Scrub playback is a really neat toy and if I’m not careful I’ll continue just playing with it all day However, after a few minutes of use I have already noticed some things:
When notes are spaced closely together, it is sometimes unclear where to put the cursor—apparently the right-hand edge of the loudspeaker is the actual location of the cursor. A thin vertical line like you get when dragging items in Engrave mode would make things a lot clearer IMO. Or perhaps just highlighting all the sounding notes.
Pizzicatos of course can’t sound continuously but it sure is jarring when a portion of the notes doesn’t sound unless you keep moving the cursor away and back. Maybe there could be options either to play them arco or just repeat them at, say, 60 bpm? I imagine the same goes for percussion instruments.
When you move the cursor backwards in time, there’s occasionally a few pixels where you can either get two sonorities sounding simultaneously, or staccato notes stop sounding.
The hot spot for the icon is actually dead in the middle of it. It is tricky in narrowly-spaced music. When we were working on it, we did look at trying to highlight the notes being auditioned, but it’s difficult to achieve, particularly because the data we’re actually using for scrub playback is the rendered playback data, rather than the notated data, and there’s often no one-to-one correspondence between the two.
However, I’m really pleased you’re enjoying using it. I worked on this feature, and even when I started working on it I didn’t think it would be one that I used myself, and now I use it all the time.
Yes, Frank, I see what you mean. It’s OK if you start with Shift held down, but not if you add or remove it while Ctrl+Space are held down. I’ll take a look at that in due course. (It works OK on Mac.)
At present the only options are ‘everything’ and ‘solo’. It might be nice if you could select some staves and hear only those when scrubbing, same as hitting P on a selection. I’d find that useful when transcribing from one ensemble to another, to check if some sonority matches.
I’m also not sure how I feel about chords playing in “dead” space. For instance, I was surprised when I got a full chord where my cursor is pictured (ie- not under any notes). Generically, I get the reason behind it as the meter change isn’t a “thing” musically/sonically speaking.
I hear a chord when I jump in cold at the spot pictured above, but also when I’m going backwards and I wouldn’t have expected to hear the notes as I drew backwards until I actually got to the notes. Not a big deal, but as you said, as I experimented more, I was confused at one point because I got a few “ugly” chords because the cursor was playing two chords at once as I drug through time signature voids. I was confused until it occurred to me what was happening.
I have another idea that only just occurred to me at this very moment:
I would love it if the playhead could appear at your current selection, and then you could use the scroll wheel to advance one voice column at a time. I already have the mouse in my hand, so my natural instinct was to start “scrolling” through the score as a more fluid alternative to dragging. I could imagine the icon snapping and scrolling just like the playhead the moment you engage the scroll wheel, and then jumping back once you physically move the mouse again.
Absolutely. Maybe today is a good time to observe how far technology has progressed. I know there are many who don’t really like the idea of playback having some priority in a product billed as producing notation. I like to have good sounding playback, even if not 100% realistic. But an often neglected benefit of improved playback is the ability to detect problems with one’s ears instead of (or in combination with) the eyes.
I know some folks have a sevant-like capability to look at the score and immediately spot the problems. I can’t do that. But I can hear problems, even if they are obscured in the mix. As the playback gets more and more realistic, it really helps me to produce better error-free notation.
I think it was only 8 years ago when a college professor of music proudly proclaimed that the only way to really listen well to the scores was to use a piano voice for everything. That was probably true in 1995, but even 8 years ago, we were well into the period when playback was realistic enough to be a valuable debugging tool when using sounds that were more authentic than a piano voice.
Today, we are so far beyond that point. It is not just about finding bad notes. I often decide to add articulations and slurs, or adjust dynamic markings, because the playback is close enough to real players to show how those elements improve the style of playing. There have been countless times that the accuracy of the playback has prompted me to adjust cutoffs to reduce clutter and produce a better musical effect. And the timbre of the instruments is close enough that we can make good judgments about the voicings across multiple instruments in the context of their various lines. The new feature to adjust dynamics as the musical line ascends and descends will make that even better.
Scrubbing is one more step in this very valuable direction. The full stereo sound is usually good enough to call my attention to problem areas, but it isn’t always obvious what needs to change. Scrubbing will help us to pinpoint the problem areas and quickly solve the problems.
Yes. Being used to Sibelius method of scrubbing, where you use the keyboard to move the playhead, I think it would be a step backwards to have to use the mouse for anything more than setting the initial location. I very frequently use back and forth with [ and ] to play a particular column over and over, and among only selected staves also.
But maybe it’s difficult to do this with scrubbing through a render. I guess I will try it out soon enough. I haven’t updated yet.
Another oddity I just found: all quarter-tones get rounded up to the nearest 12EDO pitch, regardless of enharmonic spelling. This results also in artifacts that I suspect are due to overlapping NoteOn and NoteOff commands for different notes that get rendered at the same pitch.
However, after a few minutes of use I have already noticed some things:
