Looping loops: to crop or not to crop?

Sampler and Sampling discussion (techniques, tips and tricks, etc.)
RELATED
PRODUCTS

Post

Hi all

Just wondering what the "standard" if any is.

If you are making a loop( like a drum loop), do you crop the file with the start and end points of the loop markers, or do you leave some space( not silence, just part of the original drum loop)at the beginning and end?

Do some players have issues with totally cropped looped files?..or..is it better that way?

Tim

Post

Are you talking about ReCycle? If so, then yes, crop the file to the loop start and end points.

Generally speaking, don't leave anything there that you don't want to hear in a finished track made with the loop.

If the loop is giving you problems, like an audible 'click' at the beginning or end, make sure to crossfade on both ends of the loop.

Sometimes a loop will sound fine with itself, but will click when combined with other loops. This is because the waveform of the loop meets up with itself, but not with the other loops.

Generally, the start and end points of the loops should always be at a point of 'zero crossings'. This is why crossfading is a general practice in loop making land.

Hope this helps.

Post

Sorry, Tim - are you talking about a wave file drumloop or a sampler patch loop? Like herodotus says, a standalone drumloop (or music loop) would normally be cropped exactly at its start and end points. Are you having playback problems with a host or something?

Post

If you are making a loop from a drum synth the loop points should be at the start and end of the loop. There shouldn't be any clicking. If there is that means that something is being added to the loop, like reverb, delay, etc which every1 here says should be turned off when creating loops/freezing. If pops/clicks are happening/you need the reverb/delay/etc of the loop then I recommend, recording the loop twice, instead of once. That way the 2nd repeat of the loop will contain all the wet effects - Then you just make the loop point at the start of the 2nd repeat & the end of the 2nd repeat/you could split the loop into 2 and use the 2nd half for looping (the 1st half would only play once in the mix). The only problem is knowing how long the wet effects last for the (wet) loops you are making.

Post

wow..thanks for the feedback! Yeah..I am learning how to do drum loops.
Q: to do zero crossing, you stretch the sample view, then make sure you edit the start and end points so they stop at the middle zero point? Is that what you mean?

Thanx!

Tim

Post

tconrardy wrote:Q: to do zero crossing, you stretch the sample view, then make sure you edit the start and end points so they stop at the middle zero point? Is that what you mean?
Yes.

Frequently the 'musical' place to cut a loop (i.e. the place where it grooves the best) will not have a convenient zero crossings point close to it. this is especially true for stereo recordings, because there are essentially two waveforms involved.

And so we employ the technique of cross fading. All that means is that we place a very short fade in at the beginning of the loop and a very short fade out at the end. Really short, like say, .003 to .008 seconds.

Fades this short are impossible (or at least very, very hard) for the ears to pick up, so the loop sounds seamless, especially if you add reverb and compression and the rest when you use the loops in a song.

If the loop ends aren't at zero crossings, chances are you will hear a nasty 'pop' sound, if not when you are 'looping' the loop, then almost certainly when you try to combine it with others.

Have fun!

Post

Crossfading is a bit more involved that just creating fades at the start and end of a loop. You also need to cut and paste an overlapping section of preceding or subsequent material into the portion you want to loop. I create a lot of ambient loops, and edit them in Sound Forge. I always record the initial loop longer than I need, in order to allow for the crossfade. I find that Sound Forge's automated crossfading feature suck's ass. It uses a linear fade which creates slight volume drops. I do them manually, using custom (curved) envelopes, which works much better for my purposes. I then do a very, very slight linear fade on the last few samples to create a zero point on both sides, to avoid any potential clicks. I can get seamless loops this way without have to resort to reverb to mask the looping point.

Post

John Vulich wrote:Crossfading is a bit more involved that just creating fades at the start and end of a loop. You also need to cut and paste an overlapping section of preceding or subsequent material into the portion you want to loop.


True, I was oversimplifying.

I create a lot of ambient loops, and edit them in Sound Forge. I always record the initial loop longer than I need, in order to allow for the crossfade.
Cool, though I thought we were talking about drum loops.

'Proper' crossfading is rather more laborious than what I was talking about, and only allows one to use the loop by itself. If you wanted to use more than one loop, one would have to create a new loop for the transition, and another for the transition back. Which is why the process I described is standard practice for drum loops.
I find that Sound Forge's automated crossfading feature suck's ass. It uses a linear fade which creates slight volume drops. I do them manually, using custom (curved) envelopes, which works much better for my purposes. I then do a very, very slight linear fade on the last few samples to create a zero point on both sides, to avoid any potential clicks.
This is all cool, but again, we were talking about drum loops, not complex ambient ones. The process you are talking about would add needless complexity, especially to someone who is new to the process.
I can get seamless loops this way without have to resort to reverb to mask the looping point.


So can I.

I am perfectly willing to do an A/B test wherein one drum part has an edit such as I have described, and the other doesn't. I am quite certain no one would be able to hear the edit.

I mean, this is kvr, right?

But if so, lets start another thread for it.

Now I must go. I have a bunch of loops to make.

Post

I tend to "wrap" my loops in FL Studio so that any trailing sounds wrap into the start and when looped it sounds properly cross faded. But this sometimes sounds weird when playing the first instance of the loop. If it's too much I need to reconsider my trailing sound in the end of the loop or possibly edit the start of the loop to sort of mask the wrap.
It's not like I'm a professional so I just use my ears judgment what's allowed or not ;-)
Rekkerd.org the latest news on audio plugins, sample libraries & virtual instruments, synth presets & more.
Don't click here if you can't control yourself!

Post

herodotus wrote:True, I was oversimplifying.
Yeah, I figured as much. I just wanted to clarify for TC's (and others) sake. I used the example of Ambient loops because it's what I have the most experience with. I don't edit many percussive loops but I would imagine, based on the quality of the source material, that the need to use crossfades would be somewhat rare.

Post

John Vulich wrote:
herodotus wrote:True, I was oversimplifying.
Yeah, I figured as much. I just wanted to clarify for TC's (and others) sake. I used the example of Ambient loops because it's what I have the most experience with. I don't edit many percussive loops but I would imagine, based on the quality of the source material, that the need to use crossfades would be somewhat rare.
Rare indeed.

The biggest problem with percussion loops are sustained cymbal hits ending abruptly. And this is easily dealt with using compatible one shots.

Making ambient loops commercially, in mix and match collections, is a true bitch. There are usually continuities one can work with, but it takes forever to find them and make them work in self contained units.

Post

John Vulich wrote:
herodotus wrote:True, I was oversimplifying.
Yeah, I figured as much. I just wanted to clarify for TC's (and others) sake. I used the example of Ambient loops because it's what I have the most experience with. I don't edit many percussive loops but I would imagine, based on the quality of the source material, that the need to use crossfades would be somewhat rare.

this may seem perhaps like a not so intelligent Q: but for ambient stuff, can't you take a fairly long sample( 20S or so) and first have a short fade in at the beginning, then the loop start marker, then it goes to near the end where you have an end marker, with a slight fadout at the end after the loop parts? It might be more seamless this way..provided you have a good loop inbetween the start and end points.

or

is it a fast and hard rule about cropping to the exactly the start and end points?

I guess I am thinking about regular samples ( like a string sample) that you loop to make it sustain..but perhaps it may not work with a "loop".( however..it seems to!)

Tim

Post

----Kinda depends on what it's used for/in I would think. Like to set loop points for a sampler to play back, then overlap is fine in those cases, but not for like Acid loops type stuff. There's no overlapping the sample with Acid loops, you can crossfade to overlap, but the loop points for those are just the beginning and ends of the .wav file.

Jeff

Post

tconrardy wrote:
John Vulich wrote:
herodotus wrote:True, I was oversimplifying.
Yeah, I figured as much. I just wanted to clarify for TC's (and others) sake. I used the example of Ambient loops because it's what I have the most experience with. I don't edit many percussive loops but I would imagine, based on the quality of the source material, that the need to use crossfades would be somewhat rare.

this may seem perhaps like a not so intelligent Q: but for ambient stuff, can't you take a fairly long sample( 20S or so) and first have a short fade in at the beginning, then the loop start marker, then it goes to near the end where you have an end marker, with a slight fadout at the end after the loop parts? It might be more seamless this way..provided you have a good loop inbetween the start and end points.

or

is it a fast and hard rule about cropping to the exactly the start and end points?

I guess I am thinking about regular samples ( like a string sample) that you loop to make it sustain..but perhaps it may not work with a "loop".( however..it seems to!)

Tim
Well then, let us get specific, shall we?

When making drum loops for use in a host when making a song, starting on the beginning of the beat is standard practice. This allows you to sync it up with your hosts time code, so midi parts and other loops and whatnot all work together with minimum bother.

But when making a loop of, say, an ambient texture of some sort, or of any long sustained sound or ensemble thereof, the procedure described by John affords one greater , more detailed control, helping to splice hard-to-splice parts together smoothly.

The deal with the ambient or other sustained sounds is that synchronization issues are less important than a smooth sonic texture.

The deal with drum loops is that synchronization is important, while the editing is generally unproblematic, due to all of the transients.

I do hope that this was clear and uncontroversial.
Last edited by herodotus on Wed Oct 18, 2006 8:24 pm, edited 1 time in total.

Post

most samples can be looped without crossfading, if you look hard enough - Infinity was very good for finding these points easily. crossfading , IMO, should always be a bit of a last resort, although of course sometimes there's no other option.

Post Reply

Return to “Samplers, Sampling & Sample Libraries”