Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards CASIO releases 4 new models

I was referring specifically to the PRIVIA models. We won't see the new AP series here in the US until January 2013.

So the 750 is the cabinet version of which keyboard in the new lineup? Im a college kid and having a cabinet setup would be too much of a hassle. Im looking for the best keyboard in the lineup for my money...so really, what Im asking is what is the PX-350 equivalent to? Is it just a KB version of the 750 or what?

It's pretty trivial to get 14 bits from an action, you just speed up the counters and make them wider. I believe the real questions are: where in this process is the velocity curve applied, how much resolution is left (excluding any missing codes) after it is applied, and can the sound engine make any use of data wider than 7 bits?

It's not trivial at all to have this type of resolution tied to the sound engine. So yes the sound engine takes advantage of this and it allows for a more accurate and expressive experience. This is something that you won't find on Privia's most direct competitors.

The PX-150.

Are there any studies that show 127 velocities - post velocity filtering - are insufficient? More resolution is generally better, but only up to the limits of human perception, after which the engineering effort expended isn't adding any real value.

So you're saying these 14 bits of resolution are after the velocity filtering?

i'm going to be a dork and suggest that casio consider....

making the privia line- lightweight and portable- and have gradations up in quality to where the top privia has all the sound capabilities that COULD BE available- and still a light(er) weight portable

make the Celviano line your exclusive console line- with quality grades similar to the privia.

i dont see why you need console privia's and console celvianos- i know they are different price points and are trying to address different markets, like toyota and lexus, but this gets somewhat redundant and confusing. and you are underserving your portable DP customer. seems to me you could make a better margin loading up a portable board and saving some trees.

and that's all i'm going to say about it.


PS: and yes, i understand your retail distribution channels are different and probably want exclusivity, and two brands gives you wider distribution....

'

What is velocity filtering?

Anyway, maybe I'm over-simplifying, but it sounds like the keyboard sends 16k velocity levels to the board's internal piano sound engine rather than the more typical 127. Is there more to it than that? And do you think that's not valuable? I'm afraid I'm missing something...

BRB,
We appreciate the feedback, while Privia and Celviano have similarities the sometimes the difference comes down to cabinet and the style. There are some customers that want small light and portable and other customers want something that is more traditional in style with a larger speaker system. We're giving customers a choice which I think is a good thing.

Dewster,
The limitations of the 7bit 0-127 value system has been argued a lot over the last 20 years as electronic instruments have become more capable and expressive. From pianists to drummers (using electronic drums) countless people have argued that 127 values is nowhere near enough.

Casio is not the first to implement a 14-bit system on the keyboard. You will find this on products such as Yamaha Clavinova. I'm not aware of other manufacturers providing this level of detail on products under $1000. Needless to say, the keyboard's connection to the AiR sound source that I believe takes our product to another level.

Mike, I have a question: I am planning on getting the PX-350 as I am enjoying my P130 perhaps even a little more than the P95.
I really liked the PX-830 and somewhat question why I didn't get it instead of the P95 2 years ago and it came down to portabililty.

Was any consideration given to perhaps having one of the portable models contain the 2 20 watt speakers? Or does this require a larger shell?

Also, is there any reason you would recommend a 150 over a 350?

To sense velocity from two positional switches, you need one timer per key. You start the timer when the upper switch closes, and stop it when the lower switch closes. The time elapsed is inversely proportional to the average velocity between switch events.

You have to take the elapsed time count and filter it in many ways. If the count is too high (too low velocity) the filter spits out the minimum velocity or zero velocity. If the count is is too low (too high velocity) the filter spits out the maximum velocity. In between these the filter might add an offset, apply a curve that favors the high end (harder) or low end (softer), etc.

For the filter to be able to spit out all possible values (with no missing output codes) the input elapsed time count must have more resolution than the output (since we're dealing with integers). So there is definitely a value in having more precision at the input of the filter. Is there a value in having more than 127 levels at the output of the filter? Mike is saying it is better to have more, but he's here selling Casio. I could be persuaded to believe more than 127 is good, but I would need a study. If Mike is saying it was non-trivial for Casio engineers to implement this, certainly they had some hard data driving them to commit the time, energy, and money required to do so?

I can certainly see situations where more output bits would be good, like if you are using the keys to trigger a PC sampler and want to apply a second velocity curve on the PC. I wonder if the new Casios can output this higher resolution velocity data via MIDI?

But more resolution to the internal sound generator? I'd need to be persuaded there is a real need for it.

The bottom line for me is PX-350 or used Yamaha MO8/MOX8. They're all around the same price.

I still don't quite get the distinction.

Let's say the fastest velocity (represented by the speed at which the key would travel from the top sensor to the bottom sensor if a bowling ball were dropped on it... or as close to that as they could imagine any human managing in normal playing) is 1 millisecond. Let's say the slowest velocity which they think should produce any sound at all is 500 milliseconds (i.e. anything slower than that would be slow enough that, on a real piano, it would produce no sound).

(I know those numbers may bear no resemblance to reality; I know that some DPs produce a minimum sound no matter how long you take to depress the key... but I'm just choosing a range of number to illustrate.)

So we have a range of 500 milliseconds which need to be translated into either 127 values (where 1 millisecond = velocity 127 and each additional 4 milliseconds or so decreases that value by one); or we have 16,256 values where 1 millisecond = velocity 16,256, and each additional .03 millisecond decreases that value by one.

There is not a pre-filter division and a post-filter division; there is an absolute time, and a value derived from it. Though of course, they could apply a curve so that the division of time into velocity units is not so linear. But it's still a simple equation of time on one side and a velocity value to send to the tone generator (and/or over MIDI) on the other.

No?

EDIT: I think maybe I understand, you're talking about how fine the initial time measurement is... whether it's to the nearest millisecond, or something finer (or more coarse) than that, which also has to be represented digitally as some number of bits, yes? But processors have long worked with tiny increments of time, even a millisecond is long to a processor, and MIDI has no relevance to the time calculation (from triggering one sensor to the other). So I think the only place "increased definition" makes sense is on the other side of the equation.

Dewster,
One additional detail, the new Privia's using 3 sensors in order to achieve this higher resolution.

No. For one thing you're equating subtraction with division. The elapsed time is inversely proportional to velocity.

An object traveling at 1 m/s will take one second to travel a meter, say this produces a count of 100. An object traveling twice as fast will take 1/2 second with a count of 50. And an object traveling three as fast will take 1/3 second with a count of 33.3. The velocities 1, 2, and 3 are equally spaced, but the elapsed time counts 100, 50, and 33.3 aren't. It's this cramping that requires increased input resolution to the velocity filter.

Offsets and non-linear filtering (to favor higher or lower velocity ranges) require even more. Surely you've seen convex and concave velocity curves in keyboard manuals?

I didn't see you message before adding my edit, but it looks like my edit was on the wrong track anyway. ;-) I still don't really get this (i.e. "input resolution to the velocity filter") but I don't feel I have the background to continue to discuss this intelligently, so I'll leave it here.

I really like the PX-830 as well. Other than the PX-3 it was the only model that had the matte finished keys. 20w per side could be done in a "portable" slab but it would significantly add to the weight.

In regards to the PX-150 vs PX-350, my choice is the PX-350. There are many reasons why, but my favorite is the USB audio recording. When you sit down to play, just plug in a USB thumb drive and press the audio record button twice and it is recording a .wav file.

I recorded a very informal video in my office tonight of myself playing the PX-350. I didn't have to hook up the computer, dial up software or anything. It captured all the audio which I lined up with the video later. So convenient. Not sure if I'll post this yet....wasn't really happy with the way I played. I'll watch it again in a few and make up my mind about. Either way, I'll shoot more video tomorrow.

Mike,
Will you be making a video demo of any of the new models or is Casio going to just be putting up audio samples of the new Privias? So it sounds like with the new action, the keystroke pressure will affect how the paino reproduces the note that is sounded and the nuance?

One thing I wanted to ask you was about the speakers- I feel as though the ones on the PX130 do a nice job of filling the room I am in and the piano sounds very similar with headphones on.

I also got one of those stereo miniplug cables to connect the 130 to my Mackie MR5's- and to be honest only did it once because I find that I am getting the sound I want from the keyboard itself- compliment to Casio for well designed speakers on their slab pianos.

That is my complaint about my P95 over time- it is a completely different instrument- I do not hear any attack on the internal speakers but with headphones to me I feel as though I am sitting at a grand.

Power being equal, how does having the tweeters on the 350 change the sound vs the 130 or even 150 ? Does it make it perceived as louder (with the same wattage) change the EQ?

The 850 looks great as well, but for me personally-I would be more inclined to get the 350 and put it on an x-stand and later add the custom cs-67.

Seems the 850 would come in a 100+pound box and it is just way easier for me the way my 95 and 130 came ups.

As you can tell I have the Yamaha custom stand and found it simpler to set up, easier to have shipped etc..