(Click to view full size.)

Choosing ‘Import’ from the pop-up menu in a Source content field opens the Import window.

The central portion of the window provides a simple file browser in which you can select WAV, AIFF and SFZ format sound files for importing. The AUTOPREV button toggles automatic previewing on and off. With AUTOPREV activated, you can hear a preview of a sound file simply by clicking its name. Sound files can be mono or stereo, 8-, 16- or 32-bit, at any sample rate (although rates higher than 44.1 KHz are not recommended, since they don’t yield any significant improvements in quality).

You can import sound files to any of Alchemy’s various elements: to granular, additive, or spectral, or to a combination of additive and spectral. To choose one of these import modes, use the GRANULAR, ADDITIVE, ADD+SPEC, and SPECTRAL buttons across the top of the Import window. Importing to the additive and/or spectral elements requires Alchemy to perform a special analysis of the file, which may take a moment.

A description of each import mode follows, with comments on the relevant options.

Granular

Importing to the granular element is good for drums loops, percussive sounds, and any sound to which you want to apply special granular effects. You can also use granular import and then switch the Granular element to its ordinary ‘Sampler’ mode, which is the best choice when you want to play a sound file efficiently in its original form, transposed across the keyboard.

• RootNote determines which MIDI note will play the sound file at its original transposition. If the filename has a pitch appended to it (‘MyFile C2’, ‘YourFile-E3’, ‘OurFile_F#4’), Alchemy will automatically set the RootNote parameter to match.
• The Range and TimeFreqAcc parameters do not apply to Granular import.

Note: If the sound quality is not what you expected, check to make sure the Granular element is switched to ‘Sampler’ mode, unless you intend to perform granular-specific manipulations such as time stretching/shuffling.

Additive

Importing to the Additive element allows for the most detailed manipulation of sounds, and is especially good for sound files that represent single notes (rather than chords or more complex sounds and textures).

• Good results with additive importing depend on an appropriate choice of RootNote. If the filename has a pitch appended to it, Alchemy will automatically set the RootNote parameter to match; otherwise, it will attempt to determine a suitable RootNote based on a preliminary analysis of the waveform.

Note: If a sound fails to import as well as you think it ought to, you may be able to get improved results by shifting the RootNote setting up or down an octave.

• The Range setting is also significant. For sound files that represent single notes (rather than melodies, chords, or more complex sounds and textures), choose ‘SingleNote’. For most other sounds, choose ‘OneOctave’. Again, Alchemy will attempt to determine an appropriate setting based on a preliminary analysis of the waveform.

• TimeFreqAcc determines whether ‘time accuracy’ or ‘frequency accuracy’ is prioritised during the import process. In most cases you can leave this at the default (‘GoodTime’), but for some material different settings may produce better results. (For example, to analyse speech, you should probably choose ‘BestFreq’.)

Add+Spec

When a purely Additive import fails to capture the noisy components of a sound (such as the hammer strike of a piano or the breath noise of a flute), a combination of Additive and Spectral importing may give the best results.

• After an Add+Spec import, try deactivating the Source’s Additive element and listening to the Spectral element on its own. This will allow you to hear which parts of the sound failed to be reproduced by the Additive element, and should give you an idea what kinds of sounds are better-suited to the different analysis methods.
• By default, the Spectral component of an Add+Spec import is played with the Spectral element set to ‘Noise-Resynth’ mode. Playing the spectral data back in ‘Resynth’ mode produces a markedly different effect, which you may sometimes prefer.
• By default, the Spectral element’s internal highpass control is set relatively high in order to exclude frequencies that would compete with those produced by the Additive element. If you find the default contribution from the Spectral element to be too subtle, you can make it more prominent in many cases by setting the highpass control to a lower value.

Note: If a sound fails to import as well as you think it ought to, you may be able to get improved results by shifting the RootNote setting up or down an octave.

Spectral

Importing to the Spectral element allows effective manipulation of polyphonic sounds (chords, drum loops, and other complex sounds and textures that don’t analyse well via Additive).

• Set RootNote to determine which MIDI note will play the resynthesized sound at its original transposition. If the filename has a pitch appended to it, Alchemy will automatically set the RootNote parameter to match.
• The TimeFreqAcc parameter can also be adjusted to suit different types of source material (see details in the discussion of Additive import, above).
• The Range parameter does not apply to Spectral import.

‘Import’ versus ‘Load’

The ‘Load Audio’ and ‘Import Audio’ commands available in the Source content field’s pop-up menu both bring audio files into Alchemy.

‘Load Audio’ can be used load a WAV, AIFF, or SFZ file rapidly into the Granular element for playback in Granular or Sampler mode. You can also load files by clicking the Source content field’s forward and back arrows.

‘Import Audio’ is for when you want Alchemy to analyze and resynthesize a sound file, or when you want to preview sounds before loading them. When a sound has been imported with additive or spectral analysis, the resynthesis data is stored in an ‘.aaz’ file in the same folder where the preset is saved. The first part of the .aaz filename is the same as the preset name, ensuring that the preset and its associated .aaz files will be found side by side in an alphabetical directory.

Note: It is recommended that you avoid renaming .aaz files externally (i.e. from your operating system); Alchemy may fail to locate and load these files unless their original names are preserved. While renaming a whole preset externally is less risky, it may lead to duplication of .aaz files to reflect the new name. In general, it’s best to rename presets by using Alchemy’s SAVEAS function. Once a preset is successfully saved under its new name, you can safely discard the old preset (*.acp) file.

Supported SFZ opcodes

Alchemy is capable of importing multi-zone sample data in SFZ format. SFZ is a non-proprietary file format (described at http://www.cakewalk.com/DevXchange/sfz.asp). Alchemy recognizes the <region> and <group> headers and the following opcodes:

sample
pitch_keycenter
lokey
hikey
key
lovel
hivel
loop_mode
cutoff
fil_veltrack

There is an upper limit of 100 zones when importing an SFZ into Alchemy. If you try to import an SFZ file that exceeds this limit, the surplus zones will be absent; the best workaround at present is to construct two or more SFZ files, each within the 100-zone limit, and then import each file to a different source in Alchemy.

You can create your own SFZ files for use in Alchemy by using various third-party software tools. SFZ definition files are simple textfiles, so another approach to assembling your own SFZ files is to place a copy of an existing definition file in a folder along with the new samples you want to use, and then to edit the definition file — in a text editor such as NotePad (PC) or TextEdit (Mac) — so that it references your new samples.

A Limit on the Size of Imported Sample Data

Alchemy doesn’t directly apply any limits on the length of sounds that you can import into its synthesis engines. However, Alchemy will fail to import sounds if it runs out of memory. For spectral and additive analysis, the amount of memory required is several times the actual size of the WAV or AIFF file. A general rule of thumb is that if you have at least 1GB of free memory, you should be able to import sounds up to five minutes in length with any analysis method without problems.

Note: In versions prior to 1.10, Alchemy’s import function was limited to a smaller file size. If you are experiencing problems with importing relatively large files, please make sure you are running the latest version of Alchemy.

A Limit on the Size of Sample Folders

In any single folder within Alchemy/Samples, no more than 1024 samples and folders are permitted. (This limit allows Alchemy to construct menus safely when you choose the ‘Load Audio’ popup command in a Source content field.) If you really want to have, say, 2000 samples in a folder such as Samples/User/Bass, you can organize them into subfolders within that folder. For instance, Samples/User/Bass might contain 500 samples plus five additional folders, each of which contains 300 more samples.