by) and decided to revisit that part again, even though I had already done it quickly in the HT test I spoke about a few weeks ago.
What I heard yesterday and today gives me hope there's something to be had after all. Not quite there yet but it made me positive enough to work on it for another few days/weeks.
It's kind of hard, as I gamble here and listen, without having a clear graph that tells me I'm moving in the right direction. But the signs are positive.
Some songs clearly show me an increase in depth, while bringing the central singer forward. The shuffler and it's variations I tried almost always push the center backwards, subjectively speaking.
I can move the sides wherever I want them, just by changing a couple of variables. The phantom center is much harder to get "just right". This cross talk cancellation does seem to give me that power.
And without changing the perceived balance, like adding early and late reflections do. It's hard to put into words to describe what happens. If I hear the sides I'm completely content with what I hear. The phantom part is great etc. but lacks just that little bit of that true sense of realism that I do find in side panned sounds. Well, I hope that came out more clearly, this cross talk cancellation seems to make me able to adjust the phantom part to suit the sides. I can throw any mid/side EQ at it to try and balance both similar or the same. But it only gets me half way. Adding the cross talk might just cover the other "missing" half.
It isn't easy, but I did find me a nice foam football that might let me get a picture of it with the microphone. It will still be hard as .... to interpret the measurement, but with a bit of luck it will help me fine tune what I hear.
I will never forget my first crude cross talk cancellation experiment. I was playing a Bowie song and the crude setting made Bowie pop out and sound believable. At that time, the sides were (sounding like) a mess... It was a very weird experience that made me play with this concept for quite a while.
Just this morning I've listened to about 4 songs, 3 of them improved a lot, subjectively. One failed to impress. Tonal balance wasn't quite where I want it yet. But I am positive...
Last edited:
I'm so glad I'm stubborn! Slowly but surely I feel like I'm unraveling the secrets of Stereo. So far what I'm hearing now is exactly what I've been after.
Not quite where I want to be yet, but the staging, tonality and imaging is slowly but surely clicking into place. The beauty of it is it works on almost anything I throw at it.
Don't take this as a definitive answer just yet, but I've got a feeling I'm (finally) on the right path (again). I've been here before, not always sure how I got there. There's so much room for fiddling, it's pretty hard to know or learn what does what anyways.
Time will again tell if this really is the right road to travel, if I don't succeed, will you guys come get me from here:
Hard to put into words what I'm hearing, there's a strong sense of "being there", that's for sure. There's a space created, different in each song. It's enveloping, dynamic and exiting. It lets me move along to the music, sucks me inside. The stage is wider than I've had in a long time, almost like wrapping around me, but not out of place. I've been here before, in parts and pieces. Some of the visitors certainly got a part of this.
However the more parts that fall into place, the better that place seems to get. It's still early though... But I'm loving each session.
Not quite where I want to be yet, but the staging, tonality and imaging is slowly but surely clicking into place. The beauty of it is it works on almost anything I throw at it.
Don't take this as a definitive answer just yet, but I've got a feeling I'm (finally) on the right path (again). I've been here before, not always sure how I got there
. There's so much room for fiddling, it's pretty hard to know or learn what does what anyways.Time will again tell if this really is the right road to travel, if I don't succeed, will you guys come get me from here:
Hard to put into words what I'm hearing, there's a strong sense of "being there", that's for sure. There's a space created, different in each song. It's enveloping, dynamic and exiting. It lets me move along to the music, sucks me inside. The stage is wider than I've had in a long time, almost like wrapping around me, but not out of place. I've been here before, in parts and pieces. Some of the visitors certainly got a part of this.
However the more parts that fall into place, the better that place seems to get. It's still early though... But I'm loving each session.
Reason for me to be this "exited" about it is that in previous sessions, to get to this was quite experimental and accidental. Even though I thought I knew exactly what piece or setting did what, it was more of an ever growing journey to get there, leaving it very hard to interpret what made it all gel together.
This time around I actually know exactly what it is I did to get here. Making it possible to actually be able to repeat it. The big parts and pieces aren't that difficult to get right, but these last few details are! It's much like being able to set the music free.
I'll be sure to use this weekend to document exactly what I have right now. Both in settings as well as virtual measurements. Make a "Reference copy" to fall back to and try to improve further from there. That part is always the hardest to do. Easy to get it out of whack again and even though you're sure you know what setting changed, it's quite easy to completely ruin what you had.
Where I am right now makes it exiting again to think about expanding the total volume displacement times 3. It remains to be seen if I can achieve what I want from that, but I do have high hopes to get the subs integrated and get an even better result in left/right balance (aiming for linearity of frequency and phase extending below 70 Hz for both left and right channels).
This time around I actually know exactly what it is I did to get here. Making it possible to actually be able to repeat it. The big parts and pieces aren't that difficult to get right, but these last few details are! It's much like being able to set the music free.
I'll be sure to use this weekend to document exactly what I have right now. Both in settings as well as virtual measurements. Make a "Reference copy" to fall back to and try to improve further from there. That part is always the hardest to do. Easy to get it out of whack again and even though you're sure you know what setting changed, it's quite easy to completely ruin what you had
.Where I am right now makes it exiting again to think about expanding the total volume displacement times 3. It remains to be seen if I can achieve what I want from that, but I do have high hopes to get the subs integrated and get an even better result in left/right balance (aiming for linearity of frequency and phase extending below 70 Hz for both left and right channels).
Still having fun chasing my tail with the x-talk compensation. A Stereo setup makes things more difficult than the widely adopted close speaker placement within ambiophonics etc. Think about it, a true sound coming from the left should stick to being a true sound coming from the left. However, once we start to process the phantom part, to counter-act the cross talk, you do alter the sum of left and right, meaning there's always some side info in that sum. Subtracting the part present within the sum will always have consequences too, as it leaves traces in the opposite channel. So it's another balancing act. This fact proves it will also be difficult to fight it with a real center. Unless you'd only play material mastered for a LCR setup. I must have said something along these lines before.... just trying to unload some thoughts to inspire new ideas, something like that (lol).
I do think there's much to be had with a good decent working cure, but I'm not moving my speakers to achieve it. I have lots of fun in listening sessions, which always seem to point out that this cross talk from regular Stereo somehow sits between us and really great imaging. The more you can fix it, the more a natural room or environment is presented before you within the imaging queues of songs. Even the crude mid/side EQ showed that promise.
So I'll keep at it, and if I find more time will include measurements to see if I can "find what I hear" there.
I do think there's much to be had with a good decent working cure, but I'm not moving my speakers to achieve it. I have lots of fun in listening sessions, which always seem to point out that this cross talk from regular Stereo somehow sits between us and really great imaging. The more you can fix it, the more a natural room or environment is presented before you within the imaging queues of songs. Even the crude mid/side EQ showed that promise.
So I'll keep at it, and if I find more time will include measurements to see if I can "find what I hear" there.
Last edited:
Still working on a cross talk dip, applied to the phantom (L+R) mix, while I subtract a bit of the (L-R) and (R-L) signals of that applied cross talk dip to keep the sides as is.
Sounds easy enough, right?
Currently playing around with levels of the cross talk compensation, it's frequency shape and I'm integrating that cross talk dip roughly from ~ 1000 Hz up to 7 KHz, replacing what I had in mid/side EQ earlier. I still run mid/side EQ somewhere below 1000 Hz (centred around 600 HZ) meaning a small 2.5 dB dip at 600 Hz, with a Q of around 2 (and variations there of) of the phantom while the sides get an equally shaped peak there. (a variation of that old S curve from the Phantom thread)
If I'd make the cross talk dip reach lower, I'd rob too much out of bass frequencies. There are so many variables to play with, like do I need linear phase HPF/LPF etc. The levels itself, the effects of head shading on the compensation. Balance between center and sides...
I haven't got the ideal recipe yet. Not easy doing this blind with ears only. Listening to songs where it sounds hot etc... tonality done by ear and looking for improved intelligibility. That's why I hope measurements can tell me a bit more of what's really happening. Sims only get you so far, I've run plenty of those.
Sometimes driving myself crazy, until at times, and it seems to happen more often lately, getting sucked out of critical listening mode and drowning into the music with a big smile on my face. I certainly don't mind listening to lots of music. Especially when it just sucks me right in. Makes it a worthwhile exercise even if I don't find a true answer. Even though I do hope to find it.
Sounds easy enough, right?
Currently playing around with levels of the cross talk compensation, it's frequency shape and I'm integrating that cross talk dip roughly from ~ 1000 Hz up to 7 KHz, replacing what I had in mid/side EQ earlier. I still run mid/side EQ somewhere below 1000 Hz (centred around 600 HZ) meaning a small 2.5 dB dip at 600 Hz, with a Q of around 2 (and variations there of) of the phantom while the sides get an equally shaped peak there. (a variation of that old S curve from the Phantom thread)
If I'd make the cross talk dip reach lower, I'd rob too much out of bass frequencies. There are so many variables to play with, like do I need linear phase HPF/LPF etc. The levels itself, the effects of head shading on the compensation. Balance between center and sides...
I haven't got the ideal recipe yet. Not easy doing this blind with ears only. Listening to songs where it sounds hot etc... tonality done by ear and looking for improved intelligibility. That's why I hope measurements can tell me a bit more of what's really happening. Sims only get you so far, I've run plenty of those.
Sometimes driving myself crazy, until at times, and it seems to happen more often lately, getting sucked out of critical listening mode and drowning into the music with a big smile on my face. I certainly don't mind listening to lots of music. Especially when it just sucks me right in. Makes it a worthwhile exercise even if I don't find a true answer. Even though I do hope to find it
.
Last edited:
See that's the part I don't understand. The stereo crosstalk and HRTF cause a dip or two in the phantom center - why would you add more? Seems to me that you might want to add the dip to the side signal to better match the mid.
But there must be something I'm not understanding from your posts.
The cross talk dip is a separate signal added to the phantom sum at about 0.27 ms. It is designed to block the sound "leak" from the right speaker to the left ear and also block the sound "leak" from the left speaker to the right ear. This results in less deep dips at both ears for the phantom part.
Even blocking it at one side is beneficial for perception, however it can create other imbalances. It is the fact that we perceive those dips at both ears at the same time that makes them stand out. Only happens in a room with very few early reflections.
As the phantom sum also contains L and R material, otherwise known as the sides, I subtract that part from this "counter dip" to keep the sides balance as is. That side balance is as pure as it can get. I don't want to change that. It's actually the closest thing to reality we can get with (standard) Stereo.
The blocking part of the cross talk cancellation is specifically designed to block the sound leak that wraps around your head.
Filling those dips is very beneficial for the subjective quality of the stage/imaging. It also improves intelligibility. We can't fill those dips by adding more energy, but we can change its balance with a couple of different tricks.
The shuffler caused a different sum at the left and right ear for phantom sounds, it is effective but changes the balance perceived just outside the sweet spot. The mid/side EQ shaved off the summed peaks at ~3.7 KHz and ~ 7.2 KHz and also improved the total perceived balance.
Even the added Haas kicker functions much as (late) reflections and adjusts our perceived tonal balance, filling in the dips somewhat. Getting you a more holographic soundscape with perceived depth and room for individual performers.
I find that filling those dips is much more effective than anything else I've tried. It brings the performers to life. There is room to play with levels of both phantom and sides to even out the total energy and determine the placement in staging/imaging.
While it may seem counter productive, it is the sum made in our head of the left and right signals that shapes its perception of it. The difference between a pure sound, perceived as coming from the front and it's fake Stereo representative coming at you from an angle at both left and right ear is largely responsible for our perception of it.
Helping that faked phantom and making it sound more like the real deal is the goal. Every step I've tried to get there has shown a lot of potential.
Even reflections out in the room can fill those dips somewhat. That is what creates the "they are here" perception in a lively room. I'm after the "you are there" kind of imaging and that needs to exclude those beneficial reflections.
At least, that is how I can best explain it.
I do hope it helps.... Going to fire up the speakers again for a few songs now before I have to hurry to get to my work...
Even blocking it at one side is beneficial for perception, however it can create other imbalances. It is the fact that we perceive those dips at both ears at the same time that makes them stand out. Only happens in a room with very few early reflections.
As the phantom sum also contains L and R material, otherwise known as the sides, I subtract that part from this "counter dip" to keep the sides balance as is. That side balance is as pure as it can get. I don't want to change that. It's actually the closest thing to reality we can get with (standard) Stereo.
The blocking part of the cross talk cancellation is specifically designed to block the sound leak that wraps around your head.
Filling those dips is very beneficial for the subjective quality of the stage/imaging. It also improves intelligibility. We can't fill those dips by adding more energy, but we can change its balance with a couple of different tricks.
The shuffler caused a different sum at the left and right ear for phantom sounds, it is effective but changes the balance perceived just outside the sweet spot. The mid/side EQ shaved off the summed peaks at ~3.7 KHz and ~ 7.2 KHz and also improved the total perceived balance.
Even the added Haas kicker functions much as (late) reflections and adjusts our perceived tonal balance, filling in the dips somewhat. Getting you a more holographic soundscape with perceived depth and room for individual performers.
I find that filling those dips is much more effective than anything else I've tried. It brings the performers to life. There is room to play with levels of both phantom and sides to even out the total energy and determine the placement in staging/imaging.
While it may seem counter productive, it is the sum made in our head of the left and right signals that shapes its perception of it. The difference between a pure sound, perceived as coming from the front and it's fake Stereo representative coming at you from an angle at both left and right ear is largely responsible for our perception of it.
Helping that faked phantom and making it sound more like the real deal is the goal. Every step I've tried to get there has shown a lot of potential.
Even reflections out in the room can fill those dips somewhat. That is what creates the "they are here" perception in a lively room. I'm after the "you are there" kind of imaging and that needs to exclude those beneficial reflections.
At least, that is how I can best explain it.
I do hope it helps.... Going to fire up the speakers again for a few songs now before I have to hurry to get to my work...
Last edited:
OK, thank. I can make some sense of that. You've gone much, much further with this than just about anyone I can think of in a domestic setting.
A somewhat related video popped up today, thought you might enjoy it. Walter has done a lot of good work.
Walter Murch - Film-maker - Dialogue does not sound so good in stereo - Web of Stories
A somewhat related video popped up today, thought you might enjoy it. Walter has done a lot of good work.
Walter Murch - Film-maker - Dialogue does not sound so good in stereo - Web of Stories
Thanks I guess... or does it mean I'm just nuts....
Thanks for the video btw! Very true what he says and most movies do follow that rule.
I use the movie 'Rango' a lot, even though it is an animated movie, it does have moving voices throughout the whole movie. Which makes it easier to learn if i'm on the right track with my processing in Stereo/phantom.
Even harder to get it right in Stereo than in a 5.1 setup I'd say, however movies actually do follow rules in mastering unlike music. And having the discrete channel info to work with helps too.
All I'm trying to do is find my own truth in audio. There's so much fuzziness and different opinions out there. My setup does seem to allow me to do a lot of tests, even though I'm quite aware that I don't have "the ideal" speakers nor "the ideal" room.
They are quite good at letting me hear quite subtle differences. I'm constantly amazed by what I can pick up with my ears and can prove later on with measurements and theory.
It isn't easy, but it is fun! I like the puzzle, as there are so many different answers and theories out there already, which actually makes it an open field without any true or definitive answers. I do believe that we can measure everything, if we just knew where to look for it!
It does help to have some experiments that turn out to be successful every once in a while to keep motivated.
I've often thought that if i can help one of the fellow array guys get what i have, we can compare notes and advance stuf like this together. However, most of them have a mind of their own, different expectations and/or ideas about it or simply want to do their own thing. And I can't blame them, because I'm exactly like that myself.
So that's why I do keep writing about it, so it may inspire others to experiment with what they've got. It's easier to improve what you have than to design something perfect in one go.
I had high hopes for the Phantom thread. But in the end it was just you and me carrying it. Most others just got bored with it?
or does it mean I'm just nuts.... Thanks for the video btw! Very true what he says and most movies do follow that rule.
I use the movie 'Rango' a lot, even though it is an animated movie, it does have moving voices throughout the whole movie. Which makes it easier to learn if i'm on the right track with my processing in Stereo/phantom.
Even harder to get it right in Stereo than in a 5.1 setup I'd say, however movies actually do follow rules in mastering unlike music. And having the discrete channel info to work with helps too.
All I'm trying to do is find my own truth in audio. There's so much fuzziness and different opinions out there. My setup does seem to allow me to do a lot of tests, even though I'm quite aware that I don't have "the ideal" speakers nor "the ideal" room.
They are quite good at letting me hear quite subtle differences. I'm constantly amazed by what I can pick up with my ears and can prove later on with measurements and theory.
It isn't easy, but it is fun! I like the puzzle, as there are so many different answers and theories out there already, which actually makes it an open field without any true or definitive answers. I do believe that we can measure everything, if we just knew where to look for it!
It does help to have some experiments that turn out to be successful every once in a while to keep motivated.
I've often thought that if i can help one of the fellow array guys get what i have, we can compare notes and advance stuf like this together. However, most of them have a mind of their own, different expectations and/or ideas about it or simply want to do their own thing. And I can't blame them, because I'm exactly like that myself
.So that's why I do keep writing about it, so it may inspire others to experiment with what they've got. It's easier to improve what you have than to design something perfect in one go.
I had high hopes for the Phantom thread. But in the end it was just you and me carrying it. Most others just got bored with it?
Last edited:
That's for you to decide.Thanks I guess...
Yes, well... I suspect that very few people actually notice the problem because their listening rooms aren't acoustically dead enough. It should still be a problem for dialog out of 2 speakers, but maybe no one has noticed. Tweaking the mid signal to sound like the sides (or vice versa) is kind of a top level tweak, you know. Not a lot of people have rooms that allow the problem to be heard on music. And serious HT folk should have a dedicated center channel anyway.
Keep up the good work, we do enjoy reading about it.
