Page 1 of 1

The sampling inputs and resampling..

PostPosted: Wed Sep 10, 2003 2:40 am
by DJ Broadband
Hey y'all,

I am just curious, how many of you actually use the sampling inputs on your Emu? I've never used mine before, because I picked up a SCSI card setup right when I bought the sampler. Never really even thought about using this feature until I hung out with a hip hop producer that does it all the time on his MPC.

It seems like a lot of hip hop producers use their samplers in the old fashioned way... its cool I guess, I just dont understand it.. guess I'm used to the luxury of SCSI transfer. :mrgreen:

But anyway, I'm just wondering if there are actually any benefits to this method, and if so, why? I never really liked all the work you have to do with envelopes inside the sampler and whatnot, just to get it to sound right, when it coulda been done in a sampler editor or in the VSTi it might have come from.

Also, I am curious, do the E-series of Emu samplers have a real-time resampling function? (i.e. - MIDI can trigger a sample from it and at the same time it can record a sample in the sample inputs) I wouldn't know because I'm just using a measly ESI 4000. :grumble:

And 2 more things.. I'm curious how many times you resample on average, and, can someone explain to me how to make a sample loop smoothly inside the sampler? (such as a pad sample) I've never been able to pull it off correctly.. (I realize it'll be E-series terminology but I'd appreciate an answer anyway 8-) )

Ok thats enough from me! :mrgreen: :pimp:

PostPosted: Wed Sep 10, 2003 4:30 am
by Klaseed
I resample constantly. Any given sound gets resample maybe ten, maybe twenty times? I'm not sure, but it's a whole lot.

To me, there are two advantages to this. First of all, that many trips through whatever processing you do and the Emu's A/D/A conversion process make the sample sound like a sample coming out of an Emu. Which sound I happen to love. Secondly, I find that having to go into and out of my computer all the time really gets in the way of my creativity. For some operations (fine trimming sample starts, effects I don't have in my Emu), I use SCSI, but I sort of prefer to keep it all in the sampler.

Oh, and also if you resample a lot, you can use more, and more subtle, steps in your processing than if you just bang the sample through your EQ compression, distortion, reverb, etc. all at once.

And yes, you can real-time sample.

As for smooth looping, that's an art unto itself. The best advice is to try to find a complementary start and stop point - that is to say, a point on the zero crossing, where the phase of the wave is the same (i.e. if hte wave is rising to 0 at the stop point, it should be rising from 0 at the start point). But yeah, it's realy hard, especially smoothe sounds like pads.

PostPosted: Wed Sep 10, 2003 4:46 am
by sampleandhold
okay, lets see, about using sampling inputs. i am assuming you mean sampling a from a source, like a turntable. i do this all the time. i will hook up my cdj 500 and set that on my chair and sample whatever i need. i use my dj cd players because i can cue them up better and get all the sound instead of possibly missing the start of a signal. i also us my computer's cd rom, since it has a head phone jack. i use an 8th to rca and run that through my patch bay. or i will take my md player and record stuff off of video games, or movies and then run that through my patch bay and sample that as well. i guess the advantage to doing this is that i don't have to record it on my computer and convert it or whatever. it just goes straight into the sampler.

okay, next one: resampling. in the ultra series you can resample internally. what this means is that you don't have to route your sample out to resample it. say i have a melody that i want to sample. what i would do is go into my sampler's sample manage section and set it up to resample internally. you have a few choices here, what are they... i can't remember. like 16 bit and so forth, or is it hz... jezz i can't remember, but any what they do is actualy control the level of the sample. so if you have it set at one point and it is just red lining big time, you can go down a notch to the next internal sample choice, and the level of the signal decreases. what i do is pretty much set it to sample when the thresold is exceeded and as soon as i hit the space bar, the sampler starts to record.
it is really easy to do. i actualy don't resample right now, because i have found other techniques in sound design and such that i couldn't do if i resampled a sound that i have made on my emu. as you probably figured out, i create all my sounds on my emu. i have no synth....

next one: getting loops... a while ago, i bought a sound library cd that had a bunch of old synthes and stuff like that. one that interested me was the moog strings that i had. i couldn't figure out why i could hold the note for 5 minutes and it would still play even though my sampler told me that the sample was only 1 second long. so i decided i would destroy the patch. this is where i discovered that you could use the loop function here to make things sustain, even though the sample is a few seconds long. this is where i discovered how to use the amp enevelopes and the rest. to get a loop to macth this is what you have to do. you go into tools. hit tool 1, the go into loop. you will see your sample in the window, like all the other tools screens, with a start and end point, probably set at the begaining and end... incredible huh:) you will also see that you can zoom in on the sample, autoc, this means to auto correlate, this is where the sample matches the start and end up, and doesn't really work, at least not for me. and you will see a small box that says fix, and it will most likely be checked. this is actualy a way of moving the start and end together versus moving each point seperately. now if you try to move the start or end, it wont go anywere because the start and end have no were to go since the sampler has it defaulted to fix. un check that box and now you can freely move your start and end points independantly of each other. now remember, if you have a pad, you don't have to have the start be at the begainning. the sample will play from the start then loop. so if you have a phrase like "drop the bass" and you want bass to loop, you set your loop points there and when you hit the key, it will say "drop the bass bass bass bass" so what you want to do is get your start and end points at the ruff locations you want then zoom in. now what you need to do is match the wave up. try looping a sine wave sample for practice. basically you want the start point to match the end point. so if you are looping a sine wave, the window should look like a wave but with a line down the middle. now what might happen is even though it looks lined up, it may still have artifacts. what i do is i play the sample at a really high frequency and at a really low frequency. once i have it so that the high has no clicks and the low has no clicks, my loop is smooth. you can get a way with looping a few seconds worth or you can go so much as loop one cycle of the wave. mind mech does the later. this is actualy a really good idea because you can actualy make really complicated patches that take up almost no memory. you figure one cycle is 1/440th of a second, figuring you are at a4. that is small. the only thing i had a hard time with is when i was looping just one cycle i kept getting a buzz over the wave. it is like the buzz you get if you loop a sample at a really small amount. i am sure mind mech can give you better tips on looping single cycles. he knows what he is talking about when it comes to that.

and another thing, sorry this is so long. if you have a sample of a pad that has an attack and a decay and a flange and other stuff. you might have a hard time getting a natural loop, or it might be impossible. that loop i have of a moog string patch. was basically just a tone. like a sine. all the attacks and stuff was done using the emu's envelops and filters.

so there you go :spliff:

PostPosted: Wed Sep 10, 2003 4:57 am
by MindMech
Klaseed wrote:I resample constantly. But yeah, it's realy hard, especially smoothe sounds like pads.



That's what crossfade is for ;)

PostPosted: Wed Sep 10, 2003 3:00 pm
by madmax
Resampling .... arrrghhhhhhhhhh (drools).

OK actually I take that back. I hate the actual process of resampling. All that mucking about the back of my rack (no patch bay). I wish I could just hit the tab key and the damn thing would just flip around.

Using the inputs of the emu colors the sound a lot. I love it for basslines, any synth sound really. Most kick and snare sounds too. Sometimes if I want something to sound really clear (some high freq sounds, sound fx - which are usually "soft" enough, any vocals, ) I'll resample through my MOTU2408. But it never gets that EMU sound. Anyway I'd recommend learining to use both. They will each give the sound different characteristics, and knowing when and where you want to hear each one takes a bit of time.

I find that valve circuitry and the Emu inputs really compliment each other. For example, a bassline might start in the synth then be resampled through the old Radius 30 and directly into the emu. Then perhaps split in 2, and each on a different filter then recombined (or not) back through the Radius 30 and back into the EMU. Then more filtering ... etc. Until I'm happy. Each time, the sound gets more and more whacked out. Now there is certainly a point when it starts to sound worse but I really try to push it for all that its worth. Sometimes, bouncing it back to the computer and processing in there and then sending it back to the EMU works better. It's really a matter of knowing the capabilities of each peice of kit and how it will effect the sound. Anyway I'm rambing.

PostPosted: Wed Sep 10, 2003 9:33 pm
by DJ Broadband
Klaseed wrote:And yes, you can real-time sample.


I'm pretty sure I can't on my ESI 4000, but anyway I'll give the manual another once over. Thanks for the advice folks. ;)

PostPosted: Wed Sep 10, 2003 9:43 pm
by Klaseed
Ah, I think you're right. The ESI's don't have this capability, I don't believe. In which case, a lot of what I wrote doesn't help you :cry:

However, the part about the sampling inputs imparting a certain sound to your samples holds true.