A technical support community for Apple Logic Pro users.

 
User avatar
FlowerPower
Posts: 888
Joined: Wed Jul 04, 2007 11:45 pm

 

Wed Nov 12, 2008 1:23 pm

Well, maybe what we really want even has to wait for a future version of the OS itself.

Lots of people are moving, or considering moving over from PTHD to Logic native these days, and for the money they save on not using PTHD, they could all get the fastest 8-cores and loads of RAM.

There has been great performance boosts only between sub-releases of Leopard, from one Logic version to the next, from pre-Intel Macs to Intel Macs, from using less than 8g RAM to using 8g or more (preferably with one memory stick in every slot), from dual/quad cores to 8-cores etc., and we haven't even seem what Snow Leopard can do yet.

I also tend to use the same, sample libraries over and over again in a movie - in many cues, so even if I would have lots of cues in one song, I wouldn't necessarily have a lot more samples in memory, I would just use them in more parts of the project.

If something will be easier by not having multiple timelines overlapping each other - I can live with that. I can even live with messages telling me that I can't load more sample libraries or audio files into my project... then I would have to spread all my cues over, say, maybe 3-4 projects instead of only one. The way I personally work, I do a lot of movie stuff with relatively few audio files and not too many large libraries, but I can see that the situation is different eg. if people should make realistic mock string arrangements for a whole movie in one project, eg. on a dual core PowerMac with less then ideal memory amount and memory speed.
Logic Pro X (latest version), 12-core Mac Pro, 32 g RAM, Nektar P1
 
stevenson
Posts: 457
Joined: Tue May 24, 2005 8:07 am
Location: UK
Contact:

 

Wed Nov 12, 2008 2:24 pm

it's about where the bottle neck is. on my system, i don't run out of cpu power, i have plenty of RAM (or at least enough) and i think my RAM is probably fast enough. it's simply as logic stands you can't squeeze all that much into it if you use anything other than exs. and modern libraries are all going the kontakt route because they want to protect copyright.

even if the other libraries start using mmap, my computer slows right down the bigger and more full of stuff - plug-ins, automation, regions, samples - logic gets. so get the stuff out of logic.

my next projects are going to be setup with au lab or maybe plogue bidule, depending on whether i decide to mix out of the box or not. certainly i am going to try and keep as much stuff out of logic as i possibly can.
 
User avatar
FlowerPower
Posts: 888
Joined: Wed Jul 04, 2007 11:45 pm

 

Wed Nov 12, 2008 3:00 pm

I'm not really an expert on this, but I know that the new 8-cores are said to be about 2-3 times as fast (for audio) as the Quad PPC Macs.

An 8-core with memory in each of the 8 slots is 15% faster than the same Mac without memory in each of the 8 slots.

G5s used 400Mhz DIMMs, new MacPros use 800Mhz ECC fully-buffered DIMMs. The new Mac Pros uses faster and more efficient processors than the original Mac Pro, larger Level 2 caches, the system bus is faster than older Macs, and can have 4 times as much RAm as the G5s could and so on. Intel has already announced 6 core processors (allowing for 12 core Macs), use of multicore GPUs for non-graphical purposes, better support for multicore Macs will come with SnowLeopard, then there's OpenCL and what not... and: Apple want to sell hardware. It's good for them to sell software that needs/utilizes the fastest and newest hardware technology.

Who knows - maybe Apple even could be planning to do what Steinberg did when they continued to develop Cubase but introduced Nuendo... Apple could launch a new audio app that is built on a much more modern 'skeleton' than Logic is, potentially allowing implementation of stuff many of us have been wanting in Logic for many years - things that may not be possible in Logic (like multiple timelines, elastic time, automation node edits across multiple tracks etc). Remember - Logic 8 wasn't announced as a new 'song', not even a new 'mix', it was announced as 'Logic remastered'...

Even without all these utopian ideas... for the way I personally work, and with my current 8-core and 10.5.5 I'm sure multiple time lines would make wonders for my workflow, since I don't use lots of massive string libraries or lots of instances of complex softsynths like Absynth or Reaktor. All I basically want is a way to work on a cue without seeing that other cues lose picture sync or grid alignment.

http://www.everymac.com/systems/apple/p ... .7_dp.html
http://www.everymac.com/systems/apple/m ... specs.html

There has been a massive performance improvement over the last three years only, and I'm pretty sure that Apple is designing functionality and features for future Macs today, and that the next three years will be even more "revolutionary" (in terms of what computers can handle) than the three last years.
Logic Pro X (latest version), 12-core Mac Pro, 32 g RAM, Nektar P1
 
stevenson
Posts: 457
Joined: Tue May 24, 2005 8:07 am
Location: UK
Contact:

 

Wed Nov 12, 2008 3:16 pm

but my point is that cpu speed is NOT the bottle neck. nor is RAM speed. you simply can't squeeze enough into logic's memory space to make use of either the cpu speed or the RAM speed on modern computers. logic will need to become 64 bit for that to happen, and that won't happen any time soon (i don't think).

sure you will see some improvements in speed generally, but your timeline idea while brilliant, will still push logic over the brink judging by what we have now.

i can't say it any more plainly - faster computers even with a gazillion cores and quadrazillion Gbs of RAM won't change the limit logic has on its address space. it MAY be possible to offload more of its functionality outside of that space but, in the end this is what i am talking about anyway. get everything possible out of logics address space to leave room for things that absolutely have to be there.

actually if they were to implement something like my idea you could probably have the multiple timelines, since there probably WOULD be space after that. but without fundamentally changing how logic works with its content there just won't be enough room to run something like that reliably - i am very sure of that.
 
User avatar
FlowerPower
Posts: 888
Joined: Wed Jul 04, 2007 11:45 pm

 

Wed Nov 12, 2008 3:39 pm

stevenson wrote:
but my point is that cpu speed is NOT the bottle neck.

CPU speed, memory speed, front bus speed, numbers and sizes of caches etc. all has to do with overall performance. Re. 64-bit, someone stated somewhere that Logic 8 already is 64-but where 64-bit is needed, and that making everything in Logic 64-bit would be a bad, and not a good thing. I don't know... I've just seen so many PPC users complain about Logic 8 - and 8-core users praise Logic 8 (in terms of stability, speed, performance), and I have a hard time thinking that all the improvements between these two generations of Mac hardware won't influence the 'bottleneck'... If not, why would Apple state that the new Macs are 2.6 times faster than the QuadCore G5s? That's not a minor improvement...


http://www.gearslutz.com/board/music-co ... -pros.html

In the above thread, Lagerfeldt refers to a benchmark test giving a Dual 2.7 G5 2251 points, and a Mac Pro (with more RAM) 8254 points. Maybe you are right (I'm not a coder), but seeing how Logic 8 behaves on an 8-core compared with my 'old' dual core Mac Book Pro (which apparently is close to the dual G5s in performance) is very impressive, and definitely has to do with number of the tracks, plugins, voices I can squeeze out of my system. Switching between multiple projects in RAM still seems sluggish (but that's not what I want to do anyway....)

I'm all for Apple implementing bottleneck solutions and independent cue-functions. :-)
Logic Pro X (latest version), 12-core Mac Pro, 32 g RAM, Nektar P1
 
stevenson
Posts: 457
Joined: Tue May 24, 2005 8:07 am
Location: UK
Contact:

 

Wed Nov 12, 2008 4:05 pm

sorry - flowerpower....

you're not getting this.....

the bottleneck has nothing to do with hardware. nothing. performance from your computer makes no difference. none what-so-ever. you could have the fastest computer in the world, the fastest hard drives, the fastest RAM - everything.*

it makes no difference.

logic is coded as 32bit - and i am happy for it to stay that way because there are other solutions to the problem of memory than making it 64 bit. making logic 64 bit may cause its own problems - i don't know.

but logic being 32 bit with a limited address space is simply not going to be affected by whatever computer and however much RAM you put into it.

the problem as i see it with going multiple cues in one song and expanding on the folder paradigm and having multiple timelines, is that along with everything else, you will run out of memory space to be able to do this reliably and effectively. some of the performance issues relating to speed can be mitigated by a faster computer but some will not, because they relate to timing issues or clocking issues (i am no coder either) when you have full memory.

take for example the score editor in 7, and in 8. the score editor is painfully abysmally slow in 7, especially in page mode, and much faster on the same computer in 8. i upgraded from a G4 1 Ghs DP to G5 2.7 Ghz and found there to be virtually no difference in logics GUI especially the score editor - my primary reason for upgrading at the time.

but all that is beside the point. whether the project slows down or not is largely irrelevant because there will be the same issues with lack of memory within logics address space no matter the computer you have.

* i should just add, even if you max out your macpro with 32GB of RAM, logic is only going to be able to see 3.5Gb - 512Mb so about 3Gb of it. you can fill that up with just plug-ins - no samplers.
 
User avatar
FlowerPower
Posts: 888
Joined: Wed Jul 04, 2007 11:45 pm

 

Wed Nov 12, 2008 5:13 pm

I know about the RAM access limitations.

I think the 'problem' with general discussions about and development of) software is that people work in so many different ways, and that we easily may assume that something that may work (or not work) for us will work (or not work) for others as well.

A couple of examples:

At the moment I'm helping a colleague who is writing an orchestral piece. He certainly may want to change tempi and length of certain sections within the piece, but may need only a few EXS presets to do the job + a reverb and maybe some EQs. When he's ready, he'll print the whole thing out, and hand it over to the orchestra who will perform his work. He is already working on it, and has come quite far - and have had no problems with limited RAM access at all - but would have had problems if this material would be written to a TV program or movie; problems with Logic's inability to deal with individual sections as independent units within one project.

Likewise, I've been involved in various projects where all the string arrangements have been done in Logic, basically using one patch - and later handed over to an orchestrator who is finishing the arrangement/score in Finale/Sibelius.

I can also imagine composing music to a movie using a good Steinway sample only - all in one project file - and convert my ideas (still in the same project file) to Real Book type lead sheets, and record the material with musicians.

There's no rule saying that all these attempts will fail for all people, and for those who make certain types of electronica) using external hardware synths, it would also be possible to do a lot of stuff in Logic without addressing much RAM at all.

Some generations ago people were writing large orchestral pieces only using a piano (some still do), and I know several musicians/composers who still prefer to do their main composing using piano (or another instrument) only, and uses DAWs for recording/producing only (or pays others for doing it). I'm personally sometimes distracted by the many options Logic gives me, and find it easier to compose using one sound only (preferably an acoustic piano).

So... while there are bottlenecks and limits on addressing memory space, some of us would see a lot of benefits if Logic could do what can be done today by a combination of several actions (lock events to SMPTE, edit a cue, insert 'buffer' tempo changes, re-beatmap sections that has disturbed (grid-wise) by SMPTE-locking) for us. It doesn't even have to be multiple timelines overlapping each other, the main thing is to be able to force Logic to consider material pre/post the current cue for 'holy', and either not touch it, or repair whatever weirdnesses that may come out of the necessary tweaks we perform.

I have already worked using a 'hybrid' method (a mix between one cue per project and several cues in one project) and know it can be done, and I do get what you say, but we're probably talking about so many things at the same time that things get confusing. Maybe what some of us want wouldn't help your workflow much, but it would help others (like myself).

To simplify the whole thing and focus on the most simple solution for avoiding things getting out of picture sync/grid alignment: I want a macro that does thing for me that I otherwise would have to do myself - a macro that I'm convinced that also would result in less people claiming that 'Logic's Beat Mapping is unusable' etc, simply because they haven't been diving deep enough into Logic yet to realize the side effect if what 'blind beat mapping' can do to a project.

So... in essence, the basic wish isn't even for 'multiple timelines', it's for a way to start a new bar wherever it needs to start combined with a way to lock everything (grid, picture, events, local tempo events) for certain cues.
Last edited by FlowerPower on Thu Nov 13, 2008 5:01 pm, edited 1 time in total.
Logic Pro X (latest version), 12-core Mac Pro, 32 g RAM, Nektar P1
 
User avatar
simonphoenix
Posts: 499
Joined: Thu May 11, 2006 4:30 am
Location: Melbourne, Australia
Contact:

 

Wed Nov 12, 2008 9:52 pm

Thanks for the invite to this discussion Macintonwa

This is also a bit of a favorite subject for myself.

To quote what started this discussion...

I think what you want is multiple arrangements allowed in one song project. And then a "meta arrange" window where you can string those together as "arrangement regions".

The individual "arrangement regions" of course retain their internal time signature/tempo map characteristics -- which are dynamically reflected on the (non-editable) "meta arrange" timeline.

Maybe multiple tracks in the "meta arrange" window so you can overlap/crossfade "arrangement regions" -- though you're getting into potential signature/tempo conflict issues there.

This would also be useful for music in other situations where you need a higher-order level of composition -- long classical pieces, continuous-play albums, "concept albums".


• This sounds suspiciously like using Arrangement Folders in part, the only thing missing here is it's own Tempo/Beatmap.

If you were to add that ability to Folders then you could move the arrangements around at your own discretion. This feature could take priority over a Global Tempo Map, and in the same way that one audio region takes priority over another when assigned to the same Channel, Arrange folders could work like that, including it's tempo operations. In fact, a folder level Marker Track also wouldn't go astray.

• I've taken to arranging inside Folders right from the start, this comes in handy for versioning, and the easy arrangement of sections.

However the Folder Strips are featureless in the Mixer.

I would suggest that it functions in much the same way as an Aux, giving you a volume slider and the ability to process a grouped signal. An option to "fold out" Arrangement Folders in the same way as Take Folders would be greatly appreciated, while a Double-Click will still open it in the usual way.

• I agree that timing & sync should be greatly improved, there's no excuse for problems pertaining to that. This includes syncing to External gear.

• There needs to be support for multiple videos as well, to accommodate for scene changes and versioning. in fact, now that I'm thinking, it would be great if you could tie movies to folders, while allowing for nested music only folders.

• When syncing up tempo changes I prefer working with locked regions and the Tempo Map, but like all automation, the level of detail needs to be greatly improved.

• I would also like a context sensitive Lock/Unlock To SMPTE button in my tool bar, two buttons where there should be one is a little redundant.

• I also believe Logic should be able to Import Soundtrack Projects as well as added support for round-trip editing with Final Cut Pro. Not to mention the implementation of STP's Conform feature and it's Multi-point Video HUD.

...

Logic is probably 90% there Apple would be mad not to improve on it. It's more intuitive than Soundtrack Pro, and additionally allows scoring to movies, and hosts a wide range of synthesizers that are useful for sample triggering & Sound Design.

ProTools has even had a make over adopting a lot of Logic's features, that's testament to the powerful potential of Logic.

Apple just need to observe the trends, give users what they want in an intuitive way and up the ante. After all, people would be prepared to pay a little more for a well designed program and on top of that... Logic sells Macs, often quite expensive Macs.
G5 1.8SP, 1.25GB RAM, Mac OSX 5.3, Logic 8.02, Tascam FW-1804 Interface

About time for a new computer I think!!
 
stevenson
Posts: 457
Joined: Tue May 24, 2005 8:07 am
Location: UK
Contact:

 

Thu Nov 13, 2008 12:15 am

So... while there are bottlenecks and limits on addressing memory space, some of us would see a lot of benefits if Logic could do what can be done today by a combination of several actions (lock events to SMPTE, edit a cue, insert 'buffer' tempo changes, re-beatmap sections that has disturbed (grid-wise) by SMPTE-locking) for us. It doesn't even have to be multiple timelines overlapping each other, the main thing is to be able to force Logic to consider material pre/post the current cue for 'holy', and either not touch it, or repair whatever weirdnesses that may come out of the necessary tweaks we perform.

I have already worked using a 'hybrid' method (a mix between one cue per project and several cues in one project) and know it can be done, and I do get what you say, but we're probably talking about so many things at the same time that things get confusing. Maybe what some of us want wouldn't help your workflow much, but it would help others (like myself).



i know you know that there are RAM limitations with logic. what you may not be aware is how quickly you can hit them. you may not be hitting them now but add this functionality and who is to say you won't? also, there is a general problem with dealing with this issue that is critical for film scoring. on the one hand you are saying modern computers are powerful enough and on the other you are saying not everyone is going to use a large setup.

i have also used the many all cues in the one project approach and still do to a lesser extent even now. i personally have developed a nice workflow to deal with the issue of change in timings.

where i would like to see part of the approach you are advocating is the ability to drag an entire section of a song around arrange and preserve everything within in it - markers, key signatures, tempi - everything. and then when you move that section all sections outside of it are preserved exactly as they are, meaning automatic insertion tempo, key signature and time signatures to make sure they are unchanged.

until a way is found to remove the hosting of most content we work with outside of logics memory space, developing a solution for 'some people who don't want to use large layups' seems a mite narrow IMO. you could easily say 'why not use bounced out temporary mixes?' which is what we do today.
 
User avatar
question
Posts: 127
Joined: Thu Feb 07, 2008 2:14 pm

 

Thu Nov 13, 2008 2:59 pm

Quote from Stevenson:

where i would like to see part of the approach you are advocating is the ability to drag an entire section of a song around arrange and preserve everything within in it - markers, key signatures, tempi - everything. and then when you move that section all sections outside of it are preserved exactly as they are, meaning automatic insertion tempo, key signature and time signatures to make sure they are unchanged.

This is a succinct summary of exactly what I need!!

I just finished a series of three 90 Min. films which I had all in ONE Logic project 4-1/2 hours long.
I only actually actively composed for one film at a time, and had the previous films there in order to reuse my thematic material etc..


My point is; I utilized numerous manual workarounds, lots of pencil and paper notes, knowing when to lock and unlock objects to SMPTE, using region based automation, using odd time signature bars to get things "on the downbeat", religious backups for when I messed things up, always keeping the start position of objects in a cue the same regardless of where they actually play, and lots of math to reach my goal.

If I can do this manually, Logic should be able to do this easily without a big strain on its resources.
It certainly put a strain on my brain resources, but had the advantage of being able to quickly (with the director)
try out thematic material in various places in all three films.
The mixing was also considerably easier, due to to the similarity of the sound palette within a film.

Thanks to Ski, Stevenson, Flower Power, and Fader8 who came up with great ideas and workarounds in past threads.
LP 10.4.8, 10.14.6, Intel 12 core 3.0, 32GB RAM, Apogee DA16X, AD16X
 
User avatar
FlowerPower
Posts: 888
Joined: Wed Jul 04, 2007 11:45 pm

 

Thu Nov 13, 2008 3:48 pm

question wrote:
Quote from Stevenson:

where i would like to see part of the approach you are advocating is the ability to drag an entire section of a song around arrange and preserve everything within in it - markers, key signatures, tempi - everything. and then when you move that section all sections outside of it are preserved exactly as they are, meaning automatic insertion tempo, key signature and time signatures to make sure they are unchanged.

--

This is a succinct summary of exactly what I need!!


That's also a succinct summary of what I want! But...if I drag eg. Cue#5 (bars 32-47) to a place where the end of this cue may overlap, say, the beginning of Cue#12, wouldn't this actually need separate, independent timelines? Again, I could live with non-overlapping timelines, and I'm sure the word 'timeline' by itself causes some confusion, because it could be used to describe both 'horizontal and vertical' timeline-changes: In other words, both a situation where one time-line takes over when another 'is finished', or: two timelines that exist at the same time.

I'm sure we'd all love if we could drag a an entire section around, inside a project, and at the same time preserve everything inside this (and other sections) when doing so... but like 'Question' says, since we can do these things manually, shouldn't Logic also be able to do this without a big strain on its resources (including situations where we only drag PART of a section later/earlier, inside a Cue)...?

This would of course also include moving tempo events inside a cue, and even beat map a section and still (to quote Stevenson) see that "all sections outside of it are preserved exactly as they are, meaning automatic insertion tempo, key signature and time signatures to make sure they are unchanged".

Question's and Stevenson's last posts make me a little optimistic; after all, it seems that we all more or less want the same thing. :-)

I'll post some ideas (tomorrow?) regarding how these things IMHO could be best implemented.
Logic Pro X (latest version), 12-core Mac Pro, 32 g RAM, Nektar P1
 
User avatar
ski
Site Contributor
Topic Author
Posts: 16602
Joined: Thu Dec 15, 2005 5:12 pm
Location: Los Angeles

 

Thu Nov 13, 2008 4:14 pm

And also a "thank you" to Question for suggesting some time ago the workaround for maintaining the integrity of automation events by SMPTE-locking them. That seems to work very well in Logic 8 (as opposed to Logic 7, which was the version du jour when the Question discovered this). And without meaning to put down the idea in any sense, it's still a total pain in the ass process. It's something that Logic should give you the option to preserve when SMPTE-locking regions. It's quite the mind-bender just to understand the concept. Anyhoo...

What Stevenson wrote pretty much sums up what I'd like to do too.

Right now Logic lets you lock events to SMPTE which can be thought of as a method for preserving the timing of whatever you've written and letting you change the tempo underneath it after-the-fact to suit a new cue down the line. That's one way of working with multiple cues within a song, but of course your view of note placement on the grid (not to mention your score view) goes to s#!+ in the process. Yes, wouldn't it be great to be able to "lock" (in a whole new sense) a new kind of folder -- a "cue folder" -- and let you slide it around at will while Logic preserves the integrity of everything in it, including editor views! Meanwhile, a common set of instruments could be shared by all of them, although problems can arise if the same instrument was shared between overlapping cues. Overlapping volume automation rides come to mind. Even common notes shared between cues could potentially be a problem (a long sustained note from Cue "A" be prematurely cut off by the presence of the same overlapping note from Cue "B"). Hmmm....

FlowerPower, I'm very much looking forward to the ideas you have in mind!
MacPro 8 Core 2.4 GHz Xeon | 32G RAM | Sierra | LPX10.4 | Creator of ARTzID and SkiSwitcher Articulation Switching Systems for Logic Pro X, author of tutorials on The Score Editor, Enhancing Realism in Orchestral Mockups, and others.
 
fader8
Posts: 11141
Joined: Tue Dec 13, 2005 3:46 am
Location: Tampa, Florida
Contact:

 

Thu Nov 13, 2008 5:38 pm

The concept of a multiple cue Logic project has intrigued me for some time. As a Final Cut Pro user, I'm more than a little familiar with "nested" sequences, which is a direct parallel to what is being discussed here.

So, such a thing is very do-able but would necessarily come with a few limitations. To be at all practical, each cue in the master nest has to be its own complete Logic project, with its own mixer, samples in RAM, tempo map, etc. Like FCP, the whole project as a sequence, like one big folder, would be dragged into the master sequence which would trigger an offline bounce of the whole cue, then a disabling of its associated audio objects and plugs. Kind of like freezing a track but it's the whole mix.

So then you have a new arrange, with its own mixer, and the whole film. 2-clicking a cue-region opens (unfreezes) the cue, which only references the region of the film that matches its timeline in the master sequence.

I think you'd have to dispense with tempo features in the master sequence though, but maybe that's OK?

Anyway, just some fleeting thoughts I've had on this . . .
Music helps not the toothache.--George Herbert

G5Quad2.5, LP9.0.2,OSX.5.8,RME Digi&Multiface,002r, 4xUAD-1,Duende PCIe,Unitor8-AMT8,Logic Control+4 LCXT's,Capybara320,Kurz...
 
stevenson
Posts: 457
Joined: Tue May 24, 2005 8:07 am
Location: UK
Contact:

 

Fri Nov 14, 2008 12:08 am

So, such a thing is very do-able but would necessarily come with a few limitations. To be at all practical, each cue in the master nest has to be its own complete Logic project, with its own mixer, samples in RAM, tempo map, etc. Like FCP, the whole project as a sequence, like one big folder, would be dragged into the master sequence which would trigger an offline bounce of the whole cue, then a disabling of its associated audio objects and plugs. Kind of like freezing a track but it's the whole mix.


this precisely what i was talking about too - although a tad more succinct. while i don't hold much hope that our ideas for grabbing a chunk of a song and moving it will see the light of day anytime soon, it is certainly within the realms of currently existing technology to be able to have updated bounces from song to song.

and question - using compiled folders is the strategy we have for trying out existing material in different places. by that we take demo bounces of our work and set them up in a song with minimal instruments - just moving audio files around. then you can do sound edits of the cues to see what works from where, copy and paste material from one song to the other and then use 'move audio to original position' to line up your edits to the sections of the song you want to use and then snip the rest.

like you, i generally only use RBA because it is a) more reliable b) more easily moveable and copyable c) doesn't do weird things if you need to snip the song.

either way (whether it be timelines or individual songs) we still need a solution for the lack of memory within logic's address space. performance and reliability go down the toilet when logics memory space is full. i am not sure if mmap IS the answer since it dynamically ensures the memory space is ALWAYS full. taking the audio load off the sequencer seems like a sensible, practical and easy to implement solution.
 
Fred B
Posts: 1035
Joined: Sun Apr 16, 2006 5:52 am

 

Fri Nov 14, 2008 2:40 am

stevenson wrote:
either way (whether it be timelines or individual songs) we still need a solution for the lack of memory within logic's address space. performance and reliability go down the toilet when logics memory space is full. i am not sure if mmap IS the answer since it dynamically ensures the memory space is ALWAYS full. taking the audio load off the sequencer seems like a sensible, practical and easy to implement solution.

I think the address space limitation is imposed by the operating system and can't be manipulated by individual applications. In theory it might be possible to restructure Logic into multiple applications but that looks like a problematic hack to me. Sooner or later the problem will be gone anyway due to 64-bit addressing.

In the meantime you could for example run Logic Pro and Express together on the same machine for doubling the address space, using one as sequencer and the other for audio stuff.
MacBook Pro 2017 | OS X 10.14.6 | Logic 10.5.0 | Ableton Live 9.7.6
 
stevenson
Posts: 457
Joined: Tue May 24, 2005 8:07 am
Location: UK
Contact:

 

Fri Nov 14, 2008 3:50 am

I think the address space limitation is imposed by the operating system and can't be manipulated by individual applications. In theory it might be possible to restructure Logic into multiple applications but that looks like a problematic hack to me. Sooner or later the problem will be gone anyway due to 64-bit addressing.


actually, logic NOT becoming 32 bit is for various reasons. firstly the au spec and the SDK are not 64 bit compatible yet so making logic 64-bit is a bit of anon-starter until they are. also, there are potential performance considerations with respect to track numbers and GUI that may take a hit depending on how logic is compiled. also your audio drivers and 3rd party plugs won't work, so backwards compatibility would be an issue. although, steinberg have come up with a VST wrapper for 32bit plugs in their 64bit apps so it is not insurmountable, but i don't know how well it works. the main area for making 64bit is with working with large sample libraries and there are ways around that other than 64bit logic hosts everything.

In the meantime you could for example run Logic Pro and Express together on the same machine for doubling the address space, using one as sequencer and the other for audio stuff.


well, how is that different from logic running its audio in a separate app? you may call it a hack, i call it major step forward in enhancing pretty well the only you can work with logic these days.

if logic were 64bit tomorrow, you would still have the issue of switching songs. its not terribly fast in all audio songs, but it is a major nightmare here (i have yet to set up my sounds in an external app - have to wait until after this project). my idea is to extend the 'project' paradigm to include your layup. you have a project folder and all your sounds are loaded for ALL your songs in a separate app. these may include logic native plugs.

now when you switch songs it will be as fast as switching songs when you used all external midi gear like in the old days.

calling up new sounds load these into the separate app as you go through your project. you only have to load these up once in the day, when you begin your session. such a seperate app may also allow you to handle your resources more intelligently. simple pads can stream, but harps, celestes and pianos etc can all reside in RAM.

the technique for getting sound from one app to another is really well established. i am not suggesting this as just a solution to lack of memory in logic, i am suggesting this as really really efficient way of managing your layup and aiding switching between songs.

you can do it now as you already pointed out, but if the logic guys found an intuitive integrated solution it would be so much easier.
 
User avatar
FlowerPower
Posts: 888
Joined: Wed Jul 04, 2007 11:45 pm

Top 11 / possible ways to implement Independent Cues

Fri Nov 14, 2008 4:18 am

fader8 wrote:
The concept of a multiple cue Logic project has intrigued me for some time. As a Final Cut Pro user, I'm more than a little familiar with "nested" sequences, which is a direct parallel to what is being discussed here.

So, such a thing is very do-able but would necessarily come with a few limitations. To be at all practical, each cue in the master nest has to be its own complete Logic project, with its own mixer, samples in RAM, tempo map, etc.


Why not add some sort of 'sub-project' (of course with it's own grid/tempo map), but share the same samples in memory (but not EXS preset), and the same movie (in memory)?

Try this: open an EXS with a preset using a lot of samples. Watch the Activity Monitor>Memory. Press the New Track With Duplicate Setting-button. This creates a new track, with the same EXS preset (which can have it's own automation etc), but doesn't eat extra sample memory...

The idea about 'freezing' a whole cue that's overlapping another cue is interesting.

I think what you want is multiple arrangements allowed in one song project. [...]
Maybe multiple tracks in the "meta arrange" window so you can overlap/crossfade "arrangement regions" -- though you're getting into potential signature/tempo conflict issues there.


I think that all I need is that Logic (in the background) allows us to have multiple timelines (which may or may not overlap each other), just like we can have multiple multiple polyphonic voices on one stave, independent of each other.

Option 1: However, I don't think the average user want to think of something sounding as complex as 'multiple timelines' or 'meta arrangements'. I, for one, would prefer to relate to simple terms like "Cues" (when working with... cues), or "Elements".

So: if I could tell Logic that a given area (say, everything that belongs to marker 5, between bars 48 and 72), should be an 'independent cue', or an 'independent marker', it would live it's own life, so to speak (and, invisibly, have it's own timeline; it's own grid and set of tempo changes). This would be a 'marker-based', independent element (sharing samples in memory).


Such a solution could result in one pseudo-problem: when looking at the score, there would be "untraditional stuff" going on when Cue#5 starts, because the previous timeline would be interrupted by a new "1 1 1 1" (or a new 48 1 1 1) - maybe in the middle of an existing bar. But: this would also reflect what actually happened in that project - so it would IMO be a good thing. If a musical cue is cut off in the middle because a new cue starts, I don't want to spend time trying to make it look like something else anyway. But: when printing out stuff, the user could eg. decide that the selected (of two competing regions) should be printed.

If this separate, invisible timeline should be accessed on a per marker, or per cue basis, I wouldn't really have the option of having cues (tempo maps) overlap each other. I couldn't let a marimba arpeggio from Cue 5 continue into, and overlap, a string section starting in Cue#6 without tweaking the tempi (if they originally were different from each other).

Option 2: Since I prefer not to deal with a lot of different dialog boxes, and want things as simple as possible, maybe a better way to implement 'independent cues' is a way to assign tracks (and not markers) to a separate timeline (it would probably be better to (user-interface-wise) to describe it as assigning tracks to different 'cues'...

In other words: I could assign tracks 12-18 to an independent cue (which could be given a name by the user), and this way, cues could overlap each other with no hassle, and keep their original timing, grid and tempo changes without the user having to deal with anything else than telling Logic that 'these tracks are now assigned to "Cue#5", or 'these tracks are now assigned to "Scene 5, Paul discover that Michael has stolen his money".

Option 3: Since tracks can be assigned to multiple groups, maybe something as simple as allowing a group to have it's own tempo map/grid would be one of the simplest solutions.

A fourth option would be to assign folders, and not tracks to 'independent cues/timelines'. But since Logic folders cannot be opened the same way as Finder folders can (with that little triangle), I'm not sure if I like the idea of having to open and close folders (or windows) in order to see two different cues at the same time.

Option 5: Or... maybe assigning windows to a separate cue wouldn't be such a bad idea?

Option 6: Or even simpler: maybe Logic simply should be able to have two (memory sharing) projects open at the same time, and that if they user click a button, they would all start playback at the same time, meaning that one of the windows, containing Cue # 5, would somehow sync up to the same time code, but still have it's separate grid and tempo changes? (Disclaimer; some memory adressing issues would need to be addressed, and this could easily end up with UI confusion..)

Option 7: Screen sets - one screenset could be assigned to a different, independent cue (probably not the best solution..)

Option 8: A new global track, called 'Cues', could behave like tabs - in the sense that clicking on one of the Cue names would show that particular cue in the Arrange window (with an option to show all of them at the same time)

Option 9: Split screens: upper part could show one cue, lower part could show another (not so sure about that one...)

Option 10: A new thing that behaves similar to take folders (collapsible with a triangle): 'Cue-Folder' or 'Independent Element', which would allow us to open that cue and edit it at the same time as we saw the other stuff in Arrange could be a good solution, and have some parallels to an already existing feature (Takes).

Option 11: A Wavebruner-like graphical overview at the top of the screen could show as all the cues/elements in the song (independent of Global Tracks), and clicking on on of the cues/elements could show it in Arrange.

I guess the main issue here is that we all seem to want Logic to implement something that actually deals with the need for separate cues, in a more elegant way than having to SPMTE lock all events before and after the cue we're working on in order to not see them drift (this will let them drift away from the bar grid anyway).

Maybe there are better solutions than those I suggest, and maybe what I suggest even is some kind of 'meta arrangement', but my main focus is that I want Logic to 'think' the same way I do. When making music for movies/TV, I think in terms of 'cues' (and their need to be independent of each other), and not something with semi-strange names like 'independent timelines' or 'meta-arrangements'. :-)

With an extra level of possibilities, Maybe Logic would ned some smart warning messages ("You are now having two cues overlapping each other. Cancel | Freeze the first Cue | Freeze the second Cue", or "EXS Instrument ****** is now being used by two different cues. Do you want to create a new, independent track? Yes | Cancel".)


Songs aren't called songs anymore in Logic, but projects; I guess all I want is that Logic allows me to have more than one "song" in one project, songs that may or may not need to overlap each other. For me personally, the independent-ness of each song/cue is much more important than the need to have musical pieces overlapping each other.
Logic Pro X (latest version), 12-core Mac Pro, 32 g RAM, Nektar P1
 
Fred B
Posts: 1035
Joined: Sun Apr 16, 2006 5:52 am

 

Fri Nov 14, 2008 4:54 am

According to the object-oriented paradigm of Logic I think the most natural and general extension would be to have multiple instances of sequencers in one song/project, as can be found for example in Numerology2.
MacBook Pro 2017 | OS X 10.14.6 | Logic 10.5.0 | Ableton Live 9.7.6
 
Fred B
Posts: 1035
Joined: Sun Apr 16, 2006 5:52 am

 

Fri Nov 14, 2008 5:33 am

stevenson wrote:
the technique for getting sound from one app to another is really well established. i am not suggesting this as just a solution to lack of memory in logic, i am suggesting this as really really efficient way of managing your layup and aiding switching between songs.

you can do it now as you already pointed out, but if the logic guys found an intuitive integrated solution it would be so much easier.

I'd rather suggest that the OS X guys provide an integrated solution for audio routing, just like IAC for MIDI.
MacBook Pro 2017 | OS X 10.14.6 | Logic 10.5.0 | Ableton Live 9.7.6
 
stevenson
Posts: 457
Joined: Tue May 24, 2005 8:07 am
Location: UK
Contact:

 

Fri Nov 14, 2008 7:28 am

I'd rather suggest that the OS X guys provide an integrated solution for audio routing, just like IAC for MIDI.


sure well i don't care who does it. there are many applications that do it:

audio jack, rewire, plogue bidule, rax, wiretap, vbus, sbus.

you just need another app such as waveburner, which accesses logics plugs and 3rd party aus to host the plugs and samples. each 'plug-in' within logic is nothing more than an interface scripting changes to the corresponding plug-in in the secondary app. the logic guys could do stuff like that standing on their heads.....

which probably means they won't. they only go for really exotic solutions....