We can help you _ what is the exact name (include extension) of the file you are attempting to load into your MOXF?

Thanks, Bad Mister for your help. The name of the CS80 Bonus Voice file is "CS80BONUS.n3.X3A" - shown file size on PC: 2,21 MB (2.327.507 Bytes). The file is included in the "B. CS80 Library and Bonus" folder as part of the 40th anniversary premium contents pack.

The "CS80 Bonus" Voices were created totally separately from the main Motif XF CS80 Library by Dr. Jung and were provided as "bonus" Voices for Motif XF owners – using the newly added HD Reverbs, Distortions, etc. added with firmware 1.50.

The reason you cannot load the Bonus Voices to the MOXF is because the data is coming from two separate Files made for the Motif XF _ One made by Dr. Jung (the main archive Library) which includes the WAVEFORMs and SAMPLEs, and the CS80 Bonus which was created in the Motif XF by Athan Billias using a Motif XF that had Dr. Jung's data loaded in it.

The CSBONUS data was created specifically for Motif XF and will not translate properly as programmed by Athan. At some future point – a Bonus Bank for MOXF may become available… currently one does not exist.

Here’s what we believe happened when you made attempts to load this XF data to a MOXF:
When you attempted to load from the Motif XF Bonus Voice File using the '1 Bank Voice' File TYPE, it does not have everything it needs to make the data sound properly (and it knows it) – this is because the file was saved “without Samples”. When you see the "Illgeal file!" you have made a setting that cannot be carried out, because something it needs is not within the file. If you ask the Bonus File to load "with WAVEFORM" for example - that would cause the MOXF to display an "Illegal file" message because the Bonus File does not contain the actual audio data necessary to make the data sound. Remember this Bonus bank is made with a Motif XF with the main Library (and the Waveforms pre-loaded.

Anyway that is why the error message appears.

As to using this particular Motif XF BONUS data in a MOXF: unfortunately, you do were not provided the "instructions" for a MOXF to translate (repoint the critical Waveform data properly) and load this data. Technically, speaking you would need a Motif XF ALL data file that included both the Waveforms and Samples in order for the MOXF to do any type of translation of the instructions. The "instructions" are what keep the Voice data looking in the right location for the audio to make the OSCILLATOR WAVE sound properly.

The Motif XF BONUS File (.n3.X3A) can be loaded into a Motif XF that has the Jung CS80 Library loaded in and all Voices would look to the correct locations - it is the same as creating a FILE that contains a known Waveform list. There is not enough information for the MOXF to do the reuniting of these new VOICES with the data locations from the other File. (Apparently it needs to read and verify the data from within a single file). And it is a good thing...

Why was the file saved without samples? Because even if you loaded the CSBONUS data into a MOXF, many of the Voices would not sound as programmed because the new effects are going to missing.

You would need either to have a Motif XF to combine both sets into one; or you can do this with the Melas "Waveform Editor". (I've done this...) But even if you did any of these exercises to make the VOICE data load to the MOXF, the CSBONUS.n3.X3A is really designed for those who are going to play them in a Motif XF. Voices are programmed to utilize the new Effect algorithms (added to the Motif XF in the most recent firmware update) – the "Bonus" was to have a bank of Voices utilizing the new Effects TYPES.. these are not found, at all, in the MOXF – when creating the MOXF version of the Premium Content Pack we did not include the CSBONUS bank because of this difference – it is one of those things about programming, if you do not have the same tools available some of the VOICES will simply not make sense. My recommendation: Do not worry about the CSBONUS bank for the MOXF – at a future time these Voices may be released, after being updated, not only to MOXF format, but so they make sense played from a MOXF.

Hope that helps.

The good news is that I did find a way in the meantime to fix the problem, which finally enabled me to load and to access the CS80 bonus voices in my MOXF as expected. Bad Mister, thanks a lot again for investigating. I just read your explanations above which mostly match with my personal findings I'd like to report below. However, I don't fully agree with your comment that the reason for not making these bonus voices available for the MOXF series are the "new" XF effects introduced with the XF F/W release V1.5. To my knowledge, even these effects can be achieved with the MOXF. If not, what about all the promotion claiming that 100% voice compatibility between XF and MOXF?

However, as you mentioned, no MOXF owner can solve that kind of issue without using the resources of a separate Motif XF board or without using external tools like the Melas Waveform Editor - that's something I had to find out today as well. But as I'm the happy owner of the whole Melas Suite for MOXF, I was able to fix the problem by using the Melas Waveform Editor, exactly as you proposed. I didn't even have to "combine" both, the main and the bonus library to a new one as you suggested, I only had to "tweak" an XF-specific option in the bonus library X3A file to make it also load into a MOXF.

As you explained, the "CS80BONUS.n3.X3A" includes 61 waveforms which are refering to samples which are actually not existing in the library itself, as they haven't been saved together with it. It is assumed that these waveforms/samples have already been loaded before, using the "main" CS80 library file. In fact, the 61 waveform references in the bonus library are the same as in the main file, but just pointing to "ghost" samples which are actually not there. Now, the real problem seems to be the fact that these "pseudo" waveforms have been saved to the X3A file's "FL1" memory space. In such a case and combination, the MOXF seems to be not able to load any related banks/voices from that library correcly, producing errors like "No sample data." or "Illegal file.", depending on which bank/voice you try to load. As I found out, this issue will not occur if the waveforms would have been initially saved to the "USR" memory space inside the X3A file. So, from my personal point of view, there's still much room for improvement on the current MOXF F/W, as the MOXF should show the same "tolerant" behavior when loading X3A files as the XF does.

Finally, to prepare everything in order to let the MOXF "recognize" the X3A data correctly, the following procedure has to be performed:
The original bonus library has to be loaded first into Melas WaveForm Editor application, and the whole set of "pseudo" waveforms has to be moved from the "FL1" to the "USR" memory location (there's a right-click command available for moving all selected waveforms). After that, everything can be saved back to the same (or a new) X3A file. After having performed that procedure as described, the MOXF will finally be able to "recognize" the correct waveform references and to use the related samples properly whenever the voices/banks are loaded via the "1BankVoice" or "Voice" procedures. Sorry, but this is no feature, this is a bug...

So please, Yamaha developers, maybe you can take another look at the mentioned parts in the source code, as this would really be a great improvement for any future F/W update (if such is ever planned again for the MOXF). The drawback for my described workaround is that this (MOXF-specific) loading solution will only apply as long as not more than 128 waveforms are available in the original library (as you know that the "USR" memory can not hold more than that 128 waveform "slots").

Okay, thanks again to everybody and especially to Bad Mister for his investigations and explanations.
As an enthusiastic MOXF owner I still hope that the "best selling Yamaha synthesizer" will someday get the attention it deserves, even at Yamaha!

To my knowledge, even these effects can be achieved with the MOXF. If not, what about all the promotion claiming that 100% voice compatibility between XF and MOXF?

The new Effects are NOT available and cannot be achieved with the MOXF, period. the MOXF will not read Waveform and Samples from a Motif XF file that was saved "without samples", period. I stand by my point (you are not hearing these Voices as intended by the programmer) as long as you go into this with your eyes open, that's all that matters.

As of Motif XF version 1.50 the instrument contains items not available to the MOXF.

Okay, I take it - some effects may not be included in the MOXF - you must know better than me.

But what about the procedure I explained above? Let's assume we're talking about another library, created the same way, but just as intended by the programmer. How can be explained that loading from "USR" wlll work on a MOXF but not from "FL1"? After all it's the same set of waveform data..I just want to understand it.

Thinking again about my own question I raised last, I figured out that it already contains the answer.

In each case an original X3A file was saved WITHOUT samples (meaning that only user waveforms are included along with the voice data, but no actual sample data have been saved with it which are supposed to be on the flashboard already), the following "translation" is currently performed by the MOXF loading routine if such an X3A library is loaded:

Motif XF "USR" waveforms in slots 1~128 will be loaded by default to MOXF waveform slots 1~128 (1:1 allocation).
Motif XF "FL1" waveforms in slots 1~2048 will be shifted to MOXF waveform slots 129~2176 (offset = +128).
(Of course, the flashboard limit for the MOXF is still 2048 waveforms which cannot be exceeded)

Exactly that finding matches with my observations explained above in my initial question, where I already noticed that - after loading such a library - the waveform references are pointing to a much higher waveform slot (for which no waveform is actually available on the flashboard). So, the "information" is there in the file, it just has to be interpreted correctly.

Therefore, from a MOXF user's perspective I can only stick to my opinion that it really wouldn't be any rocket science for the R&D team to consider a "-128 offset" (for compensation) in the MOXF loading routine. By interpreting the contents of the waveform occupation inside the X3A file correctly, the decision could be easily taken by the MOXF loading routine if such an offset applies and, depending on that, how to set the waveform references accordingly. This step would ensure that these "WITHOUT samples" X3A libraries could also be loaded into the MOXF without any problems. Even the possibility for a "manual" parameter choice in the MOXF "load screen" would already be a great improvement.

With all respect, this is not how it works. The "translation" of data is much more complicated than a simple offsetting of data by -128 would cure. You are looking at the illogical result of an error and trying to draw conclusions about how the entire system works... This is very much like trying to draw a picture of the entire elephant by only inspecting a part of the tail. You just might draw some incorrect conclusions.

The .n3.X3A extension tells me that SDRAM Waveforms and Samples might be in this file, but it definitely does not contain full instructions for Flash Boards FL1 or FL2 of a Motif XF... Therefore don't expect to be able to extract this important catalog information with a MOXF.

Here's what is missing and very much in play. Hope it helps (a bit) with your quest to understand all of this...(sorry about the length).

The Motif XF can place custom audio data either in volatile SDRAM(USR), the semi-permanent FL1 or FL2 boards.
You direct the data to the memory location, as you desire... By Loading data.
When you make a File, the different data is archived in Folders within that file. The TYPE of file will determine what gets saved and this TYPE setting also affects what you can read from it.
The MOXF can extract certain specific types of data from Motif XF files.

An important fact that you are not considering is that when new custom data is added, it can be instructed to overwrite the volatile SDRAM (loading an ALL file will replace SDRAM data), but it cannot, and can never overwrite the Flash installed data. New audio data can be used for normal and drum kit Voices, but audio data can also be assigned to special Sample Voices that are synchronized to play along with the sequencer. (SEQ samples). The MOXF is only Flash and data can never be overwritten... You can DELETE it or you can FORMAT the board.

Data installed to FLASH BOARD is dynamic... This means it does not necessarily, load the SDRAM's 128 Waveforms to the Flash Boards first 128 locations. When new Waveforms are created they automatically occupy the lowest numbered empty location.

Example 150 Waveforms in Motif XF
If you have 50 Waveforms in the Motif XF SDRAM and 100 Waveforms on FL1 Board.... When the MOXF is asked to read and "translate" data from this file, it will install what meets your request by creating a brand new catalog using the lowest numbered empty locations on its single FLASH BOARD. It will dutifully load only what you specify by file TYPE.

Say in the Motif XF SDRAM of the 50 Waveforms: 40 are used in normal User Voices, 5 are used as Drum kit waveforms, and 5 are samples create in the sequencer (1 lead vocal, 3 separate backing vocals, and an electric guitar solo) recorded to a Song sequence... On the FL1 Flash Board you have 100 Waveforms all creating normal User Voices.
150 Waveforms total.

You take the Motif XF ALL data file (.n2.X3A) and you load '1Bank Voice' to your MOXF - only the data that makes Voices in the one Bank you selected will be translated and loaded to lowest numbered empty locations of the MOXF. This might be all or some of the data. You could opt to load just a single Voice! (But certainly they would not load in in the same order).

Say the one bank you select is [User Drum] Bank, you load via '1Bank Voice' then the 5 drum Waveforms will be extracted from the ALL data file and installed to MOXF Flash Board's lowest numbered empty locations. As the Waveforms find their new numbered locations, the Drum Kits you are loading are dynamically updated. Every key that used one of these 5 Waveforms is rewritten to reflect exactly where it is placed (this is why it is imperative to create a new backup file) If in the Motif XF the Drum Kit Voice was pointing to SDRAM (USR) Wave Numbers 041, 042, 043, 044 and 045, the Drum Kit Voice of the MOXF (if newly formatted Flash board) will have installed these 5 Waveforms starting with location 0001-0005.

If next I decide to load 12 Voices from various normal USER locations, the MOXF will gather the Waveforms necessary for each, and dynamically write them to the lowest numbered location available on the MOXF Flash Board... and it will rewrite each Voice to reflect the new Waveform Number location. If it encounters a situation where a Voice from this same file, previously loaded, it will simply repoint the data (not duplicate the waveform) - so it looks to the catalog, compares the names of Waveforms already installed from this file to its new catalog. If the name matches and is from this same document (file) it simply updates the pointers.

Your conclusion that it is a simple 1:1 recreation of the Waveform list as found in the Motif XF is not how this works, at all. Not even close. Only data that is being used in Voices (normal or drum) or in SEQ Samples are translated by the MOXF as it reads the catalog of the Motif XF and extracts and reassembles the data to the MOXF. And they are only extracted and installed by what you choose to load.

Now again, as long as the data is coming from a full (with Samples) Motif XF ALL data archive file you will be able to properly enter and extract and correctly translate Motif XF data to the MOXF.

If some data is in SDRAM and some data is on FL1 in a Motif XF when you create the file, the data automatically finds the lowest numbered empty location in the MOXF. Even if everything was Normal User Voices and should be, as you say "1:1", I contend it would, the Waveform list would match BUT (and it is a big ol' but) you must be extracting the data from a full ALL file (with samples).. (.n2.X3A). In other words, it is not impossible to have a matching Waveform List in both products...it would be limiting, however... But thankfully, the file management system is more flexible than that. You can pick and choose just the data you want from any Motif XF library file as long as it contains the full catalog.

Using my 150 Waveform example... It still must rewrite and catalog each Voice in the MOXF. Repointing Voices that said USR Wave Number 001 in the Motif XF to their new location 0001 of the MOXF. Repointing Voices that said FL1 Wave Number 0001 in the Motif XF to the new location 0051 of the MOXF's Flash Board.

Remember each Waveform can house as few as 1 and as many as 256 sampled keybanks
It might just be "rocket science" 🙂 unseen by you are the 8,192 individual sample locations being managed on the Flash Noard... Again dynamically.

Thank you very much, Bad Mister, I really appreciate your extensive explanations and the technical "inside scoop" related to the XF / MOXF data mapping. I'm fully aware that considering all possible aspects when loading to flashboards can make the situation much more complicated sometimes, depending on what is already located on the flash, what is contained on the USB stick and what you choose to load from there.

That's why I was referring to the simplest case only, which also applies to such a library mentioned in the title of this thread, the "CS80" library in conjunction with the CS80 "Bonus" example. In this case the waveforms/samples from the first (main) library file were already saved in a single flash slot only (waveform allocation from 1 to 61 without any interleave), and the second (bonus) file is just containing the "delta" content without including redundant information again, which requires that the main content was already loaded into the synthesizer like an "ALL" file (and in case of MOXF usage to a freshly formatted flashboard). This was the starting point assumed by me in any way, as otherwise the bonus file would never point to the correct waveforms in the end, no matter how you try to tweak it. I always considered both files as a single instance in a way, which have to fit together and must be complementary to each other, of course all applied to a "virgin" MOXF and flashboard. After all, my goal was only to create a related X6A file on the MOXF for later use.

In such commercial XF libraries like in this example, you will mostly find waveforms/samples which are stored to one single flash slot only. That's why I thought it would be a good idea to have a functionality (or option) to "map" the waveforms accordingly while loading into the MOXF. For that described case my "offset" method would have worked out perfectly - but only for that special case and the related conditions I just mentioned above. Of course, I didn't figure out yet how the MOXF would behave in case you'd try to load waveform content from an XF file which was saved to the "FL2" memory space. I guess, in this case all waveform references would get lost at all in the MOXF. Fortunately I'm not aware of any commercial XF library using such a feature, as you cannot expect that every XF user owns two flashboards. But now I also understand your point that the development of such a feature wouldn't pay off for Yamaha, as all possible MOXF loading scenarios would have to be considered as well; that's why I have to accept the current workaround I mentioned, using Melas Tools. At least I made it work using that method.

Nevertheless, I'm a little bit surprised about what you explained earlier, namely that XF voices (e.g. as part of the discussed "CS80 Bonus" library) might not be 100% compatible with the MOXF, because I always thought that's a fact. Of course, I read the enclosed PDF instructions which stated that there are "also 2 bonus banks of voices available that use both the CS80 waves and the new distortions and reverbs in OS version 1.5", but I didn't pay much attention to that statement, as I didn't know that this will directly impact the MOXF usability. If there are really effects present in the Motif XF which have never made it to the MOXF - not even into the latest F/W - then how to deal with this now? I always thought MOXF F/W V1.10 was already addressing such voice incompatibility issues, e.g. in conjunction with the Yamaha CP1 / Gospelmusician / K-Sounds libraries..., so it was completely new to me that now there might be audible differences whenever XF voices are played on a MOXF.

How can I ever find out if a certain commercial library is really working "as intended by the programmer" in my MOXF? I guess not all vendors have such enclosed PDFs where they describe in detail if some of their XF voices are using V1.5 effects or not, that scares me now a little bit. After all, even though I'm a MOXF owner, I purchased the two available "Yamaha Anniversary Packs" (10th and 40th) dedicated to the XF, which have been separately sold here in Europe by certain Yamaha authorized music stores. Now I wonder which XF content could possibly include sounds which are not fully compatible with the MOXF. After all, above you pointed out that the "CS80 Bonus" library is one of those cases, even though already the very first "MOXF catalog" propagated right from the start that "The MOXF is fully compatible with the sound content developed for the MOTIF XF and MOX." Per definition, this would also include XF voice data, consisting of any included effects as part of their data set. Now, loading any XF voices using these new XF effects would also make them "sound" differently in the MOXF, right?. In fact, the whole compatibility statement doesn't seem to be true anymore, or did I get something wrong?

I haven't had any errors on the MOXF screen so far (except the one I was explaining at the beginning of this thread) whenever loading XF files into the MOXF. So, how can I now spot these affected voices in a MOXF - if they exist in a library? If there is also no indication/error during the loading process, where are any XF V1.5 effect parameters mapped to in the MOXF, instead? Will there be a MOXF F/W update in the near future where these missing effects will be also introduced, or is this not possible due to the limited MOXF OS capabilities compared to the XF's ones? Maybe you have more information on that topic.

Many thanks (especially for working on your previous replies - and all that on the weekend) and sorry for so many new questions, but now I'm a little bit worried.

But now I also understand your point that the development of such a feature wouldn't pay off for Yamaha, as all possible MOXF loading scenarios would have to be considered as well; that's why I have to accept the current workaround I mentioned, using Melas Tools. At least I made it work using that method.

Again, you have only "fooled yourself" into thinking you have worked around the issue... You have not in the case of this particular library. Many of your Voices are addressing no Effects, but I've explained that already.

How can I ever find out if a certain commercial library is really working "as intended by the programmer" in my MOXF?

If the Library was create prior to the newly added features of Motif XF 1.50 it will certainly be 100% compatible... Authors of sound libraries ARE aware and would immediately know (even if you didnt) that the new additions to the Motif XF would mean a different library would need to be made for the MOXF. Third party programmers tend to stay up with this type of specification changes, even if you don't. Again, perhaps, if Athan has time or perhaps you can go ahead and correct/finish the programming for MOXF. But if you compare your Melas converted CSBONUS with the original - you'll hear the difference. How can you find out if a commercial library is really working as intended... If it is a .X6A file you can be SURE the author intended and verified the data on a MOXF.

Question: we prepared a special version of the 40th Anniversary Premium Content Pack for the MOXF... Why do you have the Motif XF version and not the data prepared for MOXF owners?
Also to your fear: specifications are always subject to change without notice. The newly added High Definition Reverbs and Distortions are an improvement that separate the Motif XFs from the MOXF... Motif XF owners are extremely pleased by these new features and the CSBONUS Library was/is the first library utilizing any of these new algorithms. If you see a library designed for the Motif XF, that the author claims is compatible with the MOXF, then they will provide the "with samples" archive files you need to load it. By providing the CSBonus Library "without sample" Athan ensured that this file was intended for Motif XF owners only. The CSBONUS was not included (afaik) with the MOXF version of 40th anniversary PCP. Had he saved the file "with samples" it would load to the MOXF... But as a programmer he would immediately notice the lack of reverb on some of the Voices as he programmed them... So the file is "not as intended"... Nor was it intended for MOXF owners who don't have the new effect types. There is no error message, the new Reverb Effects simply come up unassigned in the MOXF... The Voice data is compatible... (while Reverb is an effect, the room it is in is NOT technically speaking apart of the instrument, an acoustic piano as an instrument is separate from the room you put it in) as a MOXF user you can simply select an available Reverb... So it does not prevent you from using this Voice data. The Voice data is compatible.

Yikes... Enough already. 🙂 you are not hearing the CSBONUS as intended, trust me - if you are at all uncertain about a Library email the author... Or ask about it here. We'll give you the straight story as we understand it.

Thanks again for your quick reply.

Also to your fear: specifications are always subject to change without notice.

I understand that statement from a company's point of view. However, such impacts and changes should be at least communicated in a better way when they become a solid fact for a certain product. Just for example, the latest brochure "MOXF Catalog" on the official Yamaha download page still contains the same slogan I already quoted in my related question above to the "full sound compatibility" - you can check it out yourself. So please let's stay fair.

Question: we prepared a special version of the 40th Anniversary Premium Content Pack for the MOXF... Why do you have the Motif XF version and not the data prepared for MOXF owners?

The answer is very simple. This pack (which I noticed lately - but unfortunately long after I had already purchased the XF version) seems to be exclusively sold in the USA, and appears to be bundled in conjunction with a new MOXF order only. Should I purchase another one just because of the desired CS80 library? No, certainly not (I have even re-sold the included flashboard because I don't need it)! There have already been many user requests to make this CS80 library available for separate purchase, but it did never happen.

Yikes... Enough already. 🙂 you are not hearing the CSBONUS as intended, trust me

I think this time you got me wrong. I already accepted the fact that this particular library will sound differently in my MOXF after you have told me. Would I have asked all the questions above to the new compatibility situation if I wouldn't have believed you? Of course, I cannot hear the library now as "intended by the programmer", but what I just wanted to point out again when I mentioned the "Melas Waveform Editor" is that the described "translation" would work out for such libraries created the same way (main+delta). However, I will have to take more attention in the future before purchasing any libraries, as this is for sure a "lessons learned" for me.

Or ask about it here. We'll give you the straight story as we understand it.

The problem seems to be that any voice compatibility between XF and MOXF is not always clear. For example, let's take the "Alan Parsons Imperial Grand Piano for Motif XS/XF". At the "YamahaMusicSoft.com/de" store this library is noted to be not compatible with the MOXF, while on the Sonic Reality store "Esoundz.com" (David Kerzner is in fact the creator of this library together with A.P.) the same product is listed to be compatible with MOXF; the title already reflects that fact: "...for Motif XS/XF/MOXF". This is exactly the same product, so whom to trust now? In fact, it's not always a "black or white" decision for the MOXF owner to determine if such a librariy will be fully usable on a MOXF (if it contains new XF effects), after all you have to rely on any available inputs and specs - and even that can sometimes be very contradictory as you see in this case. Could anybody please enlighten the situation here for this product? Thanks! The same might apply to the new "American D Concert Grand Piano". While on Yamaha's store it is shown not to be compatible with MOXF, in this forum it has been announced to be compatible in a separate thread; the vendor's homepage doesn't even mention that Motif version - there is only a pure S/W-based instrument listed. If this situation isn't really confusing for MOXF users, then tell me how I should call it...

Thanks again, I promise - after having clarity on the two mentioned libraries - I will finally shut up my mouth here (at least in this already very long thread) and I will start whining alone in my personal room without bothering others anymore ;).