Up Strum

[Surfwhammy]

You've pointed out that impracticality of using a sample library to capture all the various sorts of guitar chords. However, NOTION could fairly easily support a smaller group of chords (Maj, min, min7, sus4) and vastly improve the sound quality.

This is an intriguing idea!

For this type of activity, it works a lot better when I do as much researching and thinking as possible before deciding on a strategy, so after doing a bit of enumerating and realizing that one of the possible solutions literally could require as many as 20,000 individual samples, which makes it quite impractical, your suggestion reminds me of the way the studio drumkits in SampleTank are mapped, where each drum and cymbal is played by a different note, where for example C4 might cause a hi-hat note to be played but C3 might be a snare drum rimshot . . .

The way one of the more expert SampleTank users explained everything in replies to my questions in the SampleTank section of the IK Multimedia FORUM, when a note is appended to the file name, this tells SampleTank how to map the sound in the WAVE file, but if a note is not provided, then SampleTank analyzes the sound in the WAVE file and maps it to an arbitrary "root note" . . .

Another useful bit of information is that the sound in the WAVE file can be a single short note or a series of notes, as well as a loop, and for a longer series of notes SampleTank simply continues to play the WAVE file so long as the key on the keyboard is pressed, which in music notation maps to the duration of a note, where if I understand this correctly when there are four quarter notes in the WAVE file, a whole note in Notion 3 will cause the four quarter notes to be played, but a quarter note in Notion 3 will cause only the first quarter note in the WAVE file to be played . . .

This led me to the idea that one can play a full chord and map it to a single note, since SampleTank probably looks at the content of a WAVE file as being a generic sound . . .

After pondering your idea for a while, I think that there probably are not so many chords that are needed for what I consider to be a typical DISCO, Heavy Metal, Pop, or Rock and Roll song in a specific key, although DISCO songs sometimes have Jazz chords, but so what . . .

So what!

And for the most part, I play songs in a small set of keys, mostly {E, A, G, or C}, where I use "key" in a very simple way that probably but not necessarily maps to the actual key in terms of music theory, since I consider chords in a more absolute mathematical sense, where it does not matter to me what the lowest note happens to be, except when I can use it to map the chord to a bass guitar note, at least when there are four or more notes in the chord, even when some of the notes differ only by an octave, where so long as the chord sounds good in the song it makes no difference to me what name one uses to identify the chord, except that I prefer simple and easily remembered names for chords, really . . .

Really!

For example, if I play "Louie Louie" (The Kingsmen) with the chords A Major, D Major, and E Minor 7, then I consider it to be in the key of A, since the first chord is A Major, and the simple numerical pattern is 1-4-5-4 . . .

"Louie Louie" (The Kingsmen) -- YouTube music slideshow

If I play a slow song with {C, Am, F, G}, then I consider it to be in the key of C . . .

A lot of Beatles songs are in G and C, so I play songs in those keys, but E and A are very nice for electric guitar, since they work nicely for doing open-position stuff . . .

Based on the standalone user interface for SampleTank having a mini-keyboard with 88 keys, I think this defines the practical upper limit for notes, but it makes a bit of sense to use a smaller range of notes, since it requires too many notes above and below a treble clef to represent 88 notes, which makes two or perhaps three octaves more practical, with two octaves being very practical, since this ranges from Middle C (C4) to the High C (C5), which is on the treble clef, and then to the next higher C (C6), which is not so far above the treble clef as to be impractical to use, and this maps to 25 notes chromatically . . .

Intuitively, without actually going through the exercise, I am not certain that I can name 25 different chords in the key of E without wandering into some truly strange chords, although it depends on the chord pattern . . .

For the most part, the primary types of chords will be Major, Minor, Major 7th, and Minor 7th, and what I call the "base" or "foundation" notes in the key of E will be {E, F#, G, A, B, C, C#, D}, so with 4 types of chords and 8 "base" or "foundation" notes, this maps to 32 chords, which is not so many more than 25 chords . . .

If an F is added to the set of "base" or "foundation" notes {E, F, F#, G, A, B, C, C#, D}, then with the 4 types of chords this is more than sufficient for a lot of songs in (E, G, A, C} . . .

Intuitively, I think it is sufficient to play most of the songs the Beatles recorded, and it is sufficient to play a lot of Motown songs, as well, although to play the songs accurately it requires at least two flavors of some chords, but this can be done with a second and third set of samples, which is consistent with the way I do "sparkles" . . .

For example, I think it is practical to have three staves for a rhythm guitar part, where the first staff is focused on open-position and lower-position Barre chords; the second staff is focused on middle-position Barre and what I call low and middle "tight" chords (which are four-finger chords on adjacent strings); and the third staff is focused on higher-position Barre and "tight" chords (both low and middle for the "tight" chords) . . .

So, instead of having all the notes for a chord in the music notation, the music notation simply will have single notes that cause a specific chord samples to be played, which makes it very easy to do a three stave chord pattern based on the specific chords that need to be played at any given time, and it also makes it possible to stack chords, which is an intriguing concept and was something the Beatles did . . .

There should be a logical way to map chords to keys on a keyboard so that it will be reasonably easy to remember, but I can have a note card or something similar as a reminder of the "chord to note" mapping, although I am intrigued by the idea that there probably is a logical mapping for mnemonic purposes . . .

On the other hand, since there are only four types of chords, it might be simpler to have 4 staves for a rhythm guitar part, where each staff is a chord type and the notes on a staff are the "base" or "foundation" notes for a chord, which might be like this for "Louie Louie", which has the advantage of keeping the number of samples smaller (back to two octaves or 25 notes):

[NOTE: Most of the time I play an Em7 for the "5", although sometimes I play an E Major for the "5", but I prefer the way the Em7 chord sounds . . . ]

Code: Select all

Major        A A A D D       D D
Minor
Major 7th
Minor 7th              E E E


Intuitively, I think that there probably is a specific duration for a chord that will work nicely with several different tempos, especially if the chords are run through AmpliTube 3, where I can control amplifier and loudspeaker sustain and other effects, including slicing and echo repeats, which overall maps to needing to do 100 samples (25 samples for each of the four types of chords), which is not a lot of work . . .

And instead of having the four sets of downward strums, I suppose that there can be another four sets where the chords are played with upward strums, thereby increasing the total to 200 samples, which is within a practical range, although it requires 4 more staves, which nevertheless keeps it withing the range of "sparkling", since I occasionally spread the notes of a single instrument over 8 staves when "sparkling" the instrument . . .

By playing the chords on a Stratocaster or Telecaster in a very clean and balanced TONE setting, this makes it possible to run the Notion 3 generated chords through AmpliTube 3 once it is in Digital Performer 7, which makes it practical to create a virtual festival of different guitar TONE styles . . .

I like this idea, but I need to ponder the combinations, permutations, mathematics, and geometry for a while, since it is important to consider panning locations, which for a far-left panned set and a far-right panned set maps to 16 staves, but this covers top-center, as well, since playing a far-left and a far-right note simultaneously places the sound at top-center, which is easily heard in the current version of the "basic rhythm section" for "(Baby You Were) Only Dreaming" (The Surf Whammys), for which the link is provided in my previous post, where for reference there are two snare drum rimshot staves, where one is panned far-left and the other is panned far-right, so when you hear snare drum rimshots at top-center, this is the result of playing a snare drum rimshot on both staves (far-left and far-right) at the same time, for sure . . .

For sure!

NOTION already "understands" how to play various articulations via MIDI. The various guitar articulations (bends, etc.) certainly sound like they're implemented via MIDI.

This only appears to work for the Notion 3 guitar, but it works very nicely once you experiment with it for a while . . .

For purposes of clarity, I am referring to the Notion 3 guitar that supports having both a treble clef and a guitar tab, where for this particular instrument there is a virtual guitar fretboard that appears at the right side of the Notion 3 workspace . . .

I tried to do this with one of the guitars from SampleTank, but no guitar tab staff appeared, and the virtual guitar fretboard did not appear, either, so I think that there are some additional "smarts" that travel with the Notion 3 guitar . . .

All the standard instrument articulations and dynamics work with the SampleTank guitars, but as best as I have been able to determine the string bends and whammying articulations only work on the guitar tab staff, so I think that the Notion 3 guitar has more stuff, somewhere . . .

Yet another reason for thinking this is that the Notion 3 Bundled Electric Guitar sample file is approximately 350MB, which is nearly 6 times the size of the largest IK Multimedia guitar sample file, and it is several times the size of most of the other Notion 3 Bundled instruments, with the exception of the Notion 3 Bundled Acoustic Guitar, Drum Set, and Piano samples (.prox files) . . .

Lots of FUN!

[dcuny]

SoniVox has a number of guitar sample libraries that aren't comprehensive, but certainly sufficient for a lot of work. For example, for the Stratocaster, they have major and minor chords with down, up and picked strum options. With the acoustic guitar, they have major, minor and dominant 7th chords with up, down and muted strums.

There's also a Les Paul with major7, dominant7, minor7, min7b5, and diminished7 chords, but (apparently) only with a down strum. But the demo lacks the conviction that their other guitar libraries have.

It's worth noting that these libraries aren't particularly large - about 30Meg each. One reason for that is that a lot of the samples are reused. For example, with the acoustic guitar library, I believe the Em samples are used on all the associated barre chords (Fm, F#m, Gm, G#m) and the Am samples with their barred equivalents (Bm, Cm, C#m, Dm). Since the chords are normally barred with these chord shapes, the sound is still realistic.

Clearly Notion has the technical skill to add this sort of thing to NOTION or Progression. The addition of sampled strums would certainly raise the bar. If they went that route, they'd no doubt leave out some chords that someone considered critical. A baseline of sampled:

• Major
• Dominant 7th
• Minor
• Minor 7th
• Sus4

with up/down/muted options would really raise the bar, especially for Progression. The addition of Maj7th and add2 chords would also be quite nice...

But for starters, just adding a MIDI up strum would be helpful.

[Surfwhammy]

A baseline of sampled:

• Major
• Dominant 7th
• Minor
• Minor 7th
• Sus4

with up/down/muted options would really raise the bar, especially for Progression. The addition of Maj7th and add2 chords would also be quite nice...

I like this, and it has the potential to be mathematically elegant!

Until this discussion, my general perspective was that it was important to know a lot of chords, but after pondering it a while with respect to determining a practical set of guitar chord samples, I am not so certain that one actually needs to know a lot of guitar chords to be able to play thousands of songs in the DISCO, Heavy Metal, Pop, Rhythm and Blues, and Rock and Roll genres, provided you can focus on a small subset of keys . . .

[NOTE: If you need to be able to play a song in any key, then you need to know a lot of chords, but if you limit the key to a small subset of keys, then it does not require so many chords . . . ]

As a bit of an aside, this is one of the reasons that I do not complain about Notion 3, because from my perspective it is like reading hundreds of books about chemistry and understanding most of the information in the books but never actually having a chemical laboratory where, in this analogy, metaphor, or simile, (a) music theory is the set of chemical books and (b) Notion 3 is the chemical laboratory . . .

Having the chemical laboratory is considerably more important to me than being restricted to using only 25 test tubes at a time, since I can devise a strategy for doing elaborate experiments in such a way that I never need more than 25 test tubes at a time, for sure . . .

For sure!

I know a lot of chords, but I never use many of them, because they are so strange, although I am starting to use more of them, since most of the stranger chords are what I call "tight" chords, which are chords that are played with four fingers on adjacent strings, where one of the things I do is to work on playing "chord scales" with them, which I suppose is an exercise in working with four-part harmony but with "tight" chords instead of voices . . .

Years ago, I had a bit of free time every once in a while on a supercomputer, so I wrote a program that computed every possible combination of three or more notes (up to six notes) that that could be played on a guitar with 22 frets, and it was an interesting exercise on many levels . . .

The output of the program was a set of printed pages that had three fretboards per page, where each fretboard showed the relevant notes, and as I recall there were approximately 150 pages . . .

For example, there was a fretboard for the major triad, and there was a fretboard for a minor triad, and so forth and so on, but the program did not included chord names, since I was not interested in making sense of chord names at the time . . .

Instead, it was focused on the 12 chromatic notes, so it basically was an exercise in combinations and permutations, and the fretboards did not show how to play the chords. Instead, they showed where the notes were located on the fretboards . . .

And the program did not exclude any combinations, so for example there was a fretboard for the three notes {E, F, F#}, so quite a few of the chords were very strange, where "chord" in this context maps more to "combination", although on a grand piano, I suppose that {E, F, F#} might be a chord if the definition of "chord" is three or more notes . . .

After studying the program and printout for a while, I decided that there are a lot of chords and that it did not make much sense to try to memorize all of them, really . . .

Really!

I have been pondering the idea of doing something similar with RealBASIC on the Mac but with a different focus, and this might fit nicely with the concept of a set of simple chords, which relates to the discussion we had regarding six of the chords from the Surf Whammys song "Starlight", where there were different names for one of the chords, depending on whether the lowest note was considered to be the "root note", as well as whether four or five notes were played . . .

Noting that I definitely have a non-standard system for identifying and naming chords, which includes referring to the following pattern as the "Tighten Up" chord, I am quite curious about the idea that a chord viewed from one perspective can have an entirely different proper name when viewed from a different perspective, which intuitively refers to inverted chords, where the notes are the same but the order is different, since the lowest note is not the "tonic" . . .

So, for a C Major triad, there are several permutations, of which these are a few where the rule is that each successive note is higher than its predecessor (otherwise on a grand piano, they would be identical when constrained to one octave):

{C, E, G}
{C, G, E}
{E, C, G}
{E, G, C}
{G, C, E}
{G, E, C}

Everything is a bit simpler on guitar, because instead of having absolute notes (where for example there is only one "Middle C"), there are duplicated notes, albeit with different intonation and location, and some permutations cannot be played, for example, {open E on the low-pitch "E" string, G at the 3rd fret of the low-pitch "E" string", and C at the 8th fret of the low-pitch "E" string} . . .

Since I know more about playing guitar now than I did when I wrote the program that simply showed where notes were located without considerations for whether the notes actually could be played as a "chord", I can add some smarts to the algorithm to constrain it to chords that are practical to play, and I think that there should be a way to do a mapping of chords from the perspective of what one might call "equivalence", since I am not certain what the proper name might be, if there is a proper name . . .

My thinking in this regard is based on an attempt last year to make sense of the seven modes (Ionian, Dorian, Phrygian, Lydian, Mixolydian, Aeolian, Locrian), which makes a bit more sense than it did but continues to make no practical sense to me, with the consequence being that I decided to stop trying to make sense of it and instead switched to focusing on making sense of scales, since there are more scales than there are modes, and a lot of the scales I like do not fit with any specific mode . . .

Yet, the intriguing aspect is that one can play a major scale at a different position, and its notes are the notes of a mode somewhere else, as explained in the following formula:

Code: Select all

Tonic of Ionian scale (Major Scale) transformed to a Specific Mode =
   (8 - Specific Mode Degree Relative to Ionian Mode) + 1*

[*NOTE: The interval is flatted or diminished for the three minor modes (Aeolian, Dorian, Phrygian) and the single diminished mode (Locrian) . . . ]


For example, if you play an Am chord and play an Ionian scale (Major Scale) on G, then it is A Dorian, since Dorian is the 2nd mode relative to Ionian, where the 1st mode relative to Ionian is itself, and the notes of the A Ionian scale are {A B C♯ D E F♯ G♯ A}, so to transform A Ionian to A Dorian, the third and seventh intervals need to be diminished or flatted, which creates the scale {A B C D E F♯ G A}, which if you set the tonic to G will be a G Major Scale {G A B C D E F♯ G} . . .

[NOTE: For the Am chord, G Major scale, and A Dorian example, I am quoting myself from a post I made in a discussion about musical modes, scales, and Flamenco in the GuitarZone.com FORUM during the time when I was interested in making sense of musical modes and scales . . . ]

While I simply do not think in terms of the seven musical modes and would not play an Am chord and then play notes from a major scale on G toward the goal of being A Dorian, it is an intriguing concept mathematically . . .

And this what makes a bit more sense to me, which is focused on scales rather than modes, where the table actually has an additional column, which contains all "1", which is the octave for the various scales, but the image is clipped on the right, so the additional column is not shown:

[NOTE: Minor modes and scales are shown in red, where "minor" is defined to have at least a diminished or flatted third . . . ]



And on a related note, I did another calculation and determined that there are 1,287 scales with 7 notes and an octave of the first note, which was when I decided that focusing on scales was not so practical, either . . .

So, my thinking on this is that if there is a formula for determining how to play the notes of a major scale (which by definition is Ionian) but actually be playing notes in an entirely different musical mode, then perhaps there is a corresponding formula for chords, which is an intriguing concept, and it is something that I should be able to determine with a computer algorithm, which also can compute what happens with what I am calling "chord stacking", where one guitar plays a chord but another guitar plays a different chord, with the result being a "stacked chord", which is intriguing, because while it cannot be done in real-time by one person on guitar, it can be done on a grand piano, with a few caveats, where one can play 5 notes with each hand, for a total of 10 notes . . .

This is germane to Notion 3, because Notion 3 does not set an arbitrary six-note limit on the number of simultaneous notes that a virtual instrument can play, and several virtual instruments can play quite a few simultaneous notes that are very different from virtual instrument to virtual instrument, which in some respects is a bit like playing multidimensional space chess or something similar . . .

And while I do not understand all the music theory, I can determine when it sounds "right" by listening to it, which is the way I composed the chorus for "(Baby You Were) Only Dreaming" (The Surf Whammys) . . .

Lots of FUN!