I didn't come across any Dolby papers on the subject. Might be some. I was originally thinking of a Dolby or Hafler type matrix to EQ only what is common between left and right - I.E. the center, then mix it back in, That could work.
But the phase shuffler is different, apparently getting to the root of the problem. Or at least that's the hope.
But the phase shuffler is different, apparently getting to the root of the problem. Or at least that's the hope.
As a music producer, the stereo spectrum is generally mixed starting with the bass sounds in the middle (mono) and gradually expanding, with the high frequencies being the widest. Maybe if you have a bass sound muddying other mono sounds (like vocals), you get this problem.
Sent from my Nexus 4 using Tapatalk
Sent from my Nexus 4 using Tapatalk
My trepidation with this type of correction is how do you know what the engineer has done at the recording level? As Bogden has just stated it seems that there are as many ways to make a recording as we can think up. So what would happen if an engineer already did some of this phase shifting into the recording to create a stronger center phantom image? How would you know?
You don't know, but I've found it doesn't matter. Almost everything I listened to had the tonal shift from center to side. As I said, if the sounds never move, you won't notice it. But if they are panned, you will.
The main problem I found was that once the center was a little brighter, that threw off the overall tonal balance. The room curve had to be modified a bit to bring the tonal balance back to normal. But the center and sides now matched beautifully, it was just that now the sides where darkened a bit.
I think this is unavoidable, if you brighten the center, the overall balance will be wrong, no matter what method you use. What do you think?
The main problem I found was that once the center was a little brighter, that threw off the overall tonal balance. The room curve had to be modified a bit to bring the tonal balance back to normal. But the center and sides now matched beautifully, it was just that now the sides where darkened a bit.
I think this is unavoidable, if you brighten the center, the overall balance will be wrong, no matter what method you use. What do you think?
I disagree, tho one would think that's the case. You could think of it not as the center being too dull, but the sides being too bright. At least that's what I heard and what I tried to fix.
Most of the mixes I've heard don't have the same tonal balance when something is panned away from center. Voice, guitar, sax. Maybe they ought to, but they don't. If they did, I wouldn't be in search of a fix!
I'll post a couple more files for listening. If you don't hear the tonal shift on the unaltered files, that's great, you need not worry. But if you do hear it, let me know if the phase shuffler corrects it.
Thanks for your indulgence and your patience!
Most of the mixes I've heard don't have the same tonal balance when something is panned away from center. Voice, guitar, sax. Maybe they ought to, but they don't. If they did, I wouldn't be in search of a fix!
I'll post a couple more files for listening. If you don't hear the tonal shift on the unaltered files, that's great, you need not worry. But if you do hear it, let me know if the phase shuffler corrects it.
Thanks for your indulgence and your patience!
More files
Here are some more files for testing if you don't have a convolver.
Shuffler - Google Drive
There you will find 2 versions of two tracks. This is speech, to make it easier to notice. One version is straight up, the other is shuffled above 900Hz
On the Ralph Spoilsport Motors track both voices move around. Do you hear a tonal difference when they go from side to center?
On the Time Machine track, the voices don't move, but different voices are left, center and right. Do you hear a different tonality to any of them between the normal and shuffled versions?
Thanks for checking this out.
Here are some more files for testing if you don't have a convolver.
Shuffler - Google Drive
There you will find 2 versions of two tracks. This is speech, to make it easier to notice. One version is straight up, the other is shuffled above 900Hz
On the Ralph Spoilsport Motors track both voices move around. Do you hear a tonal difference when they go from side to center?
On the Time Machine track, the voices don't move, but different voices are left, center and right. Do you hear a different tonality to any of them between the normal and shuffled versions?
Thanks for checking this out.
Sad to see no other takers here on this subject. Just got back and checked the tread and I'm surprised to see there was no further interest in this subject.
I've briefly experimented with using EQ on the mid signal only and it definitely makes a difference. As I am using line arrays, as said before, kings of comb filtering, my results will probably not be representative for other types of speakers.
I'm using DRC processing to EQ the lines and it works wonders on the overall sound. Yet the phantom centre is less defined or tonally different from the sides. I compensated for that by using a longer processing window in DRC for the highest frequencies. But maybe I can loosen up the settings there if I EQ the mid signal separately. My first experiment seemed to work. As said, the phase shuffle seemed to affect the sides too much in my specific case. Collapsing the sound back to the speakers while without it the sound goes way beyond the boundaries of the speaker placement.
Though it did work to make a tonal improvement on the phantom centre.
It seems the higher frequencies are lacking/falling behind (in my case) in the phantom centre. More so with DRC processing than using standard IIR EQ. I'll experiment some more and report back if there is some interest in the results.
I've briefly experimented with using EQ on the mid signal only and it definitely makes a difference. As I am using line arrays, as said before, kings of comb filtering, my results will probably not be representative for other types of speakers.
I'm using DRC processing to EQ the lines and it works wonders on the overall sound. Yet the phantom centre is less defined or tonally different from the sides. I compensated for that by using a longer processing window in DRC for the highest frequencies. But maybe I can loosen up the settings there if I EQ the mid signal separately. My first experiment seemed to work. As said, the phase shuffle seemed to affect the sides too much in my specific case. Collapsing the sound back to the speakers while without it the sound goes way beyond the boundaries of the speaker placement.
Though it did work to make a tonal improvement on the phantom centre.
It seems the higher frequencies are lacking/falling behind (in my case) in the phantom centre. More so with DRC processing than using standard IIR EQ. I'll experiment some more and report back if there is some interest in the results.
Last edited:
If you have 2 speakers in the same listening plane running in mono, you're going to have the same problem..
Think of it as 2 radios transmitting on the same frequency. What happens when the two mix? Harmonics.. Cancellations.. In the FM mode, it sounds like 2 cats fighting in a back alley on a single receiver..
The stereo idea might have made music sound magic back in the day but opened up another can of worms..
Think of it as 2 radios transmitting on the same frequency. What happens when the two mix? Harmonics.. Cancellations.. In the FM mode, it sounds like 2 cats fighting in a back alley on a single receiver..
The stereo idea might have made music sound magic back in the day but opened up another can of worms..
the quality of phantom image will be mostly due to good phase response of the speaker. Time alignment of the drivers, time domain, phase alignement, ect
most speaker have some kind of phase problem imo.
I used to think phantom center problems was also due to amplifiers, but ive realized that when a speaker doesnt image perfectly, its often not caused by amps but really with the driver integration.
Also, directivity is probably very important. my amphion are the best at phantom center and is a controlled directivity speaker. pin point imaging in the phantom center is just amazing and natural when done perfectly right without any ''darkness'' of the sound.
Some speaker will also somtime image quite well with good phantom center, but have the tendency to sound thin in the phantom center. this is again due to phasing response.
Id suggest to measure the phase response of the speaker.
''It is easy to get "fooled" by a big image if the phase is off. This usually results in the phantom center being pulled to the sides and creating a "wider" image for the listener. These small delays in phase is causing the insight of the material to be blurry and undefined but "bigger" to the untrained ear. The phase anomalies is usually only measured in a few parts of the spectrum and is therefore only affecting these frequencies. If the phase response shows anomalies in the 200-500hz region you will probably have trouble separating piano , guitars, vocals , snares etc in that area . If one area is affected you can be sure another is affected when moving your head around making it even harder to take decisions when leaning over to your favourite compressor.
To be clear.. The Amphions shows excellent phase response in the entire spectrum and this contributes to the fantastic imaging , the big sweetspot and the stable audio response when moving around my desk. The 3D insight is very clear and separated. It is still very natural because the phantom center is not pushed in the middle like I have heard on many other speakers that has a deep 3d image. Normally a very deep image results in mixing phantom center sources like vocals, kick, snare and bass to loud because this area of the image is pushed back by the speakers and then compensated by the mixer. This is not the case with the Amphions. Despite the very high level of separation and depth the phantom center stays very true to the source. This again reflects back to the accurate phase response. I can not emphasise the importance of this feature strong enough''
most speaker have some kind of phase problem imo.
I used to think phantom center problems was also due to amplifiers, but ive realized that when a speaker doesnt image perfectly, its often not caused by amps but really with the driver integration.
Also, directivity is probably very important. my amphion are the best at phantom center and is a controlled directivity speaker. pin point imaging in the phantom center is just amazing and natural when done perfectly right without any ''darkness'' of the sound.
Some speaker will also somtime image quite well with good phantom center, but have the tendency to sound thin in the phantom center. this is again due to phasing response.
Id suggest to measure the phase response of the speaker.
''It is easy to get "fooled" by a big image if the phase is off. This usually results in the phantom center being pulled to the sides and creating a "wider" image for the listener. These small delays in phase is causing the insight of the material to be blurry and undefined but "bigger" to the untrained ear. The phase anomalies is usually only measured in a few parts of the spectrum and is therefore only affecting these frequencies. If the phase response shows anomalies in the 200-500hz region you will probably have trouble separating piano , guitars, vocals , snares etc in that area . If one area is affected you can be sure another is affected when moving your head around making it even harder to take decisions when leaning over to your favourite compressor.
To be clear.. The Amphions shows excellent phase response in the entire spectrum and this contributes to the fantastic imaging , the big sweetspot and the stable audio response when moving around my desk. The 3D insight is very clear and separated. It is still very natural because the phantom center is not pushed in the middle like I have heard on many other speakers that has a deep 3d image. Normally a very deep image results in mixing phantom center sources like vocals, kick, snare and bass to loud because this area of the image is pushed back by the speakers and then compensated by the mixer. This is not the case with the Amphions. Despite the very high level of separation and depth the phantom center stays very true to the source. This again reflects back to the accurate phase response. I can not emphasise the importance of this feature strong enough''
Last edited:
Here's an example of my speaker FR and Phase (grey is generated minimum phase)
(at the listening position, 3m from the speaker)
So based on your presumption I should have no problems at all? Well, the problem isn't big or huge. Re-read the original first post in this thread.
I can relate to that text and notice something similar. It's not about imaging (which my speakers do quite well I'd say) but I do spot a tonal difference between the phantom centre and the sides. Hence I took the test.
Listen to the test tracks and judge for yourself
. Should be a fun test on your Amphion speakers. Care to show a FR/phase plot of those? Imaging and a phantom centre will work if the left and right speaker act about the same in their phase behaviour. I went just a tiny bit further to get the timing as right as I could. All with the help of processing.
my intuition is that no processing is needed if the speaker is right. Also, my intuition is that the room plays a immense role in the tonal balance of the phantom center. Phase response and controlled directivity combined with a not ''too bright'' room because of first reflection treatment, bass traps, ect should help immensly (it sure help for me). I would try to fix the speaker or the room before playing with processing.
I will however try the test this week when im home but I dont experience tonal balance difference in the phantom center in my room. however, im listening 6 feet away from the speaker in my treated room.
https://www.gearslutz.com/board/11125828-post1977.html
look at the end of the post, you can find many famous speakers and their measured phase response.
I will however try the test this week when im home but I dont experience tonal balance difference in the phantom center in my room. however, im listening 6 feet away from the speaker in my treated room.
https://www.gearslutz.com/board/11125828-post1977.html
look at the end of the post, you can find many famous speakers and their measured phase response.
Last edited:
Drop by my thread: http://www.diyaudio.com/forums/full-range/242171-making-two-towers-25-driver-full-range-line-array.html
and you'll at least read I tried my best to make sure the speaker is doing well enough before 'fixing it' with processing. I'm not against processing however, it's part of the design in my case. So we don't share the same intuition. I could even argue with you that no speaker could be right without processing as crossovers are a form of manipulation as well. I have no crossover. Though I am testing a zobel network to compensate for the electrical phase.
I did 'fix' the biggest room problems to achieve my results. And I'm not complaining about those results. Just trying to learn a bit more and hopefully even get better results in the process.
The above plot couldn't/wouldn't happen with all the processing in the world if you didn't fix the room and speakers as best you could.(*) I'm not saying it's perfect though. But I'm just one of many on here to get it as perfect as I can.
One goal for me is time coherency. Look for that in a speaker and you'll only find a hand full that deliver. I can add my speakers to that list (it does at least as good as the electrostatic speaker often used on Stereophile as a reference for time coherency if I change from FIR processing to IIR. But it does way better than it with the FIR processing).
(*) at least not with the relatively short windowing used by me, striving for the least amount of fixing by making the speaker/environment better if possible
But all this does not relate to this thread's subject. I only mentioned my speaker being different from the mainstream because that might make them less of a reference to the subject at hand. Especially when I know the OP had experience with big Altec horn based speakers and noticing what I think I notice as well. Totally different speakers, though we could argue the multicell upper horns could have comb problems of their own? Basically acting like multiple smaller horns grouped together?
Pano, do you experience this on your OB cave speakers as well or was this based on the Altec's.
and you'll at least read I tried my best to make sure the speaker is doing well enough before 'fixing it' with processing. I'm not against processing however, it's part of the design in my case. So we don't share the same intuition
. I could even argue with you that no speaker could be right without processing as crossovers are a form of manipulation as well. I have no crossover. Though I am testing a zobel network to compensate for the electrical phase.I did 'fix' the biggest room problems to achieve my results. And I'm not complaining about those results. Just trying to learn a bit more and hopefully even get better results in the process.
The above plot couldn't/wouldn't happen with all the processing in the world if you didn't fix the room and speakers as best you could.(*) I'm not saying it's perfect though. But I'm just one of many on here to get it as perfect as I can.
One goal for me is time coherency. Look for that in a speaker and you'll only find a hand full that deliver. I can add my speakers to that list (it does at least as good as the electrostatic speaker often used on Stereophile as a reference for time coherency if I change from FIR processing to IIR. But it does way better than it with the FIR processing).
(*) at least not with the relatively short windowing used by me, striving for the least amount of fixing by making the speaker/environment better if possible
But all this does not relate to this thread's subject. I only mentioned my speaker being different from the mainstream because that might make them less of a reference to the subject at hand. Especially when I know the OP had experience with big Altec horn based speakers and noticing what I think I notice as well. Totally different speakers, though we could argue the multicell upper horns could have comb problems of their own? Basically acting like multiple smaller horns grouped together?
Pano, do you experience this on your OB cave speakers as well or was this based on the Altec's.
Last edited: