I have to edit a video due Wednesday morning, so this will probably be my last post until next week -- I'll do my best! I'm only pursuing this because the lyrics portion of Finale has flummoxed many people, not just on this list. My composer colleague David goes into red-faced rages over it -- and he *does* use lyrics for most of his compositions. He's just not a computer guy, and the current Finale lyrics thing is computerese to him.
At 12:08 AM 9/20/02 -0800, Mark D. Lew wrote: >At 12:19 AM 09/20/02, Dennis Bathory-Kitsz wrote: >>A hyphen or space would just >>be a marker processed by the display system, and could just as easily be >>moused in place like a smart shape or word extension. Onward... > >Yikes! I sure hope you have something better in mind down the road, >because I sure as hell don't want to mouse in every hyphen. That would be >horrible. Maybe it's late, and you're only seeing the scare words. :) "Could just as easily" -- in other words, hyphenation would not be limited any more than the methods of note entry and adjustment are limited. That answer's Christopher's question about 1st/2nd endings with no destination for the hyphen ... drag it in place. It doesn't eliminate the existing keyboard hyphen syllable break, just adds methods of its placement. Remember that the hyphen in Finale today is already not a real hyphen (nor is it in a printed score, really, because it's not something that's sung) -- it's a control code that produces a hyphen string in the display (or a behavior from the singer). I'm suggesting additional methods of placing that control code so that it's not constrained to being attached to syllables. So you're typing along, using the hyphen command as usual, and you get to a repeat ending. Where does the hyphen go? Type it, shift-click-drag it just like a slur (but constraining vertical motion with the shift), and it can then hook to the ending, or a note, or a barline, or any object. Because ownership can be shown (another key concept I've been asking for for years), you never lose either end of it. You can turn "view ownership" on or off to see the rubber-band connections. To digress to this ownership thing: The ownership of the slur was one of the best moves Finale ever made, attaching as it does to two objects. And it shows what could be done for all objects that are not a fixed size. Fixed size objects have one point of ownership, those that can resize have *two* (or can have two, in the case of, say, text expressions whose endpoint has to be fiddled with in page mode). A hyphen with ownership and lyrics with ownership means you know where they go and how far they go. Okay, now back... >Yes, I skipped that because I couldn't make heads or tails of it. I know >nothing of audio recording editing, and the analogy was completely lost on >me. Modern audio and especially video editing belong to the class of "NLE" -- non-linear editing. A typical film montage during an action scene is an example -- many angles and zooms and motions and cutaways and sound and effects and voices are used from the original multiple camera shots and retakes. They are edited today into a single film without actually cutting any film or tape. The originals stay intact and you establish a 'window' into them. The desire to have lyrics be flexible (clones, copies, mirrors, etc.) leans in the direction of NLE. To accomplish NLE requires taking one's mind and tools outside the idea that the order of information during editing needs to be linear -- which, with repetitive lyrics and syllables and melismas and seques, it isn't and hasn't been for centuries. To accomplish this, audio and video editors have a pool (and the name differs, depending on the program) of data in which clips (audio, video, titles, effects, etc.) are placed/imported. (Finale's lyrics are a little like elements in a pool, except they misbehave.) Here is an example of how a simple NLE editing session might go in Finale: The data pool entry is "The cat in the hat came back", to be distributed to some group of singing lines. It might as well be the quintet from "West Side Story", which has lots of repetitions and overlays like this: Line 1 sings: The cat in the hat came back (straight melody) Line 2 sings: The cat, the cat (melismatically) Line 3 sings: Cat, cat, cat, came back, back, back (rhythmically) Line 4 sings: The cat came back (staccato, separated with rests) Line 5 sings: c, c, c, c, t, t, t, t, k, k, k, k (percussive, no pitches) Line 6 sings: Come back, cat (pedal) (Before you jump on this example, realize that I'm only using the short phrase to compress the explanation. Also, comma, period, semicolon, hyphen, space, etc., are all markers. Implementation differs as hyphens stretch by default.) Line 1 uses the entire data pool entry. Line 2 uses the entry slipped to "The cat" (narrow the editing windows's markers to only include those two words, and click-assign). Line 3 uses individual words from the pool entry (narrow markers to individual word). Line 4 uses two separated parts of the pool entry. Line 5 narrows the markers to a single letter (not just your grandfather's syllables anymore!). Line 6 could be a new pool entry for the phrase, or just for the word "come", or just a visual addition that does not create a new entry. This same data pool entry could also be applied as the title of the composition, appear in the copyright notice, and be included in the poetic text printed at the end of the score. It could also be copied and pasted out to some other application (and vice-versa). Save for the one new possible entry in line 6, *everything* is just a window of the same data pool entry. And, should this be a dialect version that you just missed, the change from "came" to "come" is accomplished destructively in the data pool entry! Voilą! Expand this concept to entire verses and choruses, or typically repeated phrases, and you see how you can slip edit the text into myriad versions (including application of font, style, color, etc.) without needing to change your text pool entries. >I don't understand the term "thumbwheel up". GUI thumbwheels. Maybe they have different names on Macs -- for example, in the staff dialog, you can go from one staff to the other using the droplist or the thumbwheel to the right of the droplist. Thumbwheel up creates a new text pool entry. >I don't understand what "notation edit" is. Are you talking about dragging >the syllables around? You asked what would happen if someone deleted two notes, or changed them, or re-added them. This would cause an ownership issue, so the display would need to show that the text was now present but unassigned -- something it currently does not do, hence the hal-le-lu-le-lu-le-lu-jah problems. >I'm also still not following how hyphens fit into >this scheme. Don't fixate on hyphens. They're only markers you apply with any of myriad methods. You can attach, disconnect, or reattach them at any time. >Are you envisioning baseline as a feature that can be edited from Text Box >window? It's a display element that could be applied to the text box *or* type-in-score *or" subsequent editing. It depends on how significant it is for your particular use. Put it in a text box and it's only changed through destructive edit. Put it in score and it's part of the present display/print state. >Well, I'm still fixated on the hyphen because I don't understand how you're >making them work. You said something about dragging it around with the >mouse, and that scares me. I don't want to have to place all the hyphens >myself; I want the program to intelligently calculate that for me -- which >(with the notable exception of crossing system breaks) it currently does >fairly well. It doesn't really do it well at all. I did a few songs for a publisher two weeks ago and used only type-into-score, with correction of typos there. So I don't know if Finale has gotten smarter with respect to adding hyphens from a dictionary, but I didn't see it automatically adding any hyphens for me. :) The current implementation is pretty crude, stupider than ordinary text editors have been for two decades. >suppose you're in non-destructive mode. A certain syllable is attached >in more than one place in the score. If I edit it in type-in-score mode, >how does it change the one assignment without changing the other? It changes what you see. In the default state, it would remain a non-destructive display/print edit. You could leave it at that, force it to create a new pool entry, or apply it (destructively) to all occurrences of the original text that was placed. >Do you >create a new syllable and add it to the end of the text stream? It depends on what you want it to do. You could set your defaults in any way that pleases you. You could even set defaults to emulate Finale's current destructive text edit, and you'd never know the difference. >>From one >>viewpoint, you can think of (say) MSWord's round-robin editing mode. >>Changes are not committed until after the edits are made. The base text is >>the pool, and the edits are maintained separately (and differently >>colored). It can get messy, but the result is solid, and accept-edits >>cleans up everything that's visible. > >I don't understand this. I don't know what MSWord's "round-robin editing >mode" is. When you edit a document in MSWord, there is a protected editing mode in which each person's edits are displayed, but only the person who made those edits can change them. This is great for collaborative document editing. All other folks' edits are locked to that person's password. Each person's edits are assigned a specific color, so you know whose changes you're seeing. Eventually, the document gets very messy to look at (since you're seeing everybody's changes in a different color). The person who originated the document with the master password can unlock it and apply the edits selectively or all at once. >I don't see why lyrics should be >combined with every other text. Because they *are* all text. If all features could apply to all text, there would be no need to learn separate rules for text and its placement or editing. All that would change would be the mode in which you were working, and the specific features of that mode you were applying to that manifestation of the text -- a title, a copyright notice, a line of lyrics, a bounded box of text, etc. You could choose the entry method that, as a new user, you find most comfortable. The type-into-score or text-box entry/edit modes could appear to emulate current practice (except where current practice has created problems). If any trouble were to come up, it would probably be with current Finale folks used to the workarounds they've become accustomed to. Until they learned to keep hands off most of the time, they'd be destructively messing with those text pool entries! >I don't see it that way. The current method, notwithstanding its various >shortcomings, still bears a basic resemblance to how lyrics (in my mind) >ought to behave. Lyrics notation doesn't "behave" any more than music notation "behaves". Music doesn't behave like Midi, yet there's that entry method. It doesn't behave like vertical score entry, yet there's that method. Lyrics don't behave like paragraphs, except where they do (recitativ). Paragraphs don't behave like notational elements, except where they do (instructions or activity boxes). The behavior is all about what you learned, and you learned Finale's lyrics in the way they were conceived by the programmers -- as somehow entirely distinct from the text out of which they were drawn. That felt right at the time, most likely. But once you send them "home" and forget all the contortions and accommodations you've made, the bulk of the confusion drops away. >You can create any number of strings of ordered text. The >basic indirection lets me view the text separately from the music. The >existence of the separate "verses" in which I can group texts lets me make >changes to a logical group all at once. The existence of an established >order of syllable makes it possible to do all-at-once click-assigning, >lyric-shifting, and intelligent placement of hyphens. Nothing I'm suggesting changes that. You can still create ordered text, if you want to call it that. You can still view it separately. You can still assign verses to it. You can still re-group text (or split it, or join it). You can still click-assign in any manner (using hyphen markers or using a dictionary, individually or as a block, and later move what it got wrong with a select-drag of given contents). You can still shift lyrics, and your Ominous Hyphen and word extensions have even *more* intelligence because they're no longer confused by endings, by fadings to silence, etc., and will obey where you want to put them (by default 'rule' or by specific choice where the default rule cannot apply). >What I'm hearing you say is that yes you're keeping all of that but at the >same time you're changing it all (which confuses me), and meanwhile you're >mixing it all in with every other kind of text in the file (which further >confuses me). Maybe your scheme is some wonderful new thing that will >magically solve all the problems, but I just don't get it. I do encourage you to explore NLE. Everyone who uses Sonar for audio editing who also had used Cakewalk (its precessor) had the same reaction. Nondestructive editing, nonlinear editing, slip editing -- there was Terror In The Customer Base. It was a new concept. Folks were just plain familiar with the destructive editing techniques, and had invented many personal workarounds and kluges and made many accommodations in their habits. In Cakewalk, to change volume or create fade in or out, you edited the sound clip. To start later, you edited the sound clip. To add reverb, you applied it to the sound clip and reloaded the edited clip. You could really mess things up if you edited the clip badly, or deleted a part you later wanted. In Sonar, you add an envelope which you can be adjusted to change volume or fade in. To start later, you drag the end of the displayed clip so that it just 'windows' what you want; it never touches the clip itself, so you can drag it back to open the window further. To add reverb, you apply a sound effect that processes the clip in real time, without touching the clip. You can use the clip over and over, or loop it, or mirror it. And when you've got a bunch of stuff just the way you like it, you can bounce it to another track, in effect applying the changes and creating a new clip -- but *still* without touching the original clips you derived it from. And now to those folks who resisted it, nondestructive editing is the best thing since sliced bread (a phrase that had no meaning to me until we started buying local bread with hard crusts). Going back to edit old material in the original program feels like putting a great weight on your back! >OK, but how does it know what syllable counts as the "next" one? You keep >telling me that there's no such thing as a "next" anymore. Right. Trust me. :) A hyphen is only a marker. It's not part of the inherent text (even in the current Finale). You can place markers in any way that suits you -- type-in-score, text box, etc. You could easily place *real* hyphens, which now is a pain. Or you could use a dictionary to place all the markers for you, and then assign the results (adjusting for mistakes). >Huh? I've never done Midi editing, but it doesn't sound "exactly like >applying lyrics to notation" to me at all. In fact, it sounds completely >different. Try applying Midi to an audio track or vice versa. They have different clocking methods, different approaches to making sound, and their own integrity -- just as notation and text are very different in how they apply to a score. In a sophisticated studio editor, the Midi tracks can be sent via a realtime software synth to behave like an audio track. That's good, but it's still simple Midi in origin. But then what if you have a vocal track (a recording of an actual singer) to apply to that. How do you sync them? How do you get the sung syllables to match the Midi? How do you apply the sung text (or the Midi) to each other? How do you replace a badly sung bit with another bit, perhaps from elsewhere in the song? How do you apply pitch correction? Volume adjustment? Compression? Reverb? There are more variables in these than in lyrics, but they are far easier to accomplish now. >I really don't >WANT to spend an hour tossing my lyrics around like a salad. I want to >just type them up in a logical order, assign them to their proper places in >the score, and then be done with it. But I'm hearing from others who want to copy lyrics, re-use them, shift them, assign them in different ways, etc. Apparently flexibility *is* an important feature of lyrics, so they (and maybe you) really *are* tossing them around. If all you wanted was to put them into the score, you could use the Graphire method, where every syllable is its own object. Period. You could try to type new lyrics over the old ones and it would create new overprinting, objects rather than erasing the current lyrics. It was pretty simple, but it never included the concept of text. It was all a programmer's concept of objects. Finale seems to have the worst of both worlds right now. >Anyway, I think it's a safe bet that Coda isn't going to make such a >drastic change to the data structure. Ya never know. They have competition now that they've never had before, and they've never been afraid to create new versions incompatible with previous ones in behavior. I don't even know how to run Finale 2.2 anymore! Dennis _______________________________________________ Finale mailing list [EMAIL PROTECTED] http://mail.shsu.edu/mailman/listinfo/finale
