Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards The DPBSD Project!

Like a green martian stepping out of a UFO: alien.

Is SN for supernatural... or super natural? They sound more supernatural than super natural to me.

"...you really need to 'feel' the subtle changes in timbre. That's why I'm always recommending consumers play-test instruments 'in the flesh' before making any purchasing decisions. I realise this may mean require extra leg-work seeking out a dealer, but it's really the only way..."

It might also require time travel. Comparing the past models with the present (unless you have one) is a problem, and comparing anything with the MP10 would require time travel into the future.

I wouldn't mind hearing a bit more of a technical explanation. It also seems to me that a demo that shows off these 'subtle changes in timbre' in at least CD-quality isn't beyond today's recording technology, the limitations of bandwidth, or what a set of decent headphones or speakers could reproduce. Especially since retail outlets are few to find and many will be ordering them from mail-order outfits. 'Ship it back in 30 days if you don't like it' is a lot easier to say than to do.

It's understandable to me that makers' and retailers' marketing materials are intended to conceal more than they are to reveal, but as a consumer it's not much of a help--- I might as well believe Gyro. Anyone who's seen more than one cycle of marketing puffery can remember when yesterday's breathless description of the latest and greatest, turns into "We had so many user complaints about the weaknesses of the former flagship product" (insert a few damning specifics here) "that THIS latest and greatest release is so much better you can't even compare it."

It would be nice to go beyond that. Releasing the user manuals online does help. Not every maker goes so far. Of course, absorbing that information requires a kind of time travel, too... when you could be saying, "Yeah--- that sound--- that's what I want." And no printed page will ever tell you that.

"Love" is incorrect. More like "the only game in town".


No, there is a huge difference between what Roland and Kawai are doing. Roland comes right out and very boldly says that SN guarantees no layer switching, stretching, or looping. They even made videos explaining these things, with phase scope shots and everything for the technically challenged. I really don't care what they name it as long as that name actually has some real meaning behind it.

If Roland offered SuperNATURAL, "Progressive SuperNATURAL", and "Ultra Progressive SuperNATURAL" with essentially no explanation as to the differences I'd be trying to pry any and all info I could out via the DPBSD and hectoring company reps, but that isn't the case thank goodness, at least so far.

My geek side would like technical detail as well, but really, that's not so important. Lots of things sound good on paper that don't necessarily translate to real world performance anyway. But some kind of sound demo would be helpful.

Maybe James can convince them to send Dewster a DPBSD file! Eventually someone should be able to do that, and at least then we'd have some kind of both audible and quantitative way to experience at least some aspects of the differences.

You're right that trying via mail-order isn't practical for something of this size. Freight both ways can easily be over $100, which is a lot to pay to try it out.

This really isn't helping me, the consumer. Doesn't Kawai patent their technology? If so, then isnt it already out there for anyone to see? I find it hard to believe Kawai isn't proud of their technical innovations. So I don't understand this inability to put into words the relative qualifications of something I might want to pay good money for.


This is good advice when purchasing any DP, not just Kawai. But if I demoed one with PHI, what would motivate me, the consumer, to go to the trouble to seek out and demo a Kawai DP with UPHI?

I'm probably not the first person to ask these sorts of questions, and I don't mean to pepper you with them James. Could you perhaps put me in touch with someone who normally deals with this sort of inquiry?

AFAIK I've never received one directly from a company or company rep. That will most likely never happen as they wouldn't be comfortable with having so little control over the review contents (though I've been doing my best lately to run the reviews by the people who submit DPBSD MP3s).

You should make your piano tests as a standardized benchmark resulting a "Final Score" and publish the results on a dedicated website. I would probably add some kind of key quality tests as well. BTW, why don't you visit some big piano stores and get the samples of all major brands/models?
Directly comparable (consumer understandable) scores would give more power to your tests and believe me if you'd create a "pro test" for money they would buy it.

Hello dewster,



Well, if you were at a Kawai dealer, demoing an instrument equipped with PHI (such as a CN33), I doubt it would be too much effort to walk across the room and try a model equipped with UPHI (such as a CA63). After a few trips back and fourth trying both classes of instrument, I expect you would be able to appreciate the differences.



Sure, try these addresses.

Cheers,
James
x

Thanks! Who do you think I'll have better luck with, "Products and Sales - Digital Pianos" or "Technical Support - Electronics"?

OK, lacking a response from KJ I send an email to both.

Sorry, your previous post must have passed me by.

Cheers,
James
x

I guess I sounded kind of snarky, but I didn't mean it that way. Sorry James.

If I get any info from the responses I'll be sure to pass it along here.

DPBSD Version 1.9 Released

Added controller setting for reverb off, a very slight tweak to the half pedaling test, and some minor fixes and changes in terminology in the readme file.

Please use this new MIDI file (located in the root directory of the DPBSD share point) instead of all previous versions.

====================
= Revision History =
====================
v1.9 - 2010-11-14:
Readme file edits:
- "Pedal down sympathetic resonance test" is now "Pedal sympathetic resonance test".
- "Key down sympathetic resonance test" is now "Key sympathetic resonance test".
- "Pedal down silent replay test" is now "Silent replay test".
- "Quick pedal partial damping test" is now "Quick partial damping test".
- "Partial pedaling test" is now "Half pedaling test".
- Fixed "Evaluating Results" section to reflect the actual tests.
MIDI file edits:
- Added "external effects depth" (reverb send) controller cc91 = 0 @ start of file.
- Shortened the key down times in the partial pedaling test (1/4 => 1/16).
- Tests in the order they appear in the MIDI file:
- Same as v1.6 & v1.7 & v1.8

Yamaha P-95 Review

Piano World forum member "anotherscott" was kind enough to provide us with a DPBSD MP3 of the Yamaha P-95 - thanks loads anotherscott! This is the maiden voyage of DBPSD v1.9, and it seems to be effective in turning off the reverb during MIDI playback, which is something that was dogging us. Poor anotherscott had to go back and record it several times, but in the end everything worked and we now have a high quality test file for this DP.

MP3: http://www.mediafire.com/?xs9l4ty8zc5ica5
PIX: http://www.mediafire.com/?25l7jlk77d3o279

I examined the MP3 and I didn't see anything that different from the P-85, so I'll direct your attention to the previous P-85 review if you want to know how either / both fared in the actual testing. For the rest of this review I'll do a brief comparison to demonstrate that they indeed share exactly the same sample set for this voice. You are encouraged to download the MP3s of these two instruments in order to give them a listen or to analyze them yourself.

anotherscott has this to add: "... since the recorded sound seems identical between the two, to the extent that people hear a difference between the two models, it is likely because Yamaha changed the speakers between the P-85 and the P-95. (This was confirmed by someone at Yamaha who told me that the two have different part numbers, and that the speakers in the P95 had "increased sound output.")".



Spectral Phase view of the start of note C2, P-85 at top, P-95 at bottom. Amplitude normalized to -1dB to bring out detail. No significant visible or audible differences. Attack and loop sample times are the same.



Spectral Pan view of the start of note C6, P-85 at top, P-95 at bottom. Amplitude normalized to -1dB to bring out detail. No significant visible or audible differences. Attack and loop sample times are the same.



Spectral Phase view of the stretch test, P-85 at top, P-95 at bottom. Amplitude normalized to -1dB to bring out detail. No significant visible or audible differences. Stretch groups are the same with the same signature.



Spectral Frequency view of the layer test, P-85 at top, P-95 at bottom. Amplitude normalized to -1dB to bring out detail. No significant visible or audible differences.



Spectral Pan view of the start of the pedal sympathetic resonance test, P-85 at top, P-95 at bottom. Amplitude normalized to -1dB to bring out detail. No significant visible or audible differences. Both DPs fail the pedal sympathetic resonance test.

---------------
- Yamaha P-95 -
---------------
FILE & SETUP:
- dpbsd_v1.9_yamaha_p-95.mp3
- Setup: Mac with ProTools LE, Mbox interface @ 24-bit, MP3 via Quicktime Pro.
- Recorded by "anotherscott".
PROS:
- Passes the silent replay test.
- Passes the quick partial damping test.
- Responds to half pedaling.
- Timbre variation is fairly smooth with increasing velocity, though most of the effect is restricted to the mid velocities.
CONS:
- No visible or audible pedal sympathetic resonance.
- Fails the key sympathetic resonance test.
- Note decay times are somewhat short (on the order of 2/3 to 3/4 Pianoteq).
- Stretching is visible over the entire range, audible over the lows and mids.
- Stretch distances: 2,3(x28),2 = 30 groups.
- No obvious pedal up/down or key up samples.
- Obviously looped, both visually and audibly.
- Attack sample lengths are rather short.
- Attack sample lengths are (C1:C8): 2.0,2.0,1.8,1.7,1.3,1.1,?,? seconds.
- Loop sample lengths are short and fairly audible over the lows and mids.
- Loop sample lengths are (C1:C8): 0.6,0.6,0.6,0.4,0.3,0.3,?,? seconds.
OTHER:
- Purportedly a single velocity layer sample set.
- Dynamic range 47dB (vel=1:127).
- Notes played @ vel=1 produce no sound.
- MP3 levels: peak @ -0.9dB, noise floor @ -70dB.
- Tests exactly the same as Yamaha P-85.
- Date reviewed: 2010-11-14

Casio Privia PX-110 Review



PW member "Vincentimes" recently sent me a couple of DPBSD MP3 files for the Casio PX-110 - thanks Vincentimes! One file was recorded with "DSP on", the other with "DSP off". I couldn't find any difference between the two so I went ahead and performed a thorough analysis of the "DSP on" MP3.

The PX-110 visibly passes the pedal sympathetic resonance test, but it is so subtle I can't say that I can hear the effect. It fails the key sympathetic resonance test.

It responds to half pedaling, though I wasn't able to detect any key / pedal "damping", "clunk" or "loom of strings" sound effects in the test file. It passes the quick partial damping test, and passes most steps of the silent replay test but fails at the end with a note damp at pedal up.

Note decay is fairly fast, though the noise floor in the MP3 is quite high and non-white which prevents me from tracking it too far. The attack and loop samples themselves are all quite short.

This sample set is very stretched, with 25 stretch groups covering 88 notes. It is stretched throughout the entire range somewhat evenly, and I can hear the low and mid stretch group transitions.

In terms of velocity layers, this is an unblended 3 layer sample set, with both layer transitions quite audible to me. In particular, the first velocity switch results in an abrupt timbre change along with a stereo image shift which is fairly obnoxious. And the second switch is also audibly abrupt in terms of timbre. There is a progressive timbre variation with velocity over the lowest layer, which seems like it could be a low pass filter or something.

MP3 and all analysis pix here:
mp3: http://www.mediafire.com/?xs76s8rdtgyr4fg
pix: http://www.mediafire.com/?jrs7lxjjk43plzt

Some analysis pics and the text review follows. Many thanks again to Vincentimes for the DPBSD files!


Spectral pan view of the pedal sympathetic resonance test, stimulus removed. On the left the pedal is down, on the right it is up. The effect is largely inaudible to me.


Spectral pan view of the looping test, note C5. Attack and loop samples are clearly seen, cursor is located at the transition point, looping is fairly audible in the low and mid notes.


Spectral pan view of the entire stretch test. A lot of stretching going on, groups / transitions are audible in the lows & mids.


Spectral pan view of the layer test, compressed 20:1. 3 layers, both layer transitions are quite visible and audible, cursor located at highest transition.


Spectral frequency view of layer test. Lowest velocity layer has progressive timbre variation with velocity, the upper two layers are more static in terms of timbre.


-----------------------
- Casio Privia PX-110 -
-----------------------
FILE & SETUP:
- dpbsd_v1.8_casio_px110.mp3
- Audio interface integrated Intel HDA, recorded with Wavosaur on windows 7.
- Recorded by "Vincentimes".
PROS:
- Pedal sympathetic resonance is visible, but audibly it's very subtle.
- Supports half pedaling.
- 3 velocity layers visible.
- Lowest velocity layer has progressive timbre variation with velocity (filtered?).
CONS:
- No key sympathetic resonance.
- Fails the silent replay test: note damps @ pedal up.
- Visibly and audibly looped.
- Attack sample lengths are (C1:C8): 1.6,1.4,1.4,1.1,0.8,0.5,0.4,? seconds.
- Loop sample lengths are (C1:C8): 1.3,1.1,1.0,0.9,0.7,0.6,0.6,? seconds.
- Note decay is fairly short, though the high noise floor makes it difficult to see / hear.
- Obviously stretched, group transitions visible, lows and mids fairly audible.
- Stretch distances: 3,4(x3),5,3,5,5,4(x3),3(x3),5,5,3,4,3,4,5,4(x3),3 = 25 groups.
- Velocity switch @ vel=80,120 quite abrupt and audible, with switch in stereo image @ 80.
- No obvious pedal down "loom of strings" or key up/down "knock" samples.
- No obvious pedal up or key up "buzzy" damping samples.
OTHER:
- Dynamic range 43dB (vel=1:127).
- MP3 levels: peak @ -0.66dB, noise floor @ -55dB & non-white.
- Date reviewed: 2010-11-15

Yamaha NP-30 Review



Piano World forum member "anotherscott" has been quite busy with this project and has provided us with another DPBSD MP3, this time for the Yamaha NP-30 - thanks again anotherscott!

DPBSD MP3: http://www.mediafire.com/?dcsopkq0wt68ad3
DPBSD PIX: http://www.mediafire.com/?vjmx99uoak31alb

I examined the MP3 and didn't see or hear anything super different from the P-85/95, so I'll direct your attention to the previous P-85 review if you want to know how either / all fared in the actual testing. For some reason the NP-30 has approximately 9 dB more dynamic range (C4, velocity = 1 to 127) than the P-85/95. Note that this isn't necessarily a good or bad thing as it is still around the middle of the dynamic ranges I normally encounter when doing this sort of testing.

When he submitted the file, anotherscott pointed out that the tone didn't sound as nice on the lowest end of the NP-30 when compared to the P-95:


To investigate this, I made two files comprised of one second sound clips of each C note on both instruments, justaposed for easy sonic comparison. One file is of note attacks, the second of note decays (identical start times, and comfortably past the attack portions).


Figure 1. Spectral Pan view of the attack portion of (left to right) note C1: P-95, NP-30; note C2: P-95, NP-30, ..., note C8: P-95, NP-30. All samples are one second duration, with amplitudes individually normalized to -1dB peak. All are highly similar looking, though the lowest NP-30 attacks sound more muffled to me than those of the P-95.


Figure 2. Spectral Pan view of the decay portion of (left to right) note C1: P-95, NP-30; note C2: P-95, NP-30, ..., note C7: P-95, NP-30. All samples are one second duration, with amplitudes individually normalized to -1dB peak. All are highly similar looking, though again the lowest NP-30 notes sound more muffled to me than those of the P-95.

For those who want to listen to the MP3 files that correspond to Figures 1 & 2 they are located here (zipped up along with a few analysis pix):
http://www.mediafire.com/?rx6zww880bx6xrc

I agree with anotherscott, there is definitely different processing of the low end going on here. I can hear a difference between the P-95 and NP-30 with notes C1 through C3, and maybe C4. It sounds rather like there is less treble in the NP-30 for these notes, and maybe something going on with the stereo image. My theory is that this is most likely fallout from some extra processing for the NP-30 to shrink the sample set size below that of the P-85/95. The lowest notes are where the most overall size reduction can be achieved (they generally dominate in terms of ROM resource).

Finally, anotherscott wanted me to point out that the polyphony of the NP-30 (32) is half that of the P-85/95 (64) - something to consider if you hold the sustain pedal down a lot.

Many thanks to anotherscott for the DPBSD file and the super helpful analysis pointers!

----------------
- Yamaha NP-30 -
----------------
FILE & SETUP:
- dpbsd_v1.8_yamaha_np-30.mp3
- Setup: Mac with ProTools LE, Mbox interface @ 24-bit, MP3 via Quicktime Pro.
- Recorded by "anotherscott".
PROS:
- Passes the silent replay test.
- Passes the quick partial damping test.
- Responds to half pedaling.
- Timbre variation is fairly smooth with increasing velocity, though most of the effect is restricted to the mid velocities.
CONS:
- No visible or audible pedal sympathetic resonance.
- Fails the key sympathetic resonance test.
- Note decay times are somewhat short (on the order of 2/3 to 3/4 Pianoteq).
- Stretching is visible over the entire range, audible over the lows and mids.
- Stretch distances: 2,3(x28),2 = 30 groups.
- No obvious pedal up/down or key up samples.
- Obviously looped, both visually and audibly.
- Attack sample lengths are rather short.
- Attack sample lengths are (C1:C8): 2.0,2.0,1.8,1.7,1.3,1.1,?,? seconds.
- Loop sample lengths are short and fairly audible over the lows and mids.
- Loop sample lengths are (C1:C8): 0.6,0.6,0.6,0.4,0.3,0.3,?,? seconds.
OTHER:
- Purportedly a single velocity layer sample set.
- Dynamic range 56dB (vel=1:127).
- Notes played @ vel=1 produce no sound.
- MP3 levels: peak @ -1.8dB, noise floor @ -71dB.
- Except for larger dynamic range it tests the same as Yamaha P-85/895.
- Lower notes sound somewhat muffled compared to the P-85/95.
- Date reviewed: 2010-11-21

Interesting stuff.

As always, thanks for your detailed analysis, and to anotherscott for providing the MP3.

Cheers,
James
x

Some may disagree, but I actually think 32 is usually just fine, but I was also concerned because sometimes polyphony is specified as "maximum" and is cut in half for sounds that use stereo samples, which would mean, in theory, this piano could have as little as 16-note polyphony, which would definitely be limiting.

So I just did an experiment. Not only is that not an issue because it clearly has more than 16-note polyphony, but I suspect that even the 32-note limit in the NP-30 spec sheet may be a typo. I hit and released the low E, keeping it sustained only with the pedal, while I continued to play other notes. First I did a chromatic run from the top down, and definitely got past 32 and the low E did not stop ringing. I then did a whole bunch of harp-like glissandos over that low E, I could not get the low E to stop ringing. So either there is some extraordinarily intelligent processing going on, or this does indeed have more than 32-note polyphony. I don't think anyone is going to hear any note-stealing, and I suspect it may have just as much polyphony (at least for piano alone) as the P-85/P-95, regardless of what the spec sheets say.

DP note stealing algorithms are wily things to catch in the act. I presume they favor lower and louder notes, hence your low E ringing on after a flurry of pedal down higher notes. You might try low E and low F together to see if the F gets stolen, or perhaps try a note or two nearer the middle while playing notes below them softly.