Advanced Sampler Use – Question about Memory Scanning

[tomylee]

hi there,

listen to this:

http://www.filedropper.com/demomonowavetable

Im trying to browse through a sample with the start (sampe start) knob (i used a modified simple sampler patch) while the sample play is triggered by a fairly high oscillating LFO 😉

You can clearly hear the dropouts, it probably was not meant to do that, but im sure the tide is capable, I just wish there was a sapler with a sample start mod input, that way very nice experiments would be possible – I found that the clavia nord modular can do this, but its a hazzle there to program it…

So my question: how would you get rid of those dropouts changing the start point while the sampler is triggered constantly that fast… I also tried to let an lfo do the change in the start time, but it just made the Sampler go <<FILL>>

there must be a way

have a nice day everyone

tom

[IDeangelis]

Tom

I don't know what you are trying to do but it seems correct what I hear in the mp3. The LFO is continuosly triggering the sampler on and off, so I would expect what I hear.

I

[tomylee]

yes, but i am constantly changing the start time parameter knob manually, so the start time changes while the lfo still triggers (of course then it triggers always different parts of that sample) but you can hear clearly dropouts and unsmooth clicks, dont you?

I wanna get rid of that, so I can cycle through a sample without those dropouts, you can hear it, you just must ^^

any suggestions on how to control that start time parameter with something else, for example an lfo – and most importantly smoothly?

kind regards

tomylee

[IDeangelis]

Tom

again, it's not clear what you are doing and what your expectations are. At that VERY fast triggering speed talking about "dropouts" doesn't really make sense. Why would you use an LFO to trigger a sampler? Why THAT fast?

Anyway, there are ways to smooth control signals, using an hicut with c_to_a and a_to_c bridges bafore and after it, or use only c_filtered_a/a_filtered_c with a setting of 50. There's also a C_Smooth…..

best

I

[tomylee]

i think its the update rate, not the rate of change that makes problems with the start time input. The sknob surely has a different (lower) update rate as an lfo through an a-filtered-c module…cause that makes the sampler go <<Fill>> which is busy…

So is there a way to thin out that second lfo for changing the start time? Maybe a sample and hold would be the solution, I have to find a way now…

Cause in fact the knob works, just with dropouts, so maybe I just have to optimize on that behaviour a bit…

[tomylee]

http://www.filedropper.com/demomonowavetable2

here you go…

for your why questions, although im curious about you are interested anyway, cause I just wanna know how, and not get asked why ^^ but i can tell you. I found a way to control the lfo´s frequency scaling perfectly to my midi keyboard, while I send the tide no midi, the midi to cv conversion are doing other devices, cause the tides midi timing is not usable for me.

What that means? I am at creating a full blown wavetable synthesizer where the wavetables are easily recordable and changable, and it just sounds fantastic and I already did a nice track with it – the next step though would be able to scan through the sample/wavetable, and that is what you are hearing in the new demo. Just that I am scanning manually there, and I would need to let an audio rate-ramp do the scanning, cause then I can easily choose what part is to be played by cutting it in my sequencer…

Even if you dont really understand all of this, just tell me if there is a way to get the start time changing smoothed out, I dont seem to find a way, at some knob turning speed there are just dropouts, and the lfo i did not get to work with that damn control input anyway, as I already told you before.

Hey if you find a solution for me, I send you a sample of use of this program, and you will quite likely be blown away, deal?

cheeers

tomylee

[IDeangelis]

Tom

we ask questions as "why" because knowing what exactly one is doing IS the key to understanding what could be the trick in a sigfile. These things are not easy, so be patient.

The sampler may not be capable to do what you want. It was built for samples replacement in the studio. Control inputs may be smoothed with the modules I described in a previous post:

-CtoA/AtoC

-CfltrA/AfiltrC

-Hicut

-Csmooth

Unfortunately the link doesn't work here.

all the best

I

[tomylee]

the link:

http://www.filedropper.com/demomonowavetable2

should work now!

[IDeangelis]

Tom

I have no suggestions for you, other than the modules names to smooth out control signals.

I don't exactly understand what you are doing, but my feeling is that this sampler is not meant to work like you are trying to do.

all the best

I

[tomylee]

IDeangelis:

but my feeling is that this sampler is not meant to work like you are trying to do.

oh thats no issue, my gf for 1 jear also thought she was not meant to "work" for me like I tried, then I changed her mind, and the same thing im going to do with that sampler. 😉

I found an issue here that is following: When changing the start poision forward, while the sampler is playing forward, it is possible that the turning of the knob is actually faster than 44100 samples /sec and thus makes the sampler to stumble/dropout…

When I scan downward I have no dropouts – thus, there would be a half decent solution for this, When the lfo (btw I got the lfo to work ^^) scans upwards, the sample playback should be reverse, when the lfo scands downwards the sample playback should be forward.

Now I was looking for a module for about 1 hour, and I just know you got the answer: what module outputs a 1 when the incoming data stream is increasing (rising lfo) and outputs a 0 (or a -1) if the incoming data stream if falling?

This is the module that would give me a hald decent solution. To make everything aligh with the beat etc etc, I would then have to find some tricks, but I think Im going to get close to what I want if you for now just tell me that mudule, that as desscribed, gives a 1 by rising data stream, and a 0 or -1 on falling data streams…that way I can make the sampler playback direction change depending on the direction of the lfo (or ramp) changing the start position

Can you follow me?

^^

cheers and thanks!

tom

[IDeangelis]

try the following:

C_XOR

C_CMP2 (which should be better than C_COMPARATOR)

C_NOT

C_OR

all do logical operations on control signals

best

I

[haven]

by the description I think he is trying to do wavetable synthesis with the Eventide. none of the links work so I cant listen to the examples. Wavetable scanning requires fixed frequency wavetables so that as the start and end points of the scanning playback postion move through the table the start and end of the wave are on zero crossings. various comercial products allow you to use your own wavetables but are still restricted to needing a properly formatted wavetable to function correctly.

http://en.wikipedia.org/wiki/Wavetable_synthesis

using crossfades at the start and end you could get around the fixed frequency limitation but you would have to use crossfades longer than your lowest frequency cycle to avoid pops.

cool idea tommylee! I hope you share your patch when it is done.

other products that sound similar to this idea are the Harvestman Tyme Safari, Elektron Monomachine mk2 and MAX/MSP

[Author]

Posts