Obviously this isn't properly thought out but I'm sure it is all possible.

Ouch!

You're not in "management" are you?

Everytime I hear this (as I am a business applications software developer) it usually turns out as one or all of these:

a) It will take nearly X * 6 man years to do the job done (far longer than anyone thought possible, especially the users)
b) The user will not be happy cuz what they thought they wanted simply won't work, even when implemented "exactly" the way they requested it
c) It really isn't possibly without a complete overhaul of the infrustucture of the code
d) Even with <the> best possible team, the programmers don't "quite" get what the user want (cuz it isn't properly thought out)

That said, sounds like a great idea!

ORIGINAL: bapu

Obviously this isn't properly thought out but I'm sure it is all possible.

Ouch!

You're not in "management" are you?

Everytime I hear this (as I am a business applications software developer) it usually turns out as one or all of these:

bapu,

I've been a software developer, a systems and network architect. I've been in management. I'm quite sure what I propose is all possible. It's basically just GUI enhancements over the existing features in the Sonar engine. Of course I don't know how the code looks and how much the GUI code is integrated into the engine code.

It sounds a lot more complicated than it really is. It's much harder to explain than if I could just dump what I envision into a GUI mock-up. Unfortunately one thing I have never been is a graphic designer!

That said, time permitting, I'll try and make a GUI mock-up during the week-end.

UnderTow

ORIGINAL: Marah Mag

I don't know, but I wonder how possible (or practical) that is under Sonar's current development path? At what point does it become better to whip up a new dish altogether instead of serving up another helping of reheated spaghetti? (I'm sorry... sometimes, you know... I can't resist.)

Marah,

I don't know either. The backwards compatible side, from a GUI point of view, should be easy. Basically you don't show all the parameters and the tracks/buses look like the good old Sonar that we know now. As for the rest of my proposals, nearly everything is already there in Sonar.

Tracks can already take external inputs or inputs from soft synths. In other words, feeding something to a track is already implemented. The same goes for outputs. A lot of what I propose like envelope on layers, track waveform preview, nested folders are just visuals to the underlying engine that already exists or can be recycled. (If there is code for a bus or soft synth waveform preview and there is code for track layers, you have what is needed for a waveform preview of the output of a track regardless of layers). The routing matrix might be a bit more complicated.

Integrating V-Vocal is of course more complicated but I'm not too bothered about that not being fully integrated into the tracks. The basic routing, folders and layers is what interests me most at the moment.

According to Cakewalk, they concentrated on the audio engine for Sonar 8 and apparently have gotten great results. If they do the same thing with the GUI and specifically the arrange view for Sonar 9, they should be able to achieve great results with that too. And these changes should be very visible to everyone! It would be like having a totally new product but with the backwards compatibility and all the advantages that that has.

UnderTow

I've been a software developer, a systems and network architect. I've been in management.

Yeah, I knew that.

Most of that was tounge in cheek for you.

However, in the "real world" with real users and real (uninformed) managers.... well, it's another world.

OK Bapu
You got me there. I am a gear-slut.

In the last month I have
Bought a UAD2 Quad
Bought S8 Upgrade (didnt really have a choice)
Revalver MKIII
Wusikstation upgrade
Omnisphere

However, I personally consider all but Revalver upgrades that I had to do to stay current. And since I am not really a keyboard player but own tons of keyboards and plug ins, it seemed silly that I did not have most of the guitar amp sims.

This last year has not seen some of the more extravagent purchases, but I did add another KM184 to my Neumann collection, A Manley tube mic, another drum set and semi-booth thing, a bunch of drum mics...my two gretsch guitars...

I think the reason I have not bought more is three fold
(1) my house really wont hold a ton more
(b) I have been stocking up on weapons for a surprise invasion from Canada - yep - everyone is watching the Mexican border (maybe 3 people) but no one expects Canadians,
(iii) I bought two cars for cash this year - one as my daughter turned 16 and a big gas guzzler for me. The good news is that my QX56 can hold two Prius as spares with room for a smart car on the luggage rack.

ORIGINAL: bapu

Here's a useful feature for SONAR 9:

Homage Presets.

Like what Marah Mag does? Load the HCMG preset, complete routing matrix and buss & sub folders (all in pastels).
Like what yep does? Load the Yep preset, with the "extended" dialogs explaing in detail EVRYTHING you need to know and why it's done that way.
Like what space_cowboy does? Load the Maurice preset, which immediately takes you to vintage & esoteric equipment websites and automatically spends wads of your moola on more gear than you ever imagined.
Like what John does? Load the John preset (no, don't go to the john). Nearly everything you see is what you get in a straight talk mannner. Nothing subtle here.
Like what mike_maccue does? Load the I Like Mike preset. You get a free camera with that.
Like what CJ does? Load the CJ preset which immediatley nukes any sub-par soundcard in your system and configs your card for you. No more unnecessary forum posts for you, noob!

Homage Presets: Cause you don't know who you are.

This last year has not seen some of the more extravagent purchases, but I did add another KM184 to my Neumann collection, A Manley tube mic, another drum set and semi-booth thing, a bunch of drum mics...my two gretsch guitars...

I got SOANR 8 (we'll the order is online today) and a Q9550 to replace my Q6600 (not installed yet). I'm a Main Street guy (you know, broke).

SONAR was the first DAW application to have a universally flexible bussing model back in SONAR 3. Internally in SONAR's engine there is no distinction between tracks and buses - there are only buses. A track, a user bus in the bus pane, and mains - they are all exactly the same bus object. A bus in SONAR is an abstract entity that can have multiple inputs and multiple outputs (and do a lot more).

We already have the ability to do pretty much any routing in SONAR. For example sidechaining is achieved by routing tracks to an internal (hidden) track bus which then routes to an effect input, tracks feed track buses, preview streams from the loop explorer feed hidden track buses, etc... A typical project is a massive network of buses.

There would definitely be peripheral work to make it all hang together and handle delay compensation properly etc, but architecturally we could relatively easily route tracks to other tracks.
I'm curious why you think that routing tracks to other tracks at a user level gets you something that is particularly useful. What is a real world use case for this that cannot be achieved with the bussing we already have. You could end up with a routing so complex that you wouldn't understand it after a week :-)

ORIGINAL: UnderTow

ORIGINAL: Marah Mag

I agree. The strict track/bus distinction is maintained at the expense of navigational/workflow flexibility and adherence to a hardware model.

It would be very nice to be able to move buses out of their separate pane and intersperse them with tracks.

I can even imagine that a bus object could function something like a folder...

Yes. Buses and folders already share a number of similarlities. Making them more similar would be useful and would enhance workflow.

...and be able to show all the tracks feeding that bus as a means of sorting.

Totally. The inability to see at a glance which sources are feeding each destination is a major shortcoming of Sonar. At least give the buses a popup caption that shows the sources.... though that would only scratch the surface. This is software, after all.

Why not go a step further? Allow routing from one track to another and you remove the need to use buses entirely. Right click on a track to change it's colour so that it is easily recognisable. Add a MIDI layer for DXi/VSTi's (instrument tracks). Basically make a track an object with all the current possible parameters combined. Route anything to anything.

Also, allow envelope layers. The same idea as the current track layers but allow envelopes to be set on any layer you want for easy editing and clarity. Add track waveform previews that can be assigned to a particular layer (or just the lowest layer dedicated to this). This way you can see the sum output of all the audio layers within a track after processing by the FX bin.

So now a layer just becomes another attribute within the track that has audio/MIDI/envelope/waveform_preview layers.

Add a routing matrix with presets and a wizard. Add nestable folders. Each folder or sub folder can be intelligently be associated with a track or not. Whatever you want. Add ways to easily organise tracks within these folders: Right click -> organise -> bunch all tracks -> routed To current track/From current track etc. Expand the "Move to folder" option to a list of all folders and subfolders.

Integrate V-Vocal on a track or rather layer level. Integrate AudioSnap on a layer level. Etc etc etc.

Call the whole system universal track architecture.

For backwards compatibility (and for users that prefer to work that way), leave the existing bus system and have the extra layers etc as options.

Obviously this isn't properly thought out but I'm sure it is all possible.

UnderTow

ORIGINAL: Noel Borthwick [Cakewalk]

SONAR was the first DAW application to have a universally flexible bussing model back in SONAR 3. Internally in SONAR's engine there is no distinction between tracks and buses - there are only buses. A track, a user bus in the bus pane, and mains - they are all exactly the same bus object. A bus in SONAR is an abstract entity that can have multiple inputs and multiple outputs (and do a lot more).

We already have the ability to do pretty much any routing in SONAR. For example sidechaining is achieved by routing tracks to an internal (hidden) track bus which then routes to an effect input, tracks feed track buses, preview streams from the loop explorer feed hidden track buses, etc... A typical project is a massive network of buses.

There would definitely be peripheral work to make it all hang together and handle delay compensation properly etc, but architecturally we could relatively easily route tracks to other tracks.

Hi Noel. I always had some intuitive sense that, behind the scenes, tracks and buses were the same object. But it was interesting to see you confirm that. Thanks for that interesting reply.

Let me ask you: Are folders also a variation of the same bus object? They would seem to be, with their MSR buttons, and their clip icons (I forget the correct name...?) across the track pane. My reason for asking this will become more clear in a sec.

I'm curious why you think that routing tracks to other tracks at a user level gets you something that is particularly useful. What is a real world use case for this that cannot be achieved with the bussing we already have. You could end up with a routing so complex that you wouldn't understand it after a week :-)

I'll try to explain how I see it. Understand that my goal is simplified workflow, which includes easier application navigation, e.g., less scrolling, clicking, context/view shifting -- without sacrificing flexibility and power, of course!

Let's say I've got two tracks, gtr1 and gtr2, and I want to use the same fx (etc) on each, so I create a bus which lets them share the same fx bin. This creates a new object (the bus) which is located in a removed area of the UI (the bus pane).

If I was able to route gtr2 to gtr1, it would effectively turn gtr1 into a bus, which accomplishes two related things: it doesn't increase my object count, and it doesn't disperse the work area necessary to control those two parts across the track pane and the bus pane. It's all right there.

That last point gets to a related one: Independent of being able to route a track to a track, I'd like to be able to intersperse tracks with buses to avoid the "pane" split, which would give a tighter organization of the work space.

At another level, it would be useful to be able to treat buses like (or actually turn them into) folders, i.e., to organize a bus's sources "inside" the bus-as-collapsible folder. Thus my asking above if folders were also another variation of the basic bus object.

What is a real world use case for this that cannot be achieved with the bussing we already have.

Right. It's not that this kind of routing necessarily lets you achieve something that can't be done currently. It's that it lets you organize the project and its objects for a different workflow and layout and app navigation. It's a comparable advantage to a routing matrix: It doesn't actually add new types of routings, but it lets you route things more centrally and provides an overview of the signal flow. The centralized matrix reduces scrolling and pane-to-pane navigation, as would being able to intersperse buses and tracks, or, as in my first example, being able to simply turn a track in to a bus, or for that matter, turn a bus into a collapsible folder for its sources.

You could end up with a routing so complex that you wouldn't understand it after a week :-)

Indeed! But honestly, even with the current system, I'll sometimes forget what my routing is if I've been away from a project for a while, or make a dumb routing mistake that I've got to backtrace, or just happen to draw a blank at 5am. Tracing a project's routing can be tedious -- scrolling through bus and track panes... expanding track heights to check the output... or even just selecting tracks to read the output field from the Inspector. It's all about the shortest distance -- and the fewest clicks and mouse wheelings -- between two points. Project and application navigation.

In this regard, it would be very useful to be able to see at a glance the sources feeding a bus. At the most simple level, I'd hover my mouse over a bus (e.g., its trim control) and get a caption that showed its sources by track number and name. But that's also where a routing matrix would be helpful: a simple view that shows me to see what's going where, and that would let me create and delete connections, as well as control levels and other params of the path. It's all about delinearizing the workflow, and getting more info about the project and feedback from the application.

So, that's my perspective on it. Others would have different reasons or would emphasize different things. Still others would say there's no need for any of these kinds of changes (but of course, they'd be wrong! )

Thanks for asking and for reading! :)

ORIGINAL: Bob Damiano [Cakewalk]

Create an Undo Setup dialog where the user can specify which procedures will be subject to undo and which will not, and make all possible procedures available for selection, including mute, solos, and ALL track/bus param changes

In the original version of Project 5, I made all mix changes undoable. I loved it. Everyone else hated it :)

Hi Bob. I can see why some might hate it. :) That's why you need to be able to specify which types or categories of changes to exclude from the history. Like, exclude Solo, include Mute, Volume, Panning. It would be tricky, I know.

Reaper's undo records everything, including every time you change a track selection of change the track height... virtually everything you do in the program... which is way overkill. I would exclude those things, or put them in a parallel history, so that actual mix changes could be undone.

What's odd is that sometimes undo *does* cause Mute and Solo status to change. Not all the time, and I've got no idea what triggers it, but it's not uncommon.

Not sure if you are aware, but in SONAR each mix parameter does have ONE level of undo. It's a Bob D[tm] feature that came from how much liked undo for mix parameters in P5. In SONAR, after adjusting anything, right-click the control and choose "Revert to ...". It will jump back to the previous value. Really handy for A/Bing small subtle changes in a mix.

I was aware of that, but only recently. What would be cool is if each track/bus could have its own undo history.

It would also be REALLY useful if the undo history could indicate between which two changes the project was last saved (obviously the save couldn't be undone, but it would be a helpful signpost when undoing)

Too late to add something here?
Bare with me, as I am on v6, so it may be so that these features are already in 8.5 (but I don't think so)
 
Colors: make the colors on track connected to buses, so each bus has a color, and all tracks that are connected to a bus automatically gets its color. And if you change bus, the color will change too.
 
Be able to change stereo tracks to L, R, Switched or Mono.
 
Gain envelope on takes so you don't need to add one each time you want to trim the volume on a single take. Should be there as easily as fades and trim handles.
 
that's all. And forgive me if I'm ignorant of any features already implemented. Do feel free to put me right. :D

bgalvin


I nice way of keeping "notes / to do list" for each track. Make it so it is not visible when they are small but when you make it large enough you can see it. That way I can get rid of paper notes and ideas yet to do and keep it all in the box for real.

I saw there are VST plugins that you put int the bin that do nothing but just keeping notes that you enter. Transparent for all audio stuff.

I would like to see a better GUI for V-Vocal. I like to use it, but it is so painful to look at the little dots. I'd like to see it look more like Melodyne in that regard. The other thing I would like to see it do would be auto harmony, or to allow you to play the harmony parts from a midi keyboard. Kind of like the new device in Record 1.5.

Wow! With both Bob and Noel in on this thread, you just have to *know* the dev team is on schedule for the release cycle! They must have recently reached a big milestone!

I just wish "shift-click" worked like it does on every other windows program (selects things between two points) and there was some way to hold a clip/note to a vertical and horizontal line when moving it (shift/ctrl like 99% of every other windows programs). I actually though constrain worked at one point, but was nerfed in the next version.

Lunatique

Better Envelope tools/selection detection

 

Yes, this is my number one wish.  Many of us clearly would like some improvement in this area.

Faster/Easier pianoroll navigation via Photoshop’s spacebar grabber tool

Nice idea!

Susan G

eratu


Wow! With both Bob and Noel in on this thread, you just have to *know* the dev team is on schedule for the release cycle! They must have recently reached a big milestone!

What year/version are we talking about here? As of a few posts ago (#79), this thread was from Oct '08. Or were you kidding?

-Susan

LOL! Oh my, I got duped on that one. I thought this was a current thread, and didn't bother to check the dates. Well, I guess my comment was utterly idiotic in the proper context! But, having said that, IF those had been current responses from Bob and Noel, then I'd stand by my post! Oh well! Thank you for the heads up on that!