KnowledgeBase | K2600X | Sampling


A Primer - Sampling with the K2 Series

Question:

Can you provide me with an overview of the sampling process and how it works in the K2000/K2500/K2600/K2661?

Answer:

Contents

This document is quite long, but it has been kept as one long document to make it easier to print out or download. You can either read it from one end to the other, or jump straight to a specific section by choosing one from the list below:


An Introduction to the Sampling Option

Back to Contents

Sampling, for our purposes, refers to digital recording.

The K2 with the Sampling Option can digitize any stereo or mono analog signal (i.e. another keyboard's audio outputs, a computer's line-level audio outputs, a microphone, a cassette player, etc.).

Analog recordings may be made via the 1/4" TRS stereo and/or XLR ports (depending on model).

The Sampling Option can also accept digital signals from devices with digital outputs, such as CD players, MP3 players, or computer audio interfaces. The Sampling Option allow users to select the AES-EBU or SPDIF formats for receiving digital data. Coax and/or optical input ports are provided (depending on model).

The Sampling Option should be installed by an authorized service center.

Sampling Concepts

Back to Contents

The term sample has several meanings we must understand:

The most common use of the term "sampling" refers to digital recording of a sound. (i.e. I just sampled a great horn hit off of Tower of Power's album, or I just sampled Harry James' trumpet from my mother's old scratched up album). This usually includes conversion of an analog signal to a digital number.

The other use of the term "sampling" refers to the smallest increment of digital encoding of an analog signal; a single increment is referred to as a 'sample'.

In the K2 object set, we also use the term "sample" to refer to an individual audio recording in RAM, as in "sample ID#200" for ex..

A Closer Look

Back to Contents

Sampling rate refers to the number of samples being taken every second. The K2 supports 4 sampling rates for analog inputs: 48K, 44.1K, 32K, and 29.4K. Digital recording is supported at all rates up to 48K (48K and 44.1K are most common). The higher the sampling rate, the better the audio frequency response (bandwidth). At a high sampling rate, cymbals sound really crisp and instrument attacks are sharply pronounced. While the higher sampling rates offer better bandwidth, they also offer less recording time. Lower sampling rates offer poorer bandwidth but more recording time. There is a chart in the Sampling chapter of the Musician's Guide which shows RAM and Sampling Capacity. The end user must decide whether they need more time or better fidelity.

Digital signals are created when the K2 converts incoming analog signals into 16 bit words, using a 16 bit A to D (Analog to Digital) converter. A bit is the smallest possible digital representation of information. All digitally stored information is made of bits which are grouped in sets of 8; these groups are called bytes. A single byte can use 8 bits in any combination to represent some given information. Obviously, eight bits will give a clearer representation of the incoming analog signal than would a single bit. The K2 uses 16 bit words (2 bytes) to store each sample.

If we record a dog's bark which lasts 1 second at 44.1K samples per second (CD quality), we will have taken 44,100 samples which the A to D is converting and storing as 16 bit words. This will use up about 88K bytes of sample RAM. If we record the bark in stereo, it will use twice as much sample RAM.

Conversion of the digital 16 bit words back to analog signals is the final stage. Digital to Analog Converters (DACs) are used for this task. Imagine the DAC plotting points on a graph based on numeric values taken from the digitally converted data. The DAC emulates the original wave form by charting its 'important characteristics'. The DAC reproduces a signal which is not identical to the original, but is close enough to 'fool' the human ear.

Entering The Sampler - Two Different Ways

Back to Contents

There are two different methods of entering the Sampling page. Which method you choose depends on the type of sampling you are doing - how many samples you are making and if you need custom keymaps. The difference between the two methods primarily concerns the ease of accessing the keymap editor. Once you have made your samples, you will need to have them assigned to a keymap and have that keymap assigned to a layer in a program. Refer to the Sampling Chapter of your Musician's Guide, in the section entitled "Building a Keymap", for a step by step explanation of how to create keymaps.

  1. From Program, Setup, or Quick Access Mode

    The simplest way to enter the Sampling page is from Program, Setup, or Quick Access Mode.

    ProgramMode    Xpose:0ST  < >Channel:1  
    __________________  197 Doomsday
    Keymap Info       | 198 Click
    Grand Piano       | 199 Default Program 
                      | 400 Grand Piano
    __________________| 401 Bright Piano
                        402 Electric Grand
    Octav- Octav+ Panic  Sample Chan-  Chan+
    

    (Note: the Sample soft button is also on the Master page.) Press the soft button labeled Sample on one of these pages.

    SampleMode  Samples:131072K Channel=1   
    Sample:None                      Src:Ext
    Input :Analog   Time:30s         Mon:Off
    Gain  :0 dB
    Rate  :48.0kHz     L
    Mode  :Stereo      R
    Thresh:Off      -dB 60  40  * 16 * 8 4 0
    Record  Auto  Timer  Preview 
    

    This is a good method to use if you are making only a couple of samples, or if you want to assign each sample to its own keymap and program. Once you have created and saved your sample, you can press the Preview soft button. This button will allow you to quickly create a program and keymap, with that sample assigned across the entire range of the keyboard. The program is a one layer program which uses the settings from the Default program 199.

  2. From the Keymap Editor

    This is a better method to use if you are going to be doing lots of multi-sampling, or if you need to create custom keymaps in which you have your new samples assigned across the keyboard in one keymap. Call up program 199, Default Program.

    ProgramMode    Xpose:0ST  < >Channel:1  
    __________________  197 Doomsday
    Keymap Info       | 198 Click
    Grand Piano       | 199 Default Program 
                      | 400 Grand Piano
    __________________| 401 Bright Piano
                        402 Electric Grand
    Octav- Octav+ Panic  Sample Chan-  Chan+
    

    Press Edit, then Keymap. Select Keymap 168, Silence, ...

    EditProg:KEYMAP           < >Layer:1/1 
    
    KeyMap:168 Silence           Stereo:Off
    Xpose :0ST        TimbreShift :0ST
    KeyTrk:100ct/key  PlayBackMode:Normal
    VelTrk:0ct        AltControl  :OFF
    SmpSkp:Auto       AltMethod   :Switched
    <more  ALG    LAYER  KEYMAP PITCH  more>
    

    ...then press Edit again.

    EditKeyMap         <>KeyRange C 0-G 10  
    MasterXpose :0ST
    Key Range   :C 0-G 10  Lo:C 0  Hi:G 10
    Sample      :168 Silence-C 4
    Course Tune :0ST
    Fine Tune   :0ct
    VolumeAdjust:0.0dB
    Name   Save   Delete Dump  NewRng Assign
    

    This brings you to the Keymap editor. (In fact you can choose any program and keymap you want to start with, but by choosing these, you are starting with a "blank slate".) Now from the Keymap editor, press the MIDI Mode button.

    SampleMode  Samples:131072K Channel=1   
    Sample:None                      Src:Ext
    Input :Analog   Time:30s         Mon:Off
    Gain  :0 dB
    Rate  :48.0kHz     L||||||||||||||
    Mode  :Stereo      R|||||||||||||||||
    Thresh:Off      -dB 60  40  * 16 * 8 4 0
    Record  Auto  Timer  Preview 
    

    This will jump you to the Sampling page. Once you have created and saved your samples, press Exit. You will now return to the Keymap Editor page, where you can immediately assign those samples across the keyboard. Once you have created and saved your keymap, you can either exit the Keymap editor and create a program which uses your new keymap, or you can return to the Sampling page for another round of sampling

...Congratulations! You are now in the sampler.

Digital and Analog Inputs

Back to Contents

First, you will need to connect your source outputs to the sampling inputs. Connect the cables from the output jacks on your source to the jacks on the sampling option. You can find information on selecting the correct cable type in the Musician's Guide.

On the Sample page you should now select the type of sample input: Analog or Digital.

SampleMode  Samples:131072K Channel=1   
Sample:None                      Src:Ext
Input :Analog   Time:30s         Mon:Off
Gain  :0 dB
Rate  :48.0kHz     L
Mode  :Stereo      R
Thresh:Off      -dB 60  40  * 16 * 8 4 0
Record  Auto  Timer  Preview 

Analog input allows you to set:

  • GAIN of the incoming signal to 4 preset levels: 0 dB, +7dB, +14dB, and +21dB.
  • RATE (sampling rate) can be set to 48K, 44.1K, 32K, and 29.1K.
  • MODE can be selected from Stereo, Mono L, and Mono R.
  • THRESHOLD can be selected from 15 preset levels (all 6 dB apart) from -90 dB to 0dB and Off.
  • TIME can be set from 1 second to the maximum time available (depending on sampling rate, amount of sample RAM memory available, and whether mono or stereo mode has been selected).
  • MONITOR allows you to monitor the analog input through the K2's internal audio system using the master volume fader to control the audio output of the sample input. (You cannot monitor digital input signals.)
  • SOURCE allows you to choose whether you are sampling from an External Source or Internal (resampling the Kurzweil's own output).
SampleMode  Samples:131072K Channel=1   
Sample:None                      Src:Ext
Input :Digital  Time:30s         Out:Dir
Cable :Coaxial
Format:AES/EBU     L
Mode  :Stereo      R
Thresh:Off      -dB 60  40  * 16 * 8 4 0
Record  Auto  Timer  Preview 

Digital input allows you to set:

  • CABLE type Coaxial (XLR) or Optical (does not apply to K2661).
  • FORMAT (digital format) either AES-EBU or SPDIF.
  • MODE can be selected from Stereo, Mono L, and Mono R.
  • THRESHOLD can be selected from 15 preset levels (all 6 dB apart) from -90 dB to 0dB and Off.
  • TIME can be set from 1 second to the maximum time available (depending on sampling rate, amount of SIMM's memory, and whether mono or stereo mode has been selected).
  • SOURCE allows you to choose whether you are sampling from an External Source or Internal (resampling the Kurzweil's own output). On a K2600 with the DIO option or K2500 with KDFX you can also select Return which samples a 2 channel return from the KDS Out A (this requires a Kurzweil DMTi).
  • OUT appears only on K2500 (w/KDFX) and K2600 (w/ DIO) models if you also have the sampling option. It doesn't really have anything to do with sampling; it determines how the K25/6 handles its digital audio output. See the manuals for information about using the digital audio outputs.

Recording A Sample

Back to Contents

SampleMode  Samples:131072K Channel=1   
Sample:None                      Src:Ext
Input :Analog   Time:30s         Mon:Off
Gain  :0 dB
Rate  :48.0kHz     L||||||||||||||
Mode  :Stereo      R|||||||||||||||||
Thresh:Off      -dB 60  40  * 16 * 8 4 0
Record  Auto  Timer  Preview 

At the bottom of the display is a series of soft buttons including Record, Auto, Timer, and Preview.

To record a sample:

  1. Press RECORD to begin sampling (recording).
  2. Sampling ends when recording time has elapsed.
  3. STRIKE ROOT KEY appears in the display.

    
    
    
    Strike root key...
    
    
    
                                     Default
    

  4. Press the note on the keyboard which you want to play back the sample.
  5. SAVE THIS SAMPLE? appears on the display.

    
    
    
    Save this sample?  (Max = -6dB)
    
    (Current channel is 1)
    
                                 Yes    No  
    

  6. You may now audition the sample and decide whether or not to keep it.
    • a. Select YES. The instrument automatically shows you the next available sample ID#. You may rename your sample at this time.
    • b. Select NO. The sample is not retained in the instrument's memory.

To record a sample using AUTO:

Auto performs the same function as record but skips all the prompts.

Other soft buttons:

TIMER sets a 10 second count-down before recording begins (handy for that cymbal you have to run across the room to hit).

PREVIEW allows you to set the bank where the newly created samples will reside. A keymap and program will be automatically created for you.

Once a sample has been recorded, it can be edited. To begin editing:

SampleMode  Samples:131072K Channel=1   
Sample:200*NewSample-C 4 S       Src:Ext
Input :Analog   Time:30s         Mon:Off
Gain  :0 dB
Rate  :48.0kHz     L
Mode  :Stereo      R
Thresh:Off      -dB 60  40  * 16 * 8 4 0
Record  Auto  Timer  Preview 

Use the cursor to highlight the sample shown at the top of the page and press EDIT.

EditRamSample:TRIM Zoom=1/256  <>Samp:1 
S:0.000   A:0.000   L:1.315   E:1.315
| -6dB> | | |
<more MISC TRIM LOOP DSP more>

...you are now in the heart of the sample editor.

The Sample Editing Menu

Back to Contents

EditRamSample:TRIM Zoom=1/256  <>Samp:1 
S:0.000   A:0.000   L:1.315   E:1.315
| -6dB> | | |
<more MISC TRIM LOOP DSP more>

You will see a variety of menu choices along the bottom of the display including: MISC, TRIM, LOOP, and DSP. Scroll through the remaining menu choices by pressing MORE>. ( <MORE and MORE> allow you to scroll through the soft button menu choices). Your menu choices will be:

  • Zoom-
  • Zoom+
  • Gain-
  • Gain+

Press MORE>. Further menu choices appear:

  • Abort
  • Split
  • Units
  • Link

Press MORE> again. The remaining menu choices appear:

  • Name
  • Save
  • Delete
  • Dump

Edit Menu in Detail

Back to Contents

MISC allows you to see the original sample root key and adjust a sample's pitch, volume, decay and release. You may select to loop or not loop, select the type of loop, position the Alt. Sample Play Pointer before or after the loop, decide whether or not the sample should ignore release, and designate the sample rate and length.

TRIM adjusts the sample start and end points.

LOOP gives a split display, with the sample shown on the left half and the loop start/end points on the right. You can search for zero crossings by simultaneously pressing the - and + buttons (under the Alpha wheel). You can search for zeros in either direction (i.e. scrolling towards the beginning or the end of the sample and pressing the - and + buttons will search for the next zero).

DSP, short for digital signal processes, provides access to the myriad of functions for editing samples. (See DSP Functions below.)

ZOOM-, ZOOM+ displays either an entire waveform or "zooms" in on a fraction of a sampled wave, as small as a single sample.

GAIN-, GAIN+ amplifies a wave in order to see what, if any, nuances exist-as far down as -72dB. (This is helpful to see the actual noise floor).

ABORT allows you to cancel a sample dump that is in progress.

SPLIT allows you to split a stereo sample into two separate mono samples.

UNITS changes the information shown at the top of the sample editor pages from seconds to samples. This is great for people who work in the time domain vs. the sample domain (i.e. Akai users).

LINK allows you to lock together different fields in the sample editor so that adjusting one will adjust all parameters that have been linked by the same amount. This is great for moving simultaneously the start and end points of a loop, or adjusting the sample start and Alt. start pointers by equal increments.

NAME, SAVE and DELETE all are self explanatory.

DUMP allows you to send a sample dump over MIDI via SDS.

List of DSP Functions for Samples

Back to Contents

  • NORMALIZING GAIN (boosting the sample gain to just before clipping)
  • TRUNCATION (auto dB search command not to be confused with TRIM)
  • VOLUME ADJUSTING
  • CLEARING SAMPLES (leaving a space, not rejoining the materials on either side of the cleared space)
  • DELETING SAMPLES (and rejoining the data)
  • REVERSING SAMPLES
  • INVERTING THE PHASE OF A SAMPLE (useful when separating a stereo sample and inverting the left side out of phase with the right side)
  • INSERTING ZEROS (silence)
  • MIXING SAMPLES
  • INSERTING SAMPLES INTO OTHER SAMPLES
  • CREATING VOLUME RAMPS (fade in/out)
  • CREATING CRESCENDOS (similar to volume ramps)
  • RESAMPLING (from higher sample rates to lower sample rates, or vice versa)
  • TIME WARPING SAMPLES (compression/expansion-where the time changes but not the pitch)
  • PITCH SHIFTING SAMPLES (changing pitch but not time)
  • MIXING SAMPLES ON A BEATS/MINUTE TIME LINE (MixBeat - where samples may overlap)
  • REPLICATE (similar to MixBeat but samples may not overlap, and will replace any samples which overlap as opposed to mixing them as in MixBeat)
  • CREATING DIGITAL ECHOES (sample copies-MixEcho)
  • BEAT VOLUME ADJUSTMENT (adjusting a particular beat's or sample's volume in a MixBeat or Replicate sample stream)
  • CROSS FADE LOOPING (allows you to create smoother loops)
  • DYNAMICS (a general purpose compressor with a few features that make it suitable for creating smooth sample loops) - not available on K2000 series units or K2500 without KDFX.

Most DSP functions have user-selectable curves for cross fading and user- definable cross fade amounts, plus they allow the user to define the start and end points of the particular segment to be operated on.


Need more info? Check out the following links: