I was going to put each question in a separate post but decided not too.
1) What\'s the best way to record rubato music into a sequencer?
Do you just turn the metronome off? Do you record it in tempo and just insert increasing/decreasing tempo events? What\'s the best way to go about this? (Thomas_J and Simon Ravn I\'d be interested in your response becuase I\'ve heard your music and you do these particular things well) Please all feel free to respond.
2) When recording orchestral scores, how do you go about overall balancing between instruments? Do you record each instrument part separately as expressive as possible without regards to the other instrument parts almost as if this is a solo instrument performance or, do you try to record the instrument while listening to the other parts already recorded (ensemble fashion)? Personally, I think the second method would yield better results in an orchestral piece because it would \"match\" the current height of the musical line; however, when you solo the instrument part it might not sound as \"expressively convincing\".
To clarify my question: Most can agree that ensemble playing is very different from solo playing. In an ensemble you wouldn\'t play out as expressively and you would with a solo piece because you\'d stick out like a sore thumb. Should we apply this same practice when recording in orchestral parts to a sequencer since there are other instruments which will be playing with the part you are currently recording?
If the piec is a violin concerto or something, of course you can go wild with the solo violin part because it is being featured and likewise you can record it in the sequencer that way but how about ensemble playing? How should one record ensemble parts into a sequencer with regards to balance with the other instruments in its section.
Hopefully this is more clear after the above rambling.
3) After balancing each section within itself, how do you go about balancing the whole orchestra? Do you have a standard balancing method you use everytime ( string then winds then brass, then percussion etc...) or does it depend on the specific piece at hand.
4) I almost forgot: Expression (CC11) vs Filter (cc 74).
This might draw some dispute but that\'s why we have these forums For those who don\'t own breath controllers, most people grab the expression controller to shape their melodic lines. If you think about it, wouldn\'t one get more __realistic__ results by grabbing a filter controller? As you play most instruments or trashcan tops softer, the higher harmonics are reduced greatly: exactly what a filter does.
I tried this yesterday and was amazed at how much realistic the result was. Especially on something like a muted trumpet sample. The first thing I\'m grabbing for now is the filter instead of the expression pedal to shape lines. The expression controller only reduces volume and NOT the cutoff frequencies. In real life the higher frequencies are not heard at lower volumes. Only at high volumes. \"I smell the word filter\".
Strike the lowest note on the piano as hard as you can.... listen to all those frequencies coming out.... strike the same key softer and listen how only the fundamental note dominates and those other high frequencies are almost non existent to the ear. This works on anything that produces sound.
When you\'re shaping a melodic line the volume will fluctuate between notes. Shouldn\'t we reach for the filter instead of the expression pedal ? Wouldn\'t this be more realistic? It sounds pretty fake to me to hear a trumpet/french horn playing ppp when you still hear the same high harmonics as clear as you did when playing ff. As volume decreases, so does the the \"apparent\" cutoff frequencies in the instruments.
Let me know what you think. I\'m only speaking from my own expirements.
[This message has been edited by esteven1 (edited 10-04-2001).]
Your questions are really good and I\'m anxious to hear how others do these things. I\'ll answer with what I do and why, but they\'re directly related to the software I use (Logic, which IIRC you use as well). Twaeking is an iterative process and one can teak til the cows come home and still not get it perfect. It always ends up being a diminishing returns decision. You also tend to be too close and can miss the forest for the trees. It\'s very hard to maintain perspective. It helps to take the time to listen to the whole thing once in a while.
1. I use the tempo list. This is because I compose on paper and I like what shows up on the screen to match what\'s on paper. Logic\'s tempo list is pretty easy to work in and I\'ve found that a pretty effective rubato can be approximated. This may be why I don\'t write solo piano stuff because it would be much more difficult to get the high degree of expressiveness the genre requires.
2. Logic has so many tools for tweaking volume (love that Hyperdraw) that I have little problem balancing my ensembles. Where I run into problems is the overall perspective. Definitely play as you would in an ensemble, save solo playing for solos. You can always go back and tweak or rerecord parts.
3. For me this depends on the piece. I try to get all parts perfected before I record the audio parts. At that point I\'m more interested in adding room sound and reverb, but if necessary you can automate fader moves to adjust the overall balance. Once it\'s done I bounce adding stereo enhancement and dither and I\'m done.
I am very curious to hear other answers to these questions, especially to the first question. About this, I was wondering if it possible to record a part with lots of tempo changes, and then tell the sequencer which is the first beat of each bar, and then have it adjust the tempo automatically. It would be a great feature...
Regarding rubato, Cakewalk (and probably most other sequencers) will take a simple part and line up beats and measures to it. In other words, you can play a simple line, run the process and it will automatically map out the tempo changes, beats and measures. This has varying degrees of success but it can work.
Wish I could offer some advice , but to be honest I am a novice so,I am more intrested in picking up good tips too.
But here goes:
I have just started with gigastudio. I tried doing some orchestral mock-ups from Dover scores of pieces I have studied and know very well. I chose four good recordings of each piece and then sat down with the scores and listened very carefully. Currently I am doing Venus by Holst and the first movement of Scheherazade by Rimsky-Korsakov. It is surprising how different they sound. The Rimsky-Korsakov, for instance, is in fact very easy to replicate (other than the solo violin) as the orchestration is so transparent the balance as indicated in the score (dynamic markings) seems to be reproduced with ease via midi. In fact, only a little tweaking with levels (more reverb) is needed to make it sound pretty close. The Holst, however, due to the density of orchestration (French horns, flutes, and two harps all playing the same thing) is another matter. For instance, on the Dorati recording the harp is impossibly loud, obviously the result of close miking. I have been going to rehearsals at the SF Symphony with Michael Tilson Thomas to become more acquainted with how much adjustment needs to be made and I was staggered as to how much alteration to the dynamics in the score was done by a conductor. It was also very intresting comparing the pieces live . For instance the other day they did Beethovens 4th followed by Strauss,s Ein Heldenleben .
Okay that,s comparing the band the Police with Emerson Lake and Palmer but the density of sound and the dynamic fixes required were very different . So it really does depend very much on the Material as to how you balance . And iwould think how you play it two . The may be a need with some passages to be more expresive in your playing to get tutti passages to work .
So, to that end, you really have to let your ears be the judge of what to change and how. There obviously are orchestration skills in terms of what instruments blend at what levels and I suspect (certainly from looking at the Rimsky book) that there is a science to the blending of some of the instruments in the orchestra but due to the restraints of where it\'s played, who plays it, and on what instruments, this is never exact.Also with the libraries we are using how to you know the will balance without help . I agree wholeheartedly that using filters instead of the volume is a much better approach. However, with the quality of some of these libraries, most obviously Dan Dean, I find at least in Sibelius the dynamic control of the samples yields great results in terms of realism.
As for rubato passages I too agree to enter the data as written in the score and use whatever you can to affect the tempo on playback, even with Sibelius\'s rather archaic system of MIDI messages I got a pretty realistic cadenza on the solo violin passage in Scheherazade .