The possible merits of auto-tune?

[Muleya]

The other night I was watching coverage of the Download Festival from a year or two ago. One of the bands was Def Leppard and it was almost painful to watch...Joe Elliott was just slightly off pitch the whole song! Whereas part of me has little respect for auto-tune being used to keep a singer on pitch, this was not enjoyable to watch!

So what do folks here think? If Joe can't stay on pitch, should Def Leppard (or substitute one of your favorite bands here) just hang it up and stop touring?

Or does it not bother you that the singer is slightly off pitch, you'd still enjoy seeing one of your favorite bands anyway?

Or would you make an exception for aging singers who obviously could sing on pitch in their hey day, but we all know that age and hard use takes a toll on the voice, so let them use autotune so we can enjoy the performance? If so, then where do you draw the line?

Or do you think it's just a matter of him not hearing himself well enough in his monitor mix, and could/should be remedied some other way? The problem is, it wasn't for this performance!!

[Jan]

It's important to me that the singing is really good. It grates on my nerves if the singing is bad. Like Kari Jobe for example, I love her songs when we sing them in church, but I just can't listen to her singing them. I don't think autotune would help in that case - I just don't like the way her voice sounds and she doesn't pronounce her words right. But I would prefer autotune whenever it would help, and not sound fake like that dog video Brent posted. That was funny but I wouldn't want real music to sound fake like that.

[brent]

Generally, Joe can still hit the notes that Phil does not sing for him. Joe admits that he is not the singer he was though. He has offered to let someone else do it and play guitar. But if someone is flat or sharp the whole song or show, it can also be a sign that their In Ear Monitors (IEMs for short) are loud or too low. Is is the doppler effect. It is a balancing act because the point at which the ear is most linear in it's response is around 85dB. That is darn loud inside your ear canal for the duration of a concert, over and above the stage and crowd noise.

We also know from research that diet, drugs, and illness can change the hearing threshold, frequency response and sensitivity.

Autotune in a live setting is common. But not by singers. Most people like rappers, etc misuse it to get that stair step vocoder effect. Most singers will not use it because the process itself adds delay, called latency, and that jacks with the timing. So, if the Autotune process adds as much as 1300ms, which is like being 1300 FEET away, and they are using one of the average industry standard mixing consoles, which adds as little as 2.5ms and as much as 7ms or delay, and the sound system processing adds it's 7ms of delay.....the band is not going to be tight, the singer will be off, etc.

All of that stage noise bleeds into the vocalists microphone. The mic cannot discern between stage noise and vocal signal. It all gets into the channel path. If something is bleeding into the channel path like guitars, and they ARE in tune, and the out of tune vocal goes in the channel path, then the in tuned guitars become out of tune the amount of cents that the vocals are going in. It is a cluster in the worst way. So no, it doesn't usually get done unless there is a quiet stage, etc, etc....

More than you wanted to know.....

[Muleya]

After I posted I had thought about the fact it could be a bad monitor mix. Another guy I know said he saw them a couple years ago and Joe was spot on the whole night. He pointed out that at these festivals where set up is quick and dirty, these kinds of problems aren't uncommon. Makes sense.

I also hear what you are saying about latency and stage noise/bleedover. It certainly all plays a role as well. Journey was on stage for the same festival and sounded pretty tight and Arnel on key. So I'm guessing it was something unique to DLs set up. Or Joe was just having a bad night!

[brent]

I have heard him suck it up myself. :}

[Boray]

It is a balancing act because the point at which the ear is most linear in it's response is around 85dB.

Hmmm.... Isn't that related to frequency response? Not to pitch?

[brent]

At 85dB or so, according to the Fletcher Munson curves, the ear is most linear. This all goes hand in hand. Doppler Effect is what you hear when an emergency vehicle has it's siren on, and it is further away, lower in volume. As it nears you and passes, you hear it at it's constant pitch. When it passes you hear it go down again. It is not the distance that changes the pitch in your ear. It is the volume. Take your headphones and place them on the desk, turn on some music, and stand a couple of feet away. Sing to that music. As you sing, put the headphones on your head. You will be out of key. It is the same principle. When people listen at lower volumes they sing flat. When they listen at louder volumes, they sing sharp, because they are hearing it that way. It is not that the music has changed pitch. The music is changing amplitude. The reason why the FM curves and linear sensitivity comes into play is obvious. One cannot hear perfect pitch if the one cannot hear the frequencies of said pitches, especially when we are talking about those outside of our God designed range of 100 cycles through 4000 cycles. It takes exponentially more power to get over the hearing threshold of 0dB-SPL outside of that range on both ends.

[rexreed]

I have heard him suck it up myself. :}

Saw them a couple years back and he was on pitch as any other 50 year old rocker. Songs like Photograph gave him a tough time but most of the show was A OK. Kenny Rogers on the other hand.... no upper register left:(

[Boray]

At 85dB or so, according to the Fletcher Munson curves, the ear is most linear. This all goes hand in hand. Doppler Effect is what you hear when an emergency vehicle has it's siren on, and it is further away, lower in volume. As it nears you and passes, you hear it at it's constant pitch. When it passes you hear it go down again. It is not the distance that changes the pitch in your ear. It is the volume. Take your headphones and place them on the desk, turn on some music, and stand a couple of feet away. Sing to that music. As you sing, put the headphones on your head. You will be out of key. It is the same principle. When people listen at lower volumes they sing flat. When they listen at louder volumes, they sing sharp, because they are hearing it that way. It is not that the music has changed pitch. The music is changing amplitude. The reason why the FM curves and linear sensitivity comes into play is obvious. One cannot hear perfect pitch if the one cannot hear the frequencies of said pitches, especially when we are talking about those outside of our God designed range of 100 cycles through 4000 cycles. It takes exponentially more power to get over the hearing threshold of 0dB-SPL outside of that range on both ends.

The Fletcher Munson curves only describes the human ear's ability to perceive different frequencies at different sound pressure levels. How loud you perceive them.

That an emergency vehicle change in pitch only has to do with the speed of sound. When it's coming toward you, the sound waves will be shorter because of the vehicle's speed and of the speed of sound. When it's moving away, the sound waves will be longer. It should be possible to even calculate the exact pitch change you perceive at a given speed. This has nothing to do with the human ear. The sound waves are like that. It will be the same result if you record it with a perfect microphone. (If it was because of the volume, then it would also be the same change when it's coming and going. It isn't. It's the opposite.)

If it was like you think, you should hear the music change in pitch every time you change the volume of your speakers. It doesn't. If you hear a change in the test you proposed, it wouldn't be because of the volume, but from moving the headphones towards you / moving yourself toward the headphones. Because of the speed, just like the emergency vehicle.

[Boray]

If you hear a change in the test you proposed, it wouldn't be because of the volume, but from moving the headphones towards you / moving yourself toward the headphones. Because of the speed, just like the emergency vehicle.

Or even more likely - because headphones are not made to be listened to from a distance. You would only hear some whispers and your singing would more or less be a guess.

[brent]

Ok then. Try turning your car stereo way down with an ambient noise floor in excess of 80dB and you will see what I mean.

"An equal-loudness contour is a measure of sound pressure (dB SPL), over the frequency spectrum, for which a listener perceives a constant loudness when presented with pure steady tones" The most recent ISO data from Japan reveals 85 dB where the ear is most linear, where there is constant loudness at frequencies from 100 cycles to 8000 cycles. This is why the most popular monitor and metering calibration standards for studio monitoring use SPLs around 85dB.

Re: Doppler. It is not the speed of the wave alone. It is a change in relationship to the wave. What I was trying to describe was the effect. The Doppler Illusion. At the bottom I have contact info for Dr/Professor of physics who has studied it.

We can also do the same thing with delay and changes in loudness. We do it in hearing aids all the time. Shure, Sennheiser and others have white papers on this effect.

There is also some new science, new speaker systems, where there is no sound made by traditional vibration of air. There have been carbon fiber and other projection systems where sound is generated at the ear and only the listener can hear it. There is a company in Germany that created a home theater that uses the barriers of the room to become transmitters. The room is completely silent, yet perceptions in pitch can still be manipulated without having a change in relationship to the waves, because there are no waves.

Contact this guy. He has done extensive research on "dynamic change", the interaction of pitch and loudness:

John G. Neuhoff- [email protected]
Dept. of Psych., Lafayette College,
Easton, PA 18042

Michael K. McBeath
Dept. of Psych, Kent State University

Don't read too many of those old books. Old science is wrong in many cases. You should talk to someone in the hearing business as well. My brother, the Dr and audiologist would be happy to clue you in on advances in hearing research as well. There is some crazy stuff in use now. Stuff I use to deny was possible. Pro Audio is waaaay behind.

[Boray]

How big would this loudness/pitch change be? Because I can't and have never ever heard a change in pitch when changing the volume of speakers/headphones or whatever. Maybe there is something wrong with my hearing? Or maybe what that research shows is that the brain hears a doppler effect when it expects one.

[Boray]

Maybe it has to do with the ear's built in hearing protecting?

[Boray]

Doppler Effect is what you hear when an emergency vehicle has it's siren on, and it is further away, lower in volume. As it nears you and passes, you hear it at it's constant pitch. When it passes you hear it go down again. It is not the distance that changes the pitch in your ear. It is the volume.

This simply can't be correct the way it's put. The main factor must be the speed of sound. Even if there could be more factors involved, the main factor must be speed. You can stand on a hill and hear a police car drive by below. You will then hear the same volume but the pitch will still change. And that is something I have heard.

When it comes to monitoring with headphones while singing, I think the main issue is that you can perceive your own voice differently from within your own head compared to from outside. So it's a matter of getting the right level of your own voice in your headphones. Adding reverb to it helps.

[Boray]

By the way, are you among those who think it's a good idea to monitor at 80dB (or I guess 85dB) when mixing in your studio?