Recording keyboards (basic question)

  • Thread starter Thread starter chee74
  • Start date Start date
C

chee74

New member
Hi all,

I'm a beginner to home recording and I have this question:

In professional/commercial recordings, are the keyboards usually recorded as actual wave files or just MIDI? My feeling is that they should be wave since it sounds much more expressive and fluid. Am I correct? MIDI sounds robotic at best - or am I not fully recognizing its powers? When would someone need to record in MIDI?

Thanks,
Eric.
 
Midi files are the means of storing the 'commands' to the keyboard - what keys to play, when to play them, how hard, etc. It's sort of the computer equivalent of a player piano reel.

Wav files store the actual 'sound' the keyboard makes.

You play your keyboard and record the Midi commands onto your computer via the Midi interface. Only the commands and keyboard settings are saved in the Midi file.

Later on you have the computer replay the Midi file. The commands come out of the computer and back into the keyboard over the Midi cable. Then the sound (analog signal) comes out of the keyboard line out and into the computer sound card. It's converted to digital and this 'sound' is saved as a .wav file.

Hope that makes sense. :)

barefoot
 
Hi Barefoot,

Thanks very much for taking the time to explain things. They make a lot of sense. Let me go further with my queries (hope you're patient enough to read through it):

Now assuming I have this basic setup:-

Keyboard's MIDI out hooked up to computer's MIDI in
Keyboard's MIDI in hooked up to computer's MIDI out
Keyboard's audio out hooked up to sound card's audio in

I use Cakewalk Home Studio 9 to do recording.

Assume I really like the grand piano sound on my keyboard. What are the basic steps I should take to record it to my hard disk? You seem to suggest that I record it with MIDI to the computer (playing on the keyboard) first, then have the computer play it back to the keyboard, whose output goes back to the sound card to be recorded. Is this what I should do? Can a computer playback and record at the same time? What (if any) are the advantages of skipping this whole MIDI thing and record the keyboard directly as a track (wave file) with Cakewalk?

As a side, if I use MIDI, would I be forced to keep an absolute beat (i.e. all measures have to be the same duration)? How do I deal with a more ad lib style of playing?

Thanks,
Eric.
 
I do alot of keyboard recording (Korg 01/WFD and a General Music Pro 1), and I never use MIDI. I just go from my keyboard outs into my Great River pre (I used the line input) and then out to my Delta 44 Soundcard. Don't worry about the pre, you could also go directly to the soundcard. I use N-Track to record the parts into .wav files. Assuming you are happy with the performance and the tone/sample/instrument, there's no reason you need to use MIDI. From what I understand of MIDI, you can then go back and change the tone/sample/instrument used, muck with the timing, fix mistakes, etc...Sounds like you should just stick with recording the sounds directly. Good luck!
-Evan
 
chee74 said:
What (if any) are the advantages of skipping this whole MIDI thing and record the keyboard directly as a track (wave file) with Cakewalk?

As a side, if I use MIDI, would I be forced to keep an absolute beat (i.e. all measures have to be the same duration)? How do I deal with a more ad lib style of playing?
Click to expand...

I'm not expert on all of this, and I don't use Cakewalk, but I believe you can record both the sound and midi information of your playing at the same time.

Midi doesn't limit the loosness of your playing. I'm not sure what the time resolution is, but it's very small relative to your reflexes.

The advantage of having the midi file is that you can manipulate the commands after the fact. You can replay the exact same line but switch to a new keyboard sound or even a different keyboard or soft synth. You can easily tweak individual notes in time or quantize the whole file (align it to the beat), and a bunch of other things as well. It's easier than trying to edit a wave file where the sounds might all slur into one another. Then you finally record it after you have it just the way you want.

barefoot
 
Hi,

Thanks Gordone and Barefoot for the responses. You guys are making the picture clearer and clearer for me.

I guess if I record with MIDI, I can basically record a song once (assuming it's decently played) and then do as much editing as I want afterwards. It helps to know the MIDI doesn't limit my looseness in playing.

With direct recording to wavefile, instead of editing afterwards, it seems it's easier to re-record the unsatisfactory parts (kinda like recording vocals) and then cut and paste together into a nice coherent piece. I think Cakewalk allows you to do that as well.

I think probably MIDI is probably more preferable for myself because it's more portable and more flexible (can change instruments in the future).

And just out of curiosity, Gordone, what is a pre? Pre-amp? Why do you need one?

:)
Eric.
 
Assuming you are happy with the performance and the tone/sample/instrument, there's no reason you need to use MIDI.
Click to expand...

There's plenty of reasons to use MIDI. One example -- say you play a song, it comes just as you want, and then a singer comes over to add vocals, but it's in the wrong key! If it's audio, you have to transpose the fingerings and then play the whole thing over again. If it's MIDI, you can select the track and transpose all the notes up or down as much as you want.

Likewise you might have some performances that have parts you want to extract and use as horn parts, or perhaps transpose them a fifth or fourth or whatever and create some harmonized lines. Instead of playing the parts, you could select just the group of notes you want, move them to another track, transpose them as desired, and voila!

As my final example, you have a part you've played, and you want to give sheet music to a horn player who's coming over to play on your recording. If you recorded your demo synth part as MIDI data, you can print up a sheet lickety-split.

Sure, a lot of these things can be done readily by an adept keyboardist and transcriber, but being able to do all these things can save a lot of labor. Not to mention that you can put together stuff painstakingly if you don't have any keyboard skills-- or you don't have hands...

Think about the difference between typing letters with a typewriter and doing it with a computer -- being able to save, edit, change, copy and paste at will, reuse material, etc. Using a MIDI sequencer is a lot like that.

I'm not expert on all of this, and I don't use Cakewalk, but I believe you can record both the sound and midi information of your playing at the same time.
Click to expand...
Absolutely.

I'm not sure what the time resolution is, but it's very small relative to your reflexes.
Click to expand...


In most applications it's adjustable, and there's a maximum (smallest units) resolution. In SONAR, it's 960 ppq (parts per quarter note. That is, for every quarter note duration, you can divide the beat up into 960 ticks.

Let's see... at 120 bpm, 4/4 time, a quarter note lasts .5 seconds. So each tick would be 1/960th of a second. The time between ticks at that tempo and time signature would be about 0.0005 seconds, or half a millisecond. A very small instant indeed, quite negligible in terms of the accuracy of the music. The difference between timing caused by players reacting to other players several feet away is typically more than this. Sure not enough to disrupt the groove too much...

From SONAR's help system:
"In some projects you may need a different timebase. For example, if you wanted to use eighth-note septuplets (7 eighth notes per quarter note) and represent them accurately, you would need to have a timebase that is divisible by 7, such as 168PPQ."
 
AlChuck gives good advice. I've been recording in one capacity or another since way before MIDI (we actually used to record on something called "tape"). I've been using Cakewalk ever since version 1 and currently use version 9 - and at some point will convert to Sonar (I've heard several people indicate there are still some "bugs" with Sonar).

I always record everything I possibly can to MIDI and normally only record the actual audio when I'm ready for my 2-track mix. As others have indicated this gives you the ability to tweak sounds right up to the final mixdown. Even when we do choose to record audio direct if possible we still record the MIDI performance, just in case we need to make changes to the sound.

I do the same with drums (using v-drums) and have started to do the same basic thing with guitar - recording a clean sound and then feeding the clean sound into a POD to have the ability to change the guitar's amp sound at any point.

The only down side (and it can really be a problem) is that if you can delay all the decisions regarding keyboard sounds, drum sounds and guitar sounds till the final mix you can hit a wall at that point (indecision can be a terrible thing).

There are times when I miss recording to a 4-track and being forced to get a sound, record it and move on.

Regarding the question on "what is a pre-amp" A pre-amp is more complicated than this brief description, but the easy answer is - a pre-amp provides a "power boost" to an audio signal. Normally, a pre-amp is used with a condenser mic to provide power (since the signal from a condencer mic must be enhanced) However, many people prefer to use pre-amps to provide a stronger cleaner signal into a computer rather than the more limited converters on a sound card/interface. It is getting to be more common to record everything through pre-amps (rather than a mixing console) and only come through the console for the final mix.
 
Thanks for all the tips.

I have a Tascam 4-track Portastudio mixer, which is supposed to have pre-amp capability on each channel. The way I plan to record my keyboard is first to plug it into one of these channels, and then channel the output to the sound card.

Is this an ok setup? What's the difference between this and plugging my keyboard's out directly to the sound card in (can't I just use the volume control on the keyboard to amplify my signal)?

:)
Eric.
 
I suspect you only have one mic preamp on that Tascam and it's for a dynamic mic only. It's also not a very good mixer so you would be better to put the line out of your keyboard straight into the soundcard if you can. Experiment and see how much noise the extra signal path through the mixer adds.
 
If you find the direct signal from your keys to the computer to be lacking in signal strength - and if the Tascam adds too much noise, you could consider a stand alone preamp. While you can spend a whole lot on some preamps (or the current rage - "channel strips) you can buy a low cost preamp like the ART MP or the DBX mini pre for about $100.

There have been several posts in the "rack forum" about mic pres.
 
You must log in or register to reply here.
Back
Top