How do I extract a grid from a sample?

[padillac]

I am playing around with audio sequences, and would like to rearrange a beat using samples on multiple lanes, using the beat's original groove. I believe that means extracting a grid (which is super cool! if I can figure it out)

Here's what I did:

- Record a MIDI sequence and rendered to an audio part
- Convert the audio part to an audio sequence part
- Open the audio sequence part
- Create a new grid. Under "Create From", choose the sample

Rather than creating a grid from the markers that are in the sample, it just creates a single point in the middle of the grid. I've also tried dragging the sample onto the grid.

The audio sequence is clearly sliced up, and if I double click one of the sample events then it has a bunch of markers.

[mutools]

I double-checked creating a grid from sample markers and it works fine here.
Not sure where it goes wrong.
Please repeat the same steps and just before creating the new grid-from-sample, save the project, zip it, and share it here or email it to me together with a precise description of the next steps you take which result in an unexpected empty grid.

[padillac]

This is the saved project before creating the new grid: https://padillac-kvr.s3.us-east-1.amazo ... 133128.zip

(no audio FYI) https://www.youtube.com/watch?v=M5aC-Ybbjd0

[mutools]

Ok i see. The reason is that there is a selection in the sample and so only the markers in that selection are taken into account. That's meant as a feature but i understand it's a bit confusing as it is now. In the next update when there is a selection in the sample you'll get an explicit choice whether to use that selection or not. That will avoid the confusion. There will also be a couple of other improvements wrt creating a grid from a sample/sequence. Until that update: Make sure there is no selection in the sample if you want all markers to be taken into account when creating a grid from it.

[mutools]

PS Tip:
After you have created a grid from a sample, you might want to adapt the grid length. This can be done via the options button at the top of the grid editor -> Change Grid Length.

[padillac]

Ahhh okay. I remember reading something about that. I didn't realize that 1) I had one selected (which happened when I double clicked the slice to open the audio lab) and 2) that the slice is still considered selected even when dragging from the browser. Thanks!

btw perhaps you can help me with a few other things when working with loops:

1) What's the fastest way to extract a grid from a sample? I noticed that when using a drum loop, the grid is based on the length of the sample itself, which makes sense, but may not be a clean loop. So if I want to treat it as a loop I need to: 1. Import as audio sequence 2. Convert to new sample 3. Drag it onto a grid editor. Is that the best workflow to do that?
2) Is there a way to renumber the markers? Often times a sample won't have a marker at the very start (which I think it should, right?) So if I manually add a starting marker, it may be #27.
3) Is it possible to stretch / compress an audio part (not sequence) to fit on the grid? e.g. I have a long pad audio file that isn't at the project tempo. I want to start the audio part on the bar and lengthen / shorten it (timestretch) to end at a precise bar. This is like alt-dragging in Reaper. I think perhaps it's not possible, and instead it's done entirely with audio sequences?
4) Come to think of it, how does it work with a pad sample and audio sequence? If I drop in a pad sample and select audio sequence, sometimes there are a bunch of slices, which doesn't make sense. I think this may be the approach though: 1. Create a new empty audio sequence 2. Drop the sample into a lane. 3. Stretch the sample to match the desired length (length of sequence). 4. Enable auto stretch. Does that sound right?
5) Is there a way to crop / trim an audio file in audio lab?

[padillac]

PS Tip:
After you have created a grid from a sample, you might want to adapt the grid length. This can be done via the options button at the top of the grid editor -> Change Grid Length.

Cool, that answers my first question. If I have a sample that's two bars long (in a different tempo), then changing the grid length to 2 will make it a 2-bar loop like I want. Thanks!

btw two things I noticed that are weird with grids:

1) When I set the grid to 1/16 in an audio sequence, I can't stretch a sample to end at the bar line. It always ends at the 1/16 before it. I can stretch the sample to the end if I have the grid set at 1/8 or bar though.
2) If I render a sample and then drag it into the grid editor, it will appear with a single point, as if part of it is selected. I first have to drag it into the composer to create an audio sequence, and then drag it into the grid. It seems like maybe MuLab doesn't create the slice markers until you create an audio sequence, or otherwise load it into the project? Seems like it would be useful if when dragging it onto a grid editor, it created the slice markers before extracting the grid.

[mutools]

1) What's the fastest way to extract a grid from a sample? I noticed that when using a drum loop, the grid is based on the length of the sample itself, which makes sense, but may not be a clean loop. So if I want to treat it as a loop I need to: 1. Import as audio sequence 2. Convert to new sample 3. Drag it onto a grid editor. Is that the best workflow to do that?

Yes that will be the best way, but you'll need the next update for that, which will support dropping a sequence part on the grid editor and that will take the sequence part's tempo factor into account.
That next update is expected to be available in around a week.

2) Is there a way to renumber the markers? Often times a sample won't have a marker at the very start (which I think it should, right?) So if I manually add a starting marker, it may be #27.

Yes you can rename markers as you want: Right-click a marker flag -> Rename.

3) Is it possible to stretch / compress an audio part (not sequence) to fit on the grid? e.g. I have a long pad audio file that isn't at the project tempo. I want to start the audio part on the bar and lengthen / shorten it (timestretch) to end at a precise bar. This is like alt-dragging in Reaper. I think perhaps it's not possible, and instead it's done entirely with audio sequences?

Audio parts are indeed quite simple objects that simply play a piece of an audio file.
For much more flexibility (incl. stretching) best to use audio sequences.

4) Come to think of it, how does it work with a pad sample and audio sequence? If I drop in a pad sample and select audio sequence, sometimes there are a bunch of slices, which doesn't make sense. I think this may be the approach though: 1. Create a new empty audio sequence 2. Drop the sample into a lane. 3. Stretch the sample to match the desired length (length of sequence). 4. Enable auto stretch. Does that sound right?

Yes that's right.

Note there also is a preference called "Auto Create Transient Markers" which you could set to "Ask".
Then each time you drop a sample that has to be sliced you'll get the choice whether markers should be created or not. Maybe that's overkill, maybe not, that's up to you and your preferred workflow.

5) Is there a way to crop / trim an audio file in audio lab?

For Samples: Yes.
For Audio Streams: No.
(for the difference pls see http://www.mutools.com/info/M8/docs/com ... mples.html )

On Windows there is the option to open any sample/audio stream in an external audio editor, edit it there, and upon return MuLab will auto-update to the edited version. That works well and relaxes the need for a full featured integrated audio editor.

But i see you're on Mac and that "external audio editor" is not yet implemented in the Mac version. (because it takes extra Mac specific code and Apple loves to change things all the time, so if something works fine in version X then cross fingers it still works fine in version X+1. Sorry for this criticism on Apple, but it is a troublesome fact, especially for small devs where the overload of having to adapt all the time to new MacOS version incompatibilities is very big. So generally the less Mac specific code, the better)

[mutools]

1) When I set the grid to 1/16 in an audio sequence, I can't stretch a sample to end at the bar line. It always ends at the 1/16 before it. I can stretch the sample to the end if I have the grid set at 1/8 or bar though.

Pls make a video or share a project that demonstrates this issue.

2) If I render a sample and then drag it into the grid editor, it will appear with a single point, as if part of it is selected. I first have to drag it into the composer to create an audio sequence, and then drag it into the grid. It seems like maybe MuLab doesn't create the slice markers until you create an audio sequence, or otherwise load it into the project? Seems like it would be useful if when dragging it onto a grid editor, it created the slice markers before extracting the grid.

Indeed the markers must already exist before creating a grid from it. Note that you can directly create these transient markers via the sample's context menu (eg. right-click) -> Add Transient Markers. So no need to first create an audio sequence for it.

[padillac]

2) Is there a way to renumber the markers? Often times a sample won't have a marker at the very start (which I think it should, right?) So if I manually add a starting marker, it may be #27.

Yes you can rename markers as you want: Right-click a marker flag -> Rename.

Yes, but I am talking about renumbering. If I add transient markers, and it doesn't create one at the beginning, then I add one and now the first marker is numbered 27. So I'd need to rename all of them. This actually applies any time you add markers in the middle after the fact. Maybe it's not such a big deal to have markers out of order though. I'll keep playing with it and see how confused I get by it

4) Come to think of it, how does it work with a pad sample and audio sequence? If I drop in a pad sample and select audio sequence, sometimes there are a bunch of slices, which doesn't make sense. I think this may be the approach though: 1. Create a new empty audio sequence 2. Drop the sample into a lane. 3. Stretch the sample to match the desired length (length of sequence). 4. Enable auto stretch. Does that sound right?

Yes that's right.

Note there also is a preference called "Auto Create Transient Markers" which you could set to "Ask".
Then each time you drop a sample that has to be sliced you'll get the choice whether markers should be created or not. Maybe that's overkill, maybe not, that's up to you and your preferred workflow.

Cool, and I see that I'm able to edit a sample, add markers, etc directly from the browser without dragging it onto the composer first. That's super useful! It also addresses the question of dragging a sample onto the grid editor: first I can edit / add markers, then drag it to the grid editor.

---

What is the suggested workflow for processing lanes of an audio sequence separately? So far what I'm doing is creating unique copies, and soloing the lanes. They're now unique and so changes to one don't affect others, which is fine - it's just some work if I do change the sequence.

It did give me an idea, which may have been discussed before: it would be AMAZING if the lane solo / mute states were part-specific, the way Part Start Marker is. That way I could build up a single long audio sequence, then create parts that have both different start positions as well as lane solo / mutes so I can process each separately as its own clip. I've never seen anything like that before, and I actually think it would be quite a bit more powerful than even having per-lane FX or multiple outputs (unless those per-lane FX were optionally per part, which would be insane). Anyway, that's just an idea.

---

One last question for the time being: is there a way to add a transpose envelope to an audio clip somehow? So the audio clip would be time-stretched and always take the same amount of time, but I can automate the pitch during the duration of the clip. btw I am using "audio clip" fairly generically as I don't know if this would be done in an audio sequence, a MuSampla, inside of a Mux, or some other method.

---

Well I am sure you can tell by now that I am new to MuLab There are some VERY cool things in here that allow me to work in ways that I've always wanted to, but haven't been able to work out in other DAWs. There's still lots for me to figure out, and I'm starting to put the pieces together. So thank you very much for all of your hard work, your commitment to your creative vision, and your active engagement with the user base.

[sl23]

Is it possible to make markers auto re-number when adding new ones please?

[mutools]

Yes, but I am talking about renumbering. If I add transient markers, and it doesn't create one at the beginning, then I add one and now the first marker is numbered 27. So I'd need to rename all of them. This actually applies any time you add markers in the middle after the fact. Maybe it's not such a big deal to have markers out of order though. I'll keep playing with it and see how confused I get by it

Is it possible to make markers auto re-number when adding new ones please?

It's not as straightforward as it may seem. But i've added a note on the wishlist to at least add a function that sorts all markers so their auto name number is in increasing order.

[mutools]

What is the suggested workflow for processing lanes of an audio sequence separately? So far what I'm doing is creating unique copies, and soloing the lanes. They're now unique and so changes to one don't affect others, which is fine - it's just some work if I do change the sequence.

It did give me an idea, which may have been discussed before: it would be AMAZING if the lane solo / mute states were part-specific, the way Part Start Marker is. That way I could build up a single long audio sequence, then create parts that have both different start positions as well as lane solo / mutes so I can process each separately as its own clip. I've never seen anything like that before, and I actually think it would be quite a bit more powerful than even having per-lane FX or multiple outputs (unless those per-lane FX were optionally per part, which would be insane). Anyway, that's just an idea.

I'm fully with you. But i won't dive deeper into this topic at this point in time, for good reasons. The best thing i can say about this is: Please stay tuned and check out the next main steps in the MuLab adventure. (M9, M10, ...)

One last question for the time being: is there a way to add a transpose envelope to an audio clip somehow?

No, not yet, sorry.

Well I am sure you can tell by now that I am new to MuLab There are some VERY cool things in here that allow me to work in ways that I've always wanted to, but haven't been able to work out in other DAWs. There's still lots for me to figure out, and I'm starting to put the pieces together. So thank you very much for all of your hard work, your commitment to your creative vision, and your active engagement with the user base.

Thank you for your kind and encouraging words. I'm eager to further emerge MuLab as a creative and inspiring tool. I'm aware that, while there certainly already is a nice amount of magic in it, there also are weak points here and there. Always feel free to post feedback. Each bit of constructive feedback is always appreciated. Eventhough i may not always reply on it in detail because of time management (i have no choice), but rest assured i read all and do my best to select and combinate the best resonating suggestions.

[sl23]

Thanks Jo