I think there is a misunderstanding how the Reverse function is supposed to work. It does not work in a way of mirroring the events (“optical reverse”). When you perform an optical reverse, the musical result is wrong, if the length of the notes is not quantized. We had complaints, that the output would not sound in time on a optical reverse. (NoteOff-position becomes NoteOn-position). So we use the start-position of the last note as an “anchor” and only reverse the start-positions of the rest of the notes. This gives the best musical result.
What do i want to reverse…? the order of the notes and the place they start, the relative distance between them…? If everything is quantised (starts and ends) then all is well… but on notes that have no length quantise applied, if I just want to try an ‘optical reverse’, it may not turn out too good…!?! Then that’s my problem (not the programs problem).
Or, its the relative timing (rhythm) between the notes that’s attractive… so how should a reverse in this case work…? The devil I guess is in choosing the ‘anchor point’ and how to deal with ends of notes becoming the start of notes.
So, its a bit of trial and error - in that some pieces will work and others absolutely won’t…!! Trying to wrap my head around how the start of the notes works when calculating its reverse (if you’re ignoring its end point…).
I’m not sure the answer is to tell us that the new version of reverse is more musical. For me it’s not. It’s not as inspirational or different as the optical reverse technique.
Please can you reinstate optical reverse, for creativity’s sake!
Or, its the relative timing (rhythm) between the notes that’s attractive… so how should a reverse in this case work…? The devil I guess is in choosing the ‘anchor point’ and how to deal with ends of notes becoming the start of notes.
The two main ways to reverse are firstly, simple note reversal which doesn’t affect any other midi data so A C E becomes E C A with note lengths, pitchbend, expression etc. left positionally intact and
Tape type reversal like tape played backwards. This is what I call “mirror” reversal and where the starts of notes becomes the ends.
May be a case for an option. Simple note reversal can be set with the Transformer however.
You’ve got me there! How? (the closest I can see is to calculate the average between the lowest and highest selected note and use that as the center value from which to “Mirror”… but I’m pretty sure that doesn’t give simple note reversal)
Oh! Surprise to me. I would have thought that was basic to it. Just a reversal of note sequence end to end as in: 12345 becomes 54321 given a range of course so the average would fall between the first and last notes including the two as in 123456 becomes 654321. I know, and can’t remember if it’s still a preset for the Input Transformer that it will do that for the keyboard input so you can reverse the keys which I used to find quite handy for inputting drums. As it can reverse the keys I assumed it could also reverse the notes.
It’s a while since I looked at the transformer so I’ll check what I’m talking about. An maybe called for.
I know simple mirror reversal can be done both ways, as in the movies “Mirrors” where the gaffe at the end is that evrything gets mirrored (sh… gave away the ending) but they just turned the stencil around on a police car so it read ecilop but in a real mirror the “E” would still be at the end. The letters would be backward and not the whole word.
So I guess I’m talking about the whole word ie: between the first and last or a given range of notes.
See what you mean.
I’m talking a different mirror then. Note event reversal I suppose? Where just the notes are reversed leaving the langth values, say, untouched so a
A [quarter note at position 1.000] and a B [8th at position 2.000] reverse and ends up as B [quarter note at position 1.000] and A [8th at position 2.000]. Cubase playing it just as you would play a piece up and down yourself. Note lengths left intact and at the starts they were. Daaah dit dit dit does not become dit dit dit Daaah or tid tid tid haaaD etc.
What I was asking for was regular retrograde, which means that the pitch and duration of the last note becomes the first, the second to last note becomes the second, and so on. This is only works out in Cubase when you quantize the lengths of all the notes, and it can’t be too complex. (Really I never tried using it again after making this thread, and forgot the quirks.)
So, there really is no retrograde function in Cubase, and that’s what I wanted. Finale has a very nice plugin that lets you do simple retrograde, or add in a transposition or inversion, all in one dialog box.
I am using Finale more now, the whole note-based thing is a whole different mindset and I am into that, plus I am only writing anything for live musicians these days. (I can see Conman nodding…) (Hi vic)
I think that Cubase would do the Time Mirror correctly if it simply didn’t bother to swap note-ons and note-offs… just reverse the positions of the original note-ons and give them their original lengths. Of course this doesn’t sound like a reversal (i.e. compared with the reversed audio)
I totally understand… (Logic Pro gives you two ways to reverse since Logic Notator) When Cubase VST was around the pitch and position reverse worked just like :
We need options for this function…I repeat… We need options for the “REVERSE FUNCTION” please.
hint: New name for the reverse function"PITCH AND POSITION REVERSE"
Thank you
(without having to blow the cobwebs off my old Atari, as suggested in that link )… I’ve just had a quick look at how Digital Performer handles it, and it is perfect! It offers two options… Retrograde (which is the “optical” mirror), and Time Reverse (which is similar to what Cubase does, but not identical)…
Observe the following screenshot. The 2nd group of notes is the Retrograde of the 1st group, and the 3rd group of notes is the Time Reverse of the 1st group (Cubase would have placed that whole group one beat earlier)…
Thanks for the screenshot.
If those two should be considered options then I propose a third where (and I’m sure it worked that way on the Atari but I might have lost a brain cell there), in the second example the seen events would not change but the notes would. Just as a player would play them himself.
As you say, Cubase would put that third set a beat earlier but reversing the event lengths, to my mind, should not be the first choice or only choice option.
Players would probably prefer my example while music programmers might want something a bit more radical in the palette and prefer the other types.
My way would mean that you would not have to have all the notes the same length to get a true reversal.
If I read correctly what KDEF2004 said above, the straight-forward retrograde process found in Atari Cubase was still the way it was done in Cubase VST. If so, and if you still have that available, your Atari’s cobwebs could stay undisturbed. (I’ll try to remember to try Cubase VST, the next time I want to reverse a MIDI part that way.)
Haven’t quite understood exactly what you mean there, but when I’ve done reversals in Atari Cubase, I’ve got the equivalent of a mirror image in the key editor (ie a true retrograde, and what I wanted).
How? (the closest I can see is to calculate the average between the lowest and highest selected note and use that as the center value from which to “Mirror”… but I’m pretty sure that doesn’t give simple note reversal)
maybe called for.
See what you mean.
When Cubase VST was around the pitch and position reverse worked just like :
)… I’ve just had a quick look at how Digital Performer handles it, and it is perfect! It offers two options… Retrograde (which is the “optical” mirror), and Time Reverse (which is similar to what Cubase does, but not identical)…