Don't know - tell us what you are doing.

I load VST Editor in Cubase 7.5.3, synchronize it with the Motif XF, menu VST Editor do import library: File - Import - library file. Then choose a category Bank. Note in the left and right column Normal User 1. I press Import - after the process closes the dialog, open the bank in the selection window presets, choose presets and all imported from [injuring as VST Instrument.
After trying to re-download this tool in the VST Media Bay, he shows that loaded, changing its name in the XF and VST Editor, but a number of instruments do not sound or sounds wrong.

When you are working with a library that contains new Samples and Waveforms, you must be concerned with the installation of that data onto your FLASH Board to ensure that when VOICES are recall they address the correct Waveform location. Since in your post, you make no mention of having loaded/installed the Waveform data necessary to make the CP1 (Sample) Voice Library sound properly.

The Editor does not move the AUDIO (Waveforms and the Samples that they organize) - It only handles the Voice parameters (pointers).

You cannot skip the important step of actually adding the Waveforms (and the Samples that they organize) to your Motif XF's available Wave ROM. This requires that you actually use the LOAD routine first.

Using the VST PRESET librarian requires that your Motif XF's Waveform Lists contains a consistent position for each Waveform. All the VOICES that you Save as VST PRESET, will require that your Waveform List is a "known" quantity.
_ The Preset Waveforms, all 3,977 of them, are always listed in the same order - so this makes it easy for VOICE data to access them.
_ The User Waveforms, 2048 of them per FLASH BOARD, that you install will be listed on the FLASH Waveform List ([UTILITY] > [F6]) in a consistent order - so that the VOICES you install to the MediaBay's SOUND BROWSER will always "know" where their Waveforms are located.

Without having a consistent position for your Waveforms, the Sound Browser function will not work for you. This is why it is important, to add ONLY the Waveforms that you want to have access to. In other words, once you purchase a Library and decide what VOICES are will wish to add to your personal set... you should install only the Waveforms to make your favorites work, never install a Waveform that does not support a VOICE you want to have access to.

You must think each of your Flash Board's 2048 Waveform Locations as 'semi-permanent' locations to support the VOICES you think you will use. There is never a reason to install Waveforms that just take up space.

So specifically, what you did wrong in this case was: You did not LOAD the Waveform data from the FILE. This cannot be loaded via the Editor. The Editor deals with VOICE parameters only, (not the actual audio data of Waveforms/Samples). You must LOAD from a USB drive in the normal fashion and install the custom Audio data to support the CP1 VOICE Library. Then once that is done, then and only then, will SAVE AS VST PRESET be logical and supported.

Remember, this is what happens: The VOICE data is dynamic, when it comes to the Wave Bank and Wave Number. If you have a VOICE and its data says to "look" in the Preset Wave Bank for Waveform Number 0001 - the result will be the Grand Piano Waveform that occupies Preset Waveform 0001. If the data says "look" in the USER SDRAM Wave Bank for Waveform Number 0001, then it is always going to necessary to make sure you restore that Waveform to the USER SDRAM - which is volatile memory. If the data says "look" in the FL1 or FL2 slot for the Waveform Number 0001, then it will always find the Waveform that YOU decided to install in FL1, number 0001.

So the Voice contains "pointers" - they literally direct the rest of the VOICE parameters to trigger the audio stored in the target location... if you do not ensure that the correct audio data will be found there, then you will not get consistent results - because you have not satisfied a fundamental to this concept: You want to create a consistent location for the Waveforms that you deem "keepers". This way you can recall a Voice in seconds, not wait minutes while the large audio data is restored.

And the key to this all is - no matter where you load the individual VOICE data, it points to the same Wave Bank and Wave Number. So you can load the Full Grand Piano in any USER VOICE location - it always "finds" its Waveform. You can load any VOICE from the CP1 Library - but you must make sure you INSTALL the Waveform data to a FLASH BOARD. Because what you want to establish is a "semi-permanent" location for the Waveform data. So that no matter where you import one of the CP1 pianos (whether you load the Voice data to "USER 1, A05" or "USER 4, H16", it will always know where to look for its Waveform data.

Hope that helps.

Thank you! I figured out! You need to load the library in the XF with the stick, then save as a VST instrument. Kind regards!

Also if you Store the Voices in your Motif XF you may want to add a "Category" and "SubCategory" to the Voice data before you execute the "SAVE to VST Preset" routine. This information allows the Sound Browser to place your Voices in these divisions when it transfers the data. If you leave the Voices set to "Category = No Assign"; and "SubCategory = No Assign", that is how it will be placed in the Sound Browser.