Hello. In the Peter Jung Symphonic Library I downloaded, many of the samples seem to have peculiar transpositions & keybank distribution. For example, a sample will be transposed up 6 or 7 semitones, and the key bank will be a group of notes about an octave to half-octave below the actual sample.

I could fix this myself. I am asking here to see if anybody else has encountered this, and if so, if they knew why. Perhaps it’s intentional and I am missing the reason.

I am not complaining. I am grateful to both Peter Jung & Easy Sounds for the hard work involved in providing these high quality samples.

Thanks,

Whilst I can’t comment on the key distribution, and as a total stab in the dark, could it be that the transpositions that you’ve encountered are deliberate, because they are transposing instruments and not concert pitch instruments?

For example, an Alto Sax in Eb sounds one Major 6th lower than written and a French Horn sounds one perfect 5th lower than written.

dsetto - 30 November 2016 12:15 PM

Hello. In the Peter Jung Symphonic Library I downloaded, many of the samples seem to have peculiar transpositions & keybank distribution. For example, a sample will be transposed up 6 or 7 semitones, and the key bank will be a group of notes about an octave to half-octave below the actual sample. [...]

I might be able to shed some light on this, but knowing exactly which samples you’re referring to could help. Would you give some specific examples?

Thanks guys. … The waveforms in my collection are mostly orchestral strings, and one choir, a few synths, and percussion.

The particular waveform that seems will require some editing to make it sound conventional is “So. Syn. Str.”. When I played a simple C major scale on the Melas Waveform Editor (MWE) on-screen keyboard, that is not what sounded. I considered the possibility that somewhere, somehow, something was changed on translation from one format to another. (I got this as a promotional; and I think they sound terrific. And, I am not complaining; I am grateful.)

The peculiarities I am observing is from within Melas Waveform Editor ver. 2; specifically from the on-screen keyboard, and the Keybank Parameter windows per sample. Moreso, I have not loaded these into the Motif yet (Flash nor SDRAM). And, I have not tried to play them from a MIDI keyboard, controlling MWE. It is my assumption that the on-screen keyboard in MWE mimics what a connected MIDI keyboard would do (other than velocity levels).

I haven’t checked every sound, but what I am observing is found in the following waveforms:
BSS L
Em. Str. L
Choir L
So. Syn. Str.

So, I dug deeper, and checked out the keybank parameters. Then, I noticed the intriguing way a sample was mapped out on the keyboard. From memory, I believe “BSS L” sounded conventionally, as expected, when triggered on the MWE on-screen keyboard. When I reconciled the different approach of key mapping, but, a proper key-trigger to key pitch correlation, (i.e., a pressed C sounds a C, pressed C# sounds a C#) - I assumed it was an intentional technique to achieve a certain goal. Perhaps, to create uniformity of timbre, spread out on a keyboard, reconciling a variety of different string instruments, over a wide range. So, in BSS L, it appears there is never an instance when a key is pressed, and the sample for the pitch typically corresponding to that key is triggered. In other words, it appears to me, that one never hears the actual sample, in its raw, non-transposed form. I haven’t loaded or played BSS L-R yet, so I’m not commenting on the result. I’m sure it’s terrific; even if mapped unexpectedly.

--
Perhaps the “programmed” coarse tunings are compensating for incorrectly assigned roots. Perhaps the roots were assigned automatically, for expediency.

I noted the following behavior in Melas Waveform Editor v.2 on Windows that may explain the oddity I observed. And moreso, it may very well make this a non-issue, once these waveforms are loaded into & played from within the Motif.

- I changed the coarse tuning on a keybank that was part of a waveform that behaves as expected. (By expected, I mean normal with regards to correlation between MWE’s on-screen keyboard presses & correlated pitches (key-press C -> sounds a C, etc…).
- However, when I changed the coarse-tuning of a keybank, and pressed “ok” for it to take effect, the key-press did not register the coarse-tuning edit I just made; which was present when I opened up the keybank parameters window again.

--
So, my current assessment is:
for some reason, in my experience with Melas Waveform Editor, coarse-tuning alterations (from 0), are not registered by its on-screen keyboard.
And as such, the on-screen keyboard is not always an accurate representation of what to expect on how a keybank will playback once it is loaded into the Motif.

This is all my guess. I just figured I’d add it here, because it may clear up my initial confusion with the Jung Symphonic library.

The most accurate/realistic-sounding sampling tends to occur when every note is sampled. Unfortunately, unless the samples are very short, a large amount of memory may be needed. Therefore, samples are often taken at several-note intervals, and the samples stretched to cover the gaps. Stretching them too far (which sometimes isn’t terribly far) can result in unnatural sound, possibly including very unmusical artifacts. Also, some things might sound better when stretched upward, but others when stretched downward.

In addition, samples that include modulation (tremolo or vibrato) can sound wrong when stretched because the speed of the modulation also changes.

“So. Syn. Str.” is synthesized, so what’s “natural” for it is impossible to know unless we knew the original source. From what I’ve seen in the library, various Parameters were used to modify the sound of the Waveforms in several Voices, as might be expected. For curiosity I loaded the PJ Symphonic library to my XF, and determined that USR4:064(D16), “*PJ Syn.Cl. Vox”, uses “So. Syn. Str.” in Element 2. However, when I soloed Element 2, it sounded quite different than the raw Waveform. One reason was that significant low pass filtering was applied in that Element.

If you want to use the Waveforms for your own purposes, and they don’t quite meet your needs as-is, they could certainly be similarly influenced at Voice Element level. Of course, the Waveforms themselves could be edited - different note ranges, loop points, etc., are easy enough to set.

Thank you.

dsetto,

You should just load the library and try it out. The voices are mot transposed.

Not sure why you are poking about in the waveform editor.

I suppose it depends on how you define “transposed”. For example, in the KeyBanks of “So. Syn. Str.”, Coarse Tune ranges anywhere from -15 to +25.

Thanks guys. Good to hear that once the waveforms & samples are loaded into the XF, the Motif will register the keybank-level transpositions & mapping; and a Voice with these PJ Symph waveforms will sound as expected. I.e., playing a C scale will sound a C scale.

If I find that, I’ll assume that I can’t expect the MWE on-screen keyboard to register MWE Keybank Parameter Window transpositions in the MS Windows, ver 2.

If possible, could someone corroborate that Melas Waveform Editor’s on-screen keyboard does not register MWE’s Keybank Parameters “Course Transposition”?

If this truly is the case, I would graciously request if it’s possible to add this ability to Melas Waveform Editor. I don’t want to be a pest, however, as I am very grateful where it is now.

I believe the value of this abilit would be:
A more comprehensive software environment for prepping Waveforms for the Motif. (Incredibly cool on a laptop; untethered from the Motif.) And it’s a non-performance job that can be done well, even away from a physical, external keyboard.)

-- as a portable alternative, a super tiny external midi key-like controller may do the job; assuming MWE will take MIDI input and register keybank parameter “coarse transposition” alterations.

The following might help to clarify the situation…

Waveform #25 in PJ Symphonic library is “Solo Str.” It contains 30 KeyBanks, with Root Keys designated as C3 through F5. However, the sampled notes apparently don’t directly correspond to the Root Key designations. Here are a few examples:

R-K > Note
C3 > C1
C#3 > C1
D3 > D1
D#3 > E1
E3 > F#1
F3 > G#1
F#3 > A#1
G3 > C2

Playing that with the virtual keyboard in MWE, using a mouse or a connected MIDI keyboard, doesn’t result in a chromatic scale. (The Motif itself can play a correct scale with the Waveform.) Most sample libraries have a better correspondence between the designated Root Key and notes, and MWE can then play a chromatic scale.

By the way, I find Waveforms in the PJ Symphonic library to be of inconsistent quality. Some of the samples are good, but others have quite audible loop points and artifacts.

great. got it. thanks.