Logic Pro 9 On the verge of giving up
- Thread starter Rafa
- Start date
Rafa
Logician
Guitar track completely messed up
Flex messed up a guitar track completely. The first screen-shot shows how it starts.
From that point on the audio gets progressively out of sync and I have no idea how to put it back in its original position.
Of course I've tried to insert flex markers over the transient markers, and then try to set them to neutral position - it didn't work.
For some mysterious reason, at some point the audio stops expanding and starts compressing - all that without one single flex marker! The second screen-shot illustrates just that.
Flex messed up a guitar track completely. The first screen-shot shows how it starts.
From that point on the audio gets progressively out of sync and I have no idea how to put it back in its original position.
Of course I've tried to insert flex markers over the transient markers, and then try to set them to neutral position - it didn't work.
For some mysterious reason, at some point the audio stops expanding and starts compressing - all that without one single flex marker! The second screen-shot illustrates just that.
Attachments
Although I cannot see where the audio is out of sync, what happens if you turn Flex off? Flex works non-destructively, the signal should go back to its original state.
You did not answer the question in the other thread yet: Are your regions quantized? If yes, what happens if you turn that off?Rafa said:
More questions: Do you have tempo changes? Does the audio follow some tempo for whatever reason? Which Flex mode do you use?
---
If you suspect a basic problem, try with a new project: open the default empty Logic template, record a short piece and, without doing anything else, turn Flex on and tell us what happens if you see anything strange.
Rafa
Logician
The screen-shot shows two grouped, phase-locked, guitar tracks, which are perfectly in sync with one another. But as you can see, they coloured orange, so they are stretched and behind the tempo of the song. Turning Flex off puts it back in place - that was a precious piece of information! - but it really doesn't help me fixing this problem.
No regions are not quantized.
Yes I have tempo change.
I didn't understand the second question.
I use only slicing mode.
I will do this I soon as I hand in this job I'm on, the problem is that Flex is not helping meet my deadline!
Last thing, could you make anything from the screen-shot in the first post in this thread?
Rafa
Logician
Good news! (sort of)
I figured a way to workaround it!
If you look at the screen-shot named "outcome" you will see waveforms being stretched with no Flex marker on them (I can not understand how this happens). Well now, I've placed flex markers over each transient marker, and set them individually to neutral position and it worked. Screen-shot named "Good news" shows exactly that. If you got time compare both screenshots: "outcome" (previous post) and "good news" (this post).
It also took forever x2, because, even if the tracks are grouped, setting to neutral position is not a "grouped action"
But "Mr.X da Flex" didn't let get away so easy, oh no, there is still one note that is impossible to set to neutral position. I kindly ask everybody to look carefully to the colouring pattern on this waveform, do you see a pattern? Screen-shot: "same thing"
I figured a way to workaround it!
I tried the same thing again, this time more systematically, and with a fresher mind, and it did work!
If you look at the screen-shot named "outcome" you will see waveforms being stretched with no Flex marker on them (I can not understand how this happens). Well now, I've placed flex markers over each transient marker, and set them individually to neutral position and it worked. Screen-shot named "Good news" shows exactly that. If you got time compare both screenshots: "outcome" (previous post) and "good news" (this post).
It also took forever x2, because, even if the tracks are grouped, setting to neutral position is not a "grouped action"
But "Mr.X da Flex" didn't let get away so easy, oh no, there is still one note that is impossible to set to neutral position. I kindly ask everybody to look carefully to the colouring pattern on this waveform, do you see a pattern? Screen-shot: "same thing"
Attachments
Rafa
Logician
You can't catch a break with Flex
News, more news...
The more I work with Flex, the more I'm convinced that you can't mix both, "scissors & crossfade" with Flex editing techniques. They simply don't go together.
The simple act of slicing a region, mind you, I did not move a thing, makes Flex stretch everything from the slicing point onwards.
This is one thing, and I kinda knew that this was happening already, the new bug I encountered is "disappearing waveforms" in Flex view.
The first screen-shot, shows the original recording, the second shows the same region with Flex view on. The second screen-shot shows two bugs, the first and more obvious one: waveforms disappear (but the audio sounds fine, except for the fact that everything is out of place). The second one, is the fact that everything is coloured orange, even though the audio is in its original recording position! I didn't move a thing, and yet everything got stretched just because I sliced that region.
In my experience Logic provided a tool that is just impossible to use.
News, more news...
The more I work with Flex, the more I'm convinced that you can't mix both, "scissors & crossfade" with Flex editing techniques. They simply don't go together.
The simple act of slicing a region, mind you, I did not move a thing, makes Flex stretch everything from the slicing point onwards.
This is one thing, and I kinda knew that this was happening already, the new bug I encountered is "disappearing waveforms" in Flex view.
The first screen-shot, shows the original recording, the second shows the same region with Flex view on. The second screen-shot shows two bugs, the first and more obvious one: waveforms disappear (but the audio sounds fine, except for the fact that everything is out of place). The second one, is the fact that everything is coloured orange, even though the audio is in its original recording position! I didn't move a thing, and yet everything got stretched just because I sliced that region.
In my experience Logic provided a tool that is just impossible to use.
Attachments
Rafa, I think your handling of tempo is the problem. You record something, then you change the tempo and then you want to use Flex. But both deal with time, they cannot be independent. Coloured regions without Flex markers can only have one reason and this is tempo. Please follow the tests below to understand the dependencies.
1. Raw recording
I have two raw tracks from an acoustic guitar. The upper track is a close mic, the lower track contains both mics of a M/S setup. The "Follow Tempo" checkbox is visible, Flex mode is off and the piece was recorded at 107 BPM.
2. Switching to Flex
Now I engage Flex for the upper track. The "Follow Tempo" checkbox goes away, gets replaced by "Flex". This indicates that Flex and tempo are not independent which is no surprise since tempo settings and Flex work in a time-based manner. Nothing happens in the flex'ed region because there is no processing yet:
3. Changing the global tempo
First I make the piece faster by changing the global tempo to 110 BPM. As you see in the screenshot, the unprocessed region becomes longer as usual. But the flex'ed region stays in place and shows green borders because, relatively to time, it gets compressed. The two tracks are no longer synchronized:
Then I reduce the global tempo to 105 BPM (slower than the recording tempo) and the opposite happens. The unprocessed region becomes shorter and the flex'ed region, with unchanged lenght, turns orange because relative to time it is now expanded:
The above tests tell us that, once engaged, Flex takes over the tempo handling. This means, tempo changes and Flex are not "compatible" under all circumstances. There may be ways to sync flex'ed and unflex'ed regions for tempo changes but the target of this test was only to show the dependencies.
4. Partial tempo changes
For the final test I switch back to the original tempo of 107 BPM and insert random tempo changes. Flex behaves as we can expect from the previous tests. The whole region stays in place and the audio signal gets compressed and expanded in time, according to the tempo:
The situation above shows what you (Rafa) see as "random coloring" but actually it tells how tempo changes influence Flex: Slower tempo expands and faster tempo compresses the audio signal relative to time. If the tempo is at its original value, nothing happens.
Conclusion
When Flex is engaged, tempo changes work on audio material as if there were Flex markers. And the last test above shows us how complex the new function is. Flex cannot ignore the tempo because it is supposed to control it. But the tempo information comes from three sources:
How to solve the problem?
I don't know yet, I do not use the global tempo track for audio and have no more time left to do more tests. Maybe you can flex all tracks to get an equal situation everywhere in your arrange. Maybe you have to switch Flex off and bounce everything in place to go back to a neutral state, because Flex remembers a lot and might not let you repair a big mess. But I think Flex will always react to your tempo changes. The problem might get harder if you have virtual instruments too. From my view using Flex and tempo changes is not a good idea. At least it can easily become a source of trouble because both functions deal with time.
A couple of hints
1. Raw recording
I have two raw tracks from an acoustic guitar. The upper track is a close mic, the lower track contains both mics of a M/S setup. The "Follow Tempo" checkbox is visible, Flex mode is off and the piece was recorded at 107 BPM.
2. Switching to Flex
Now I engage Flex for the upper track. The "Follow Tempo" checkbox goes away, gets replaced by "Flex". This indicates that Flex and tempo are not independent which is no surprise since tempo settings and Flex work in a time-based manner. Nothing happens in the flex'ed region because there is no processing yet:
3. Changing the global tempo
First I make the piece faster by changing the global tempo to 110 BPM. As you see in the screenshot, the unprocessed region becomes longer as usual. But the flex'ed region stays in place and shows green borders because, relatively to time, it gets compressed. The two tracks are no longer synchronized:
Then I reduce the global tempo to 105 BPM (slower than the recording tempo) and the opposite happens. The unprocessed region becomes shorter and the flex'ed region, with unchanged lenght, turns orange because relative to time it is now expanded:
The above tests tell us that, once engaged, Flex takes over the tempo handling. This means, tempo changes and Flex are not "compatible" under all circumstances. There may be ways to sync flex'ed and unflex'ed regions for tempo changes but the target of this test was only to show the dependencies.
4. Partial tempo changes
For the final test I switch back to the original tempo of 107 BPM and insert random tempo changes. Flex behaves as we can expect from the previous tests. The whole region stays in place and the audio signal gets compressed and expanded in time, according to the tempo:
The situation above shows what you (Rafa) see as "random coloring" but actually it tells how tempo changes influence Flex: Slower tempo expands and faster tempo compresses the audio signal relative to time. If the tempo is at its original value, nothing happens.
Conclusion
When Flex is engaged, tempo changes work on audio material as if there were Flex markers. And the last test above shows us how complex the new function is. Flex cannot ignore the tempo because it is supposed to control it. But the tempo information comes from three sources:
- Global song tempo
- Individual tempo changes
- Flex markers
How to solve the problem?
I don't know yet, I do not use the global tempo track for audio and have no more time left to do more tests. Maybe you can flex all tracks to get an equal situation everywhere in your arrange. Maybe you have to switch Flex off and bounce everything in place to go back to a neutral state, because Flex remembers a lot and might not let you repair a big mess. But I think Flex will always react to your tempo changes. The problem might get harder if you have virtual instruments too. From my view using Flex and tempo changes is not a good idea. At least it can easily become a source of trouble because both functions deal with time.
A couple of hints
- Do not see Flex and tempo changes as regular methods to squeeze a standard audio recording into another time pattern. Record with the right tempo and rhythm as good as possible.
- Flex is a post production tool, mainly thought for corrections. You can of course use it creatively but this is another story. Normally you want to use it to tighten several instruments or to correct rhythmical phrases at certain points.
- If you must use tempo changes, turn Flex on for all audio tracks because it will respond to tempo changes and you should force it to take the tempo from the global tempo track before you change the time of audio by Flex markers.
- If you have audio tracks and virtual instrument tracks and tempo changes and need Flex, don't start with such a song before you tested the dependencies of all elements.
- In former Logic versions we did not think much about time for audio regions. This has changed, we can now control the time. But that means that we cannot longer ignore the time while recording. Time is now an issue to think about.
Jay Asher
Logician
In regards to Peter's ajswer here:
I did an experiment the other day. I brought some drum tracks from a ProTools session into LP9, Detected the app. tempo (there were tempo changes), cut then into smaller regions and quantized them. Within 10 minutes they were locked spot on. I then created a tempo map with a gradual ritardando over the last 8 bars and it all tracked perfectly.
I did an experiment the other day. I brought some drum tracks from a ProTools session into LP9, Detected the app. tempo (there were tempo changes), cut then into smaller regions and quantized them. Within 10 minutes they were locked spot on. I then created a tempo map with a gradual ritardando over the last 8 bars and it all tracked perfectly.
You may have a reason for that but I didn't get good results yet with slicing mode on acoustic guitar. Maybe I do something wrong, for me polyphonic is the only mode that works well for this instrument. Very well. But I use Flex only for corrections. I set two Flex markers before and after the part I want to time-shift and then move the audio signal between them.
Well, think about what's happening: You have one audio file in one track. You cut a piece out, move one region above the other and apply a crossfade. But you have still only one file, the regions are like two windows showing different parts of the file. If you use Flex in the first region and push the audio forward in time, you would actually move the whole remaining file. This does not happen, the movement gets stopped at the regions end. Similar if you use flex in the second region. But what happens where the files overlap, when the left regions gets time-compressed and the right region expanded? What happens to the inaudible part and to the "partly audible" part at the crossfade? It works for me but I wouldn't bet that it works in every situation. This situation is far from trivial and I am really happy that the programmers managed to handle it. But I will not be surprised when I get some glitches from such a construct.Rafa said:
I think the safe way is to get the crossfades right and when the track sounds right, bounce in place and start with Flex. This does of course require a workflow with some milestones where you say, "This is ok, le'ts bounce and continue with the result." And it should of course be done before the actual mix.
The more functions we get, the more we need a plan. We cannot go back to the past from every point. We all cried for more "professional" functions. Now we got some of them and either have to accept that we need a "professional workflow" too, or do without these functions. Fortunately we have the choice.
Rafa
Logician
You're a life saver
Peter, thank you very much for your interest, you're really a life saver. Lately, the Flex discussions were looking like a no man's land, no useful info and almost no responses. Just to able to share experiences is already all I could ask, not to mention the quality of your replies!
Although post #8 showed very interesting characteristics of this tool, it doesn't really apply to my case. The hints were certainly very valuable, but let me clarify some points about that particular session I started on this thread.
Peter, I really appreciate your effort, and it was really helpful indeed, but one big issue got overlooked. What about that weird looking waveform showed on screen-shot "same thing" on post #5 of this thread? What do you make of it?
Once again, thanks a lot.
Rafa
Peter, thank you very much for your interest, you're really a life saver. Lately, the Flex discussions were looking like a no man's land, no useful info and almost no responses. Just to able to share experiences is already all I could ask, not to mention the quality of your replies!
Although post #8 showed very interesting characteristics of this tool, it doesn't really apply to my case. The hints were certainly very valuable, but let me clarify some points about that particular session I started on this thread.
That was exactly what I did, the tempo changes were there since the beginning, the audio was recorded that way.
Again, this is precisely what I have in mind. I maybe young but I admire the old school approach to music: study hard, rehearse to exhaustion, record the very best take that is possible for your level of playing, use post production tools just to tighten the screws a bit, but the general tightness was already achieved in the rehearsals and in the raw material.
Peter, I really appreciate your effort, and it was really helpful indeed, but one big issue got overlooked. What about that weird looking waveform showed on screen-shot "same thing" on post #5 of this thread? What do you make of it?
Once again, thanks a lot.
Rafa
This should be an ideal situation. I have never tried it but if you record with tempo changes and switch Flex on for all recorded regions, the waveforms may get coloured according to the tempo. This is not ok in my opinion but we don't hear the colours - how does it sound? Is the playback tempo ok then? If not, this is a bug.
I have seen this a couple of times, it's display problem. Maybe changing the zoom factor helps. Bad if you need to work in this area though. If you see this often, report it to Apple.Rafa said:
Peter Ostry said:
Not sure if this is necessary. You say that your recording was done with existing tempo changes and that it was ok. Maybe you just messed it up because you got nervous about the weird colours. Give it a try: Switch off Flex, check your crossfades. Bounce in place without effects and volume/pan automation. Switch all regions to Flex and listen. If it sounds ok, you are done and can start correcting some parts with Flex.
What I haven't tried yet is to flex first and cut/crossfade later on.
Rafa
Logician
Flex handles this situation just fine, turning flex view over the original, unedited recording, does absolutely nothing, as we'd expect. The waveforms are grey all over, even under the tempo changes, and the audio sounds O.K.
. I know you brought this up on post #8, but this was never an issue for me.
The "colour changes" which were obstructing my workflow had no apparent connection with the tempo changes in the session.
This happens a lot, and some of these occurrences have nothing to do with a graphic glitch. I've seen these glitches, and I'm fine with them, the problem are these other notes, like the one showed in post #1 of this thread which stay that way and are impossible to set to neutral position. I guess I'll report it.
That sounds great!
I gotta tell you, I was getting seriously psyched out about this and now I really believe that things are not half as bad as I imagined them to be, thanks for the tip, If I could I would send you flowers!
If you do that, I want original flowers from Rio!
Rafa
Logician
Peace at last
Hey Peter I must say that my life improved 97% after this paradigm shift. I was placing Flex tool on the first step of my editing process and that alone was the cause of all my troubles. Of course it would be great if some of those bugs could get fixed, but just changing the order of things really did it for me.
Now I'm editing everything "old school" first, bouncing in place, and just then I'm using the Flex tool to give that final tweak, and fix those gritty places that are really hard to do with scissors and crossfade.
I'm a happy camper now!
(original flowers for you man!)
Hey Peter I must say that my life improved 97% after this paradigm shift. I was placing Flex tool on the first step of my editing process and that alone was the cause of all my troubles. Of course it would be great if some of those bugs could get fixed, but just changing the order of things really did it for me.
Now I'm editing everything "old school" first, bouncing in place, and just then I'm using the Flex tool to give that final tweak, and fix those gritty places that are really hard to do with scissors and crossfade.
I'm a happy camper now!
(original flowers for you man!)
honestly, no one else has this problem with flex?
So Rafa, I had the same issue as you in the picture at the top of this thread.
Wave form changes color between markers.... It happened several times, even with changed zoom modes.... Quantize not on.....in a brand new project ....no tempo changes.....
Do you know if 9.0.1 has fixed this?
You wre able to work around it by turning flex on at the end of editing.
--------------
here was my "case study"
1) new project.
2)import 2 mono wave files.
3) add a 2nd mono track to project
4)drag in files to tracks.
5)turn on flex view.
6)set each track to polyphonic
7)looked in dismay to multi colored wave files both redish orangish & green.
I didnt change the song tempo
I do not have quantize selected.
my colors on wave form change between markers...
and you mean to tell me no one else but you and me, has this problem?
i must have a corrupt install of LP9, or SL
hopefully 9.0.1 fixes this, I didnt see it in the list.
So Rafa, I had the same issue as you in the picture at the top of this thread.
Wave form changes color between markers.... It happened several times, even with changed zoom modes.... Quantize not on.....in a brand new project ....no tempo changes.....
Do you know if 9.0.1 has fixed this?
You wre able to work around it by turning flex on at the end of editing.
--------------
here was my "case study"
1) new project.
2)import 2 mono wave files.
3) add a 2nd mono track to project
4)drag in files to tracks.
5)turn on flex view.
6)set each track to polyphonic
7)looked in dismay to multi colored wave files both redish orangish & green.
I didnt change the song tempo
I do not have quantize selected.
my colors on wave form change between markers...
and you mean to tell me no one else but you and me, has this problem?
i must have a corrupt install of LP9, or SL
hopefully 9.0.1 fixes this, I didnt see it in the list.
Rafa
Logician
It is quite a mystery, this whole flex business really took me over the edge with rage in the recent past.
The worst thing is that no one seems to talk about it, no one has a clue of what's going on, the only decent reply I've had was Peter Ostry's on this thread.
I've changed my editing workflow dramatically and it seems to be working fine now, but your case is definitely the weirdest of all! I had the same multicoloured waveforms but in a project with heavy editing on it, but I really made things difficult for flex on that one, and so I kinda understand weird things happening. Your case however is completely different because it is happening on a brand new project.
In my experience the multicoloured waveforms don't affect the audio, it's just a huge graphic bug. It really creeps you out, plus it makes the multicoloured note uneditable, but for the rest it doesn't mean that the audio has been modified.
It's all really frustrating, and I'm sorry if I don't have much to add to your case, let's hope LP9.1 fixes it!
The worst thing is that no one seems to talk about it, no one has a clue of what's going on, the only decent reply I've had was Peter Ostry's on this thread.
I've changed my editing workflow dramatically and it seems to be working fine now, but your case is definitely the weirdest of all! I had the same multicoloured waveforms but in a project with heavy editing on it, but I really made things difficult for flex on that one, and so I kinda understand weird things happening. Your case however is completely different because it is happening on a brand new project.
In my experience the multicoloured waveforms don't affect the audio, it's just a huge graphic bug. It really creeps you out, plus it makes the multicoloured note uneditable, but for the rest it doesn't mean that the audio has been modified.
It's all really frustrating, and I'm sorry if I don't have much to add to your case, let's hope LP9.1 fixes it!
Eli
Logician
I think I know what your problem is. You've imported two files into a fresh project and don't have any tempo changes in the project. But is this new project the exact same tempo as the tempo of your imported audio files? Probably what is happening is that when you import the files, Logic is importing their imbedded tempo information. And then when you turn on flex mode, it is adjusting based on the discrepancy between the source tempo and the project tempo.
Thanks
Eli, thanks for the reply. I will check that out asap.
however I was under the asumption that Flex does not alter the file until either you physically stretch it or you turn on quantize??
Also, the 2 audio files were recorded in PT's and are just raw wave files?? I think?? I thought raw wave files do not have imbedded tempo information?
anyways I will check out....this sounds promising
thanks,
Again
Eli, thanks for the reply. I will check that out asap.
however I was under the asumption that Flex does not alter the file until either you physically stretch it or you turn on quantize??
Also, the 2 audio files were recorded in PT's and are just raw wave files?? I think?? I thought raw wave files do not have imbedded tempo information?
anyways I will check out....this sounds promising
thanks,
Again