|
Welcome to the Piano World Piano Forums Over 3 million posts about pianos, digital pianos, and all types of keyboard instruments. Over 100,000 members from around the world.
Join the World's Largest Community of Piano Lovers
(it's free)
It's Fun to Play the Piano ... Please Pass It On!
|
|
68 members (aphexdisklavier, bobrunyan, anotherscott, AaronSF, apianostudent, beeboss, brdwyguy, benkeys, 17 invisible),
2,192
guests, and
372
robots. |
Key:
Admin,
Global Mod,
Mod
|
|
|
Joined: Feb 2011
Posts: 117
Full Member
|
Full Member
Joined: Feb 2011
Posts: 117 |
D, However.... I might just ping you once in awhile and definitely look forward to the flaming I am about to receive!!
No real talent for the Piano.
|
|
|
|
Joined: Dec 2009
Posts: 4,675
4000 Post Club Member
|
4000 Post Club Member
Joined: Dec 2009
Posts: 4,675 |
A while back I did a back of the envelope analysis of a single NAND commodity Flash part here (though I have no direct experience with the interface so grain of salt). Some followup here. For some reason I tend to get a lot of push back on this, but it's never specific enough for me to make anything of. NAND Flash seems almost a though it was designed specifically to hold DP sample sets. IMO.
|
|
|
|
Joined: Feb 2011
Posts: 117
Full Member
|
Full Member
Joined: Feb 2011
Posts: 117 |
D, started reading your Hive document. Very interesting. On the software side similar challenges exist. We use the Actor model (isolated distributed multi threaded processes), no global memory/data ( different to Hive), with isolation and avoidance of race conditions by message passing. Lots of concurrency, fault tolerance, dynamic updates etc thrown in for good measure... Erlang / OTP is my weapon of choice.
I will read the balance of your document with some interest. Thanks for the link.
No real talent for the Piano.
|
|
|
|
Joined: Dec 2009
Posts: 4,675
4000 Post Club Member
|
4000 Post Club Member
Joined: Dec 2009
Posts: 4,675 |
D, started reading your Hive document. Hive is my pride and joy. It's turned out much better than I had initially hoped, though I haven't used it for very much yet. On the off chance anyone is interested in the System Verilog code, or the Excel based simulator (unfortunately not updated to v6 yet) here is the archive: https://www.mediafire.com/folder/tm63l3qkiethp/hiveErlang / OTP is my weapon of choice. I don't know Erlang, but the arbitrary arithmetic precision seems like it could be inefficient regarding sample playback for a DP (I started a Hive sim in Python and this actually made the ALU model much more complicated). And I can't say I'm a fan of VMs in general. ===================== Making anything new (much less a DP) from the ground up is daunting. Manufacturing it, distributing it, maintaining it, etc. is double daunting. The easiest route might be to take an already existing DP with acceptable keys, rip out the brain and rubber key switches, and install the good stuff that should have been in there in the first place. But that would make everything highly dependent on the production of that particular model of DP, which could disappear at any point. And you would be paying for a bunch of stuff you end up tossing out, and who would really want to pay for all that extra labor, particularly when economies of scale have gone out the window? I see a place for the improved product in the market place, but probably not much demand at the achievable one-off reworked price point. It's too bad companies aren't generally more forthcoming when it comes to helping fellow engineers get a bit of a leg up in markets they dominate. We have similar goals and the results could easily be synergistic, but secrecy for its own sake and NIH syndrome would likely be the order of the day.
Last edited by dewster; 01/23/15 07:16 PM.
|
|
|
|
Joined: Apr 2014
Posts: 262
Full Member
|
Full Member
Joined: Apr 2014
Posts: 262 |
Need to watch out for the ARM chips as pure GHz is not the only point of measure. The family (A7, A9, A11, A15) matters but unfortunately, you can't assume larger numbers are faster. What ARM did was introduce a bunch of lower-powered families and assigned bigger numbers to them. Today's latest smartphones use A15s which are twice as fast as A9s (the above link) clock-per-clock.
Of course, the biggest problem with using an ARM chip is you can't run any of today's existing virtual piano software on them (other than recompiling free versions of Timidity or Fluidsynth). So instead of using off the shelf software (or shipping an empty box and letting users decide), you'd have to be hardware developers, software developers and audio engineers. That's quite a task to take on in your spare time.
There are similar Bay Trail products but they will be more expensive/hotter.
|
|
|
|
Joined: Feb 2011
Posts: 117
Full Member
|
Full Member
Joined: Feb 2011
Posts: 117 |
D, and MossySF,
I would stick with standard parts for now then migrate if we want / need to. The Pianoteq folks (for example) are specialists at what they do - no need to do anything in this regard but to find an execution engine (SBC) to house and run the VP.
No need to build a keyboard - will use Fatar or one of the others (very few vendors for this). OS - Linux or Embedded Windows - SBC (somewhat interchangeable so long as the power / interface logic is there).... and so on.
The general approach will be COTS - put together a plan and prototype and see what happens. The trick is integration and packaging... and function - warm / fast boot, flexibility in software supported VSTs etc.
No real talent for the Piano.
|
|
|
|
Joined: Dec 2009
Posts: 4,675
4000 Post Club Member
|
4000 Post Club Member
Joined: Dec 2009
Posts: 4,675 |
IMO, but Fatar is something of an anagram for Fail. You may need to take the bull by the horns here, though I have no clue how (beyond my own crazy notions). Physical design is a huge hurdle where everyone seems to end up punting on the keys and suffering for it. DP manufacturers are probably distinguished more by their key technology than their sound technology.
|
|
|
|
Joined: Oct 2014
Posts: 69
Full Member
|
Full Member
Joined: Oct 2014
Posts: 69 |
LoL
A software piano and a controller is no genuine substitute for a grand accoustic piano.
“To a mind that is still, the whole universe surrenders.” ~ Chang Tzu
|
|
|
|
Joined: Feb 2011
Posts: 117
Full Member
|
Full Member
Joined: Feb 2011
Posts: 117 |
leafhound, odd I tried to reply to your post but it got deleted. The general idea here is not to replace a grand AP but to have a better integration of DP components... to produce a relatively better product. Might be doable... that is what I intent to find out.
dewster, you are correct about the keys - but I suspect the more responsive the human-piano interface and the better the sound produced, the better a keyboard feels. So latency, sound, tactile response, etc. might be areas where we can improve things. A realistic goal should be set... something like "is this better than xx% of the AP pianos one is most likely to play".
No real talent for the Piano.
|
|
|
|
Joined: Jun 2013
Posts: 5,751
5000 Post Club Member
|
5000 Post Club Member
Joined: Jun 2013
Posts: 5,751 |
LoL
A software piano and a controller is no genuine substitute for a grand accoustic piano. ....But it is an alternative.
|
|
|
|
Joined: Jun 2013
Posts: 5,751
5000 Post Club Member
|
5000 Post Club Member
Joined: Jun 2013
Posts: 5,751 |
but I'm to dumb to understand such complex terminology. Evidently. Please judge the instrument after you've played it. Kind regards, James x "Dumb and Dumber to." Hello? Seriously, it was a typographical error. I should of (should have) checked for errors, but I guess it's 2 late 4 that now. I was only kidding about the CA97. I just couldn't help myself regarding the improved "black keys;" Oh, and I was pissed off -in general- about NAMM.
|
|
|
|
Joined: Dec 2009
Posts: 4,675
4000 Post Club Member
|
4000 Post Club Member
Joined: Dec 2009
Posts: 4,675 |
When all's said and done, let's not forget that the computing power of each of our DPs is probably greater than that of all the combined (and massive) computers that sent Apollo 11 to the moon. OT, but 100% awesome (DIY Apollo Guidance Computer): http://klabs.org/history/build_agc/Early computers are like Cambrian Explosion fossils - all kinds of weird stuff abounds.
|
|
|
|
Joined: Feb 2011
Posts: 117
Full Member
|
Full Member
Joined: Feb 2011
Posts: 117 |
No real talent for the Piano.
|
|
|
|
Joined: Feb 2011
Posts: 117
Full Member
|
Full Member
Joined: Feb 2011
Posts: 117 |
Came across a couple of things in doing research. The old folding VAX people are rebooting. I am a great believer in doing something, learning, and iterating. They have a new approach and are on kickstarter (unfortunately I do not think they are doing a folding portable version)... https://www.kickstarter.com/projects/1200817609/vax-midi-keyboard-controllerA New approach also originally on Kickstarter, DP is ipad with portable keyboard that can be joined for more octaves... http://www.miselu.com
No real talent for the Piano.
|
|
|
|
Joined: Dec 2009
Posts: 4,675
4000 Post Club Member
|
4000 Post Club Member
Joined: Dec 2009
Posts: 4,675 |
The VAX, while a noble project, was IMO much to gold plated to be a successful product. I really liked their continuous sensor approach, but I would have done it capacitively rather than with Hall effect devices. The folding thing was pretty stupid IMO introducing much higher cost and fragility, and magnesium vs. aluminum was weird. Here they don't have a 7 octave model, which I don't understand as 7 octaves is the sweet spot for portability with playability IMO. I would have gone with 8, 7, and 6 octaves rather than their 8, 6, and 4.
|
|
|
|
|
|
Piano
by Gino2 - 04/17/24 02:34 PM
|
Piano
by Gino2 - 04/17/24 02:23 PM
|
|
Forums43
Topics223,408
Posts3,349,457
Members111,637
|
Most Online15,252 Mar 21st, 2010
|
|
|
|
|
|