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

Bring it on!

It will be interesting to see what different manufacturers do.

I'm looking at the V-Piano manual, and it will definitely play an SMF (Standard MIDI File) on a USB flash drive plugged into the front USB connector. The instructions starting on page 64 in the manual are pretty clear on how to do that.

http://www.roland.com/products/en/_support/om.cfm?PRODUCT=V-PIANO

Download the dp_bsd_v1.3.mid file from my share directory at mediafire, copy the file to your USB flash drive, then plug the flash drive into the V-Piano. Navigate to it in the V-Piano, set the tempo to 120, and see if it plays OK. Please turn off the reverb as it makes analysis more difficult.

However, I don't see a way to record to wave or MP3 (you can play these files on the V-Piano, but probably not record them). The manual discusses ways to record and save songs but I assume these are MIDI files (?) - which aren't useful for this case.

So I think you will need to run line-out from the V-Piano to your PC, and record it there. The S/PDIF connector is the best as there is no digital/analog/digital conversion going on, but use the analog line level outputs (outputs A) if you can't easily connect S/PDIF to your PC. If connecting analog, try to get as hot a signal as you can without clipping. For the file you are recording to, specify 44.1kHz sample rate, stereo, and either 16bit PCM (good) or 32 bit float (best). After recording, save as a 192kbps CBR (Constant Bit Rate) MP3 file. If you can't save as MP3, save as a *.wav file and I can probably convert it for you.

Then put the output file somewhere that I can get at it (if necessary, create a mediafire account, it is super easy).

You should be able to play the MIDI file on the V-piano simply by copying the file onto a USB thmb drive. Try the just to see if it plays. Get that to work first.

OK you've solved half the problem. Next comes recording the sound to the computer. The intrnal sound in the MBP is good enough. You will need a cable. Use a fiber optic S/PIF if you can otherwise you will have to use "Line In".

The Mac uses a special 3.5mm round SPDIF connector. The normal cone is square. You can buy a S/PIDIF cable with the right ends but a square to round adaptor cost under $5 and will let you use a standard and cheap optical cable.

Line in uses the same 3.5mm jack (It is a combo optical/electric jack) Yu need a 3.5mm TRS splitter cable to takes the stereo jack into two mono cables. Then you connect these to the left and right line outs on the Roland. Either draw a diagram of what you need and look for the parts are Sam Ash or Guitar center (both have a good selection of cables) or call Sweetwater and just tell them what equipment you have and what you want to do.

Once you have the cables plugges in recording audio is easy. Inside GarageBand make a real instrument track, asign the input to it, click "record" then start the v-piano playing. Later you can highlight the silence in the track and click "delete. then save the file

I assume you have Garage Band on the Mac.

Ok guys, thanks. I'll see what I can do tomorrow. Steve

dewster, check you pm.

I just uploaded a Yamaha P155 recording.

One more technical suggestion for the next version of the MIDI file. I had some trouble setting the recording level. I thought I had it set right and then later far into the track I noticed clipping so I had to re-record. So the suggestion is to add something that near the front, maybe a simple chord progression that could be used for settig the level. It should be the loudest part of the file. You could set the velocity to 127 but you'd have to guess which notes have the loudest sound, I'd guess the bass. or cords with the damper pedal down.

I used Garage Band to make this. Left everything at it's default setting and the only equipment I use was a cable to plug the P155 directly into the line-in in the Mac. GB wants to "normalize" the volume on any MP3 file it exports. But all this really does it set a constant in the MP3 file header.

The file is at BSD_1.3_P155.mp3

Hi Dewster,

Kawai MP-5 is here:

http://www.box.net/shared/oc1cxg9pmz

Excellent suggestion!

Yes, that is a good idea to put the highest level up-front in the recording.

ChrisA, I took a quick look at the P155 and I must say that Yamaha is doing a very good job of blending their velocity layers, I can't hear or see them, and timbre variation seems good with velocity. Typical sample lengths (<1 sec to ~3 sec) and stretching for Yamaha, though the looping is a bit better than most of what they've done in the past. Nice levels in the recorded file.

Rob, this is another one of those linear phase files, my first from a non-PC based sampler, which makes analysis more difficult. Also has nice blended timbre variation with velocity. I clearly hear stretching and sample lengths in the neighborhood of what Yamaha tends to do. One strange thing is the G7 note, which has so many harmonics it sounds more like G8!

ChrisA, and Rob, thanks much! I'll get on these in-depth tomorrow morning.

http://www.mediafire.com/file/gmjnzdtmm1z/01 V-Piano DP BS Detector.mp3

Dewster,

Above is location of the file on my shared mediafire account.

I tried using garageband but I am useless with technology. There is no optical digital in on my MacBook Pro because it is the 13" version.

In the end I recorded using the V-Piano's coaxial digital out to my stand-alone hi-fi CD recorder. The only problem with it with a digital input is that you cannot adjust rec level and it is slightly quiet. At least it was in the digital domain throughout. I uploaded it from CD into iTunes as a 192kbps mp3 file. I hope you can work with the file. I'm interested to see how the V-Piano performs in your test.

Cheers,

Steve

Thanks again for that file ChrisA!

I reviewed it this morning, and, other than the nice seamless blending of the velocity layers (I couldn't see or hear velocity layer transitions), there are no obvious technical improvements over the P-120. The samples themselves are the same old short 3 seconds max, looped & stretched. Sympathetic resonance sounds unimpressively identical to the P-120.

I'm quite frankly rather shocked at this. And I must admit this is starting to seriously damp my hopes for the new CP1/5/50 products.

One strange thing I notice is the note G7 (second-highest G) sounds strangely like G8! I hear this in my VintAudio sample too. Very odd.


----------------
- Yamaha P-155 -
----------------
FILE & SETUP:
- dp_bsd_v1.3_yamaha_p155.mp3
- Garage Band.
PROS:
- A very smoothly blended 4 layer (reportedly) sample set with no detectable layer switching.
- Huge dynamic range (~62dB, vel=1:127).
- Sympathetic resonance.
- Responds to partial pedaling.
CONS:
- Obviously looped, though not too badly done as these things go.
- Samples are 3 to 1 seconds (lo to hi) with 1.5 to 0.5 seconds of crossfade to loop.
- Somewhat fast decay of notes.
- Obviously stretched.
- Stretch distances: 3,3,4,4,4,2,4,3,3,3,2,3,3,3,2,2,2,2,3,1,3,3,4,2,2,3,5,10 = 28 groups.
- No key-up or pedal samples.
- Sympathetic resonance fairly fake sounding - echoy and reverby.
- C9 sounds a bit flat and short.
- G7 so bright it sounds more like G8.
OTHER:
- Volume in MP3 file good.
- Date reviewed: 2010-02-08

R0B, the file is in mono for some reason. Any chance you can get me one in stereo?

Flat-line phase plots = Mono recording

Please ignore every idiotic comment I've made about linear phase. All of the flat-line phase plots were due to mono recording.

I accidentally recorded my VintAudio C7 in mono, and the Windows Media Player piano is a mono sample. The VintAudio sample has been re-recorded in stereo & updated in the archive.

Bright G7 = Headphone resonance?

Please also ignore my earlier complaints about G7 being bright, I think this is due to my headphones.

Hi Steve,

Thanks for that! I know it was a fair amount of effort on your part.

Nice to hear what the V-Piano sounds like up-close. Strange that there are velocity discontinuities in something modeled, I wasn't expecting that. And the sympathetic resonance is unfortunately very weak sounding. But it beats all sampled DPs I've heard so far, and is on par with PC samplers.

------------------
- Roland V-Piano -
------------------
FILE & SETUP:
- dp_bsd_v1.3_roland_v-piano_01.mp3
- Recorded direct to CD recorder.
PROS:
- Beautiful long note decay.
- No looping (modeled).
- No stretching (modeled).
- Sympathetic resonance.
- Responds to partial pedaling.
- Good dynamic range (~34dB, vel=1:127).
CONS:
- No key-up or pedal samples (could be down in the noise floor of this sample?).
- Something like velocity layer switching (abrupt timbre changes) going on at vel=28,40,66,74,100.
- Sympathetic resonance effect is very subtle, almost undetectable.
OTHER:
- MP3 file recorded a bit low, equivalent to ~14 bit resolution.
- Date reviewed: 2010-02-08

Thanks Dewster, very interesting. My ears may not be quite as sharp as they used to be but I can time the sustain in the lower registers to 40+ secs after a good slap on a key. The pedal sample is there (if you mean the little whoosh noise when it is depressed and returned...the level of it is adjustable, I have it fairly discrete), but you might not be hearing it due to the quietness of the recording. Likewise, surprisingly most presets in the vintage section of the V-Piano have sympathetic resonance set pretty low, although you can have as much as you like. I had initially recorded the BS detector in Garageband but through analogue input at the perfect volume but then inadvertently deleted it...The Mac nearly went out the window at that point!

Cheers and thanks on behalf of us all for the work you are doing with this project.

Steve

Sorry you had trouble Steve - this stuff is kind of easy once you set it up, but getting it set up in the first place might drive you crazy if you run into any HW/SW issues.

If you or anyone else wants to submit what might be a better sample, I'd certainly take a listen to it and update the review. I'd like to hear what the key and pedal samples sound like, and I am always very interested in what the sympathetic resonance sounds like.

Interesting points have been raised by the V-Piano test.

1) the DEFAULT Grand Piano 1 (whatever that means!) patch should be used - not a version that has been edited by the user. In the case of V-Piano, I guess Vintage Piano 1 might quality. (However, in terms of the sound quality aspect, V1 Impactance (with the hammers softened) sounds MUCH better.

2) V-Piano's output should be boosted. The default setting is very conservative.

Utility / Output Gain / (I have mine set at +9dB) Experiment to determine what settings work best.

Lawrence

Dewster -

Anything in your tests to suggest that there is any sampling going on in the V-Piano? (A suggestion that has been made by some).

Lawrence

What headphones are you using?