Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards Midi Controller/digital piano for beginner

No worries anotherscott, I got it too!

This is great information. I will have to find a source that actually knows how to present electric pianos. I have the impression that guitar center isn't a very good place to buy a keyboard...

(On Reddit there are discussions - from people who said they had worked there) that guitar centers sells returned items as new - can't tell if this is correct or fired employee revenge...

The problem is that in the LA area only guitar center has Rolands on display.

Yes, absolutely.

5pin-DIN midi to USB convertors add latency.
Cheap ones can add massive latency.

Even the more expensive, higher quality ones like Roland UM-ONE (47eur on Thomann), will add 1-2ms of extra latency.
Which isn't much, but it all adds up!

If your audio interface has 5-pin DIN midi inputs, then that's fine, but getting audio interface which has good drivers that don't add extra latency isn't easy (if you just dont shell out for a RME).

So I would highly recommend to avoid introducing anything into the signal chain which adds extra latency.

I have a M-Audio Midi sport UNO:

PC-> Midi-Sport -> piano -> Midi-Sport -> PC

The piano is used with the thru port then add no latency. The round trip time for a short event is 3ms. The MIDI path is 1ms (given the baudrate and the number of bits). Then the Midi-Sport only add 1ms one way and 1ms in the other way (roughly). Then yes, there us an added latency but not that much.

Cheap MIDI-USB interfaces (typically $5) should also be avoided because they are not reliable : some flip some bits, some can’t send SysEx reliably, some omit the required opto isolator.

The question is not whether using a 5-pin MIDI DIN interface adds latency, it's whether it inherently adds MORE latency than the alternative, i.e. using a USB connection.

MIDI itself has a 1 ms per event latency that can't be avoided. ("Round trip" latency is not an issue if you're just using a board to trigger a piano VST.)

I didn't think using the 5-pin connection would introduce any more latency than using a USB connection, but after googling, I think that's not the case. From what I can quickly discern I believe the answer is:

* 5-pin from the keyboard into 5-pin in another device is best... you can't get any lower latency than that... but the computer doesn't have a "native" 5-pin input, so that isn't an option for a VST.

* USB from the keyboard into USB port of computer is next best from the latency perspective... latency isn't necessarily worse, but it can be worse.

* inserting any conversion between the two can add a small amount of additional latency.

That said, using the 5-pin MIDI connection of the keyboard avoids other potential problems of USB. It eliminates the possibility of noise from ground loops (though there are ways to address that problem with USB if you run into it), and you can run much longer cable lengths without running into problems (so if you have long distances and must do the conversion, run long MIDI cables, and keep the "USB converted" part of it short). The physical connections are also more rugged.

Surely, the round trip time is not the issue, but I can’t measure easily a one way delay. The RTT can however gives some hint about the input latency.

We have to take into account that a USB device can’t emit anything to the USB link until the host (PC) gives the right to emit. On USB1, the PC ask every millisecond if there are something new. Then we can have a 1ms delay because the device wait the PC signal.

Where have your read that?

A USB 2.1 cable has a data transfer rate of 480 Mbps thats Mega Bits per second or 480,000 Mega Bits per millisecond. The USB over MIDI protocol takes advantage of this and is hundreds of times faster than MIDI's original speed. And with a bandwidth so wide that you could put multiple midi ports in one USB. If you can still perceive any delay, your probably hundred times more sensitive than the rest of us.

What? 480 Mbps = 480,000 Mega Bits per millisecond?

No! You've overstated the data rate by a factor of 1,000,000 !!!

480 Mbps = 480 Mega bits per second!

But for MIDI purposes it just doesn't matter much.
The old spec using the old DIN cables could send at 31250 bps, which means one complete (3-byte) MIDI note in just under 1 msec.

There can be some benefit by going a bit quicker. But there's no need to get to even 1 Mbps, never mind 480 Mbps.
Your fingers can't play that fast. Even Liszt could not!

Don't fall prey to the "more is better" mantra.
It's more appropriate to recognize that more is better ... until it isn't better anymore.

I had one such from Sweetwater :-) so this is probably true.

It was not a piano but a sonic bar. The moment I opened the package, I knew this was a repackaged one. The wrapping was torn in places and taped up with some non-standard tape. The user manual was taken out of its protective cover and replaced. The power cord didn't have the usual wrapping. I took photos and immediately sent it to the "sales engineer." I got a call and they were very apologetic and acknowledged a mixup, though they couldn't give any specifics. The money was refunded immediately, even before they could send me a return label. The last part was very professional of them, although the mixup is not their finest moment. I suspected the mixup was putting them up as Open Box/B-Stock vs. New.

Perhaps you need to decide just how ‘MIDI-capable’ you need your digital piano to be. I have an ES110 (originally with only a single pedal) bought specifically for recording. It’s brilliant, but I discovered its limitations as a controller when I first started delving into MIDI settings with Pianoteq Standard. I was able to achieve the extra control I needed for a specific project by buying the 3-pedal bar, but the ES110 is primarily a piano, not a controller.

I’d recommend you check the manual for each DP on your short list to see whether it meets your MIDI requirements before making your decision.

Thanks, MarieJ, I don't have a MIDI requirement list because I'm pretty much at the beginning of this acoustic journey.

I discovered a certain liking for high quality midi keyboards with hammer action Fatar keybeds which are all a bit out of my current budget.

I'm thinking of maybe getting a very simple 88 keyboard so I can continue my piano lessons and buy something good later. Spending 750$ on something that has serious downsides... maybe I should rather buy a semi-weighed keyboard and buy something like the Yamaha 515 or a Kawai equivalent or a Studiologic Grand in a year or so.

The ES110 is on my list even though I haven't been able to find a demo sample yet.

People seem to have bought up everything - in six months I predict a wave of used, low-priced digital pianos when the Covid holidays are over.

Hello,


Out of sincere curiosity, within the piano context, what specific limitations did you discover with the ES110? Other than it having dual instead of triple sensors, I can't think of any from the top of my head.


This is truly sound advice to anyone on the purchase trail. Luckily I did this, for some of the pianos on my list failed the moment I consulted the 'MIDI implementation' pages in their respective manuals.

Cheers and happy playing,

HZ

What exactly do you check in regards to MIDI requirements? Throughput numbers? I don't have enough information to place this recommendation.

Hello,


The really long answer to this would be... well... really long. So I'll try to give you some hints (already long-ish).

First of all, be clear on what you require your controller to do well. Note: if you buy a digital piano as a controller, and want to also be able to send MIDI to the piano to be played through its internal sounds, there is a two-way stream to consider.

MIDI is a stream of messages (a.k.a. 'events'), each containing a 'command' (a.k.a. 'status'), a key number and a value. Generally speaking, that is. So you want your controller/piano to be able to send (and in the two-way case also receive) the commands and values/value ranges as you need them to be. Manuals have, or should have, tables (usually in an appendix) showing you the 'MIDI Implementation' for the respective controller/piano. If applicable, you see columns for the outbound as well as inbound MIDI messages (which by the way can differ quite a bit).

So for instance. A modern MIDI implementation uses separate commands for note-on and note-off. Older implementations (e.g. as found on the Yamaha P45) use the note-on command for note-off messages as well, with a velocity value of 0 (zero) indicating the note-off. While this usually still works today, in such a configuration you could never send a note-off velocity to your VST. With the separate note-off command, all velocities can be sent (only if actually implemented!), such as e.g. the Roland FP series do (and many others nowadays, as I assume/would hope).

Concerning pedals, does the controller/piano send on/off only? Or a mid-position value as well? Or a proper continuously variable 0--127 range? Which pedals are implemented (sustain a.k.a. damper/sostenuto/una corda a.k.a. soft/and possibly others)?

Concerning aftertouch (although seldomly found in digital pianos, if at all), is it 'per key' or is it on/off for the whole keyboard at once?

So with this I hope to convey that there are many details (the ones I mention here are just a few, albeit important ones) to carefully consider when it comes to each specific controller/piano's MIDI behavior. It really demands and deserves careful study and consideration.

One last caveat: The implementation tables can't and/or don't tell the whole story. Again with the P45 as an example: the table may tell you the velocities are in the range 1--127. Unfortunately that doesn't mean that the piano actually sends out neat gradual velocity messages using all of the fine grain this range would/should offer. Instead, the range may be limited in reality, often not exploiting the upper part of the range and sometimes coarsely 'stepping' through a limited number of velocity values that are actually used.

The latter can only be learned through testing and/or, as I did, from... drumroll... fellow Piano Worlders.

Cheers and happy studying,

HZ

OTOH... if you're just using it to play piano, anything you like the feel of will work fine. ;-)

I think this MIDI sidetrack is s bit overblown, and should not scare you off. Just to play a software-based piano, you don't have to know anything about any of it.

If you want to have three pedals, buy a piano that has an attachment for three pedals. The MIDI will take care of itself.

If you want to be able to do partial sustain pedaling, buy a piano that has that feature. The MIDI will take care of itself.

IOW, buy the piano you want, that has the features you want. There's no need to be analyzing MIDI implementation charts for this.

The only point above that I think could have relevance is whether a keyboard sends release velocity, there are piano VSTs that will take advantage of that... though I'm not sure which if any low-cost board can send it anyway. But that brings me to...

It doesn't "usually" work today, it always does. It's part of the MIDI spec and is required. It is not newer or older; the equivalence between Note On of 0 and Note Off has existed since the dawn of MIDI. Though you are correct that a board that sends release velocity needs to use the Note Off approach. For a board that does not send release velocity, it doesn't matter which they use, neither is older/newer or better/worse than the other. You're right that the P45 send Note Off by sending Note On velocity 0... the P125 sends it as Note Off velocity 64... but the two will create the same result. Neither sends different release velocity information based on how you play, and absent that, the two approaches are equivalent.

Whole keyboard ("channel") aftertouch is at least available on some hammer action controllers, even though it doesn't do anything when simply using it to play piano. Howver, AFAIK, there are no hammer-action "per key" (aka "polyphonic") aftertouch keyboards currently in production. I don't think there have been any this century!

Thanks. So basically there aren't any true tables and you need to study the implementation technology. Someone had mentioned the P-45 being outdated so I took it off my list (even though it was on it because I like its keyboard)

Thanks, I had replied already when your post came in.

So would you advise against a p-45 (the keyboard, not the world war II fighter airplane)?

Don't the Arturia Keylab mark II and the Komplete Kontrol S 88 have aftertouch?

I really like the Arturia but I couldn't get to try it out so far and I heard it has a very "slow" or mushy keyboard.

If you like the feel of the P45, I think it's fine. For triggering piano VSTs/apps, I think the only advantage of the P-125 is that it has an optional triple pedal.

Yes. Channel aftertouch (affecting all notes equally), not polyphonic aftertouch which (is activated key by key). Neither applies to piano playing, but can be useful for other things.

As always, great post! Thanks for this, now I could probably have a another reason to justify purchasing an FP-30

Hello,

@MarieJ quite sensibly advised to check MIDI implementations and I seconded that for, from my own experiences, I know how important that is.

Then, wanting to answer @Archipelago's follow up question about this, I gave quite a bit of my time this evening to write a useful answer. The information in that comment is based on real-world findings; for one thing I'm glad that I did that MIDI homework last year, otherwise I may well have ended up with an instrument that had disappointed me.

I certainly do not second the way @anotherscott waves most of this away as unimportant.

A year ago I knew not very much about MIDI, but as said I did the homework and am glad I did, for now I have a piano setup that acts exactly as I want it to, and I have a proper understanding of how it all works and I can rest assured I know what I am doing and how to fix issues should there be any.

This will do for this conversation as far as I am concerned.

Cheers,

HZ