Get started with Raspbian Jessie and RPi2

Posted on by

The Raspberry Pi 2 Model B (RPi2) is a computational gem. For $40 USD, you get a 900MHz quad-core ARM processor, a built-in graphics core, 1GBytes of RAM, 4 USB ports, HDMI, an Ethernet port, audio output and a Micro SD card slot. (The RPi2 does not have a built-in audio input.) This platform can handle most of the every-day tasks that people can sling at it and could easily replace platforms costing 10 times as much. Add in the cost of a keyboard/mouse, display and Micro SD card, and the total package price tips the scale at a little over $100.

When the RPi2 was introduced in February 2015, the Raspberry Pi Foundation released Raspsian Wheezy, the first Linux distribution supporting the RPi2. I installed the first release last February, and it felt, well, kind of wheezy.

I just installed the latest release, Raspbian Jessie (February 2016) and all I can say is “Wow.” This release feels finished. If you tried Raspbian on RPi2 before and were disappointed, it’s time to come back into the fold. (Download Raspbian.) This is the release that should have been there at the RPi2 launch! (See a quick introduction to the Jessie desktop.)

With a foot of snow on the ground, it seems like an opportune time to see what RPi2 and Jessie can do for musicians. I intend to try the RPi2 as a synthesizer and will post my experiences here. In the meantime, here are some tips for getting started with RPi2 and Jessie.

Linux requires just a little more work to get started than Mac OS X or Windows. However, if you put in just 10 or 20 minutes, you can have a quad-core music making machine for cheap. Shucks, even OS X and Windows 7 need to know your account name, etc. and Jessie doesn’t require much more information than that. So, what follows is my personal checklist for getting started with Raspberry Pi 2.

Hardware

Of course, you need the hardware. I bought a Canakit Raspberry Pi 2 kit last year. The Canakit includes most of the accessories that you need to get started. I imagine that future Canakit’s will include the latest Raspbian release (Jessie) pre-installed on Micro SD card.

My RPi2 lives in a cheap plastic case. That’s good enough — no fans, no heatsinks. I use an old HP monitor with an HDMI input and an even older Logitech wireless keyboard and mouse. The Logitech wireless interface takes up a single USB port, leaving three open USB ports. I connect the RPi2 Ethernet port directly to our router since I like to have the network up and running right from the start. The Canakit package included a USB Wi-Fi interface, but I never felt motivated to bring it up. Cables work good, too.

Once you have everything connected, it’s time to move on to software.

Download and install

Since our house is littered with computers, I first downloaded Raspbian Jessie to a Windows 7 PC. I followed the installation guide for Windows and wrote the disk image to an 8GByte Micro SD card. I do not recommend using anything smaller than an 8GByte card since you will need room for all the applications, samples and stuff for your projects.

The installation guide recommends the Win32DiskImager utility from Sourceforge. This utility works like a champ. Just be sure that you write to the correct destination disk!

If you’re installing from Linux or Mac OS X, there are installation guides for you, too. I do not recommend upgrading an old Wheezy system to Jessie. I read through the process and it is far easier to do a clean install.

First and second boot

Plug the Micro SD card into the RPi2 and apply power (i.e., plug in the power supply). It takes a few moments until the RPi2 boots the kernel (AKA “the OS”) and starts the X Windows system. The stock RPi2 boots into the desktop. The default user name is “pi” and the default password is “raspberry”.

At this point, it’s important to get a few housekeeping tasks out of the way. These tasks are similar to the ones you need to perform after installing OS X or Windows. These tasks use the raspi-config utility.

First, launch a terminal window by clicking on the terminal window icon in the task bar at the top of the screen. This action brings up a textual “shell” where you enter commands. Enter:

    sudo raspi-config

to launch the raspi-config utility. The sudo tells Linux that you want to use administrator (“super user”) privileges when running raspi-config. Linux will prompt for a password. Enter “raspberry”, the default password for the default user, “pi”.

raspi-config displays a rather 70s-ish interface with several options. Use the arrow keys to move between items. Use the ENTER key to select an item.

The disk image which you wrote to the Micro SD card probably didn’t use all of the available space on the card. So, your first job is to extend the Linux file system to use the entire Micro SD card. Use the arrow keys to move to the “Expand filesystem” item. Hit ENTER. When prompted to reboot, choose OK. You need to reboot to get access to the full capacity of the Micro SD card.

After reboot, you should be back in the desktop again. Launch a terminal window to get the shell. Enter sudo raspi-config to perform a few more housekeeping steps related to internationalisation. Use the arrow keys to move to the “Internationalisation options” item and hit ENTER.

It’s called “Internationalisation,” but it’s really configuring your RPi2 to your local country or region. raspi-config displays a short list of options. Choose the “Change timezone” option, follow the on-screen directions, and set your local timezone.

Next, choose your locale. The locale controls language and formating of date, time, currency, etc. The default locale is set for Great Britain. If you’re in the United States, for example, select one or more locales for the US, e.g., “en_US.UTF-8 UTF-8”. The text interface is a little weird here — use the SPACE key to mark one or more locales and hit ENTER when finished.

Then, change the keyboard layout. Follow the on-screen directions to find a close match for the keyboard that is connected to your RPi2. When you see questions about “compose key,” etc., fake your way through the menus. You probably won’t be doing this stuff anyway.

Finally, reboot. Rebooting the system at this point makes sure that the locale and other internationalization changes kick in.

Explore and browse

After reboot, the RPi2 should again return to the desktop. Now it’s time to explore the desktop a little bit. I recommend taking a tour through the start menu in the upper left corner of the desktop. When you find a menu item for the browser, try it out! If all is good with your network connection, then you should be able to access the Web. This is my lifeline to helpful information about Raspbian Linux and the many tutorials and HOW-TOs on the Web.

Also, check out the File Manager. This is a graphical way to browse through your files. Linux uses a hierarchical file system where absolute path names begin from the root, which is symbolized by the slash (“/”) character. More about this in a minute.

Just a few more things

I recommend creating a new user for yourself and keep the default user “pi” around for emergencies or administration. The Raspberry Pi folks have a nice introduction to user management. It’s a short read and now that you have the browser running, why not?

If you’re too lazy to read the guide, then use the following command to create a new user:

    sudo adduser XXX

where “XXX” is the name of the new user. The system prompts for the new user’s password. This part is up to you! You can remove the password for a user by entering the command:

    sudo passwd XXX -d

where “XXX” is the name of the user. The passwd command can be used to change your own password, too. If you want to remove a user, then run the command:

    sudo userdel -r XXX

where “XXX” is the name of the user to be removed.

The guide to user management describes “sudoers” and how to grant permission to a user to execute the sudo command. This process changes an internal user privileges file, so you must be careful. Enter the command:

    sudo visudo

and find the line:

    root	ALL=(ALL:ALL) ALL

Create a new line using this line as a model, except replace “root” with the name of the user who is to be a new sudoer, e.g.,

    XXX		ALL=(ALL:ALL) ALL

Save the changes and exit the editor. Oh, the editor here is nano, which is one of the pre-installed applications.

Users have their homes in the directory “/home”. If the user’s name is “XXX”, then their home directory is named “/home/XXX”. Here are a few commands that you can use to navigate through the file system via the shell.

    ls          List directory
    cd          Change the working directory
    pwd         Display the current working directory
    mkdir       Make new directory
    rmdir       Remove directory
    cp          Copy file (or directory)
    mv          Rename a file (or directory)
    rm          Remove file
    cat         Display file contents
    more        Display file contents
    nano        Edit text file
    date        Displays the current date

If you need help, you can always enter the desired command with the “–help” option. Or, you can display the manual page for the command, e.g., “man ls”. All of these commands have many options, making them quite flexible and powerful.

You can find more basic information about using Linux on this page.

Install applications

Speaking of new applications, you can install a new application from the command line if you know the package name. This is the most straightforward way to install a package (application). For example, I like the emacs text editor. The following command installs emacs:

    sudo apt-get install emacs

The apt-get command searches for the package on-line, downloads it and installs it. The command also installs any other packages which the target package needs (e.g., libraries). In Linux terms, it resolves dependencies. Installation is sometimes slow, so please be patient.

There is also a package manager with a nice user interface. Find the package manager in the desktop start menu and browse through the available applications. I’ll revisit this subject in future posts when we discuss specific music-oriented applications.

USB flash drives

The need for a USB flash drive is sure to come up. I recommend this guide to adding and using USB drives. Here are a few quick commands for reference. Create a mount point:

    sudo mkdir /mnt/usb_flash

Mount the USB flash drive (after inserting the drive):

    sudo mount -o uid=XXX,gid=XXX /dev/sda1 /mnt/usb_flash

where XXX is your user name. Unmount the flash drive when you’re finished using it:

    sudo umount /mnt/usr_flash

You can display the currently mounted file systems with the command:

    df -h

This command also shows the amount of used and available space in the various file systems and drives.

Raspbian Jessie is smart enough to recognize when a USB drive is inserted. It displays the File Manager automatically. If you are a File Manager type of person, definitely go this route. You must eject (unmount) the drive before removing it. The EJECT button appears in the upper right hand corner of the desktop.

Boot to a login shell

Raspbian Jessie boots into the desktop as the default user “pi”. You probably want to boot into your own account instead. At the moment, you need to dig into some system files to make the change and I simply don’t recommend diving into that, especially if you are new to Linux.

Instead, you can easily change the boot behavior using raspi-config. Launch raspi-config and choose the “Enable Boot To Desktop” option. Then, choose to boot to the command line. The next time you boot, the system will display a login prompt where you can enter your user name and password. Once your identity is validated, Linux puts you directly into a command line shell. If you want to, you can enter any Linux command into the shell and do some work. When you want to start the X Windows system and the desktop, just type:

    startx

Read the on-line documentation about respi-config for more information.

Shutting down

All operating systems like to shut down in an orderly way. OSes often times keep data in temporary buffers that need to be flushed to disk or flash memory. An orderly shutdown helps keep data in a consistent, correct state.

You can shutdown the system through the desktop start menu. (Yeah, that sounds oxymoronic.) You can also shutdown the system via the command line shell. Just execute the command:

    sudo shutdown -h now

The -h option asks Linux to halt the processor after shutting down. The shutdown command has other options for rebooting and so forth:

    sudo shutdown -r now

Here’s another way to force a reboot. Just enter:

    sudo reboot

on the command line.

If you enjoyed this introduction, you might want to check out the Raspberry Pi tips and tricks page that I wrote for the first generation Pi.

Well, that wasn’t so bad, was it? Good luck and have fun with Raspberry Pi 2!

All site content is Copyright © Paul J. Drongowski unless otherwise indicated.

NAND flash data retention

Posted on by

In my previous post, I raised the issue of NAND flash data retention. Even though vendors describe NAND flash memory as “non-volatile,” data can and will eventually be lost. The issue is “When?”

If this question peeks your interest — and it should — please read the Spansion Practical Guide to Endurance and Data Retention . This is a very well-written, easy-to-read white paper.

Basically, reprogrammable non-volatile memory is rated for endurance and data retention. Endurance is the number of sector erase operations that may be performed before failure. Erase operations are performed on a sector basis (i.e., not per-byte or per-bit.) Software can alleviate this problem by distributing erasures evenly across sectors, marking out failed sectors, data refresh, etc.

The Spansion 8Gbit NAND flash device (Spansion S34ML08G1) is rated for 100,000 erasures. Fortunately, we don’t write (and thereby erase) locations in wave memory very often. Further, we don’t write (erase) locations in voice, performance, pattern, etc. memory very often either. (Please don’t forget that user data is often stored in reprogrammable non-volatile memory, too!) As given in the table in the Spansion article, you could write (erase) a sector every 53 minutes (average erase frequency) or 27 times per day and still attain a 10 year device lifetime.

Data retention is the period of time for which data are reliably retained and retrieved. Data retention is affected by “temperature and voltage, electrostatic environment,
exposure to radiation, cumulative erase cycles, etc.” That’s right, data retention decreases with the number of erase cycles, too. Erasure introduces minute defects into memory cell structure and these defects accumulate. Fortunately, again, we’re looking at relatively low erasure frequency for wave and user data memory.

Quoting the article, “Spansion single-bit-per-cell floating-gate flash devices are designed to provide 20 years of data retention after initial programming when exposed to a 55°C environment.” Data retention time is 10 years when the cumulative erase cycles per sector is 1,000 erasures or less.

So, should you worry that your synth or arranger workstation will lose its contents? Probably not. I would worry more about pressing the wrong button at the wrong time and accidentally losing my work! (You are backed up? Right?) However, as a manufacturer, I would definitely anticipate some data failures in the long run and have a means to restore original factory programming through a field service program. The PSR/Tyros products, for example, do a quick wave memory self-check at start-up. One or two such failures have been reported on the PSR Forum. Presumably, the memory devices can be replaced and reflashed by a qualified service technician.

Sleep tight, but don’t forget to back up your data. Overall, you are your own worst enemy when it comes to data loss!

Montage wave memory

Posted on by

Folks are speculating about the wave memory in the new Yamaha Montage. Without the actual service manual in hand, it’s impossible to be definitive. However, I think it’s reasonable to assume that:

  1. The Montage uses the new SWP70 tone generator, and
  2. The wave memory interface is the same as the PSR-S970.

Here’s a few details about the SWP70 and wave memory interface in the PSR-S970 arranger workstation. If you buy into the two assumptions above, then these details should apply to the Montage as well.

I realize that my earlier posts dive deep and cover many aspects of the SWP70. This blog post concentrates on a few specific aspects of the wave memory interface in the PSR-S970 instead of the whole she-bang.

The SWP70 has two 8-bit memory data ports — HIGH and LOW — and a common set of wave memory control signals. The interfacing standard is the Open NAND FLASH interface (ONFI). One flash memory device plugs into the HIGH port and a second flash memory device plugs into the LOW port. The two flash memory devices share the control signals, that is, the same control signals are routed to both memory devices.

SWP70_wave_memory

The PSR-S970 memory devices are Spansion S34ML08G1 8Gbit NAND flash memory devices. The S34ML08G1 is a dual-die stack of two S34ML04G1 die. Spansion currently produces the S34ML16G2, which is a quad-die stack of four S34ML04G1 die.

Thanks to ONFI, the 16Gbit (2 GByte) S34ML08G1 is pin compatible with the smaller S34ML08G1. Thus, a tone generator complex with twice the wave memory capacity can be built in the same printed circuit board (PCB) footprint.

The ONFI bus is not the same as the old flash expansion memory DIMM interface as provided in the later Tyros and Motif/MOXF products. The DIMM expansion memory interface consists of two, full-parallel memory channels with separate address and data signals for each channel. An ONFI memory device, on the other hand, has a single bi-driectional (tri-state) data port. The memory address, data and control information are sent to the memory device in byte-serial fashion. (The bus is time-division multiplexed.)

The tri-state electrical interface supports expansion by plugging multiple memory devices onto the same 8-bit multiplexed bus. The control signals and protocol choose the device that drives (or reads) the tri-state bus at a given time.

Yamaha may not have found a convenient way to make the ONFI bus user-extensible. Or, Yamaha have simply decided to not provide end-user wave memory expansion in the field. Yamaha accrue several benefits by dropping the DIMM expansion slots:

  • The cost of the DIMM connector(s) is eliminated including the cost of mounting and testing the connectors.
  • PCB size is greatly reduced.
  • The access cover and chassis hole are eliminated.
  • The cost of stocking another part/SKU is eliminated.

The disadvantage to the end-user is “All the sample space you get is built-in right from the start and no more.”

Yamaha’s new approach to user waveform memory is to reserve space for user samples in the physical wave memory. In other words, the user expansion memory is contained in the same physical package (48-Pin TSOP 12mm x 20mm x 1.2 mm) as the factory waveforms. The Montage specifications describe wave memory as:

Preset: 5.67 GB (when converted to 16 bit linear format), User: 1.75 GB

The compressed factory waveforms occupy 2.835 GBytes of physical memory (assuming a 2-to-1 compression factor). Compressed user waveforms require 0.875 GBytes of physical memory. These figures point toward a 4 GByte physical wave memory size, which would reserve some space for Yamaha’s own future use. BTW, if the actual effective compression factor is higher, then user samples could be stored uncompressed.

For reference, here is a terse summary of the Spansion S34ML08G1 device that is used in the PSR-S970:

    Spansion S34ML08G101TFI000
    Density: 8Gbits (4Gbits x 2)
    Random access: 30us (max)
    Sequential access: 25ns (Min)
    Block erase time: 3.5ms
    Program time: 300 us (typical)
    Data retention: 10 years (typical)
    100,000 program/erase cycles (typical)
    Pricing: $7.84 USD (quantity 250 up)

The data retention time (10 years) should raise a few eyebrows. NAND flash is volatile and charge (data) is eventually lost unless it is refreshed. I wonder how many manufacturers have planned for the day when keyboards, phones or whatever lose presumably “permanent” data stored in flash? Mask-programmable ROM never had this problem… I don’t think Hank done it this way.

Update: Read more about NAND flash data retention.

All site content Copyright © Paul J. Drongowski unless otherwise indicated

Nord Stage 2 ex: Test Drive

Posted on by

The Yamaha Montage announcement got me thinking about the kind of “all-in-one” keyboard that I would like to play. I still enjoy playing my Nord Electro 2, but the NE2 falls short as an all-in-one. My all-in-one needs to be strong in B3 organ, pipe organ, acoustic sounds, and to a lesser extent, electric pianos. Ideally, the action would be a waterfall keyboard or a good quality “synth action” keyboard. I do not need the weight or expense of a hammer action keyboard. And speaking of weight, the all-in-one should be as far under 20 pounds (about 10 kg) as possible.

The current Nord Electro 5d has gotten very favorable reviews. As one would expect of Nord, it is one of the leading clonewheels, has very good electric pianos, and plays back sampled acoustic instruments from the Nord Sample Library. The 5d has a waterfall keyboard, sliding drawbars, and a nice clear OLED display. The 5d can layer and split voices with a few limitations. Finally, musicians can create sample-based voices of their own using the Nord Sample Editor.

Looks great and the on-line demos sound good! Now, where can I find one to try? This is a dilemma faced by many musicians today and it’s not only trying to find Nord products on display in store. Brick and mortar stores cannot afford to keep a wide spectrum of keyboards on the floor just in case someone feels the urge to try out a new ax. Keyboard sales are not that hot — guitars out-sell keyboards by 5 to 1 when measured in dollar sales volume. Plus, pro-level keyboards are expensive and that’s a lot of money to tie up in inventory.

Fortunately, the nearest GC (the store whose name I dare not speak) had a Nord Stage 2 ex 88 on the floor. So, I grabbed my audition folder and took a drive. I’m glad that I did. (Wednesday night at 8PM is a good time. No shredders and head cases.)

Most NE5 reviews focus on the clonewheel and electric piano sounds. Nord Stage reviews put the synthesizer section to the test, too. My review is different because I decided to concentrate on the quality of the sampled acoustic instruments. One leap of faith is needed: the acoustic instruments on the Stage are not doctored up by the synthesizer when compared to the NE5. Still, a favorable response to the Stage has encouraged me to look for an NE5d to try, possibly by going to the downtown Boston store. (A day trip for me.)

I scrolled through the Stage’s presets and pulled an appropriate lead sheet from my audition folder whenever I found a voice that I wanted to try. I played mainly hymns and liturgical service music from our repetoire: contemporary hymns, traditional hymns, gospel hymns, etc. Yeah, some B3 got in there. I’m weak.

Without being long-winded, here’s a quick rundown.

  • The handfull of pipe organ sounds (big church and chapel) are pleasing and useful. The big cathedral sounds are not overdone, one of my biggest complaints with typical synth “church organs.”
  • Strings? You got ’em. Big, small, sections, solo. The majority of the string voices are very playable. Big strings that are rich without getting screechy in the high end.
  • The orchestra brass ensembles are generally darned good. The trombone section is too loud and brash for church. Softer French horn voices are needed, too. The few horn voices are borderline bright and loud — I need mellow. The pop brass ensembles sound terrific. (“Knock On Wood,” anyone?)
  • Woodwinds, too, are a mixed bag. The woodwind sections are good and playable. The orchestra solo winds (except the flute) are terrible. If I bought an NE5d today, I would cobble together my own solo oboe and clarinet. Although it wasn’t a focus, I played one sax patch that was pretty decent and I wouldn’t be embarrassed to play it in public.
  • B3. Nord groovy as usual. The B sounded darker compared to my memory of the NE2. The Stage has the fast/slow switch on the left where it should be. Nord needs to make the switch BIGGER as it is really difficult to find and hit. (I switch speeds via foot pedal normally, so this is a minor niggle.)
  • Electric pianos, thumbs up.

The Nord Stage 2 ex 88 has a hammer action keyboard. I was pleasantly surprised to find it easy to play organ with this action. The keys did not cut my hand when doing palm swipes and I didn’t have too much trouble playing with a legato touch. Nice work, Nord.

You might reasonable ask, “Why use sampled pipe organ when the NE5d has modeled pipe organ?” The modeled organ solely consists of principal pipes. I think I could use the modeled organ to lead congregational singing as principals are a clear, supportive voice. However, after listening to the demos, the principal pipes alone get “same-y, same-y” fast. I hope Nord continues their work on modeled pipes as the current implementation needs a more varied sound (e.g., reed voices, and so forth).

Overall, the Nord Stage 2 88 left me with a very favorable impression. Despite the shortcomings mentioned above, the acoustic instruments are pro-quality and suitable for liturgical music. I will seek and find a Nord Electro 5d for trial. It’s worth the effort. The Nord Stage 2 ex Compact (73-key waterfall) has a street price around $3,600 USD. The Yamaha Montage 7 (FSX action) has a street price around $3,500 USD. I see a shoot-out on the horizon…

What you might have missed…

Posted on by

Here’s a few small items that you may have missed in the deluge of Montage-related videos and forum comments.

Stephen Fortner — former editor of Keyboard Magazine — shot an interview and demo with Yamaha’s Nate Nate Tschetter back in December 2015. It’s a sneak peak. Here’s the link:

Montage sneak peak with Keyboard Magazine

Dave Polich is a sound programmer and musician who has contributed voices to many keyboards including the Motif series and now Montage. Dave made an interesting post to the Gearslutz.com site:

I’ve had one in my studio since January of 2015 (because I’m on the Yamaha sound design team).

It sounds great. It does not sound “thin”. You can make it sound thin with EQ or filters if you want. It’s very loud, about twice as loud as the Motif XF. Bottom end is warm, full, midrange is punchy and present, highs are detailed and clear. It’s the only true 8-operator FM synth on the market now. it is not the old DX7 style of FM sound, but very high-definition, the engine is similar to the FS1R but no formants, however, you do have more than just sine-waves as carriers/modulators, every operator can have resonance and its own amplitude envelope, plus you have global filter and amplitude envelopes available for an FM “part” (which is the same as a “voice”, or single sound, in the old Motif series), as well as DSP effects. There are loads of new samples. Effects are the usual superior Yamaha quality. The motion control features are insane. It has an envelope follower that works. Do I care about the lack of sequencer? Absolutely not. I just like a synth that sounds good. Montage is a true synthesizer (FM) with sample playback…great combination.

The factory sounds are a good balance between bread-and-butter and EDM. The electric pianos feature adjustable balance between the main tine sound and the mechanical noise. The organs feature adjustable overdrive, leakage, chorus/vibrato and percussion. Yes there is a 9-drawbar B3. Many of the sounds are hybrids of FM and sample playback, including string and brass sounds. There are FM bell and percussion sounds, guitars, sitars, dulcimers, basses, electric pianos, clavs, and tons of FM pads, soundscapes, and edgy EDM sounds too.

Don’t judge anything based on YouTube audio, that audio sucks. SoundCloud audio sucks. The only way you should hear it is in person.

I hope that Dave doesn’t mind that I quoted him here. He is a knowledgable, reliable guy.

In a separate comment, Dave mentioned that the sounds have not yet been finalized. I guess we won’t see the Data List any time soon! That might also explain why the Yamaha demonstrators seem to rely on the same Performances…

A Montage of quick thoughts

Posted on by

Video demos

Finally had a chance to watch a few of the longer video demos and interviews with Blake Angelos (Yamaha). He’s got that nice guy from the Midwest charm and is a heck of a player.

I’m beginning to get the idea of macro control and the Super Knob. One needs to understand the flow of values from the physical controllers, through the Super Knob, to the destination parameters within the synthesis engines. (That’s what the complicated looking flow diagram is all about.) From Blake’s comments, up to four parameters can be assigned per knob or slider. The eight knobs/sliders are routed through the Super Knob — the macro controller — to the synthesis engines. When Blake spun the Super Knob, the knob and slider indicators changed, too. This is going to require some study.

montage_mc_diagram

The developers simpified the user interface by removing SONG, PATTERN, VOICE and MASTER modes from the Motif. Now there is just Performance mode. (“Don’t mode me in.” Larry Tesler) You can still edit the voices within the context of a Performance. It’s possible to operate the entire instrument from either the touch screen or buttons. The latter is an important capability for vision-impaired players.

Much has been made about the sequencer. This is part of the simplification, too. The Montage sequencer is similar to Performance Record in Motif/MOX. It’s good enough to lay down a basic 16-track song. Then, transfer the song to a DAW for the real micro-editing. This is my workflow already, so I’m not complaining. I presume that arpeggios operate and record in the same manner as Motif. More to study now that the Montage Owner’s Manual is available.

Blake mentioned improved strings, organs, pianos, and choirs. He played a really nice FM-X voice — American Garage — that sounds like a natural jazz guitar. If this is the new age of FM, then count me in!

The Montage8 is physically ginormous and weighs 63 pounds. I cannot possibly move such a large instrument by myself anymore. I’m more interested in the Montage6 which tips the scale at 33 pounds. Still kinda heavy for me…

Sound-wise, I’m still forming a preliminary opinion. The acoustic piano has terrific dynamics from soft to loud. (Producers who compress the snot out of a song in search of “loudness” probably don’t care.) The motion control examples are appealing: changing between string section size, orchestral brass swells and morphing through choir, string and organ layers.

For people complaining about “thin sound” or whatever, please give it a rest until you have a chance to play a Montage in person through decent monitors. The videos, etc. all suffer from crap compression. Kind of ironic because…

Pure Analog Circuit (PAC)

PAC seems to be the least understood and appreciated feature. Essentially, Yamaha paid careful attention to the design of the post-DAC circuitry. Here is a quote from the PAC developers:

During the development of the MONTAGE hardware we focused on improving the performance of the audio output circuitry. We devoted a great deal of time and effort to its design, seeking advice from a variety of people as we built prototypes and conducted audio tests in a repeating cycle that let us establish the direction we wanted to take. The resulting circuitry delivers more natural harmonic content and a greater clarity in the mid- to high- frequencies, allowing us to realize a distinctively crisp, clear tone. Reviewing the pattern used on the printed circuit board produced enhancements that led to a better signal-to noise ratio, which, in combination with the low noise connections from the MONTAGE (made possible with a TRS balanced audio output circuit) improves the sound conveyed to the listener still more.

PAC demonstrates considerable attention to detail and analog audio quality.

The PAC engineers in the developer interviews mentioned that the “analog power supply for the DAC was the real key to the sound quality.” A clean signal begins with clean power. From the quote above, they carefully designed and analyzed the signal traces (wires) on the printed circuit board to eliminate stray signal effects like crosstalk. The engineers also specified high-grade capacitors (“condensers.”) This is old school engieering!

I’ve heard and seen a lot of schmutz at the outputs of various keyboards. Manufacturers, you know who you are. Higher-end Yamaha products are dead quiet. (PSR-E443 needs help, tho’.) I hope Yamaha keeps an eye on its suppliers, especially if they contract with China. Vendors get tempted to replace high-grade components with cheaper components of dubious quality. Apple got burned this way with the first batch of iMac G5s. Vigilance.

Platform

One simply cannot forget that there are real breathing human beings behind these products. Please be kind.

The developer video gives the impression that the engineers were given the time to make the instrument right. The engineers may disagree with this perception, of course. 🙂 Viewed through the lense (mine) of an engineering manager, the group seemed a little too “ponderous.” I doubt if an American manager would have had the patience to give the group time.

Given the relatively long time between product releases, Yamaha needs to view its flagship products as platforms which evolve via periodic updates. The Motif XF version 1.5 update, for example, was a good product extender. I recommend annual updates at the very least. These products are enormously expensive and Yamaha cannot risk good will by making customers wait for major product generations (e.g., XS to XF) to get improved functionality and by making customers buy a whole new keyboard on top of it! Many customers want Montage to step up to the Kronos, for example, and an evolving platform strategy might let Yamaha get there with modest resources.

Peace

I’m continually shocked by some of the negativity in Web forums. Debate over features is acceptable; personal, derogatory comments are not. How can we address war, poverty, hunger, homelessness and disease if we cannot even accept another person’s style of music? Or merely their age?

Please, let’s strive for peace.

Site content is Copyright © Paul J. Drongowski unless otherwise indicated.

Yamaha Montage: It’s here

Posted on by

Here are a few links to start your day and mine. The best new resource on the Montage is the latest issue of the Yamaha Music Production Guide.

Well, we have to get the question of money out of the way sometime:

    Montage 6    $3,499.00  USD MSRP
    Montage 7    $3,999.00  USD MSRP
    Montage 8    $4,499.00  USD MSRP

One major on-line retailer is listing $2999.99, $3499.99 and $3999.99, respectively. The street price will probably remain high until the current Motif XF inventory clears. The Motif Fully Loaded offer is good to the end of March 2016. Yamaha is bundling the Bösendorfer Imperial Premium Grand Piano sound library as an introductory promotion to drive early adopters. Oh, Montage will ship in May 2016.

But, wait! There’s more! “Be a Montage Early Adopter and get a Free Montage Motion Control Pedal Pack!” All of the details are on this page at the US Yamaha site.

Yamaha also announced a Bösendorfer 280 VC concert grand piano (MSRP: $219,999). I don’t think Yamaha is bundling one of those with the Montage. They better fix up some of the ambiguous language on their Web pages.

Sud Claviers post some of the best demo videos around and were the first ones out of the blocks this morning.

Operational Overview
Acoustic sounds
Synthesizer sounds

Hope you understand French!

The demo’ed acoustic sounds are named: CFX Concert, Seattle Section, Seattle Violin, Organ AMP, B3 Curved, Folk Guitar, Electric Guitar, Modern Guitar, Choir Ensemble, Drum & Bass, Trumpet (muted), and saxophone (untitled). The synthesizer sounds are named: Side Chain, Journey Within, EP Goes, Tech FM, FM Lead, Universe, Couleur Life, 4 Pads, Soft Pads, FM Pads and Astral. A little French might have slipped in here.

There are now tons of Yamaha sponsored YouTube videos, too. One of these videos is the “Montage Development Story.” It’s about eighteen minutes long, showing the developers while they talk about their creation. There is not a lot of technical content, but the video does provide a back story on motion control and motion sequencing. The idea is to sync control to the beat. Control changes are synchronized with the Super Knob and they are synchronized with keyboard performance. The idea is to set up motion sequences that are triggered by whatever the musician plays on the keys such that the sound is transformed and evolves. The goal is to “rhythmically change tone.” Thus, it’s possible to play tones from the keyboard that would ordinarilly be created by tediously drawing the controller information into a DAW.

The Montage can sync to an external, live performance. It derives the beat from an incoming audio signal. Remember all that stuff about patents on beat detection algorithms? We’re going to see more of this from Yamaha.

The segment with the industrial design guy was a stitch. Our son started out as a graphic artist and views the world as an artist like this gentleman. The Yamaha designer spoke about the concave shape along the part of the keyboard that faces the audience. He spoke about the energy of the audience pushing against the concave shape. It reminded my of Kev. It also reminded me of why we would fight like cats and dogs about certain concepts and ideas! “You left, just as you were becoming interesting.” 🙂

Part of the retail price goes into non-recurring engineering (NRE) costs. The NRE cost needs to be recovered sometime or Yamaha would eventually go broke! There are other non-recurring costs such as set-up for manufacturing and test, and pre-launch promotion. The per-unit, recurring cost is the money spent on parts, manufacturing, quality assurance, distribution, promotion and support for a single keyboard.

I counted 50 people in the group photo of developers at the end of the video. A typical engineer’s salary in Japan is ¥ 5,500,000. The estimated annual burn rate for the Montage development group is ¥ 275,000,000 or about $2,340,000 USD, not including overhead for capital, benefits, etc. The so-called “burdened” cost of a developer is higher depending upon the overhead.

Let’s assume a $3,000 average selling price (ASP) per unit and that 10 percent of a retail sale ($300) covers NRE. Yamaha knows this number; we don’t. If development took four years, then Yamaha needs to sell:

    (4 years * $2,340,000) / $300 = $9,360,000 / $300 = 31,200 units

to cover NRE. Due to the assumptions made, this is a lower bound on the break-even point! For sure, it does not include costs incurred by the international members of the team (USA, UK, etc.)

If you read my cost analysis yesterday, that was a first draft of a first draft. (Updated: January 22)

AC Hamilton and Anderton’s — who are also European retailers — posted a few hi-rez product shots. Please click on the images below for higher resolution.

montage_tilt

montage_top

montage_rear

And, finally, a short list of specifications. Please click the image to display in full size.

montage_specs

I’m sure the debate and discussion is already underway! I will update this post with further thoughts and comments.

BTW, I noticed that the Yamaha UK site is advertising the following promotion: “For a limited time only, purchase a state-of-the-art Tyros5 (61 or 76 key model) and receive a free sensational Yamaha MCR-B043D HI-FI.” Hmm, guess what’s next?

Apple Music Memos: Snap Review

Posted on by

Just had to give Apple Music Memos a try.

Apple Music Memos records your performance, figures out the tempo and chord progression, and plays back the original audio plus a bass and drum accompaniment. Pretty neat.

After installing Music Memos on my iPad, I went up to the studio, turned on my Yamaha PSR-S950 arranger and launched Music Memos. I decided to try a 4/4 song with a simple chord progression — Van Morrison’s Brown Eyed Girl.

What did I learn? First, Music Memos needs a little time at the beginning of a track to figure out the tempo. It displays a vertical line which presumably shows the downbeat or chord change. If you try to keep good time, it will follow your changes. If…

Well, I ain’t no Hal Blaine when it comes to timekeeping! Turn on the S950’s metronome, set it for 110BPM and try again. Unfortunately, the loud sound of the metronome seemed to throw off the chord recognition, etc. Music Memo provides visual feedback of sound level by sending out waves (like a pebble in a pond) around its record button. I decreased the metronome volume so that it wouldn’t trigger Music Memos’ sound detector. The much louder piano sound still picked up.

The third try was the charm. I played a four bar lead in to give Music Memos a good, solid beat before kicking into the tune. I laid down a basic track by playing a piano chord on each downbeat. The second thing I learned is Music Memos does a good job with chord inversions, but doesn’t always recognize a dominant 7th. It got the root and major/minor triad OK, not the flat 7.

With the basic track in place, you can edit the chord progression and make fix-ups. I couldn’t find a way to straighten out timing which is why I resorted to the metronome method. Music Memos desperately needs a metronome even if it is a silent, visual metronome. Apple shouldn’t have any trouble adding this feature to the app.

Now for the fun part. There are two icons at the bottom of the screen: bass and drums. Tap the icons to turn on the accompaniment. If you hold an icon, Music Memos pops up a chart that let’s you change the overall characteristics of the bass or drum accompaniment. You can choose simple vs. complex, etc. I left the settings on the defaults.

OK, plug in the Boss MicroBR recorder, hit RECORD on the MicroBR, hit PLAY on Music Memos and lay down the organ part. True confession. The rehearsal was nice and loose. With the red light ON, things got yippy. It took six takes to get an acceptable demo track. (Hats off to the Wrecking Crew, again.)

Music Memos can probably do more tricks and it will take a little time and practice to work with it. (Drum accompaniment when laying down the initial track?) Here’s the quick demo track just to give you a taste. Total cheesiness is intentional! Musical Doritos. Cha-cha.

Some other goodies to appear at NAMM 2016

Posted on by

Hey, hey, just flipping through a few media press kits (NAMM 2016) and wanted to jot down some notables. This site is not just about Yamaha after all.

The Roland A-01 controls MIDI and CV/Gate-equipped analog gear, has a 192mm X 40mm LCD screen, onboard tone generator, built-in speaker and 16-step sequencer. The A-01 is battery or USB powered and snaps into the (optional) K-25m keyboard. The pair are shown in the image below. (Click images for higher resolution.) Connectivity includes standard MIDI IN/OUT and wireless MIDI via Bluetooth LE. Wonder what the BLE latency is like?

roland_a01

The tone generator is described as an “internal 8-bit CPU sound generator.” Your guess is as good as mine. Seems like a logical extension of the Roland Boutique line. Totally tiny and portable. (Tinny, too?)

The Roland JC-01 Bluetooth Audio Speaker is a mini Jazz Chorus. The device purports to have a clean sound, an integrated, rechargeable battery, passive radiator, and a three-band EQ. No effects. You can even use it as a speakerphone!

roland_jc01

I’m a sucker for these little amps. The Blackstar Fly 3 Bass has my eye, too. I just hope that the Roland doesn’t carry its usual exorbitant Roland mini-amp price. We’ll have to put a keyboard through it and see if it’s up to the job. A nice beefy B3 patch usually reveals flatulence and boxiness in a hurry. The Fly 3 Bass might just carry the low end better than the mini-JC.

The Roland System 500 is a thing of beauty. I keep my old Electrix modules just because I like the light show in the rack. But, this is visual art.

roland_system500

As Will Smith said in Independence Day, “I have got to get me one of these!”

Boss has a few new, fun toys, too. The VO-1 Vocoder Pedal could do some serious damage to things other than guitar. And, it’s all in a stomp box. Speaking of stomp boxes, Singlular Sound released the BeatBuddy Mini with a lower street price than its bigger, er, buddy. The Mini does away with all of the external interfacing and extensibility — kind of a one-off pedal. Yes, folks, all those darned connectors and connections cost money and need to be tested in QA. I think we’re going to see more and more tech in pedals, especially accompaniment and backing. There’s money to be made in those darned [guitar] hills!

Maybe not at NAMM, but Apple announced its free Music Memos app. Here’s a few quick quotes:

Sometimes the best ideas come when you least expect them. When those moments happen, open Music Memos to record high-quality, uncompressed audio through the built-in mic in your iPhone, or connect an external microphone. Music Memos is optimized for acoustic guitar and piano, and it works with other musical instruments, too.

Music Memos automatically analyzes the basic arrangement and suggests chords you played for each track. Then you can hear how your music sounds with a realistic virtual drummer and bass player jamming along. If your performance speeds up or slows down, so does the band. You can even tweak the energy and performance of the drums or bass with a few simple controls.

To build in additional tracks and instruments for a professional-sounding mix, move your recordings into GarageBand on your iOS device, or GarageBand or Logic Pro X on your Mac.

Hmmm, looks like Yamaha Chord Tracker has competition! Yamaha have been very busy with audio analysis algorithms to detect beats, chords and the like, so we haven’t heard the last from them. Another app to consider — although it is currently pitched toward classical musicians — is Cadenza. Cadenza and Music Memos both track your tempo. All of that academic research on music analysis is beginning to bear fruit!

The Korg nanoKONTROL and nanoKEY Studio are too cute for school. Honey, I shrunk the Taktile. USB and Bluetooth LE. (Again, latency over BLE?) Both controllers are bundled with software.

korg_nano

I still like my Triton Taktile 49. The Korg software bundle with the Triton Taktile was a really sweet deal for the cash. Sometimes it’s worth it to buy a piece of kit just for the software. I bought an M-Audio Keystation Mini 32 (once) just to get Sibelius First. The bundle was cheaper than buying Sibelius First alone and I desperately needed a notation tool at the time. (Still like Sibelius First, too.)

Of course, there’s the Minilogue. It’s micro-analyzed to death in the forums, so no need to dive into that here. I like the inclusion of an oscilloscope, but I think one of the larger, inexpensive ‘scopes like the Xminilab Portable Small Mixed Signal Oscilloscope would display more information, have two inputs, be portable and be ready for action anywhere in the lab or studio.

Finally, there’s the Korg Kronos Platinum. What a handsome beast! I noticed that the back panel is perforated which should help with airflow through the unit.

korg_kronos

Only 300 units will be sold in the U.S. market. I wonder if John Batiste will get to rock a Platinum on LSSC?

Dessert topping? Floor wax?

Posted on by

Ah, we’re starting to hit the silly phase of keyboard micro-analysis and Web discussion vis a vis the not yet, not quite announced Yamaha Montage.

So, is it a floor wax or a dessert topping? Should we follow the holy sandal or the holy gourd?

Any other comedy bits I can rip off? 🙂

Frankly, I will probably never touch the bits which are touted to be for “EDM.” I’m not angry or even mildly perturbed that Yamaha would put such “useless” features on my keyboard. It’s a big world. Thank heaven for those features because more people will buy the Montage making it easier for Yamaha to offer this product at a lower street price. And gasp, I may even come to learn, use and love some of those “EDM” features. At the same time, I want to thank all those guys doing covers at the Holiday Inn…

I’ve been drawing up a shopping list of sounds and features which I would like to see in the Montage or whatever keyboard that replaces/upgrades my MOX6. What are those sounds and features? They’re personal. Not secret — personal. Even the Motif XF6 is a candidate, because heck, the discount is only going to get sweeter after the Montage drops. So, I also drew up a list comparing the MOX6 against the XF based on the desired feature set.

After a day of God (gig) and football (play-offs), I took a fresh glance at the high rez screen shots. Some these pictures are making more sense to me now. (Click on any of these images to get full resolution.)

montage_motion

The Motion Control Synthesis Engine concept is a more concise and powerful way to think about all of those control assignments and parameters in the current Motif/MOX UI. The Super Knob, Motion SEQ and Envelope Follower sections twitch and tweeze parameters in the AWM2 and FM-X tone generators. The Motion SEQ and Envelope Follower add a dynamic aspect to the twitching and tweezing. The Motion SEQ can sync these changes to tempo — something that you don’t always get in Motif/MOX.

This afternoon, I was busy deconstructing the programming behind a few of the drawbar organ patches on the MOX. On the current products, you can tweak individual parameters using the mod wheel, assignable function buttons or assignable knobs. Conceivably, one could control a group of parameters from the mod wheeel, for example, but setting this up through the current UI is an exercise in tedious menu diving. If Yamaha got the Montage UI right, then it should be easier to assign multiple parameters to the Super Knob (or Motion SEQ or Envelope Follower).

montage_superknob

Further, the Super Knob appears to support morphing between two scenes where a scene is a particular configuration of one or more parameters, kind of like the old AN200/DX200. Let’s say you want to be able to morph from one organ drawbar setting to another. (Thanks, Bad Mister, for this idea.) If my interpetation is correct, then it should be possible to set up the first drawbar setting as scene 1, set up the second drawbar setting as scene 2, and then morph the drawbars between the two scenes. Cool. Maybe not as flexible as moving individual bars, but workable and low stress during the gig.

Ew, did I just use one of those “EDM” features? 🙂

BTW, parts of the screenshot to the right of the Super Knob should look familiar to Motif/MOX users. The Motif/MOX UI separates Voice and Performance editing into COMMON and per-PART (or per-ELEMENT) tabbed pages. This screenshot shows the COMMON page. The first vertical column of six touch buttons on the left-hand side of the screen replace the physical function buttons that selected tabs. The second vertical column of six touch buttons replace the old physical sub-function buttons. The sub-functions in the second column depend on the what’s selected in the first column. It’s all contextual. If you know how to navigate the Motif/MOX, this should be a breeze. It looks like this page edits the knob control assignment for an FM-X voice.

montage_best_of

The Performance selection screenshot started to make more sense, too. The touch button in the upper left corner let’s us select the Performance bank. The touch button in the lower left does category search. No surprises. Each voice is tagged with up to four icons: AWM2, FM-X, MC, and SSS. The icons indicate the kind of Performance, that is, the tone generation method, Motion Control and SSS (Seamless Sound Switching).

On Saturday, I was trying to figure out the meaning of “All 9 Bars!” and how individual drawbars might be controlled. Given what I’ve inferred about the Motion Control Synthesis Engine, this Performance most likely morphs from one drawbar setting to another via the Super Knob. SSS comes into play when switching from one drawbar setting to the next such that the sound is not interrupted. Of course, this means that the number of parts is limited to eight maximum. Current Motif/MOX voices make use of waveforms like “Draw 1+3” and “Draw 2+4” to cover more than one tonewheel footage per tone generation element. It may be necessary to exploit such waveforms on the Montage, too. We’ll see.

The Seattle Sections Performance might be fun, too. Maybe it morphs from quiet strings to fortissimo? The crescendo could be gradual and tempo sync’ed — a musical effect that is difficult to play live through velocity (key dynamics).

montage_sequence

I’m rather surprised that folks are debating whether the Montage has a sequencer or not. Or whether the Montage has arpeggios or not. These features appear quite clearly in the screenshoots. There is even a physical “ARP ON/OFF” button on the front panel. Good thing it’s physical because you wouldn’t want to dig for that switch in a menu somewhere! The “lane” notion is just a way of dealing with limited vertical screen space. The UI probably uses lanes that fold up and hide rather than displaying all horizontal lanes and endlessly scrolling up and down. I’ll bet that the UI designers drew from Steinberg’s experience with Cubasis on iPad.

If one assumes that the screen captures are one-to-one, then the Montage native screen resolution is 800H by 480V pixels.

In closing, I must say that the graphic design is clean, modern and inviting. Yamaha have definitely been applying their experience with Cubasis and Mobile Music Sequencer.

Well, that’s it, folks! Like you, I’m waiting for the demo videos, manuals and a test drive. In the meantime, put on whatever moves you and chill. Maybe “Wicked Game” by Groovy Waters. Or “Lean On Me” from 20 Feet From Stardom and listen to Darlene Love testify. Peace.

Extra! Extra!

Here’s a blast from the past — quotes from the old AN200 manual about its Scene and FreeEG features.


AN200 Scene

While playing back a Pattern, turn the [SCENE] knob slowly, back and forth. Notice how the sound gradually “morphs” between 1 and 2 — in real time as you work the knob! Do this in sync with the rhythm and create your own shifting textures!

an200_scene

The exciting and powerful Scene feature lets you create and use two different Voices within a single Pattern. Most importantly, it lets you instantly switch between them or gradually “morph” from one to the other — all in real time. Now, it’s time for you to create a Scene or two of your own.


  1. Press SCENE button [1]. This is the Scene you’ll be working on first.

  2. Work the controls and get the sound you want for Scene 1. Any and all of the Synth knobs can be used.

  3. While holding down [STORE], press SCENE button [1].

  4. Do the same operation for Scene 2 — repeating Steps 1 – 3 above with SCENE button [2] this time.

AN200 FreeEG

The AN200 has so many real-time control features, it’s hard to get a grip on them all. Feel you need an extra pair of hands? Or maybe an extra two? No problem. Just use the amazing Free EG feature. The AN200 is packed with a lot of powerful recording functions — but none are quite as impressive as this. Free EG gives you up to four tracks for recording your knob moves — letting you incorporate real-time sound changes and knob moves as a part of the Pattern. So every time you play the Pattern, your knob changes play right along with it — just as you recorded them. We call this “Free EG” because it allows you to create unique, complex, continuous parameter changes that would be impossible to achieve with conventional EGs.


  1. Call up the desired Pattern, and press the red Record button.

  2. Select the Free EG track you want to record.

  3. Start the Pattern, and make your moves.

  4. To stop recording, press the Start/Stop button.

  5. To hear your Free EG recording, make sure that the appropriate track buttons are on, then press the Start/Stop button to play the Pattern.