Timpani rolls

[Irish28]

Hi,

How do you make timpani rolls the best way ?
(I know GPO offers left and right hand).

I use GPO with Sonar.

Thanks

Alex

[Styxx]

Hi,
Being a percussionist, I actually hate doing any percussion or drum tracks with the keys on the keyboard. Interfaced to my system is a Roland SPD20 that I use to play in rhythms. If need be I can go in and tweak or quantize sections if something is off. You can always use step record method or the mouse to click in your notes in the score or piano roll.
I am positive there are other useful methods others can share with you.

[DPDAN]

I just bang away at the correct keys on the keyboard to create a realistic roll.
I just finished the timpani track on the theme from Jurassic Park and noticed how sharp the timpani samples are in GPO, however this is not a probelm since we have the option of tuning the instrument. I set the tuning to -12 on the Kontakt player for that instrument. Although that setting does not make all of the samples absolutely correct, it makes most of them closer in tune with A440 being used as a reference.
Please bear with me for this next part. It never ceases to amaze me when I hear timpanis that are way too sharp in professional concerts and recordings. Here is my theory that I came up with years ago using only common sense. When I watch a timpany player tune the drum, they are bent over with their ear real close to the drum head, and they tune with their finger very quietly as to not interfere with the rest of the music. This is where I believe the problem is. When a mallet strikes the head of the drum, the internal volume (space) is less, (hmmm? a smaller drum). Smaller drums typically create higher pitches. If the timpani needs to be "in tune", then the player should tune the head slightly flat, so that when the drum is struck hard, the pitch is right on. Please forgive me for what appears to be sarcasm here, but this is the first time I have been able to vent my feelings about this continuing annoying dilema with timpanis, well.....most of the time.
I want to apologize to all the percussionists out there, but if you are aware of this phenomenon, then it shouldn't offend you, because not all timpanis are sharp when I hear them. Ultimately if you do teach, and you teach percussion, then this is something that your students should be made aware of. The sad thing about this problem and why I believe it is not going to go away any time soon, is that teachers are not teaching this element of tuning.

I'm done ranting :)
Dan

[DarwinKopp]

Hi Dan,

You're right about the quiet tuning tending towards sharpness, but it isn't because the volume of the drum is smaller when playing loudly, it's because the head is momentarily stretched tighter when playing loud passages. This is not much of a problem if the head is already stretched tightly, say an E-natural on the 26" drum, but becomes very noticeable on a loose head played very loudly, e.g. the G-natural on the 29" drum.

I find the GPO timps to be about 20 cents sharp overall, but it varies with the note/sample. Instead of dropping the timps an octave to halve the tuning error, you might want to consider using the pitch bend controller to drop the samples into tune instead.

[DPDAN]

instead of dropping the timps an octave to halve the tuning error, you might want to consider using the pitch bend controller to drop the samples into tune instead.

I'm confused by your statement, but curious
"dropping the timps and octave to halve the tuning error?
I didn't drop the timps an octave....
I considered the pitch bend, but it's faster on the Kontakt player to just set it and forget it, you know like Ronnco :)
Of course the beauty of the pitch bend is that each individual note can be altered to exact tune, but that's not that important to me, unless one of the notes is way off, it would simply take too long. You have to admit, part of GPO's beauty is that it allows the user to make music pretty fast.
I assumed that the sharpness of pitch was created by the internal volume being ever so slightly smaller, creating a higher pitch,,,, no? Of course a loose head would be more prone to out of tune notes, yes?
Remember, I was just using common sense, but that doesn't always prove to be right :)
Dan

[DarwinKopp]

When you said -12 in the Kontakt player, I thought you meant 12 semitones (i.e., an octave)...never mind :)

The volume of a drum's air cavity (if any) has zero to do with its pitch, though an air cavity will create resonance waves if the cavity is partially enclosed (e.g., bongoes) or fully enclosed (e.g., timpani). It's the size of the head on the drum and the tightness of the head that determines pitch. For an extreme example, rototoms have no air cavity at all (rototoms have no shell, either closed or open), but they vary in pitch over a two octave range by simply rotating/tightening the head. Bigger rototoms have a lower fundamental due to a larger head, and smaller ones have a higher fundamental due to a smaller head.

[DPDAN]

Darwin, I believe you are correct and that makes sense, however, a big bass drum in an orchestra would not sound the way it does without the drum cavity. In other words, a smaller drum could be tuned to the same note, but the larger of the two will sound much deeper even though the pitch is the same. I think of it like this... 30 hz is 30 hz. That is approximately a full octave above a low "C" on a pipe organ pedal. This note can be heard on a 3" speaker as well as a 30" speaker, but the 30" speaker will more faithfully (reproduce) the fundamental as opposed to the harmonics, simply because it can move more air. If it weren't for harmonics, we would never even hear the 30 hz note on the 3" speaker.
I agree with you, it is largely due to the tuning of the head.
So do you suppose that timpanis are almost always flat because they are tuned quietly?
Just for fun, this is a picture of a 30" Electro Voice speaker that I use for special effects on remote location sound jobs, earthquake rumble, thunder, pipe organ notes, this amazing speaker is flat from around 24 hz down to 11 hz where is starts to drop off. You have no idea what 20 hz sounds,, feels like, AWESOME! It makes all my double 18" sound like a joke.
Sorry dial up users,,, this is a large file, I mean speaker :)
http://www.mydocsonline.com/pub/DPDAN/EV%20W30

[KevinKauai]

(Warning: take-off on a Chinese restaurant joke ahead): With three you get timpani rolls!

But serious folks (apologies to Styxx for indulging in the humor department for a bit): The IRC discussion hit on a notion for timpani rolls a few weeks ago, which was to have a "timpani roll" set and let the Mod Wheel (or other similar control device) dictate the speed of the roll. Although I am primarily a keyboard kind of guy, I actually played timpani years ago and while I was confortable with playing rolls, I find the GPO approach (of left and right fingers two octaves apart) to be fine, too.

I'm glad to understand the tuning question, though. I thought my ears were tricking me. Someone could write a fairly large tome on the use of Var 1, Var 2 and the "Tune" control.

Learning every day ... KevinKauai :)

[DarwinKopp]

Dan,

That EV is something else. I never knew cone woofers could be that large! Years ago I owned a pair of Altec 15" duplex drivers (where the horn is embedded at the center of a 15" woofer) that sounded amazing. There's nothing like bass that "gets physical". :)

Interestingly, a bass drum struck with the mallet vibrates exactly like a large woofer in motion. A head that is somewhat smaller than a bass drum's head might reproduce the same pitch, but the bass drum sounds the way that it does because the heads and cavity resonate the pitch. Think of it sort of like a bass transducer mounted in a speaker cabinet. No cabinet, and the speaker will still reproduce the pitch, but will sound very thin. Furthermore, there is an opposite head is in sympathetic motion on the concert bass drum. The opposite head is usually tuned slightly higher or lower than the batter head, causing both another frequency vibrating and creating a very low difference frequency between the two heads. This gives the concert bass drum that very low, rather indistinct pitch we all know and love.

Yes, timpani can well end up sounding sharp because they are tuned quietly with almost no pressure on the head. Again, it depends on what pitch is being played on what drum, and the force with which the upcoming notes are played. You can see the exact same effect with the toms on a drum set. Many drummers spend hours trying to get their toms to "scoop" the pitch just right when struck hard. The trick is to tune them a little flabby. Then when they are stuck hard, the impact of the stick momentarily raises the pitch, which then immediately falls off, producing a sort of "dooonnhhh" sound.

Also, sometimes the timpanist just can't get the pitch right...the pedals are sticky or jerky (or the whole tuning mechanism slips) or the drums haven't been acclimated to the room yet (cold and they start flat and keep changing until their temperature is normalized, hot, vice-versa). Certain timps have discrete lock points as the pedal is moved up or down, which sometimes cause the the head to be locked slightly sharp or flat. In this design, there is usually a fine tune mechanism to compensate, but it does the timpanist little good if he needs to change pitches quickly and then play that note for several measures. Also, if calf heads are being used, the drums will play flat if the head (or the room) is humid, and will constantly change until the humidity in the room and in the heads have equalized.

Come to think of it, I'm surprised timpani can be played in tune at all. :)

[Irish28]

Hi guys,

I must say that my first post was to ask how you made timpani rolls ?
I meant in what speed you rolled, for example.
In a mesure with a four-beat note, how many midi events do you insert.
how many notes do you play.
Do you play them the same way, or some have more velocity ?

And last question : Do you use ModWheel ?

Thanks
Alex

[DarwinKopp]

Hi guys,

I must say that my first post was to ask how you made timpani rolls ?
I meant in what speed you rolled, for example.
In a mesure with a four-beat note, how many midi events do you insert.
how many notes do you play.
Do you play them the same way, or some have more velocity ?

And last question : Do you use ModWheel ?

Thanks
Alex

Oops, sorry Irish28, didn't mean to hijack the thread.

The short answer is to play whatever number of beats sounds good, neither too many nor too few. It depends on the tempo of the music, and whether the tempo is strict or rubato, but a rough guide would be to use 6 midi events per beat at 100 beats per minute and match any change in tempo roughly inversely proportional to this ratio. So for a measure of 4/4 at 100bpm, you would insert 24 midi events. Increase the velocity for crescendos and decrease the velocity for decrescendos. I don't believe the mod wheel does anything at all for the timpani.

[KevinKauai]

I think the philosophy (GG -- correct me if I'm wrong) on the GPO Timpani is that it's like hitting a drum head. The analog is with how much velocity you hit it. I mostly "play" the timpani with two fingers, mimmicing how I would produce a roll. I think the only thing that the GPO model potentially leaves out is the different type of drumsticks that you might choose to use. (I have seen many pro players with various symphony orchestras select from 6 or more different types of 'sticks' -- with varying amounts of softness or hardness.)

The only thing I would add is that I have also resorted to using the "line" tool for velocity (in the Cubase SX piano roll editor on the Velocity control strip) to get an absolutely smooth crescendo or decrescendo.

i hope this adds to the discussion ... KevinKauai :)

[Tom Hopkins]

Kevin has it right: Play rolls with the "two finger" technique (timpani is a velocity sensitive instrument – no mod wheel). It's very intuitive. Strike the key harder to simulate striking the timpani harder. The timbre changes with the strength of strike. Higher velocities are brighter. A crescendo will increase in velocity over the course of a roll. When I was designing the programming I had a great deal of fun playing the timpani. Of course, you can alter anything you play by editing the data. I've always gotten the best results playing the parts in rather than step entry. Those small timing and velocity inaccuracies of playing help realism enormously. The update will include some new enhancements to the timpani.

Tom

[Haydn]

I find playing rolls in by hand works better than using canned rolls. Mainly as the rolls should change in speed over time. I usually speed up the roll as I get louder. Listen to orchestra recordings for ideas.

[DPDAN]

I too play timpani with the midi keyboard, I use a Yamaha KX88. Since we are talking about timpani stuff here, I thought I would post the timpani track that I have finished for my Jurassic Park theme. In the background is the original track that I am copying for fun. This is the most difficult thing I have ever copied by ear, I have just the timpani and 1st violins done.
Hopefully this will enhance the discussion. Who knows :)
the timpanis are dry (no reverb) for this demo

timpani demo (http://www.mydocsonline.com/pub/DPDAN/timpani.MP3)

[Irish28]

Thanks,

This helps, nice rolls, good luck for your work !
(That's a lot of time in front of your computer).

Alex

[Styxx]

I find playing rolls in by hand works better than using canned rolls. Mainly as the rolls should change in speed over time. I usually speed up the roll as I get louder. Listen to orchestra recordings for ideas.

Never, absolutely never. Speed up as the tempo speeds up overall but to “speed up the roll as I get louder", sorry, common misconception! Possibly you mean to move into the subdivisions of the meter during a crescendo. But if I am to roll on 32nds and crescendo and the tempo calls for say quarter = 136 or whatever, there is no speeding up. I’ve noticed in plenty of midis over the past ten years two egregious errors in programming timpani. 1. Out of range either TOO LOW or TOO HIGH in pitch. The first being the worse of the two because the low notes get muddy and distorted and the second only due to the stretch of the skin can only go so far! Not to mention the fundamentals becoming distorted and losing the “sweetness” of the bowl(s).
2. Programming a roll or rolls out of context of the composition and the given tempo and or articulations. Rolls too fast in meter and or quantized perfectly even with any note value or its subdivisions.

Thankfully, most of the people on this forum have NOT been anywhere near what I have described. Then again, if speeding up suites your desires in your composition or marked within a composition, then by all means go for it.
Then again ... I'm not a musician just a percussionist.
:D

[DPDAN]

Sorry Haydn, I have to agree with Styxx, I'm not trying to gang up on you, :)here is a pic of the last timpani roll on the Jurassic Park theme that I am working on. Click on the link that I posted for the MP3 previously, and you can hear the definite increase in dynamics,..... this pic shows that there was no increase in tempo.
Dan
http://www.mydocsonline.com/pub/DPDAN/timpani.jpg

[Styxx]

Make you own Timpani rolls using this recipe shared by Styxx.
INGREDIENTS:
· 3 timpani bowls
· 1 cup of granulated sugar
· 1/3 cup of water
· 1 teaspoon of vanilla
· 3/4 cup all purpose flour
· 1 teaspoon of baking powder
· 2/3 cup of timpani material or jam
· Powdered sugar
PREPARATION:
Heat oven to 375.
Line a timpani roll bowl, 34 1/2 x 23 1/2x1 inch, with aluminum foil, then grease with shortening. Beat heads in a large bowl on high speed until crescendo or decrescendo. Gradually beat in tempo changes. Beat in 3/4 and 4/4 on low speed. Gradually beat in 16th, 32nd and 64th notes until smooth. Pour into timpani roll pan; spread to corners.
Bake until wooden timpani sick inserted in center comes out clean, 12 to 15 minutes. Generously sprinkle written score notes a little larger than the timpani head.
Immediately loosen timpani head from pan and turn over onto the stage. Carefully remove foil. Trim off any crisp edges of head. While hot, carefully roll timpani and foot pedal from narrow end. Let cool on wire rack at least 30 minutes.
Unroll timpani and remove quantinization. Beat timpani head with a stick just enough to soften; spread over bowl. Roll up notes and spindled with powdered sugar.
:D

[Haydn]

Oops, was watching the olympics while typing that. Normally, I will speed up slightly after the initial attack if the attack is Sfz. I usually slow down a tiny bit at the end of a loud crescendo depending on the piece. What happens many times though is that I will speed up at the end if I know I will be adding a ritard in later when fine tuning the tempo as the ritard will slow down the roll too much.

[KevinKauai]

Loved your timpani roll recipe, Styxx! But if I'm out of "baking powder" could gun powder be substituted?

I, too, prefer fresh timpani rolls to the canned (which I find way too loaded with syrup from the canning process).

KevinKauai :)

P.S. Here's to the late, great Julia Child for a wonderful laugh. (She was full of `em!)

[rwdelling]

Never, absolutely never. Speed up as the tempo speeds up overall but to “speed up the roll as I get louder", sorry, common misconception!

Sorry Styxx, but I have to disagree with you on this one. In order to make the most realistic cresc. timp rolls, one shouls always speed up the strokes of the roll as the volume increases. After all, this is how a real timpanist plays cresc. rolls. It will sound more realistic. Also, a simple volume envelope "cresc." helps a lot as well.

[Styxx]

Sorry Styxx, but I have to disagree with you on this one. In order to make the most realistic cresc. timp rolls, one shouls always speed up the strokes of the roll as the volume increases. After all, this is how a real timpanist plays cresc. rolls. It will sound more realistic. Also, a simple volume envelope "cresc." helps a lot as well.

Yes, but I disagree in return. Then again, I can easily "sequence" a timp roll to go faster during volume increases. However, realistically speaking, how fast do you expect a live timpanist to speed up to during a lengthy crescendo. And, if this is indeed the technique, then one would expect a timpanist to "slow down" during decrescendo's right?

[FlyingRon]

Not familiar with that sort of Timpani roll. I'm more of a Timpani-Maki kind of guy.

[Skysaw]

Would be nice to have concrete analysis on this, but my guess is that a typical timpanist will actually have decreased speed during a crescendo. (Disclaimer - I am not a percussionist...)

If you watch a timpanist do a crescendo roll, you will notice the sticks getting further and further away from the drum head. The player needs the increased distance to work up to the required velocity for a harder hit. So yes, his hands are moving faster, but because of the increased distance (up and down), the notes themselves become further spaced.

Try it yourself on your desktop. It's pretty easy to rapidly alternate taps between your right and left hands if the hands stay close to the surface. Now try to do it just as rapidly, but from further away from the surface.

Another analogy, if you play guitar. Try holding a pick close to the strings, and make rapid mandolin-style pluckings. The louder you get, the harder it is to keep it rapid.

[Styxx]

A professional timpanist feels the rebound off the head and thus can use it to his or her advantage to keep the tempo steady during a crescendo. There are also specific “sticking” techniques that are learned and applied that help the player obtain continuity of speed at any level.
I have contacted Mr. Vic Firth and had asked for his professional advice on the proper technique. One thing we must all remember is that we are using electronic devices; sequencers and our PC's obtain the desired result(s). In other words, where is the human factor?

[FossMan]

Now, maybe those of you who are purists and like to do things by hand will shoot me because of this, but can't the next version of the GPO include predone timpani roll patches? I remember several years back playing on an Alesis keyboard and the timpani roll on that machine was very nice. I don't see why the GPO couldn't include something like that, with the Mod Wheel controlling dynamics which would automatically increase/decrease the "speed" of the roll.

[Irish28]

Hi everybody,

I posted this topic a few weeks ago, and I must say that my question was on that !
I think that Gary could offer us timpani "already rolled", sometimes, that's very handy, and for just a demo, we can have something quite realistic very quickly.

[DPDAN]

When a percussionist performs a snare roll, the duration between "hits" is the same whether it's quiet or loud and doesn't change.
Get out some old records or CD's and listen to recordings. I am no theorist, but I have quite a few timpani rolls in projects that I have done before this topic started, and none of them have any noticeable alteration of tempo change during a "roll". Of course a timpanist can create a roll with an increase or decrease of tempo, but I have never done that when I program timpani rolls, it just feels very natural to only increase and decrease the dynamics.

just my .02

Dan

[Haydn]

There are 2 problems with the canned rolls in other percussion libraries I have.

1. The generic rolls are too consistent with no dynamic and timbral changes. If you crescendo the roll the drum doesn't resonate more like the real thing. The speed of the roll may not fit the piece.

2. The crescendo/descrendo rolls are set to fixed lengths which usually won't fit the piece your working on (at least from my experience).

[DPDAN]

I like GPO timps just the way they are. Very easy to play and create very realistic personalized sound.

[Skysaw]

I think rolled samples could work quite well, using a looped sample with dynamic cross-fading controlled by our friend the mod-wheel (same way the bass drum roll works). It certainly couldn't cover all musical needs, but for those it did, it would be both user-friendly, and realistic.

The problem with building your own rolls is that hitting a drum head sounds different, depending on whether or not the head is already vibrating when it's struck. It's not a huge problem, as I haven't had too much trouble getting good results, but it's not perfect. More than sufficient for orchestral tuttis, though.

[Tom Hopkins]

It bears repeating: There will be no new sample material added to the coming update. The update concerns only programming changes. On the subject of timpani rolls: There will be some new playable options.

Tom

[DarwinKopp]

Sorry Styxx, but I have to disagree with you on this one. In order to make the most realistic cresc. timp rolls, one shouls always speed up the strokes of the roll as the volume increases. After all, this is how a real timpanist plays cresc. rolls. It will sound more realistic. Also, a simple volume envelope "cresc." helps a lot as well.

Actually, a timpanist rolls faster on a tighter head and slower on a slack head. A tighter head requires a faster roll because the tighter head is much less resonant than a slack head. A slack head requires a slower roll because too many hits tend to interfere with the natural oscillation of a slack head already in motion and would tend to decrease resonance. Again, remember that timpani are more about maximizing resonance than percussiveness.

One would think that objectively higher pitches are always tighter and lower pitches always more slack, but this is not necessarily so. What is true is that notes at the low end of the range of a given timapno are generally rolled more slowly than notes at the high end of the range. However, what is true for any given ptich is not true for the set of timpani as a whole because the ranges of the drums overlap. So, a Bb forced out of a 32" drum requires a faster roll than a Bb played in the middle of the range of a 29" drum, which requires a faster roll than a Bb at the bottom of the range of a 26" drum. These variations in speed are not great but are there nonetheless.

The hardness of mallets also changes the roll speed. Very hard mallets are typically rolled much faster than very soft mallets, again because the hard mallets produce a much less resonant sound than soft mallets.

So as something of an extreme example, a high F-natural with hard mallets on a 26" drum is often a noticeably faster roll than a low F-natural with soft mallets on a 29" drum.

There is not really a direct relation between roll speed and volume. It depends on what the timpanist is trying to achieve. If he wants a relaxed, full volume, sustained crescendo, then the rate of rolling stays pretty constant. If he wants a dramatic flair into a tutti ritardando, then the roll may actually slow a bit as volume increases. If the timpanist wants a hard, biting, no nonsense crescendo, then the roll may speed up towards the end with rising volume.

Yet, some composers routinely notate exactly how many semi-quavers the roll will contain. Beethoven was notorious for this in many of his works, and all conscientious timpanists will do their utmost to honor his directions (see the first movement of the Ninth Symphony for a prime example). So, in these cases, there is no variation in speed at all in the roll as volume changes, except where the timpanist has technical difficulties in execution, problems following the conductor, or is fighting against too great a tempo than that which Beethoven had indicated. Incidentally, I believe this style of timpani part helps impart to Beethoven's music a very definite, incisive quality.

Also, roll passages notated with heavy accents and fp rolls don't even roll at first. Typically, the accented initial note is struck, and the timpanist then pauses, perhaps even as much as a full second, before commencing with the roll. This allows the pitch of the initial accent to be clearly established before the roll continues.

These are, for the most part, generalities, and most timpanists do not very often even consciously think about much of it. Beethoven aside, the timpanist just instinctively tends to roll in a way that sounds best in a given setting. :)

[bmonroney]

In DP, I have the best results using step-time entry for timpani rolls. Once I have the note duration I want, I switch to ticks, and manually vary the ticks slightly as the roll progresses to make it less mechanical, usually speeding up or slowing down depending on the phrase...after that I draw in the velocity curve I want with spline tool, and bump the velocity of the first hit. I have a few of these rolls stored as DP clippings files to copy and paste into a piece...it's a big time-saver. You can can use the time-scaling features to change the speed of the roll. This is handy if you already have a speed change but the roll is just a bit too short.