> I like the idea about the mixer replacement. You could make a button "advanced"
> which could switch the mixer to an avanced mode.
I'm not sure yet how possible this is, but I'm looking into it. It's obviously possible to make the mixer a separate app, but that wouldn't be as elegant as replacing the media kit mixer.
If the mixer was reasonably flexible (e.g., multiple busses, aux sends, and direct outs on each buss), would there be any need for a Cortex-like app? Seems like it'd be easier to do all the routing on the mixer, with just a few mouse clicks, rather than pulling virtual wires all over the place.
Of course, the existing Cortex would still work. I'll see if I can make the mixer and multitracker reasonably friendly towards being disconnected at random -- some existing applications tend to lock up and crash when you do that in Cortex.
> "Recording multiple takes " Maybe it would be handy to have a feature similar to this;
> you select a region you want to re-record, the seq. will record but not overwrite your
> input but save it in seperate ... (tracks whatever) so you can choose wich take you like
> best by means of a switch.
I see. Using sequencers like Cakewalk, I tend to use lots of tracks for this purpose. Since Cakewalk is entirely keyboard-navigable, I can record multiple takes by muting all the tracks I want to record, hitting R to start recording, then when I've finished a take I hit space, down, and R again, then play the next take. It'd be even more fluid if it supported MIDI control of the transport, but sadly Cakewalk 3.0 doesn't.
If the sequencer was at least this fluid in terms of the workflow, would there be any need for automated multiple-take features at all? I'm mainly thinking in terms of reducing feature creep here, but if it's still a desirable feature, I could implement it.
> "displayed in sync " With that I mean that the line showing the current notes being
> played should be accurate and to accomplish this you should be able to adjust the
> display of notes being played to avoid a difference between what notes you see
> being played and heard by your midi gear (caused eg. by latency of that gear).
Displaying the line in sync shouldn't be any problem; this isn't windows after all!
Given that it's only a single column of pixels, I'll be updating the line at the same refreshrate as the monitor and vsynced, so it'll hopefully move very smoothly. I might try and highlightthe notes in some way when the line intersects them.
As for latency, I'm not really sure if that's a problem. Real hardware is extremely responsive; even slow gear like the JV-1080 is still quite quick -- within 2 milliseconds or better, so invisible to the eye.
Software synths under BeOS are also fairly responsive; even with poor soundcard drivers and badly designed synths, I've never seen worse than 20 milliseconds, and with good ones it's usually closer to 5 - 10ms. That's still less than a single video frame, so again, invisible to the eye.
There will probably be a time +/- parameter for each track, to allow sliding of the track ahead or behind in time. This could be used to compensate for latency, though it's most useful for producing echo effects, and making sure pad chords trigger slightly ahead of each beat to avoid bogging down slow MIDI gear.
> Transport controllable by MIDI is an excelent feature but standard logic like play/stop
> by pressing space etc. is also very wise. what I don't like is 'page-flipping' Sequiter does it
> too, I think it's better to have a fixed line and the notes and grid moving underneath it.
I did consider this, but there are two problems: 1) It gives the user the false impression that the visible range is somehow tied to the transport (which isn't the case; the sequencer can quite happily be playing bar 100 while they're editing bars 2 - 10), and the big reason 2) Smooth scrolling means updating the screen 50 - 100 times per second. Invoking the blitter chip sometimes causes slight timing delays (a millisecond or two), so doing this frequently might upset MIDI timing, or even cause audio stuttering.
It might be available as an option; I'll have to look at the issue a bit more to see if it's possible to take any shortcuts; e.g., just drawing the edges of the notes as they move, rather than redrawing the entire note in its new position. Perhaps the timing problems can be addressed in the video drivers; I may have to whine to Rudolf about it!
> 3Dmix; hold down, shift/alt/ctrl/windows key and drag with the mouse
Well, that feature's not available to me then -- I don't have a windows keyboard!
> "phase-adjusted panning" What is that? I don't recall hearing any stuff I didn't kike
> and I use Genelec 1029A monitors.
It tries to simulate what happens when you've got two mics arranged as a stereo pair. A sound source in the centre arrives at both mics at the same time, but a sound source on the left arrives at the left mic slightly ahead of the right mic (the delay depends on how far away they're spaced; usually not very far). The delays are in the millisecond range, which means that phasing problems can result if the two channels were ever combined. Fine on headphones (no L-R leakage), marginal on monitors (the right ear can hear the left monitor slightly, and vice-versa), and terrible if it's ever played on a mono system.
I imagine this reason is why mixers still use simple pan pots (simply attenuating the left or right channel) for stereo positioning rather than fancy phase adjustment.
> By 3D Mixing I mean the actual 3D displaying of the channels, could be used for surround IMO as well.
That might be a lot of work to get going, though if it was instead of a traditional mixer rather than supplementing it, it might be possible. Would 3D mixing be preferable to slider-based mixing?
In some ways I think that an approach like this might be better than trying to exactly mimmick a hardware mixer, although the latter has its merits in terms of proven usability and features.
I really hate the current "virtual studio" fashion where every piece of software is designed to look exactly like its hardware counterpart, down to photo-realism. That leads to a usability nightmare. Even a small physical rack unit is larger than the biggest computer monitor, and most hardware controls (knobs, etc.) were never designed to be manipulated with a mouse. This is more of a problem with things like synths and effects units than mixers, but there's even a problem with mixers sometimes; compare the Pro Tools mixer (not *too* cluttered though not great) with the Cubase mixer (I can't express how I feel about that one, for fear of breaking the keyboard).
If something like 3D mixing were implemented instead of slider-based mixing, it probably wouldn't need to be true 3D. Everything 3DMix does can be represented on a 2D top-down view, and probably easier to use too. Less impressive to look at maybe, but this isn't the demoscene. =P
|
Re: New MIDI Sequencer for BeOS
[re: nutela]
