Looking for anyone using Johns software for the MX and any comments.  I have the Motif bundle and it is one of the best buys I’ve made.

With the significant differences between the two boards, however, I’m wondering which elements of the bundle you’re finding most helpful.

Gurus are invited to chime in if you’ve tried this software.  I did reach out toJohn but I like real time feedback as well.  Thanks

The Melas Tools are, as all his programs are, very cool. The VOICE EDITOR actually allows you to get at some of the parameters you cannot get to from the front panel or via the FREE “VycroMX Performance Editor”.

You can open VOICE Libraries from the Motif XS, the Motif Rack XS, S90XS/S70XS, Motif XF, MOX6/MOX8 and of course, the MX.

It’s a no-brainer!!!

I am a beginner (in Motif and synth) and I have a question related to Melas tools:
I installed the demo version for MX and imported a free XS/MOX library containing 384 voices. But I can only view 128 voices… I suppose the 128 voices of the first bank. How can i view and work with the other two banks ?
I know that the MX series has only one user bank, but how to import XS/MOX sound library containing several banks if only one bank is usable ? or how to import a bank as the first bank ?
I hope a XS voice editor is not required to manage the source library, for example to save each bank in a specific file.

And how Melas voice editor manages the case where some samples/elements of a voice are not present in the MX ?

Melas Tools are quite expensive (relatively to the MX price, 10% of the price), so I prefer to be sure. Thanks.

I agree that this kind of tools should be provided for free from Yamaha, as other competitors do (I have written this in other posts). Anyway, with Melas tools you can do things that are impossible from the onboard menu:
1. Full Voice editing with lots of parameters and access to elements level
2. DrumKit editor
3. Performance editor (but there is also the free VycroMX Performance Editor). Maybe Melas editor is more complete.
4. Full Performance/Voice/DrumKit librarian
5. Import and upload of other Motif libraries
6. Still, no VST support, only standalone apps.

Have in mind that the demo version has a lot of limitations, but I don’t think that there is a limit in the number of voices to open in the librarian. The limitation to 128 is in Performances, User Voices, User DrumKits (you cannot store in MX more than 128 Performances, 128 User Voices and 8 User DrumKits). But I think that you can open all the compatible files and in worst case test and load one voice per time. Since I don’t remember this part exactly, can you provide me the link to download the 384 voice library so I can test it?

Regarding the missing elements (waveforms) parts, I think that that the tool leaves those missing wavesforms empty.

Here the link: https://dl.dropbox.com/u/73668948/XF/Music_and_artman/PSC SongWriters Collection MOX.zip

J. Melas has sent me some answers:
- unlimited number of voices in voice library… but I can see only 128 voices even in voice library… strange ???
- Melas tool uses a mapping table to replace missing waveforms by the closest ones in MX set… but results are often terrible… unusable… again strange !

I download the file and I will let you know about the results.
For the replacement of missing waveforms, I am sure that I have seen empty elements, but at that time I was using the demo version. I will recheck with the full version.

I verified that Melas’ librarian loads only 128 voices from the PSC SongWriters Collection MOX library, which includes 384 voices. Maybe, this is relevant with the hardware limitation of MX to store internally only 128 user voices, but the tool could offer the user the option to load another bank from the library file.

Also, I noticed that in some cases, the loaded voice has fewer active elements that the description in the included pdf document of the library, having also some other elements disabled but I can not tell for sure if the editor makes a guess for the missing waveforms. The situation is even worse in the case of MOTIF XS Dance Xpanded library, where there are cases that the opened voices have no active elements at all and produce no sound.

By the way, the specific PSC library is not very interesting to me. I have found much better voices in the MOX Electro library, in the MOTIF RACK XS Pop&Dance;library and in the MX Dance Voices library. I have opened also the converted ES_MO libraries (all of them downloaded for free but I don’t remember how).

You mean that you successfully imported MOX Electro library and MOTIF RACK XS Pop&Dance;library in the MX user voices using Melas Tools ?
Because with the both free libraries i have found (from moessieurs site) the result shows only that the importation process doe not work as expected. The most important problem is how the voices from XS/MOX sound in the MX : there is clearly a big problem in 70% of voices… and even in 30% which seems to work, many of them sound the same.
The first waveform of the MX set (number 000) appears in many bad voices : a default setup when the importation process can’t map voices for different reason ???
In some voices some elements are clearly missing or disabled, in other voices there is a (too ?) long list of random waveforms (random because not very closed together to give audible sounds).
Many voices are simply unplayable because each note plays a different sounds… or alternate same sounds… or the sound is saturated (bad parameters).
So yes there is an importation function… but it does not mean that importation is possible.

If you have some good feedbacks… I am curious !
And if you have some other free sound libraries from XS/MOX (that you known they sound good) I can test them and record some sounds to compare.
For now I am checking the waveform that have disappeared from the XS: for the moment the piano set is simplified a lot, in the guitar section it’s a carnage, in the strings and brass section it is the same as for piano… in fact many articulations have disappeared. But it does not explained the importation problems because missing waveforms are certainly small modification of reference tones.
I continue to investigate…

PS: I confirm that the problem for importing more than 128 voices is really a bug… the next version will fix it.

Yes, I managed to open the whole libraries (128 voices) that I have mentioned above with Melas librarian and then I randomly used the voice editor in order to hear the sound of some voices. I also stored some of them in the MX user bank (but not the whole library).

From what I can tell, I found some very nice voices in the MOX Electro library, in the MOTIF RACK XS Pop&Dance;library and in the MX Dance Voices library, played very well (in some cases in a very loud level, almost distorted) but without the problems you mentioned. The only thing that I have noticed is that in some cases there seem to be some disabled elements but since I don’t have the documentation of the libraries I cant’t tell for sure. Only in the case of the PSC library that there is a description document I have verified that in many cases there are is one or two fewer elements active. Anyway, this library does not sound interesting at all, maybe due to bad matching of waveforms. Only in the case of MOTIF XS Dance Xpanded library I have some voices that play no sound at all.

Anyway, this is an extra feature of Melas tools that we couldn’t have it otherwise. Having in mind that MX has 166MB of internal waveform memory while XS and MOX have 355MB, I think the tools do a pretty good job.

If I have time (not very likely), I may record some of the imported sounds and upload it in my utube channel (https://www.youtube.com/channel/UC0zue3cGy6ph3wZErfycnTg)

John Melas has explained me the problems:
- Arpegio is missing in some voices for the MX, and so on these voices become useless… and strange.
- no high pass filter in MX
- only one insert effect by voices for the MX (sure this makes the difference)

But I found a bug and John found also a bug in the MX firmware by testing the Xspand library: in some cases voice do not produce sound at all… need to select manually the insert effect to listen the sound !!!

And I will certainly buy its tool.

John knows better but I think Arpeggios is a Performance feature in MX. Anyway it seems there are plenty of parameters that might be missing in MX and so the voices can’t be loaded properly. By the way, I have uploaded a video in utube, showing the import of voices from MOX Electro library in MX61. I think the sound quality is pretty decent.
(http://youtu.be/KFdSCilaQb8)

To close the discussion (before new experimentations with not free banks), is there someone who has compared MOX/XS voices in some library (with no sample) with MX conversion ?
I can note that no MX bank exists yet… two answers: 1) the automatic conversion is immediate and give very good results 2) the adaptation requires a (too) long and manual process for obtaining approximative results.

KostasT - 05 September 2013 05:05 AM

MOTIF XS Dance Xpanded library I have some voices that play no sound at all

If you are using the version that includes new samples, those samples do not exist in the MX so you would get empty elements, in some cases no active elements at all. There is an alternate version of most Easysounds libraries that use only factory waveforms, try that version instead.

The samples are certainly not the correct explanation in this case, it is related to a bug that John Melas has discovered recently in the MX Firmware:
In a situation (not clearly defined for the moment), some voices containing an insert effect transfered using John tools do not produce sound at all. But if the insert effect is forced using the MX interface (by browsing into the different voice menu and setup again the same insert effect), everything sound correctly… magic… but it is a firmware bug.
For sure yamaha has never expected official tools to modify voices in depth… only performance for the moment.
And perhaps other similar bugs exist in the firmware 1.0 and could explain some strange sounds when I have tested free library.

It is difficult to give conclusion without a solid comparison between MOX/MX for the same library. For sure, I will never buy a MOX for this… and i suppose that no MOX owners will buy a MX for testing purpose. And all of this to save 189Mo and some DSP power… Yamaha is always in the 90’s.

I cannnot verify castorex’s post but it seems more likely meatballfulton’s point. I tested two additional MOTIF XS libraries ("Oriental Instruments & Percussion” and “European Loyalty V2"). All of them include new samples in them (huge files). In all the cases that I hear no sound from the loaded voice, I can see in Melas editor that no element is activated. If I activate the elements myself (with any factory waveform) I can hear the sound with the loaded XS voice parameters (meaning that the voice parameters are transferred, but not the waveforms). In some cases the ARP cuts the sound (you have to disable it). In all cases that at least one element is activated, I can hear a sound. The problem seems to be more often in MOTIF XS libraries because they include new samples. I think I haven’t find this kind of problem in “MOX Sound & Groove” (electro) libray. In some cases I have noticed also that some elements are disabled but NOT with the default waveform selected (piano strech hard st). Maybe this is the case that Melas tool makes a guess for a missing internal waveform. Only John can verify all these. We must have in mind that this is an extra feauture that we cannot have it otherwise and that the tool tries to make a guess for any missing parameter. It is logical, that all common internal waveforms must be and sound identical between all MOTIF AWM2 synths. By the way, MOTIF XS “Oriental Instruments & Percussion” library includes some new Performances that are not loaded in Melas librarian.

Importing sound library with samples is for sure impossible… or could be considered only for some voices which do not use specific samples, the same for voices using arpegio and chain of effects… the MX cannot handle the parameters for these ones.
But I continue to think (and by reading your comments) that some strange behaviors exist. We find two bugs for now:
- one in the Melas librarian tool
- one in MX firmware
But there are maybe (certainly) other bugs because no so many tests have been produced.
It is important to report all strange situations to John Melas… so he will explain to us if it is normal (something we misunderstood) or it is a bug that it will correct in a next release.
For the disabled waveforms, don’t forget that the MX engine does not manages the note-off, and all waveforms related to note-off are missing (and disabled in this case).

Could you give me a link to your Dance Library (a small sample of this one with some good and worse voices, because I know this library is not free) for my own tests. I could prepare my questions for John. Thanks a lot.

Last question: Is there a way to load a library and to view all voice parameters without a MOX/XS hardware? With Melas tools the Yamaha hardware has to be connected… but it could be useful to identify voice parameters in the original context (before importation for the MX).