Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards Is the midi standard a limiting factor for Software Pianos

I've seen lots of questions like 'why don't DP manufacturers just put a cheap PC running Ivory II(are other software piano) in their pianos?

I don't know the full extent of midi data fields used by Software Pianos, but they must be limited and finite.

Therefore if a DP manufaturer adds sensors that capture data within the piano action and pedals that is above and beyong the Midi Standard, is there not an opportunity for manufacturers to develop software to use that data to create a better playing experience?

Any thoughts?

Maybe, however, IMHO if they did put a PC into a DP, it would still be a worthwile improvement, just doing what is already possible using standard MIDI.

Greg.

There's no doubt in my mind that the best Software Piano's are at least as good as, if not better than the native DP sounds I've tried in shops. Having said that I have limited experience so am not the best judge.

I suppose there's a flip side to this question. What, if anything would Software Piano developers want transmitted from a controller in order that they could develop their products even further?

PC hardware is cheap, but it's a terrible market to intimately tie your products to. One production machine tool developed by a company I used to work for had a PC for brains - it was of course effectively obsolete by the time the machine hit the market, and upgrading it was problematic due to the level of integration necessary with the motion control electronics. What seems like a slam-dunk on the surface can be a tiger by the tail thing down the line.

The big embedded space revolution going on now is the ARM processor lashed to a Linux operating system. That should finally trickle down to the DP market backwaters sometime this century. When it does, DPs could potentially change explosively and for the better. (I believe clunky old proprietary processors and OSs are the main reasons DPs are such overpriced underperformers - that, and lame programmers.)


MIDI is quite limited, but it works pretty well for piano (though it's right on the edge). If we were developing the standard today it would undoubtedly have more bandwidth and more resolution, and be inherently bi-directional. I think MIDI has held synthesis in general back quite a bit, but that's perhaps another topic.

I don't think midi is the issue. There are 16 midi channels per port, each has numerous different controllers and each of those can have 128 different values.

There are people who think that more than 128 velocity levels would make the sound more realistic. I am not one of them. You could argue that having more samples within that range would result in less switchting/blending of samples, but the benefit of that past, say, 20-30 different samples is debatable when the blending is well done.

dewsie: I think you're right about the PC vs. ARM point. But either way, the situation still depends upon the manufacturer's perceived need to do things differently and better. I don't think they see any such need, though Kawai's recent progress might kick Yamaha in the pants.

Or maybe not. I looked at the "specs" for the new CVP-600 series ... and they're sticking with their "Real Grand Expression" engine. Same old, same old.

There is an extension to MIDI that uses high resolution velocity that allows for something like 16000 levels if I recall correctly. I know at least one keyboard (VAX infinitre response) and one software piano (Pianoteq) that are compatible with it. I've not tried it... I find the VAX too expensive and it does not suit my needs.

I think that having more velocity levels would be an improvement... not a huge improvement because having 128 levels is reasonably good, but is not going to worsen things, specially for modeled pianos where it has more sense to have more resolution in velocity. For sampled pianos the difference would be minimal to none as they usually have far less than 128 layers. I am not a huge fan of blending, but is better having it than not having it.

I believe that V-Piano uses more than 128 velocity levels internally... but not sure about that.

Regards,
Kurt.-

To me the crux of the matter is knowing what needs to be achieved.

From the comments so far I get the sense that both native DP engines and Software Pianos use the velocity and pedal data captured.

I'm guessing that work is done within the sound engine to select and process samples from sample banks. The modelled result is then output to a sound processor (Forgive my simplistic view - I'm IT literate but have no specialist knowledge in this field).

So, to determine whether we have the right data we have to fill in the following gap - "What's missing is .........." i.e the DP doesn't do this paticular thing and if it did that would make it better. Perhaps Kawai said 'what's missing is sound board sesonance' - so they added a soundboard.

Once we think we know what's missing we can determine what we need to do to remedy the situation.

We may decide that our sound engine needs to do something, but it can't 'cause it hasn't got the data. In this case we need to look at data capture and data transmission.

Alternatively we may decide that we have all the data we need but something different needs to happen in the engine or physically in the case of the DP.

So what's missing that would take the whole DP experience to a new level? or have we gone as far as we can barring refinement?

What are you talking about? Look at the new cabinet design! One of the models even has the lid, just like a real grand piano! I mean really, it doesn't get any more realistic than that.