The problem of stereo keyboards into mono

When you sum a stereo source into a mono amp it causes some out of phase problems. If only there was some way to have the left channel go to the lower pole and the right channel to the upper pole...

For example: My Yamaha P120 and P90 digital pianos have only stereo piano samples, there are no mono piano samples in the keyboard.
When play theor stereo pianos through any mono amp the tone becomes thin, edgy and almost all attack. Other keyboards' stereo patches behave in a similar manner. When the left and right samples are summed on top of each other into mono there is a out of alignment problem that causes a chorusing and phase cancelation effect.
It's not the equipments fault, it's just what you get when you sum a stereo signla into a mono amp.

You may think, but most keyboards have a L/R mono output. Yes, but it sounds crappy because it is also suffering from out of phase problems.
Original Post
Phase, part 1
http://www.music.columbia.edu/cmc/courses/g6630/Phase.html.

Phase is a time relationship between two sounds, waveforms, or signals in a circuit. When the time relationship is coincident, the two signals or sounds are "in phase." And their amplitudes are additive. When the time relationship is not coincident, they are "out of phase" and their amplitudes are subtractive. This has ramifications both acoustically and electronically.

Example - Close miking a piano with two microphones. When you strike a key, the sound from the string will reach the closer mic sooner then the mic that is farther away, creating a slight timing or phase difference in the microphone signals. If those two signals are played on a stereo system the effect should be a stereo "image" of some sort. Our ears capture the same kind of timing and phase differences and the brain uses these differences for sound localization.

Now if the two signals from the two piano microphones are summed to mono, the phase difference between them could result in phase cancellation. In that case, certain frequencies, depending on the music content and the microphone distances, are reduced in volume as a result of the subtractive amplitude of the out of phase signals. (Conversely, if frequencies are in phase, the volume will increase.) The result could be a change in timbre of the piano, a decrease in volume of certain notes, or even a kind of distorted, dirty sound.

Whenever you are miking in stereo, especially close miking, it's a good idea to listen, then sum the two channels to mono and listen again. Then reverse the phase of one channel and listen again. This should tell you what frequencies are being canceled when you sum to mono. If you don't like what you hear, reconsider your microphone placement - phase is working against you.

ProTools has a Reverse Phase effect in the audio suite.

Sound reflections in a room can create similar cancellation effects, which recording engineers must consider in the room setup for a recording session.
petros,or bose. If this is a phasing problem (which it appears to be) would it make matters worse, or be worth experimenting with changing the polarity of the cord coming from one channel of the piano or the other? could it possibly correct anything? or would it just be a waste of time experimenting.
so i was thinkin petros....i think i'll try it and just try the cable in the left, in the right,,along with the other cable wired correctly to the other piano jack...heck not much work and it will either sound better or it wont....certainly no elbow grease involved...
i tried it with an emu module i have here.....it really made a difference here....i have been having trouble with the middle register of the piano sounding somewhat hollow and dinky if you know what i mean. switchin the polarity of one of the cords coming from the piano, and then running them both into a stereo mixer channel and then to bose...quite a substantial difference...i like the tone much better..let me know what you think. i dont have my yamaha here now as i was going to sell it maybe...but i will get it back over wekend, and see how it reacts. im sure it will make a vast improvement also. but wont know till i try.
glad i read your post, as i wasnt able to put my finger on how to descirbe the tone I had completely forgot about phasing issues. And that is to me exactly about as near a description as i could have thought of.
im going to try this cord and scenario also with my computer pianos which seem to suffer the same syndrome and see if i like better.
if you try it, let me know if you could hear the difference. it does seem to warm up that middle register for me. at least with the emu.
well at least on this emu module , it makes a huge difference. I do notice maybe a little less in the bass register, but its too me almost a good thing which can be compensated with eq.in th middle range of the keynoard i now can really hear the attack, and the string resonance, which i couldnt before. I am much happier now...been playuin all day through the bose like this, and i think its a great sound. much much more like a piano should sound i think.
quote:
Originally posted by wfs:
well at least on this emu module , it makes a huge difference. I do notice maybe a little less in the bass register, but its too me almost a good thing which can be compensated with eq.in th middle range of the keynoard i now can really hear the attack, and the string resonance, which i couldnt before. I am much happier now...been playuin all day through the bose like this, and i think its a great sound. much much more like a piano should sound i think.


Hi wfs,

I'm sorry, I don't exactly follow you.
What element are you referring to that makes a huge difference?

Thanks
I'm assuming you are talking about using a mono speaker setup for playing. I guess that makes sense in the context of a band and a single musicians amp. I wonder what other pianists do in this situation? Maybe Yamaha isn't the best gig keyboard for this reason alone. You don't see a lot of them in live bands.

Anyway, the Yamaha outputs are unbalanced. That means they have a signal and a ground. If they were balanced, they would have a positive signal, a negative signal and a ground. In that case you can play with reversing the signals. In the unbalanced case, I'm not sure you would get any sound at all. You would be feeding ground into your signal and your signal into ground. There must be some external component that takes a stereo input and allows you to mix them into a mono signal with the ability to adjust the individual channels phase delay and time delay and such when mixed.
ill try and explain petros but it may take a few lines , or a couple of posts
basically in the beginning i found my yamaha p80 to be very pleasing to me when i ran it through my own home stereo. very enjoyable to listen to and play.using both the left and right outputs of the keyboard into my stereo and having the left and right signals go to the respective left and right speakers. a wonderful sound.
when i bought the bose system. which was mono. the only way i knew to hook up was either run the left and right outputs of the yanaha into the ch1 and ch2 or ch3 and ch4 of the bose. basically summing the individual keyboard outs to the bose and setting the volume on the bose channels the same.
the first difference in the sound i noticed when going through the bose, was i had lost the stereo spread. the sound was very direct coming from one speaker column and not my stereo monitors.
then as far as tone. the bass register of the piano seemed ok, as did the high register of the keyboard. but the middle range of the keyboard seemd very hollow and unpronounced. especially when comparing it to the stereo sound i was used to.
i would try all kinds of eq, basically found i was lookin for somethin in the high mids to control. i tried every bose preset. tried cutting, boosting etc....but just couldnt hone in on a good sound.
ithen tried coming out of the yamaha, and going into a stereo mackie mixer channel. i liked the high eq control better. then id use an aux send from the mackie to one channel in the bose. it helped, but the middle range of the keyboard still sounded hollow, and just some odd artifact that just didnt sound as if the keys wee attacking the note right.
then i read your phase post and got the idea to try reversing just the leads on one end of a 1/4 in unbalanced instrument cord to see what would happen
you are correct. in doing this i essentialy have the ground wire in the 1/4 in going to the positive tip, and the positive wire is connected to the ground. I expected to hear quite a bit of noise if it even did work
but no noise, and there is a signal. im not sure why it is working.
but the difference i hear is that there is no hollowness at all now in my middle keyboard register. and the attck and the string resonance sound that was barely heard before is now much cleaner and the whole sound is a lot warmer. floyd above has recomended a product that im sure does whatever ive done or in a better way more eficiently and more electronically correct. of course if bose tells me that this may harm my system, well i wont be able to do it. but right now i am tickled pink in the difference and more warmth of my piano sound.
again basically my setup now is on my emu sound module is
one reg 1/4 in cord from the right output from the sound module. one cord where i reversed the one end and put the positive wire to ground, and the negative wire to the positive tip of the 1/4 in cord from the other sound module out. from the sound module i run these 2 wires into a mackie stereo channel. I then send the signal from that mackie channel to the bose using one of the mackie sends. I dont have my yamaha keyboard at the moment, but thats the way id hook up the yamaha to try it
i had a friend stop by and listen to the comparison to make sure it just wasnt somethin else that i thought i was hearing, and actually wasnt there. but he immediatly with no knowledge of what i did or was doing , also immediatly said he enjoyed the sound with the one channels reversed wire much better. hope that explains at least sorta what i did and why and what i noticed in the sound improvemnet.
agian, i think my biggest concern with this modification to the one wire is that the actual signal is now very suscepteble to noise. here in my apt there is no problem. but in the real world if this reversed cord i have made was next to anything that may cause interference like a power cord, some strange bar flouresent etc....it may not work with out catching alot of interference noise...basically i guess it really becomes like an unshielded cable. someone here may be able to tell you why i am hearing the difference that i describe in tech terms. but im just not sure. i am just going by my ears and what i hear. it just sounds better. probably the best you could do would just get an old cord and do what i did and see if you hear a difference. i would like to get the same results another way. but im not sure it would be very cheap. and would at least involve buying somethin a bit more complicated as the above mentioned by floyd.
also petros when i speak of the keybords middle register, im basically talking about middle c up and down an octave. tah particular range. hope all this explained it better? the emu sound module im using at the moment for this experiment is also unbalanced outs stereo left and right.
one last post...i still have yet to experiment with the actual yamaha keyboard, but wont get it back till sunday. however this morning i did try the technique with my computer piano samples.
coming out of my audio interface left and right i did the same type of comparison. but i really didnt hear much if any difference.
i'll let you know as soon as i have the oppurtunity to try it. I did find it strange that i noticed very little differnce in mny computer stereo samples. so i am guesing a lot is dependant on how the stereo samples were actually recorded in the first place ( mic placement, types of mics etc.) but i think its worth experimenting with. just as sometimes when playing my accoustic guitar. sometimes i would be in envirnments where bass was a real problem ( too much)....on occasion i would reverse the speaker wire polarity, and it would help significantly on my old system.
so it may be a hit or miss type deal. But like i said i couldnt figure it would hurt to try. it will either sound better or it wont, or maybe not make much difference at all. But i certainly wasnt pleased with the sound i ws getting so what did i have to lose. but now a total of 3 people have definetly heard the difference in my emu module. and all love the sound of the hookup with the reversed wire from one channel. i'll keep you informed.
"Interesting post.

What's happening is that you're reversing the signal (phase) of one of the channels. Suddenly the signals that were out of phase (and therefore cancelling each other out in the middle octave) are now in phase and the signal strengh is increased instead of being weakened which would boost the middle octave.

I would have thought though that the opposite would also apply, ie waveforms that were previously in phase would now be out of phase (presumably the octaves that sounded ok before swapping the wires). In fact I would expect the middle octave to sound unnaturally loud and the other octaves to fade away.... "

fancyfingers
My engineer buddy writes:

"Not with my mixer it isn't. Here is one of the input stages from my cheap mixer (posted below). As you can see there is a common ground that is shared! This means you will be taking your reversed channels signal to ground. Maybe that guy has some different sort of mixer circuitry. Maybe Mackie does a better job at isolation. I don't think Yamaha is going to warranty you shorting out the signal output.

You could take the unbalanced signal from the Yamaha and drive a balanced input and just float the ground. A balanced input has two signal inputs and a ground. An unbalanced input has one signal input and a ground. If you took the unbalanced Yamaha output, signal and ground, and connected those two wires to the two inputs to a balanced input, you wouldn't short out the Yamaha. You could also reverse them without damaging anything."
guys -

here is a comment from one of the yamaha guys posted at motifator.com re what you're talking about....

You should never sum the Left and Right output to a mono amplification source. All Yamaha synths are designed that if you are going to hook up in mono use the LEFT output. The jack is labeled on the back of your keyboard.

"L/MONO" - it is written in your manual: "For monophonic output, use just the L/MONO jack."

If you connect both the left and right outputs to a monophonic system, the voices will be degraded - its physics. If a cable is plugged into just the LEFT output, the Motif will electronically sum the signal and maintain phase. If a cable is inserted into just the right output the signal will automatically be split out and assume you are connected to a stereo sound system.

If you absolutely must prove it to yourself do the following experiment. Connect just the Right output (if you are connected to a mixer - one that you can pan a channel - set it to CENTER). If you call up a VOICE in the Motif ES and then use the COMMON/ F2 OUTPUT Pan parameter to pan left, the signal will disappear (except for the stereo FX return from the System Effects).

If you plug a cable into just the LEFT output and do the same pan experiment you will notice that the signal is coherent throughout. In neither case will you get the benefit of stereo but at least you will have a coherent signal. You only degrade the sound by duplicating both Left and Right outputs in a MONO system.

Stereo when you can, if not...just connect the L/MONO output. Period.
as did i petros. and it does to a certain extent with the bass register. but no where near as much as i thought. when comparing the 2 hookups, the bass is definetly not as strong with the wire rversed. but not extreme at all. and it still blends well with the high and mid rregister. I would have expectd the same as you, for the other 2 registers to suffer. but at least with this particular sound module it doesnt. i pretty much understood why the middle range was coming out better with the emu module,( because of the reversed phase etc ) but what i was confused about was why i could barely hear any difference when messing with other stereo samples. i would have thought i would have heard at least some differences. If they are there in the puter stereo samples, they are for the most part negligable which ever way it is wired. im not saying that there isnt a difference, as there has to be. but it certainly is no where near the pronounced effect when the technioque is applied to the emu sound module. perhaps because with the puter samples every note is sampled 8 times. 4 for pedal up, and 4 for pedal down. its like a 2.5 gig sample. im sure the emu maybe has 10 notes sampled or so, and the rest of the notes are made off those samples. it is just weird.hopefully sunday try the yamaha. and be able to report.
I might just try that reverse cable thing on my P120 and Motif 6 later today with my PAS. I tried it with my GEM RP7 (Which sounded pretty good any way) and a Kurzwiel Mini-Ensemble and I got a clicking noise. Must be some interference. I went through a Alesis Mutimix FX6. Maybe I'll go straight in whilest I'm at it this time. Or try one of my Mackie boards. Meanwhile I got an E-mail from a guy at Little Labs and said thier gadget would help. (Of course, what else would he say?)
wfs,

Did you try a rewired cable with a Yamaha P series digtal piano? I am not sure which unit yu are talking about her than your Emu.

Klaus,

The problem is the L/R Mono output jack on the Yamaha P series pianos have problems. The L/R Mono out is thin, edgy and a over emphasized attack sound. It's probably a result of some phase cancellation.

petros

[This message was edited by Ken-at-Bose on Sun February 15 2004 at 09:37 AM.]
well petros, this phasin thing sure is odd. i got my yamaha back and did some experimenting with the reversed polarity cord.
basically on the yamaha there is very little difference. when the reversed wire is hooked up to one channel, it does maybe seem a little softer( understandably) but the basic tone of the sample doesnt change or improve to my hearing..
why this made such a drastic improvement on my emu sound module, but has little affect on the yamaha, or my computer samples is beyond me. just thought id give ya a report on what i found with the yamaha p80. not much luck. but my emu sure sounds nice now.
Yamaha P120, P90 and P80 get phase cancellation if played in mono

Phase:
Audio waveforms are cyclical; that is, they proceed through regular cycles or repetitions. Phase is defined as how far along its cycle a given waveform is. The measurement of phase is given in degrees, with 360 degrees being one complete cycle. One concern with phase becomes apparent when mixing together two waveforms. If these waveform are "out of phase", or delayed with respect to one another, there will be some cancellation in the resulting audio. This often produces what is described as a "hollow" sound. How much cancellation, and which frequencies it occurs at depends on the waveforms involved, and how far out of phase they are (two identical waveforms, 180 degrees out of phase, will cancel completely).
quote:
Originally posted by wfs:
so i was thinkin petros....i think i'll try it and just try the cable in the left, in the right,,along with the other cable wired correctly to the other piano jack...heck not much work and it will either sound better or it wont....certainly no elbow grease involved...


Hi. Just reading these posts and I'd like to correct a misconception. Reversing the wires at one end of a 1/4 in unbalanced cable will simply ground the input that is connected to that cable. If you draw a diagram with the connections as "ground" or "signal", you will see that the source (keyboard channel 'X') has a shielded output signal, with shield connected to the sleeve of the 1/4 jack and signal connected to the tip. When a cable that has shield and signal reversed at one end is connected to the source it then provides source "ground" to the destination input which prevents any signal from being amplified on that channel. This results in hearing only the signal from the normal cable of a stereo pair, so the phase cancelling is no longer taking place (which may account for the changes in sound), but the modified cable is not passing a phase-reversed signal. It is electrically impossible.

In contrast, reversing speaker wires on an individual speaker will flip its phase - the speaker is not ground referenced (floating) except through the speaker cable, and will quite happily work "backwords" and reverse the phase of the sound-wave in the air relative to another speaker.
wsf

What's the latest? Did you try using two PAS systems with your Yamaha P piano yet in an attempt to maintain the full bodied tone that the left and right stereo channels provide. We know that the Yamaha P pianos tone is degraded when summed in mono or if just one channel is played. By the way which P model do you have? P250, P120, P90 or P80?
I have a Yamaha S80 and it sounds terrific if I use only the right output instead of the left mono output. The listener in the audience will not be missing any of the piano tone if they are hearing only the right output. This is beacause the only difference between left and right piano samples are some phase and minor timbre variances that you can hear through a good set of headphones. However, in a performance situation where the piano is mic'd and put through a mono PA system,the audience would never hear these things anyway.
quote:
Originally posted by petros:
...
I would have thought though that the opposite would also apply, ie waveforms that were previously in phase would now be out of phase (presumably the octaves that sounded ok before swapping the wires). In fact I would expect the middle octave to sound unnaturally loud and the other octaves to fade away.... "


I suspect what is really happening with the samples is a constant value phase delay as part of the original stereo sample, not just a simple phase reversal. Recording with two mics gives a constant delay between them which is not a function of the frequency. At any rate, when you reverse the wires on one channel, you reverse the phase of ALL the frequencies -- but only those frequencies which (sort of) match the original recording delay will get the benefit of the phase-reversal effect.

So, while reversing one channel's leads (reversing its phase) may help with some samples, it's not the same as dealing with the "source problem", which is that the original sample is recorded with a pair of highly isolated mics which introduce a constant time-delayed between the left & right channels. To make a truly mono reconstruction, you need to use a constant delay before summing them.
Agreed. Very nice presentation (in the other thread, too) and welcome aboard, Dan!

I need to do some more research. When panning my Kurzweil K2500X piano patch all the way left or right I hear very clearly (in my studio monitors) the "reediness" that's so objectionable. I have not plugged this instrument into the PAS.

My normal gig instrument is a Kurzweil PC2X, and it definitely does not have this problem going through the PAS. I use just one out, and the keyboard has a global setting for stereo or mono, which of course I set to mono. I do not use chorus or 'verb when playing live.

I also do not hear this sound from the Yamaha PF500 I play Sunday mornings. I have the effects (and the internal speakers) shut off, and take just the L/MONO out to the PAS.

Last summer I had use of a Yamaha P250 for a series of outdoor concerts; I don't yet know if we'll have that instrument again this summer, but if so I'll see what if anything changed between the P250 and the PF500. The soundsets are virtually identical, so unless they improved the phase coherence, I'm baffled why there'd be so many complaints about the P250.

In the meantime, I'll try taking a single out from the K2500 and see if I can lose the reediness. I'll also fool with the global stereo/mono switch; I'm pretty sure that instrument has one.

Edit: By the way, unless I'm going for a special effect, I always use the best, most drool-inducing piano patch on the keyboard.
Well, I finally got around to doing a "partial" test of introducing a delay into one channel of a stereo keyboard. I say "partial" because I was actually positioned BEHIND the single L1 rather than in front of it, so it was a little difficult to judge "true" tonal quality.

However, I could hear relative tonal changes, and introducing a 12-15ms delay into just one channel of the L & R output of a digital keyboard can make a qualitative difference in the sound "balance" (low-mid-high) of the keyboard. (Details on the setup below.) I'm not promoting a particular setting, but I am saying that you may be able to, to a non-trivial degree, "fix" a stereo piano sound which sums "poorly" to mono by introducing a delay into just one channel.

I tested this effect with both a Roland RD-700 and a General Music RP/X Piano Module, with similar qualitative changes (FYI: those two piano generators sound distinctly different to my ears).

I put one (L) keyboard output into Channel 1 of a T1, and the other (R) output into Channel 2. Utility (flat) Preset, and no effects (to start with) on both channels -- and all the keyboard EQ controls off/flat.

I then added a Delay to one Channel. I set the Mix to 100%, the Fdbk to 0% (i.e.: only one "delayed" sound, not multiples), and then (click the Fdbk button) adjusted the delay Time. By pushing the FX Mute button on/off it was very easy to compare Delay/NoDelay sounds. (In fact, I found it necessary to compare that way, because just turning the Time delay knob introduced 'zip-like' artifacts into the sound which made it difficult to "adjust" by just tweaking the Time delay knob; I had to use the FX Mute to really compare sound changes.)

I also tried using Analog vs. Digital delay; there wasn't any significant difference I that I could tell for this purpose -- and both had a minimum of 10ms of delay.

I found the most interesting differences were when the delay was between 12-16ms. That amount of delay caused the "balance" to shift between the low and mid-range of the keyboard, tending to somewhat reduce the lows and "brighten" the mid-range. FYI: Delays over 30ms start sounding really strange. I would have liked to try something less than 10ms, too ... seems that may be a v1.6 limit on the T1.

While I'm not sure which I prefer (I'd have to really judge from being in front of the L1 !), I can see where in certain rooms and/or groups that having the delay option (assuming a spare T1 channel!) would be another means to adjust/control the digital piano's sound.

It just occurred to me that I did NOT compare the use of delay to just using the zEQ!?!?! I guess the next time I have an opportunity, I'll have to compare the EQ vs. Delay.

I also did not, in that same setting, attempt to compare the piano modules' "built-in-sum-to-mono-using-just-one-output" vs. using two T1 channels to sum-to-mono. I will say that I usually use both of those modules with "just-one-output-connection-mono" with great satisfaction -- with the RP/X being my piano-sound of choice.
Adding a fixed-time delay copy of a mono-source will cause some frequencies to sound relatively "thicker" while other will sound relatively "thinner" -- changing the amount of delay changes which frequencies.

To get a tiny bit technical:

A "frequency" has a time, a "period", it takes for one wave cycle. For example, a 1000 Hz signal (roughly 2 octaves above middle C) will have a period of 1/1000 of a second == 1 ms, whereas a 100 Hz signal (a little more than 1 octave below middle C) will have a period of ... do the math ... 10 ms. A 200 Hz signal (~1/2 octave down from middle C) will have a period of 5ms.

If one adds two identical signals that are exactly 1/2 a period apart, they will "cancel" each other out. On the other hand, a small amount of difference (delay), say 1/10th or 1/20th of a period, will "thicken" the original 'pure' sound.

One final understanding: delaying a signal by multiples of the "period" has the same effect as a "single" delay. That is, delaying a signal by 0.5 of a period is the same as delaying 1.5, 2.5, 3.5, 4.5, ... of a period.

So, using the example frequencies from above, a delay of 5ms will cause the 100 Hz (10ms) signal to virtually disappear, but not the 200 Hz -- it will sound louder ... but not necessarily any "richer". So,also, will the 1000 Hz signal be louder -- if the delay and frequencies are exact. A 4.7 ms delay, on the other hand, will probably make that 100 Hz signal sound real weak -- but present, whereas the 200 Hz signal will become "richer" (more complex).

Since each frequency of a mono signal is affected differently, adding a delayed exact copy of the original signal can shift the tonal character of a sound predictably ... but very differently than using an EQ control.

As noted in previous posts in this topic, the "problem" with some stereo sounds is that the use of two microphones spaced apart during recording introduces a "delay" between when the same sound is picked up by the two microphones. That can be an advantage for stereo reproduction, but can hurt certain frequencies when the two stereo channels are electronically added for a mono output.

So, to answer your question, more directly, Kramster:

I didn't try any other sounds; for a mono source, you can not only play with the delay time, but also try reducing the "mix" (so that the 'canceling' frequencies don't 'cancel as much') ... which is likely to have the effect of making the "tonal shift" more subtle than with a 100% Mix setting.
Interesting to me how these keyboard companies spend so much time and space trying to record the perfect "stereo piano" when the only one that might here a real live piano in stereo would be the person playing as the audience would never hear it that way anyway unless sitting directly or close enough to hear the lower notes to the left and..well you know..... silly me to think such things.

Add Reply

Likes (0)
Having trouble signing in?

We recently updated our sign-in procedure and if you have old sign-in data cached, this can create a problem. Please:

  1. Clear your browser cache and cookies
  2. Then close the browser (not just the window)
  3. Open the browser and try again
Thank you

Please make sure that your profile is up to date
×
×
×
×