I am a Roland Fantom G user and know Motifs only from friends and from shop checking. There is a lot I like about my present workstation, but I am extremely dissatisfied with the missing Roland support for constant OS development (virtually absent since two years, but necessary for curing essential weaknesses of the present OS).

Now I am gathering information about the XF, because a one or 2 Gb Flash user soundset seems very attractive to me, if the rest of the workstation has a high usability.

My question concerns
a) the new soundmaterial on board (I know many standard preset sounds of the XF from listening).
- I heard the XF has new promising D6 Clav sounds, but does it also contain new Rhodes and Wurly material?
Demo sounds of available EPianos are essential for me.

b) the availability of additional sampling material
a first overview in the motifator shop showed me some preset libraries, but I didn’t see sample libraries at first glance.
Where do I find them with demos, and/or are new libraries planned?

c) the quality of the sampling engine, especially:
- does it allow at least 8 velocity layer sampling?
- does it allow release samples?

Excuse me if I ask questions which experienced Motf users don’t have to ask, but I need plenty of information before considering a workstation change.

I’m not sure by what you mean by release samples, perhaps sample triggering? If you’re looking for great rhodes then you should be fine. I have used many Fenders and the Motif’s rhodes do a great job, especially with the VCM effects that can be used on them. You can find more info on the upcoming XF on the home page of this site under the “XF” tab, articles.

The EPianos seemed farily unchanged to me between the classic and ES. However, the XS EPianos were completely reworked to take advantage of the upgrades to the 8 element sound engine that was introduced in the XS. Since the XF is still using the 8 element engine, and since Yamaha hasn’t announced any new EPiano waveforms or effects, I expect that, in terms of EPianos, the XF will sound like the XS, which is really good. EPianos is an instrument category that the new Motifs handle very well! Most of the XS EPianos have three or four velocity layers, plus a tine release sample. The modeled effects, particularly the phaser and compressor, can really pump up the sound of them.

The digital pianos (FM DX-like voices) didn’t seem to be changed too much in the XS, though they were tweaked. There are certainly a good many DX waveforms in the XS that you can blend to come up with most DX7 piano/comp voices that you’d want.

Clavs in the XS had an added key release sound (muting clav strings on key up), thanks to the new 8 element engine. That added a fair bit of realism. The XF’s better clav waveforms will prob help that out quite a bit.

For most of the EPiano voices, you prob should listen to XS demos. Those are probably very close to what that category will sound like on the XF.

About your velocity layer question, it isn’t cut and dry. Yes, you can make voices with 8 velocity layers. However, depending on how you create the voice, you can make the voice with even more velocity layers than that. You get 8 elements to use for building a voice. Each element plays a single waveform. Each waveform can contain up to 256 samples, called keybanks, that can be set to cover a specific key or velocity range. Unlike the Roland engine, you don’t need to use separate voice elements to define velocity ranges.

For example, take a voice of a pitched drum. Suppose that you sampled it sparingly (just once per octave, for four octaves), and you sampled it at 4 velocities. That gives you at total of 16 samples to map.

On the Roland engine, you can map samples to key ranges inside one of their waveforms. Their engine provides four elements per voice, and each voice plays one waveform. So, you need to use all four elements for our pitched drum example, and set their velocity ranges so that each of the elements triggers with-in each of the 4 velocity ranges to cover.

On the Motif, you could stuff all 16 samples in to a single waveform, mapping them to key and velocity ranges, and then play that waveform with just 1 of the 8 available voice elements. That would leave you another 7 elements to use for layering other tones, if necessary.

Suppose you had a piano, and you sampled each key. That’s 88 samples. Two layers of samples would then be 176 samples total. Since that is less than the 256 samples per waveform limit, you could put two full layers of the piano sample in to a single waveform. Since you could have 8 elements, and therefore 8 waveforms, active in a single voice, that gives you the potential for a 16 velocity layer piano, sampled on every key. Of course, that would require a huge amount of memory, and you wouldn’t have any elements left over to play key release sounds, to simulate the resonance on the high keys, etc. If lots of layers are your thing, though, you can get them.

One last idea. You could use 128 of the 256 available samples in a waveform to stack on a single key. That way, one of the 8 synth elements, triggering a single waveform, could give you all possible 128 levels of MIDI velocity, but just on a single key. This is possible all day with waveforms that get assigned to keys in drum kits. You could have a drum kit full of keys with many sample layers.

At some point, though, you’ve got to call it quits on layers. Four might sound a good bit better than 2, but 6 won’t sound as big of an improvement. Certainly, when you get to 12 or 16, then you’re getting less and less help from layers, while eating drastically more memory. At that point, it might be better to focus on synthesis tricks inside the elements, subtle processing with effects, etc to improve the sound.

defrigge - 11 August 2010 04:02 PM

I am a Roland Fantom G user
I am extremely dissatisfied with the missing Roland support for constant OS development (virtually absent since two years, but necessary for curing essential weaknesses of the present OS).

Defrigge,

Customer Support and OS update support are great for the Motif series.  Yamaha is planning on releasing a substantial OS update for the XS (the keyboard just prior to the XF of course) that will give it as many of the new features of the XF as are possible.  Yamaha has stated in one of the recent posts about the XF that they want users to feel confident in buying the new XF that it will be supported for quite a while (like the other keyboards they have made).  This is a very strong case for owning a Yamaha.  I’m’ sure Yamaha is looking at your post and very happy indeed that they have decided to continue support for the XS.  It does indeed make future customers more sure about spending $2000-$3000.

As far as sample libraries, go back to the shop and look again.  Some of the libraries are just preset libraries but many of them contain new sample material.  You can differentiate between the two because the ones with sample libraries will tell you that you need to purchase sample ram in order to use them.  Many of them are quite good.  I own most or all of Dave Polich’s stuff (DCP productions I believe).  Very top quality stuff including some nice electronic/vintage keys stuff.  And the ability to load 2 GB of sample libraries into instant access flash memory...forget-about-it!  Go ahead and get the XF...you won’t be disappointed.  Is Yamaha perfect?  No, but in my opinion they are really on top of their game and the best of the big 3 for workstations hands down.  I had to decide a couple of years ago between the Roland, Korg and Yamaha workstations and after a lot of research came to the conclusion that Yamaha sounded light-years better than the other two.  And the onboard effects (VCM, Rev-X and such...wow!) Not that the other two are bad...I like things about them as well.  But with a budget to only afford one...Yamaha. 

Hope that helps.

God bless

@bsmart:
thanks a lot for your elaborate response!

I asked for layers and release samples (yes, samples triggered when releasing a key), because good Epiano samples need at least 8 velocity layers (16 don’t hurt either) to render the dynamics of the original instruments without audible steps (a Rhodes has even bigger dynamics than a grand piano).

So if a device offers 2Gb of Flash memory, it is good to know the space can be used by quality sampling material, which could close the gap towards the software solutions (Kontakt and the like) at least a bit.

@dimitry
the Fantom G has some very strong sides. You named the big display, which makes your life a lot easier in editing or working with the sequencer. The live mode also works fine, with seamless effect switching between complete live sets (each with up to eight single patch sounds). And yes, the user interface could(!) have become the most user friendly by far. You have a mouse to work on the big screen, but the mouse has only been implemented in the OS in part, and you find yourself having to use buttons at some places instead.

Same with the sequencer: principally it is much more powerful than any I know on a workstation: you named the specs. But the seuqencer is built on a silly phrase concept which is hard to use for linear sequencing. For example, if you simply want to copy a phrase and edit is as new variation, you can’t, because the sequencer does not contain the midi notes, but only references to phrases. So you have to double the single phrase and only then can edit the copied version, because otherwise you would change the original reference phrase.

Phrases are written sequentially into available free slots, just like samples, and are numbered instead of named within the directory root on the usb stick (which gives you a lot to guess in manual project handling, see next paragraph for projects). In the screen directory lists they have names with too few characters for big directories: you are abbreviating your way through that. Now imagine working with all the phrases for different tracks and different bars/parts without even having the material of one song in one folder (which would still be quite a mess). Instead all phrases (including drum patterns) are scattered all over the place in one big directory.

The project idea would be nice, but the Fantom G loads ALL data belonging to a project when loading it (all samples, all phrases, all songs ...) and streams nothing when needed. That leads either to longer prject loading times or to allocating your material to different projects. Now once more the bad news: you can’t just transfer all neccessary data from one project to the other, because the Fantom G is uncapable of any form of relative data management: you can only work around this in part by copying for example phrases or samples into the exact same absolute slot numbers one by one manually , only then enabling the song or isample instrument definition to refer to them in a proper way. Same with sampling material. The whole Fantom G file system only works within narrow project boundaries and is more like a big chaos, working only as long the device writes all successively into one project and then can use the working pointers.

I could make a long list of these kinds of flaws, but I won’t, because I don’t intend to bash the G with it’s great basic features, especially those for live play and for editing sounds.

The really incredible thing from my view is that the Roland management never found it worth to fix the many G OS shortcomings and design flaws within two and a half years, or at least to begin somewhere in a substantial way. Every musician on each modern workstation has to live with some compromises, and that is normal from my view. But Roland’s total carelessness for a steady and substantial G OS development is outstanding in it’s negative stubbornness as well, and completely intolerable, with far reaching consequences for my future buying decisions, from my view as a user.

To me neither Roland nor Korg nor Yamaha are flawless, and all workstations let us musicians rather wonder day and night, why they are bulit around totally outdated stone age technology, compared to modern PC hardware, for example with hard to get DIMMs with bad, slow specs from ten years ago. They all seem a bit like DOS machines in windows 7 age, or like Pentium 3s in core i7 age.

So I will never become a fanboy of any one manufacturer, and don’t want to fight silly fights, but I want a workstation with a very good soundset and high usability, connected to a manufacturer who cares, and allows and gives user feedback.

defrigge - 11 August 2010 07:29 PM

t the seuqencer is built on a silly phrase concept ... you have to double the single phrase and only then can edit the copied version

imagine working with all the phrases for different tracks and different bars/parts without even having the material of one song in one folder

you can’t just transfer all neccessary data from one project to the other, because the Fantom G is uncapable of any form of relative data management

Thank you very much for your long and detailed answer.

I once owned a Fantom G, which I sold to get a Motif XS and Korg M3. I miss the Roland sound character, but MotifXS/Korg are kings in their categories too.

I must give Yamaha 1 thumb up for good support. While certain features I wanted (eg: grid recording) have not yet been implemented in XS after 3 years since its launch, I feel overall support is there. If these little workflow issues were solved, I will easily give Yamaha a 2 thumbs up.

Soundwise, you will miss Roland - Yamaha has a distinctive Yamaha sound - it won’t sound like a Roland. I am considering getting a Roland again for this very reason. But yeah, Roland US is not as strong as Yamaha US. There is a guy who runs Roland Clan - he seems pretty good at what he does and keeping in touch with Roland Corporation in Japan.

To answer your specific questions.

The XF has 8 elements per Voice so it does allow 8 velocity levels in a Voice.
It also has note off capability so you can have samples triggered at note off.

Regarding sample libraries on top of the libraries already available in the shop, there are a large number of new libraries in the works for XF/XS from Sonjc Reality, Garritan, Digital Sound Factory, DCP productions , K Sounds and Easy Sounds.

nchrist - 12 August 2010 04:31 AM

Yamaha has a distinctive Yamaha sound - it won’t sound like a Roland.

IMHO Roland and Korg sounds are vastly overprocessed and the patches often heavily rely on the effect processor. If you want this kind of sound on the Motif XS, just tweak the effects to sound over-the-top, that’s all. On the other hand, Motif sounds are programmed rely on the effects to a far less degree, which IMHO is far more appropriate in a studio.

Thanks to all of you for the informative replies.

One thing is obvious already to me: support and reactions from Yamaha staff seem to be light years more direct and better than I have experience it from Roland, who don’t even maintain any forum for their synth flagships. There is a private forum, and you get useful technical hints there, but else that seems to be rather a pure fanboy site, not tolerating or supporting critical user feedback, and they don’t seem to like that kind of debate, even if it comes along calm, unoffensive and thoughtful.

Today I spent some time getting a closer impression of that basic “Yamaha sound”, and got a very ambivalent picture. My evaluation is, of course, purely subjective, and from the view of a curious outsider:

- I heard some really good natural sounds (stunning nylon guitars etc.)
- I hardly found any demos really demoing certain sounds. Most demos show off whole song arrangements, and that does not at all help to judge the quality of patches
- many sounds leave - for my ears - a very strong “middle-of-the-road” pop sound where I miss character. I caught myself quite often getting this “jingle and elevator music feeling”
- on the other hand, there seem to be some treasures buried within the mass amount of sounds which DO or at least do have the potenial to sound more distinct and edgy.
- just like on the Fantom, I am quite irritated about the amount of - judging from my ears - as I call them, outdated 90s looped (normally much too) short, small samples. Like on the Fantom (and the M3, and most of all the Kurzweils, to make the picture complete), a lot of recycled sound material of that kind seems to have flooded the sound ROM for the XS series as well.

To be frank, I have completely other sound expectations towards a 2010 device (concerns all major three), and I could easily live without two thirds of ROM banks filled like that, in exchange for a tenth in numbers, but well chosen, well editable (variations) high quality sounds. People today are used to hear high quality Kontakt samples, and while I understand that this can’t just be carried to hardware workstations, it is still this kind of sound which is definitive today. The beig manufacturers seem to grasp that and react to that fact only hesitating and slowly.

So I end this day with a mixed result from my view:
- as far as I can see Yamaha support seems to be much better that what I met from Roland
- concerning sounds all three big (and Kurzweil even more) still seem to be based on much outdated material in their ROM (at Roland it’s the then not bad, but today much less convincing JV/XV series stuff, and the other manufacturers seem to have ancestor material of comparable age and limited quality)

I will visit a shop and take the time to listen to the XS onboard EPiano and analog synth patches, as to many other areas, carefully, and then once more when the XF appears. Maybe I see an additional chance to get firstclass Scarbee sounds and the like resampled in the XF flash memory, and the inclusion of release samples is good news to me. Perhaps it is possible to convert for example a Scarbee Rhodes with 16 velocity layers to an XF Flash memory preset with 7 layers plus release samples, or whatever.

Once again, thanks again for the helpful feedback, and I will ask again when I got more to ask.

P.S:
The “nice only because overprocessed” argument is funny, because I hear it in each user forum towards other manufacturers (never towards the own gear) ;-)))
I rather think that all this outdated wav material from all manuaturers needs overprocessing, because the naked loops sound close to ridiculous, just not from our time.
That’s one reason I am curious for the doubled XF ROM and hope it is not just used 80% for one or two piano*s samples.

DmitryKo - 12 August 2010 03:37 PM

nchrist - 12 August 2010 04:31 AM

Yamaha has a distinctive Yamaha sound - it won’t sound like a Roland.

IMHO Roland and Korg sounds are vastly overprocessed and the patches often heavily rely on the effect processor. If you want this kind of sound on the Motif XS, just tweak the effects to sound over-the-top, that’s all. On the other hand, Motif sounds are programmed rely on the effects to a far less degree, which IMHO is far more appropriate in a studio.

After playing Roland boards for years.. Its not really “overprocessed” with effects… I think it is more of a finished and mastered sound… which still sucks because all of the character is kinda taken out and you cant get the type of character you are looking for without undoing all the mastering that is done for you…

I will say ... if you are recording straight into a daw and then mixing and mastering with very little knowledge.. then the Roland gear will make your content sound more professionally mastered… from a mainstream perspective…

if you are playing live… it feels like you are constantly trying to cut through the mix… very frustrating… I ended up tweaking almost every patch so that I could cut through the mix…

in regards to programming patches… hands down the Roland is easier...the GUI is way more straightforward and somewhat simpler… but it also seems like the actual architecture is simpler.  I have yet to wrap my mind around Yamaha’s sample based architecture and their terminology is a bit confusing at times...keybanks and waveforms are backwards or something… its weird.. and honestly havn’t gotten into too deeply yet… still hacking around in voice design with the rom waveforms… the beast is definitely deep…

Anyways… when the new OS comes out… I will probably actually buckle down and learn the thing… I suppose I will keep it since I am getting my beloved tap tempo back :)

DmitryKo - 12 August 2010 03:37 PM

nchrist - 12 August 2010 04:31 AM

Yamaha has a distinctive Yamaha sound - it won’t sound like a Roland.

IMHO Roland and Korg sounds are vastly overprocessed and the patches often heavily rely on the effect processor. If you want this kind of sound on the Motif XS, just tweak the effects to sound over-the-top, that’s all. On the other hand, Motif sounds are programmed rely on the effects to a far less degree, which IMHO is far more appropriate in a studio.

also… technically I would rather see a basic sound that is handled through effects.. this gives you a much broader palette to play with… I don’t want a voice that doesn’t need effects because then I can’t dial back the character of “phasing” “slicing” or other wave effecting stuff… I like to turn the effects off and hear what they were doing… it challenges you to be more creative with simple sounds… IMHO its what makes the Roland patches shine so much. they are very well designed…