Lets say that the CD is blended to mono below 100 Hz. You LP filter from 120 Hz. There will still be some visible separation of the channels. Unless you match your filter to the actual blend filter the X-Y technique won't work.
Summing phase reversed signals is a better way, but still not ideal because any phase shift will cause this to not work. The cross correlation works no matter what the phase shift is.
In your test above you cannot tell if the falloff below 100 Hz in both displays is reduced bass or correlated bass. Only the cross-correlation method is completely accurate.
Last edited:
Which one? With the lowest note on the double bass being 41Hz. I'd be curious to know what you were hearing down there on the left side of your graph . . .
Although granted if there's any genre where the low bass (what there is of it) isn't mixed to mono it would be classical (where it's most often naturally mixed by mic placement anyway).
Thanks for the explanation. A few more questions:
How did you do the cross-correlation? Custom software, or is there a tool I can try? FWIW, when I ran Bill Waslow's Diff-Maker software on the Left vs Right track, the resulting difference file shows the same spectrum.
True that if the low pass is above the blend area, we'd see a difference. But even with a 4th order low pass at 80Hz, it's still there. Isn't that well within the range of a normal subwoofer?
Isn't the phase shift part of what makes stereo? Amplitude can be the same, but phase shift or delay between the channels is a fundamental part of stereo, right?
Agreed and that puzzled me at first glance. We'd have to compare it to a normal mono sum of the same track. But if any part of the signal is mono, shouldn't it simply cancel out? It does on the mono tracks I have tested. The analysis should show no content where the signal is mono.
How did you do the cross-correlation? Custom software, or is there a tool I can try? FWIW, when I ran Bill Waslow's Diff-Maker software on the Left vs Right track, the resulting difference file shows the same spectrum.
This discussion is getting to the usual "absolute" extreme. In other words you guys are answering the question "Can I find a single CD that has stereo bass?" - Probably so, that was never the question. Prior to the late 90's it would be very difficult to find one. Today I have no idea. One guy posted that he heard bass separation on a Beatles CD - no Beatles CD ever had stereo bass. None that I ever tested.
And just because you can find a stereo bass source does not mean that you can hear it. To prove that would be a pretty difficult test - one that I am sure no one here is going to bother to do. And, as I am sure you know, I do not accept your "Well I hear it!" as evidence. Sorry.
Small phase shift at LFs would not be detectable because the periods are so long. Greisinger reports that random LF phase differences at the ears causes a "pleasing LF effect" - "Spacious bass" I think he calls it. So is some small phase shift at LFs because of multiple spaced sources an issue? Not in my mind. In fact I am in Gresinger's camp, I want my bass de-correlated (completely random phase) just like it is in a large auditorium.
I simply don't understand why people think that aspects of hearing that are clearly true at HFs must therefor also be true at LFs. Things like source localization from phase shifts or amplitude variations at the ears. Or that early reflections are an issue at HFs. These things are not true at LFs like they are at HFs. Rooms are different at LFs, the way we hear is completely different at LFs. Basically nothing is common between LF perception in a small room and HF perception.
Cross correlation is not too tough, my SpectraPlus software can do it, but it is kind of esoteric and many do not know what it is or how to use it. But there are cases, like this one, where it is really the only thing that is going to do the trick.
If I get the chance I may run an older CD and report, but in reality, I am not all that interested. I have better things to do - like clean my room, or shovel snow, you know fun things like that.
Last edited:
"Diffuse and directionless" is what it is in a large hall, and it's a reasonable goal. Hard to get in a small room, though, what with all those "modes" messing things up. Perhaps you can approximate it with multiple subs (without the "room sound" messing things up too much), but if your room permits it I think you're better off (and get "cleaner" bass) if you can avoid the additional resonances, either through leakage or damping (or both). In any case you work with what you've got and make the best of it.
In live performance in real halls it's often difficult to tell which side of the stage the basses are on from "orchestra" seating, and hardly a chance from the balcony. What is it that we're trying to "reproduce" here, anyway . . . an "experience" that we don't have in the real world?
Every single recording I have made for the past 30+ years has had stereo bass.
So, for me the mono option is not an option.
Also I find the time/phase relationship between the sub and the higher frequency drivers is rather important in terms of ultimate fidelity and spatial reproduction.
Try some Dorian recordings, afaik many if not all of their catalog was minimal mic'd and not summed in the bottom end. I know this since I knew one of their primary recording engineers, and have a few raw "demos" as well a quite a few of their commercially released versions of the same performances.
I suppose that there are various levels of progressive refinement that folks are interested in exploring, as well as different compromises that are acceptable.
I find that dead flat sub bass response is secondary to proper time/phase relationship - both is of course about as good as one can get.
YMMV.
So, for me the mono option is not an option.
Also I find the time/phase relationship between the sub and the higher frequency drivers is rather important in terms of ultimate fidelity and spatial reproduction.
Try some Dorian recordings, afaik many if not all of their catalog was minimal mic'd and not summed in the bottom end. I know this since I knew one of their primary recording engineers, and have a few raw "demos" as well a quite a few of their commercially released versions of the same performances.
I suppose that there are various levels of progressive refinement that folks are interested in exploring, as well as different compromises that are acceptable.
I find that dead flat sub bass response is secondary to proper time/phase relationship - both is of course about as good as one can get.
YMMV.
With ORTF or even a Decca Tree the bass is summed "naturally" at the mics (and by the hall), so there is no need to do anything more in the mix. Just how much "bass separation" do you think you get with a Schoeps MSTC 64 U? Even if there are some "highlight" mics they're more likely to be on the oboes or other winds (and soloists, of course, who will get mixed center anyway) than on the double bass. And there's really not a whole lot of subwoofer-low bass in most "acoustic" orchestral recordings . . . so what are we talking about, anyway ? ? ?
Last edited:
A Decca tree sends the center to sum with either side, it still has a "left" signal that is different then the "right".
What we are after is natural reproduction, not a stationary glutenous mass of monotonic dullness.
What you do experience in the hall, comes from all over the place, a choral effect that has "movement" in the air for lack of a better term, reverberation, that cannot be reproduced from a single source.
I suggest when the weather warms up you take a trip to a outdoor music fair, multiple venues simultaneously playing. Locate yourself in a reasonably central location to all the stages.
Then come back here and tell me you can't hear where the bass is coming from.
If you point a 21" subwoofer towards one side of your head, you should be able to localize it by measuring what one of your earlobes is flapping the hardest. But then again, that's cheating. Disclaimer: I'm not responsible for any hearing damage. 
Other than that, the answer is the same as with most psychoacousitcal questions: a loud, resounding.... it depends. Just to throw one more variable in the mix, it should be said that it's probably (certainly) easier to localize 120Hz, 180Hz, 240Hz, etc, than 60. Many subs have rather high levels of harmonic distortion, and this is further compounded by the lower threshold of audibility as the harmonics go up.
Other than that, the answer is the same as with most psychoacousitcal questions: a loud, resounding.... it depends. Just to throw one more variable in the mix, it should be said that it's probably (certainly) easier to localize 120Hz, 180Hz, 240Hz, etc, than 60. Many subs have rather high levels of harmonic distortion, and this is further compounded by the lower threshold of audibility as the harmonics go up.
Oh yummy. A vacation. That would be nice. I haven't done Vancouver Folk for a couple years . . .
I also stage manage for a Mozart Festival orchestra. We do four different venues each of two weekends each year, including outdoors. Over the past decade I've heard this same orchestra, up close and personal, in over a dozen halls, from 200 seat to 1200.
But thanks anyway for the "advice" . . .
Assuming one uses closely spaced mics the arrival times are similar. But are they similar enough to consider that to be "mono"?
Afaik Dorian did not use closely spaced mics in most instances.
Their Pictures at an Exhibition, on organ, this recording most certainly shows bass coming from different organ pipe lofts, witnessed by the action of the left or right subs.
But the idea is somewhat irrelevant, since if one uses a single mono sub, OR one uses subs placed at different points in the room *other than* in a timed relationship to the mains, will in my view cause somewhat unclear spatial relationships. Of course without the ability to compare the two cases (timed vs. randomly timed) one would be hard pressed to notice.
If this is not important to you, or you believe it can't be heard, fine.
I disagree, as do my ears.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.