Lua: Stretch marker guard

[mpl]

Stretch marker guard is a lua Reaper script.
It allows to protect area around existing stretch marker. Could be useful if you have stretch markers at drums and you want keep some transient area unchanged by stretching algorithm. The idea is taken from this FR.


Installation:

• if you haven`t ReaPack - install ReaPack extension, restart REAPER
• if you have ReaPack - go to ActionList -> Syncronize packages
• go to ActionList -> ReaPack: Browse packages... -> navigate and install Stretch marker guard package
• go to ActionList -> ReaPack: Browse packages... -> navigate and install ReaImGui extension if its not installed/updated yet

(version from 2016 before 23.11.24 was completely based on existing markers and allowed to set only once, and look like this:
)

[vanhaze]

mpl, Do you know this is really a revolutionary feature come true ?!

Jeeeeeezzz, i HAVE to try this NOW !!!

EDIT: I find "Transient Guard" a wonderful name, exactly hitting the nail.

[X-Raym]

Very impressive, congrats !!

[heda]

excellent stuff as usual mpl.

An idea to improve it, could be to store the data of the stretch markers as they were before. So the reset button could work after closing the script.

[James HE]

This looks really sweet

any thought to making the surrounding stretch pairs snap to grid divisions instead of ms? i.e 1/128, 1/64, 1/32, 1/16... etc?

this might be a useful creative thing for some sources.

[mpl]

heda, I thought about that. That is goes not so complex, but maybe "Delete all non-1.0x markers" internal action is a better/simplier solution?

[musicbynumbers]

Awesome! Will try when I've got time but this could be very useful thanks!

[heda]

Originally Posted by mpl

heda, I thought about that. That is goes not so complex, but maybe "Delete all non-1.0x markers" internal action is a better/simplier solution?

yes that is even better I think.

[Jaybee]

This looks amazing but I'm not sure I've grasped how to use it in a real world situation. I've recently been looking at quantizing audio so explored Stretch Markers but I'm getting artifacts (especially on bass) that aren't there via the old-school split/quantize methods of correcting timing.

Here's what's happening:

1. Recorded bass line
2. Use Dynamic Split to detect Transients and write Stretch Markers
3. Set Grid resolution
4. Snap Stretch Markers to Grid

Now the bass line is perfectly in time *but* I'm getting lots of clicking/artifacts which were not there before the Stretch Markers moved. I gather this is because the Transients have stretched?

...and so by searching I found this thread...

Could anyone kindly explain for me the workflow on how using MPL's tool can eliminate my artifacts by protecting my transients? I have no idea what values to use inside the tool? I've taken my quantized Stretch Markers, applied MPLs tool and added a marker left and right of the transient (say an arbitary 20ms each). This just makes the artifacting/clicks even worse.

If I apply the tool before the quantize then I can't snap to grid as I can't define the Stretch Markers to use for the snap (i.e. just the transient) or can I?

Any help gratefully received! I can see by the thread responses this tool is very much appreciated, I just can't figure out how to use it properly

[mpl]

Now the bass line is perfectly in time *but* I'm getting lots of clicking/artifacts which were not there before the Stretch Markers moved. I gather this is because the Transients have stretched?

No. You'll get artifacts always when you stretch bass with any algorithm in any situation. Use dynamic split to split bass, then quantize small pieces. You cant just stretch any tonal (and especially low-freq) audio without artifacts. Even if you stretch slap stuff you can get this script to work but only for a start of every piece, other body of every note will be dirty and glitchy as hell after even a small stretching.

[Jaybee]

Originally Posted by mpl

No. You'll get artifacts always when you stretch bass with any algorithm in any situation. Use dynamic split to split bass, then quantize small pieces. You cant just stretch any tonal (and especially low-freq) audio without artifacts. Even if you stretch slap stuff you can get this script to work but only for a start of every piece, other body of every note will be dirty and glitchy as hell after even a small stretching.

Now I understand! Thank you for the explanation

[heda]

I agree mpl, but also, to stretch tonal bass you can try to tweak the parameters of Elastique audio mode for that item. Preserve formats lowest pitches, and also activate "optimize for tonal option". Sometimes it works.

[mpl]

Year I guess this algos use zero-crossing detection to define optimal grain size. But results almost always was bad for me. Well, anyway it depending on how it sound in the mix If coefficient not so big, within final mix bad sounding stretch algos can be masked sometimes.

[daxliniere]

Originally Posted by mpl

Stretch marker guard is another useless killer feature for REAPER. LOL.

Holy shit, man, this is amazing!!

Originally Posted by mpl

("Transient guard" sounds not exactly correct)

Nope, I think it's a perfect name, actually.

[daxliniere]

The Russian ReaScript contingent is blowing my mind, lately!

[mpl]

Heda, added to ReaPack both internal action and external script for removing non-1x markers. Hope didn`t broke something.

[lolilol1975]

Hello mpl, I think this is a question for you:
https://forum.cockos.com/showthread.php?t=201295

How can we use your tool to do time stretching of drum parts ? Thank you.

[hopi]

Originally Posted by daxliniere

The Russian ReaScript contingent is blowing my mind, lately!

must be the better vodka

mpl wow what a brilliant idea... thanks yet again for what you do!!!

[pipelineaudio]

This seems to have a bug that it brings back any stretch markers you deleted, but aside from that, pretty damn awesome!!!!

[Thonex]

Very slick!!!

Fantastic! This is a real gem!

[shady]

took me a few rounds of the licecap, and then i got it. dis is sick.

[foxAsteria]

Ah cool man, thanks for this. I'd seen it in the list but couldn't figure out how it worked. Very helpful!

[shake_puig]

How can I use it by slightly adjusting? it seems to snap all the way to the right or left.

[Scrumpy Jack]

Originally Posted by shake_puig

How can I use it by slightly adjusting? it seems to snap all the way to the right or left.

I just ran across this too, and it looks like this is caused by some naming collisions that happened since it was written. If you find & replace the script's functions math_q and SetExtState to use different names, the sliders work again as they used to. It's a very handy script!

[mpl]

Originally Posted by Scrumpy Jack

I just ran across this too, and it looks like this is caused by some naming collisions that happened since it was written. If you find & replace the script's functions math_q and SetExtState to use different names, the sliders work again as they used to. It's a very handy script!

This is no longer needed, the script updated to some modern API implementation, it uses now printed take markers as a source for central markers, most of the code rebuilded for more stable usage.