It's time to have more than 4 style variations

[Gunnar Jonny]

I mean, we already have features that 90% of owners don't use -however, the remaining 10% sure appreciate that they are available.

I'm aware, that for majority here, these exotic features have no importance. Some say "I don't need that", others "I can live without that" and some maybe even "no idea what's all about". For those, let me just say that nothing would change if we would have these features -but remaining 10% would appreciate having them.

+1

To me it's simple. Those who are more than satisfied with the things as they are, don't have to buy any more complex and expensive keyboard, because what they have is already more than enough. For those of us that like to see progression and new features, keep 'em coming!! 

[DerekA]

It's all about what is a reasonable compromise that suits the majority of users. That is always going to be the case for anything that is hardware-based.

The end result can be achieved using multiple style files and registrations, so I can't see it happening anytime soon.

[pieterpan]

I'm satisfied with my T4, the only thing I whised to be changed is the fact that all breaks and fill-ins are just 1 measure more measures is a big wish.

[mikf]

Not just a compromise on usability, but on cost and development time. Bogdan says that the additional style 'slots' could be unpopulated, and left for customization. But Yamaha would be scratching their heads and thinking isn't that exactly what the combination of custom user styles and registrations do, and registrations are way more powerful? They would see a possible mega cost hike here because they would know that adding extra style variations but leaving them empty would not look sensible. It would get heavily criticized. And the retro cost of changing/extending every style would be huge, maybe push the cost of a Genos 3 close to 5 figures. Would anyone pay $9000 for a Genos just to get extra style variations.
It's a fun discussion, - especially when there is not a lot more to talk about for another week or two - but I just don't see it even being considered by Yamaha for the reasons above.
Mike

[BogdanH]

Mike, I think you worry too much about how Yamaha could do this or that and keep costs reasonable -it's Yamaha's job to figure that out! On the other side, we users, are legit to express our wishes and needs. When you say $9000 for Genos, it's like you try to justify (for Yamaha) why we can't get certain features.
If we keep thinking that way, then that means, that we can't expect for next Genos to have any important new feature, because the price is expected to be similar as for current Genos when it came to market. But speaking for me, I do hope and expect, that this will not be the case. Progress makes that possible: getting something better at the same price.

Bogdan

[mikf]

Bogdan, I don't worry about this stuff at all. Nor do I care abut Yamaha. I am not a Yamaha enthusiast or supporter. I have owned many musical instruments in my lifetime and the arranger was a Yamaha first for me. I am not even an arranger enthusiast, I am a lifetime musician who found some years ago that the arranger was a new and fun instrument for me to play. As a member of the forum I am just expressing my opinion of what seems most likely to happen, and my reasons for thinking that, just like everyone else.
I appreciate that a lot of people may want ( or think they want) these additional features. But the trouble  is that they can drive the cost and complication of operation in a direction that doesn't always benefit everyone and can be counterproductive. The original success of the arranger is that it is a simple instrument to learn to play, that appeals to a certain market.   
Mike

[Patt22]

I'm satisfied with my T4, the only thing I whised to be changed is the fact that all breaks and fill-ins are just 1 measure more measures is a big wish.

Hello,

You can do this by using Ending 1,2 or 3 depending on your song, and interrupting it before the end of the 1st or 2nd bar with the Fill ( or not ) of the next desired Variation ...

This use of short endings can also be applied to registrations.
For example, in an extreme case, you'd like to program 4 totally different pads on each 1~10 registration!

At the end of your current registration, you launch your adapted Ending before moving on to the next registration and you interrupt it before the end of the desired bar with the next registration and its 4 other pads ... provided that the Stop Pads Ending in Setting is set to "On" and you have your Stop Pads using the Ending while keeping the 4 Pads playable!

Patrick

[J. Larry]

I've enjoyed reading all the griping, wishful thinking, complaints, and ballyhoo about Genos 1 versus what may appear on November 15.  None of that may matter, unless Yamaha has been listening over the past few years.  After the smoke clears, the dust settles, any the many demos appear on youtube, it'll be interesting to see if any of that verbiage has any affect on a new arranger, should it appear.  Surely, somebody keeps notice on requests that make it to a final product.

[Enildo]

Mike, I think you worry too much about how Yamaha could do this or that and keep costs reasonable -it's Yamaha's job to figure that out! On the other side, we users, are legit to express our wishes and needs. When you say $9000 for Genos, it's like you try to justify (for Yamaha) why we can't get certain features.
If we keep thinking that way, then that means, that we can't expect for next Genos to have any important new feature, because the price is expected to be similar as for current Genos when it came to market. But speaking for me, I do hope and expect, that this will not be the case. Progress makes that possible: getting something better at the same price.

Bogdan

Hello guys!

On this point I agree with you.
Whenever we suggest something new, someone will appear saying:
* Everything is already good.
* Why change?
* I'm already satisfied... etc
I always say, if I were to listen to these people, we would still be playing on a PSR 510. For those of you who don't want changes, that's great, continue with your old models, but the industry needs to innovate to stay in the market.
Everything we think is unnecessary today will be used by someone in the future. Some things will happen in the next model, others will happen in 10 or 20 years and others will never get off the ground.
But as Yamaha consumers, it doesn't hurt to ask and put our interests on the table about the product we want to consume.
I believe that Yamaha really likes these suggestions. We are helping them think and make money!

Enildo

[pjd]

You can do this by using Ending 1,2 or 3 depending on your song, and interrupting it before the end of the 1st or 2nd bar with the Fill ( or not ) of the next desired Variation ...

Patrick

Hi Patrick --

I've used INTRO 2 and 3 in this way by replacing the long, mostly unusable (to me), song-specific, orchestrated introductions. If one misses the switch at the end of the INTRO, then, at least, the INTRO kicks into a regular MAIN section.

All the best -- pj

[pedro_pedroc]

Well. I create professional styles. People order styles from me to be built. I have been doing this for the last two years.

Instead of more variations (4), I prefer to have more style channels/tracks, which is actually 8 to 12 or more. Sometimes, I need to ignore voices, because all the 8 tracks are full.


That's what I need (just my opinion)

Best wishes.
Pedro

[pjd]

As to innovation, Yamaha haven't made significant improvements in the auto-accompaniment in a long time (over two decades). They are recycling the same old features, the same 34 chord types, the same chord detection, etc.

It's like they are afraid to crack into the auto-accompaniment and sequencing code. Maybe Ichiro -- who understood musical time signatures and the auto-accompaniment code -- retired in 2005 and was never replaced.

I don't think it's unreasonable for customers to expect more from a vendor. I can't imagine forums for other computer-based products are docile and quiet. 

As long as we treat each other with respect, we're good -- pj

[BogdanH]

hello Pedro,

...
Instead of more variations (4), I prefer to have more style channels/tracks, which is actually 8 to 12 or more. Sometimes, I need to ignore voices, because all the 8 tracks are full.

I must say that so far it never happened to me, that I would need more than 8 tracks. I think it depends on music for which style is created and personal taste on how the style should sound. But 12+ voices simultaneously can sound quite "busy" .
But I understand your wish and at the end, if someone doesn't need that many tracks, they can be left empty.

The main difference between your wish (more tracks) and my wish is, that more variations (and other stuff that I mentioned in my 1st post) don't really require any hardware change. On the other hand, if number of tracks is increased, that automatically involves the need for more polyphony (I imagine at least 320 voice polyphony). Means, the need for more advanced & faster chipset. Is it possible? It always is. But usually it's polyphony that separates expensive keyboards from cheaper ones. And so I can't imagine seeing midrange (PSR-SX) keyboards with 320 polyphony anytime soon.

Greetings,
Bogdan

[DerekA]

Whenever we suggest something new, someone will appear saying:
* Everything is already good.
* Why change?
* I'm already satisfied... etc

I think there is a difference between wanting to see genuinely new innovation, and making it easier to do something that you can already do. From Yamaha's point of view, if there is a way to do it - even if it is a little clumsy - well, then that's enough to satisfy the small number of people who want to do it.

[pontuseuxinus]

I had this thought for a good while now, maybe someone already knows how to implement it.

In the context of what I am describing, variations are not necessarily new song sections, but variations that could be triggered independently on select instruments, at any moment during the performance. This helps breaking the repetitive nature of the style, and would provide complex, always unique departures for any instrument (melodic or rhythmic), while the rest continue as before. This would be far more valuable than additional sections, because it's more natural and closer to how a real band performs.

The multi-pads could be ideal for this, as it already has everything needed. But instead of layering new sounds on top of the orchestration, it would affect an existing instrument/style track. It's not a build-up, it's instrumental variation. The implementation should be relatively simple as it only requires one parameter: specifying to which track the Pad should send the midi notes to.

To me this would be a game changer, as it would produce much more natural performances and provide unlimited ways to bring new, interesting variations throughout the composition. As the song progresses, I can have one or multiple instruments perform subtle changes or even progress independently from the rest, all within the same song section. And if the pad phrases are loop-ed, you effectively created a new section.

If there is a way to do this currently, I would love to hear it. Been thinking about this as a project, such as using external midi automation (muting tracks when certain pads are triggered), but it's quite involved and would be very song specific.

[Graham UK]

Remember, Yamaha added some Adaptive Styles on the DGX670.
Variations auto change depending on your playing pressure and style of playing.
This works well and should be developed further.

[Joe H]

Everything mentioned can be accomplished with registrations.  Registration banks is where we can "arrange" complex compositions.  Registrations are the centerpiece of arranger keyboards.  Some people want to change the way arrangers work. Yamaha has made a big investment over decades of time.  Arrangers have come a long way.  Until MIDI 2.0 is implemented (with backward compatibility), I don't see any major changes in the architecture of the SFF GE style file. It's easy to say Yamaha could do this... or that... but it's not that simple.

Our keyboards can accomplish a lot... if we know how to use them.  There is no creativity by just pushing a button. The more "Smart" devices get... the dumber WE get. 

IMHO



Joe H

[pontuseuxinus]

Everything mentioned can be accomplished with registrations.

I am not sure if you were replying to my post, I would certainly be interested to learn how registrations can be used to create momentary variations to a select midi track in a style.

It's no secret that the main shortcoming of arrangers are their repetitive aspect, regardless how many section variations are available in a style. It's the same pattern, over and over again until the next section. While this works fine with certain genres, a bit of creative liberty would go a long way to allow for more personal and natural sounding performances.

I can currently make pad libraries and record dozens of different drum fills, bass variations, etc. and trigger them as needed. The issue is of course that they would go on top of the drums / bass performances played by the style. While I could try to manually mute and unmute these parts while pads are performing, a more simple and elegant way would be to direct the pads notes directly to those channels instead. I can then use these fills and variations on any style, regardless of the voices used.

From an implementation standpoint, it's a low hanging fruit - send the notes to track x instead of whatever track the M.Pad is using. It doesn't require modifying the architecture of the style, since it's an external function.

[soundphase]

You choose a style,

you choose a variation, you change an instrument in the style, you mute some tracks, you change effects, EQ,.... you store the configuration in registration 1.
You choose another variation, you change instruments, mute/unmute, effects, ...you store in registration 2



.... Until registration 10

You have 10 variations for the same style for your song.

[ton37]

A question for clarification: is a 'style track' a track that runs in the 'background'. Something like a midi track? Or keyboard in combination with a DAW, which runs a track on the computer in the background? So in fact a kind of 'karaoke effect'. If I'm wrong, please give me a link based on an example on the internet. So if that is the case, then you are 'obligated' to continue following that track, right? So if you join in too late with your solo instrument (or vocals) you have a 'problem'. Am I seeing that correctly? In other words: when you play live it can cause problems if you want to 'correct'/improvise something. It can be an advantage for a (studio) recording (although there are other options too?)

[BogdanH]

You choose a style,
you choose a variation, you change an instrument in the style, you mute some tracks, you change effects, EQ,.... you store the configuration in registration 1.
You choose another variation, you change instruments, mute/unmute, effects, ...you store in registration 2
.... Until registration 10

You have 10 variations for the same style for your song.

That's not 10 variations, it's still only 4 variations! Changing only voice parameters in some variation doesn't make a new variation. Basically, it's just a copy of the same variation sounding a bit differently.

Anyway, when I started this thread I was mainly targeting those who create custom (song specific) styles.

Bogdan

[BogdanH]

@ton37
Your thinking is correct. Let me try to summarize:
Nowadays we can get almost any song as audio backing track. That's an audio (wav/mp3) file where leading vocal is missing (has been removed). By having such file we only need to play main melody (with any voice of our choice). Obviously, result will sound stunning -as original, so to speak.
But there are big drawbacks:
1. We must play the music in the same key as original
2. We must keep the timing perfectly
3. We can't repeat (or cutout) certain part of music while we're playing -in short: we can't improvise.

It's quite similar midi track. Although in this case we can change tempo, key, voices, etc., the main drawback remains: no improvisation. That is, we must play the whole song as it is. Obviously, we need to keep timing as well, because if we miss start of some variation in the middle of the song, then there's no chance to correct that.
In short: imagine audio and midi files as singing on "playback".

We can imagine a style as a midi file divided into many smaller parts (having a length of only few bars), where each part is independent from another and any part can be repeated until we choose another part. No need to say, that it's practically impossible that we would miss the start of next variation (because we decide when it starts). Obviously, we can also decide to go to ending part at any time -if we notice that audience is getting bored  . Or we can decide to repeat the last third of the song if we see that audience is enjoying dancing on it.
I think, that's the main purpose (and benefit) of the styles.

Bogdan

[ton37]

Thank you for your explanation @Bogdan. I rarely used a running track, but I am curious enough to investigate the (im)possibilities of the application. In the past I have created song-specific styles several times with varying degrees of success. But that required a significant investment of time and was quite laborious to achieve a good result. That's why I stopped doing that. Sometimes I try to archieve it by using the registration memories (with possibly slightly adapted or other styles or style assembly) in combination with pads (midi and/or audio). So to each his own.
I now understand the essence of this discussion better. 

[soundphase]

A question for clarification: is a 'style track' a track that runs in the 'background'. Something like a midi track? Or keyboard in combination with a DAW, which runs a track on the computer in the background? So in fact a kind of 'karaoke effect'. If I'm wrong, please give me a link based on an example on the internet. So if that is the case, then you are 'obligated' to continue following that track, right? So if you join in too late with your solo instrument (or vocals) you have a 'problem'. Am I seeing that correctly? In other words: when you play live it can cause problems if you want to 'correct'/improvise something. It can be an advantage for a (studio) recording (although there are other options too?)

A style contains 3 intros, 4 variations (Bogdan would like 8 ), 4 fill-in, 1 break, 3 endings that are each, a predefined combination of 8 midi independant tracks: 2 drum tracks, 1 Bass, 2 chords, 1 Pad, 2 phrases with their own choice of instrument, EQ, effects ....

Some tracks follow chord changes, some other not. It depends on the style. => You are not at all "obligated" to continue following the track. You decide the chord you want to play when you want, and notes are automatically adapted to follow your chord choice.

You can associate a new instrument, different values of EQ, DSP, ... as you want.

You can then store the current result in a registration.
The registration stores the reference of the original style you chose, and all your current configuration changes.

You can also program a completely new track (the notes) if you want with the "style creator" or by importing midi tracks (experts). And you create a new "original" style when you save your result on disks (with "style creator").

When you play, there is absolutely no difference of latencies, when a style is only the original one or a style you changed some configuration parameters which are stored in registrations.

There is no latency at all as long as you only reference the same style for all registrations.

There is a small latency (max 4 measures) when 2 registrations reference 2 different styles, because the style is reloaded from disk.

You can find information on internet and youtube, test the style creator, or simply go to the "mixer" to play with the 8 tracks of a style

There also is an advanced mode (not available with the style creator) with 16 tracks (instead of the 8 available with the mixer). Tracks are then associated with the kind of chord (maj, min, min7 ...), so different notes can be heard according with the kind of chord you play.

[soundphase]

That's not 10 variations, it's still only 4 variations! Changing only voice parameters in some variation doesn't make a new variation. Basically, it's just a copy of the same variation sounding a bit differently.

Anyway, when I started this thread I was mainly targeting those who create custom (song specific) styles.

Bogdan

Personally I use registrations for one song. So for me, it's 10 variations. (I can change styles between registrations)

[pedro_pedroc]

Yes. Registrations can solve "almost" everything - I said almost, because they can not solve the max 8 style tracks problem. I know you can start multipads to add more 4 parts, but it's not the same thing.

Let's wait to see what genos 2 will bring to us.

Pedro

[Amwilburn]

When you play, there is absolutely no difference of latencies, when a style is only the original one or a style you changed some configuration parameters which are stored in registrations.

There is no latency at all as long as you only reference the same style for all registrations.

There is a small latency (max 4 measures) when 2 registrations reference 2 different styles, because the style is reloaded from disk.

The latency is actually quite noticeable depending on the USB device (I program registrations based on USB stick locations, as there's no guarantee going from instrument to instrument that the style exists on board) and depending on if the 2 styles are in the same folder or not, so while I use registrations to the fullest (I think?) there are a lot of things I can't quite eliminate, *especially* when simultaneously recording audio to the USB stick

Mark

[DrakeM]

I confess I have not read every post in this thread but ...

I do use a registration to tie 2 styles together to get extra variations. But in most cases I just compromise and keep the song to the 4 variations (5 would be better).
For songs with different timings I also tie 2 style together. You have to do that and don't see any way around achieving that with the current style creator.

I would like Yamaha to include many more styles with style parts that are 8 measures long. They would serve, for me, as examples of how to go about creating them for myself. I have examined the few 8 measures that Yamaha includes with the keyboard and some Korg converted styles I have. From listening to these styles I have created several of own 8 measure parts to my custom styles.

   

[mikf]

There are always restraints and compromises in music. If I play solo piano I am free to do anything, change tempo, change key, repeat sections etc etc. But it is inherently constrained - one instrument only 10 fingers! So you choose what you play appropriately.
But as soon as I involve other people other restraints and compromises appear. Playing in a large group needs full written score to avoid chaos. Then it can be a very elaborate arrangement, but the freedoms disappear, because now my job is to play exactly what is written, no deviation. 
Playing in small bands is in the middle - you get some freedom, but can be much less elaborate.
In most ways the arranger is no different. I can make a multi track midi, play along with it and have a very elaborate arrangement, but cannot deviate. And it is a lot of work to prepare. Or I can just sit down, use a generic style and have quite a bit of freedom, but cannot exactly reproduce an elaborate arrangement, and so the accompaniment can be a bit repetitive.
If you make more elaborate styles, then some of the freedom also disappears. 8 bar styles for example might offer more room for an elaborate arrangement, but it introduces some other constraints and might work great for a custom style, but not be very generic, which is why Yamaha largely avoid them. Having more variations available is somewhat in the middle - it is more work and obviously more cost for the commercial suppliers, but does remain fairly flexible because you don't have to use them all. But there are already ways to get similar results, so that becomes pretty much a viability question.
When I played in small live bands, it works more like a style than people here seem to think. The accompaniment is fairly repetitive, and the main thing you think about is that the accompaniment is good enough not to distract - ie doesn't mess things up for the vocalist or lead instrument - which is what people really hear. In a good band, everyone is competent enough to play their part but not get in the way. And the accompaniment might pay lip service to well known recordings of the song without trying to become a carbon copy.
When I sit down at the arranger, it seems very much the same to me as the small band. As long as the style is appropriate and well made, it will sound reasonable, and what really matters is how well I play my lead - things like ; is the overall arrangement nice, is it a good song, have I added some individual touches like chord alterations or extensions, is it interesting. My enjoyment and satisfaction comes mainly from what I play, not what the 'machine' plays, as long as it doesn't spoil what I play. 
When a new arranger is on the cards there are lots of posts about how to make the voices better, styles better, more easily editable, more realistic. For some people that is important, especially if you get paid to play and want to produce the illusion of a big sound from a OMB, but not use backing tracks.
But the unfortunate truth for most of us is that the voices, styles etc on the arranger are already pretty decent and to sound significantly better you mostly need to play better, and worry less about styles and whether the piano voice is sampled at a high enough rate.
Mike 

[BogdanH]

Perfectly described, Mike!

Yes, I see arranger as a small band and I believe, as long I use it sounding "small band-ish", my playing remains believable (in sense "yeah, that's what one person actually can play"). Because as soon music becomes too complex/busy, it might raise the question if it's a playback or not -especially to those who don't know about arrangers.

Bogdan