Not unless the artists playback system had this tweak in it. But I don't know any recording studios that do this. Maybe some do and some don't, but there in lies the problem does it? With no standards there is no right or wrong, but I do feel that the great producers do things right - that's why they are great.
Depends on the timing. Signals > 10 ms. delay will have little to no coloration. You use 20 ms, which is a little long IMO. If you look at the impulse responses of Gammatone filter sets you will see that 10 ms gets down to about 500 Hz without convolution of the reflection with the direct (the usual cause of coloration.) To me this is low enough (which is consistent with Griesingers conclusions.)
That time of 20 ms is chosen based on previous papers on Haas Kickers in Studio use.
It's deliberately late to let possible imaging queues in the recording come before it.
The Haas limit doesn't clearly state what would be acceptable at which frequency. The signal I use is limited at the top rolling off second order at ~5500 Hz for phantom centre.
The level is as low as I can get it while keeping the benefits. Granted these are subjective terms. But being in a real living room it's the best I can come up with to make it sound unlike that small room I am in. Hiding the room, really. Walls and speakers disappear and all that's left is the music.
As far as I know it's not used in current Studios anymore, but it's interesting to read why it's not in use anymore. Those reasons made it that much more interesting to play with in a listening room, at least for me and a few others.
It's deliberately late to let possible imaging queues in the recording come before it.
The Haas limit doesn't clearly state what would be acceptable at which frequency. The signal I use is limited at the top rolling off second order at ~5500 Hz for phantom centre.
The level is as low as I can get it while keeping the benefits. Granted these are subjective terms. But being in a real living room it's the best I can come up with to make it sound unlike that small room I am in. Hiding the room, really. Walls and speakers disappear and all that's left is the music.
As far as I know it's not used in current Studios anymore, but it's interesting to read why it's not in use anymore. Those reasons made it that much more interesting to play with in a listening room, at least for me and a few others.
Wesayso - I can appreciate your position, what you are trying to achieve and claim you do is exactly what I am trying to achieve and what I claim to as well, but mine is done without any of the "tweaks" that others are using. What you are doing is a small modification and the impulse responses that you posted were impressive, much like what I would expect mine to be. I am always leery of "tweaks" and claims thereof. So I bristle when people make these claims. The change in center channel timbre is real, no one doubts that, I just don't think that my listening room is where it should be corrected because I believe it is likely caught and corrected by a "good" producer.
As said many times in this thread, it's NOT the center tonal balance that is the problem, or something we wish to change. It's the sides that don't match the center, so they might be brought to a closer match.
If this unequal tonal balance is fixed with a shuffler, then the center will be too bright. The overall tonal balance EQ would need to be adjusted to fix that.
If the tonal imbalance is fixed with M/S EQ, then one need not touch the center channel at all, simply using EQ on the sides to bring them closer to the tonality of the phantom center.
Both approaches have their advantages and disadvantages. Those have been discussed here and are also addressed in the paper linked to in the top of the thread.
To see what I've got would depend on what one would call RFZ:
Here's a picture of RFZ:
Here's my Haas Kicker setup:
Showing a clear RFZ...
And button pressed to turn it off:
As I've read it, that 'termination' has been discussed a lot for studio use. But I do have a clear reflection free zone, or rather reflection reduced zone (it's still a living room, I'm not at -25 to -30 dB). That gap is set to 20 ms and is user adjustable.
Here's a picture of RFZ:
Here's my Haas Kicker setup:
Showing a clear RFZ...
And button pressed to turn it off:
As I've read it, that 'termination' has been discussed a lot for studio use. But I do have a clear reflection free zone, or rather reflection reduced zone (it's still a living room, I'm not at -25 to -30 dB). That gap is set to 20 ms and is user adjustable.
As I explained before, the record companies have to make their final products compatible with headphone, car audio, mono, home stereo etc. Good producer should not correct this issue, because the number of the listeners who sit very center of home stereo is very small today. If they correct, the ordinary listener today has to listen to strange sounding music. And there is no way to correct this issue at the production side anyway, as far as I know.
Last edited:
It could be done with M/S EQ tweaking the side channel to match the center. But as outlined in paper on this subject, that doesn't work for all systems and listening positions, so probably not a good idea to do in mastering. As you mention, the mix has to sound good across a wide variety of speakers and spaces. If a few audionutz like us notice an imbalance, that's our problem.
Floor to ceiling line array to take care of floor and ceiling reflections.
Yes I've looked carefully into the spectrum of reflections (and can change it).
I find that hard to believe. Although I have not measured any such speaker, I would expect a mix of reflections due to each driver. Perhaps it would be clearer looking at long impulse data.
Sent from my iPhone using Tapatalk
I'll make a clearer mental picture for you. I was wondering if I should move this to my own thread but more people might wonder how this works.
Basically you've answered the question already. Think of one driver and its reflections. Measure it with an array of microphones at differing heights. What happens? Each reflection unique for that measurement position will not show up in the next measurement, taken at a slightly different position. This is how the noise floor seems lowered because we are averaging out the reflections by summing this string of measurements.
Now we go back to the array. We have the opposite here. We now have a row of drivers, each with a unique reflection pattern compared to the next one (at different height) as seen from the microphone position (or ear). The reflections that stick out are what all the drivers have in common (big vertical planes in parallel with the array). Look at that room picture again, what you see are the arrays, two curtains that each hide a (pretty huge) damping panel behind it and a big damping panel posing as a "poster" behind the listening position. No floor or ceiling absorption is needed.
I already showed the ETC graph covering a much longer time frame and this should tell you what you asked for (the IR past 20 ms). I'll post it to satisfy your curiosity:
You could have seen the same, but more visual in my APL_TDA plot:
Get the demo version of APL_TDA and test your own setup. The above example is with L+R speakers tested simultaneously.
You can see the timing but also the reflections out in the room after that first wave front. I could tell you what each bump belongs to as I've been around the room to check
. It's almost like a map of reflections for that microphone position.If you have more questions, please jump on my thread and I'll answer there.
I don't want to derail this one.
Has anyone listened to these?
http://hiresaudiocentral.com/dirac-...like-a-hi-fi-system-and-then-theres-3d-audio/
Sent from my iPhone using Tapatalk
http://hiresaudiocentral.com/dirac-...like-a-hi-fi-system-and-then-theres-3d-audio/
Sent from my iPhone using Tapatalk
My perception of the problem with the center phantom image is that the path differential from each speaker to your ears is off one way or another, regardless of how far off center you are, which causes a comb filter effect on the center phantom image that doesn't necessarily occur with sounds coming out of only the left or right speaker (same problem as with the Carver Holographic Generator inter-aural cancellation circuit).
The thing that could fill in those cancellations in the phantom image could be more room reflections in the frequency range where that is an issue (upper mids and high treble). I wonder if adding extra horizontally displaced tweeters to each side might even things out (?). I wonder if rear firing tweeters would help this as well (?). The DBX Soundfield speaker system had a hexagon shaped tweeter tower on top that had a tweeter on each panel (6 total), shooting treble in every direction, which would cause many comb filter effects that may perceptually fill in each others cancellations... I wonder if that helped this issue (?). Some things to try anyway.
The thing that could fill in those cancellations in the phantom image could be more room reflections in the frequency range where that is an issue (upper mids and high treble). I wonder if adding extra horizontally displaced tweeters to each side might even things out (?). I wonder if rear firing tweeters would help this as well (?). The DBX Soundfield speaker system had a hexagon shaped tweeter tower on top that had a tweeter on each panel (6 total), shooting treble in every direction, which would cause many comb filter effects that may perceptually fill in each others cancellations... I wonder if that helped this issue (?). Some things to try anyway.
The most important thing to remember is: in a setup without (early) reflections the comb filter pattern at the left ear and the right ear will be the same. That's why it will be more obvious under those conditions in that position. The position that should be the best place to listen in.
In all other positions (slightly off centre) the comb dips occur at slightly different places for the left and right ear. Here the left and right ear will make up for the perceived imbalance compared to that centre position where both ears get that same dip pattern.
In other words, I did not look for a solution that needs to fix anything slightly off axis.
The phase shuffler discussed in this thread makes the comb patterns slightly different for the left and right ear at that exact sweet spot. This part works! Though off axis there are differences between left and right side. It sort of favours one off axis side. Personally, I preferred this to the unprocessed stereo setup, where both ears get the same dip pattern.
I don't want to fix the centre problems by adding (relatively) early reflections I purposely got rid of. But that may just be me. I gained too much by getting rid of those reflections, I don't want to loose or get rid of that advantage in perception.
The biggest and most obvious dip is the one at ~1750 Hz. The one at ~5500 Hz stands out less. The higher dips probably don't make a huge difference no matter what you do due to head shading.
In my room, just fixing the perceived tonal difference with later reflections consisting of the phantom centre but diffused (but within the Haas limit) was the most satisfying solution.
The sweet spot continues to work as the best spot to be in, but no obvious tonal differences when moving to the left and right side even slightly. Equal performance off axis to left or right. Pleasing sound over a large area, best spot is in the exact sweet spot.
Everything you fire from different locations will create new problems, unless perhaps when fired from straight forward. But getting that centre channel to sound right with stereo material would be a whole new quest. But interesting enough to try?
In all other positions (slightly off centre) the comb dips occur at slightly different places for the left and right ear. Here the left and right ear will make up for the perceived imbalance compared to that centre position where both ears get that same dip pattern.
In other words, I did not look for a solution that needs to fix anything slightly off axis.
The phase shuffler discussed in this thread makes the comb patterns slightly different for the left and right ear at that exact sweet spot. This part works! Though off axis there are differences between left and right side. It sort of favours one off axis side. Personally, I preferred this to the unprocessed stereo setup, where both ears get the same dip pattern.
I don't want to fix the centre problems by adding (relatively) early reflections I purposely got rid of. But that may just be me. I gained too much by getting rid of those reflections, I don't want to loose or get rid of that advantage in perception.
The biggest and most obvious dip is the one at ~1750 Hz. The one at ~5500 Hz stands out less. The higher dips probably don't make a huge difference no matter what you do due to head shading.
In my room, just fixing the perceived tonal difference with later reflections consisting of the phantom centre but diffused (but within the Haas limit) was the most satisfying solution.
The sweet spot continues to work as the best spot to be in, but no obvious tonal differences when moving to the left and right side even slightly. Equal performance off axis to left or right. Pleasing sound over a large area, best spot is in the exact sweet spot.
Everything you fire from different locations will create new problems, unless perhaps when fired from straight forward. But getting that centre channel to sound right with stereo material would be a whole new quest. But interesting enough to try?