Archive

Posts Tagged ‘software’

Intermezzo (case study)

April 26, 2013 Leave a comment

Cavalleria Rusticana (~1880). The stairway, lying (resting?) guy and ladies panicking are reminiscent of the last minutes in the Godfather III.

The song

Warning: If you haven’t seen them yet, the links that lead to youtube videos can be spoilers.

I wouldn’t have been familiar with this song if not for the movies or TV shows that used it. It is probably best remembered for the tragic ending in The Godfather 3. I have actually seen that scene in a black and white furniture television when I was very young, but my dad, having watched it already changed the TV channel, so I did not remember the song from that movie. Next, I’ve seen my friend playing a black and white video wherein there is this boxer warming up in slow motion. As I eventually found out, that was from the opening credits of Raging Bull. It definitely has cult appeal and  I partially remembered the tune, but I wasn’t interested enough to dig further. It was not until the 31st episode of Rurouni Kenshin, that I really got interested with this song. By that time it has reached 31 episodes, the characters and story must have grown on me, hence making Intermezzo one of my favorite classical pieces. The unfolding story, artistic animation and the music, simply made a very powerful emotional combo (that could probably make a normal person cry).

The score

Since this is an old classic, it is very likely that a computer playable sheet music is available out there. PDFs are available at IMSLP, MIDIs are available from various sources, but the best I was able to obtain was from musescore.com, creater by MaestroMoi.  Since majority of the transcription has already been done, I had the luxury of neat picking further, almost obsessively, trying to make the score look  identical to the one available at IMSLP, and incorporating the additional woodwinds parts in a much recent transcription, also from IMSLP. I could have just used one of the google-able MIDI files, but that would have spoiled the fun and learning process.

From the score, I learned that the 2nd violins and A clarinets are played divisi (divided further into smaller groups playing different parts, it would look like double stops when notated). Hence, I should balance the volume to avoid making it sound that there are twice as much second violins or clarinets. Articulations markings in the score would also help me decide how to modify the MIDI.

Anyway, here are the files:

  • MuseScore score. Size A3.  I find the A3 size compact enough but still legible when scaling to an A4 printout.
  • MuseScore exported PDF. Note: some symbols are not optimally arranged.
  • MuseScore exported MIDI.
  • Skaiju edited/tweaked MIDI tweaked and articulated with Sekaiju.  This is 4.3 times larger than the “un-articulated” version.

The complexity of the score also helped me learned more of MuseScores features like putting notes on the next staff instead of using more ledger lines, and putting beams over notes in different bars (which starts to look odd if the bar is in a different line or page, revealing room for improvement in MuseScore). I also found it helpful to modify the score layout to make it the page as long as possible so I don’t have to navigate to different lines or pages when editing an instrument part.

Divide and conquer

At some point in the working process, it is easier to use sheet music for the individual instruments, instead of the full score.  Fortunately “parts” were also available from  IMSLP.  Since I focus at one instrument at a time when tweaking MIDI, it is convenient to have the full instrument part in one page. Extracted parts also speed up work by avoiding confusion or distraction from other instruments. It is also helpful to put the bar number in each bar, not just the first bar in each line  (you may want a print out). Since parts are not always available, re-writing the score in MuseScore would also give you this advantage.

Software

The usual suspects plus AAMS:

  • MuseScore: converting the visual score to something that can be made into MIDI
  • Sekaiju: Further MIDI tweaking (expression and articulation)
  • Synthfont: Rendering the MIDI with soundfonts
  • Freeverbtoo: For reverb
  • AAMS: For EQ mastering based on a reference recording

Samples

The samples used are mostly from the Sonatina Symphonic Orchestra, with a few exceptions

  • Organ. Jeux d’orgues
  • “Low harp” aCoUsTicBaSs from the Jazz Page. Since the lowest harp notes in the pieces are not audible with SSO.

I’m not very particular about these other instruments since the strings dominate the sound. The strings and woodwinds are also from SSO worked fine. I’m not quite happy with SSO’s harp, but I just let it be since it is not as loud as the rest. For reverb I used a cathedral preset in freeverbtoo. I actually thought whether I should go for an IR convolution reverb, but I probably made a “mistake” by starting out with the snappy and convenient “go-to” freeverbtoo, that it became difficult to make the piece sound the way I like with other reverb VSTs.

MIDI tweaking

Since I’m very familiar with how the song, this proved to require more effort than my usual re-arrangements. In short, I had higher standards because I had an easy and definite way of benchmarking, i.e. listening alongside an actual recording or the my mind’s “ear worm”. The score helped me decide how to tweak the MIDI by:

  • Slightly overlapping slurred/legato notes
  • Modifying the MIDI velocity depending on the note’s dynamics (e.g. p pp ppp). Actually MuseScore will take account of dynamics when exporting to MIDI, but you would still want to make adjustments to get it sounding right.
  • Separating notes that stick end-to-end. If the same note/pitch is played in succession without a gap in between, weird buzzing sounds would result sometimes. Also, it is likely that musicians in the real would also make a short pause in such cases. Can anyone bow the same note twice without pausing in between? Probably not, but the natural reverb and decay of the sound will feel in this gap, just like when a pianist steps on the sustain pedal.
  • Modify Expression (MIDI CC11) to control loudness, e.g. in notated crescendos and decrescendos, and where ever I feel like changing the volume. CC7, volume, also works similarly, but I did not tweak this part. I also used this to minimize the piercing or ringing sound from high notes at the end (from oboe and 1st violin). This also explains how the MIDI file got more than 4 times larger (still small at 47KB) since many data points are used to draw expression curves.
  • Randomize harp note starts by a slight amount to make it sound “human played.”
  • Add the low harp work-around. As I mentioned earlier I couldn’t hear the first low notes of the harp, so I added a supposedly pizzicato contrabass for that part, which was then rendered with an acoustic bass guitar.

I also used CC11 to apply a longer fade out on long notes. Without modifying this volume, the long note will only start fading out close to the end. This release time is about 0.5 seconds for SSO. It may not always sound natural if a long note is near full volume for most of the time then, just fade out for a  short time only towards the end. Musicians may play a long fading note. Especially since, unlike MIDI, musicians know beforehand how long the note should be. This may not always be the case though, so you should trust your ears in the end. I think this applies more if there is an anticipation of more silence after the long note, or if the long note ends a phrase.

Results

We’ve had enough talking, let’s now hear the music.

One difference I notice from real recordings is that there are more pronounced solo instruments. I.e., my ears can distinguish a solo violin on top of other violins. The ensemble sound provided by SSO is somewhat more “homogeneous”, or there is no dominating solo instrument within the ensemble. This may have to do with the mic placement in real recordings.

Automatic mastering using reference recordings

I also tried out this new cool tool, the Automatic Music Mastering System. AAMS adaptively applies EQ settings on an audio file based on references or even based on other recordings you have. Unlike the usual EQ with loads of presets, AAMS first analyze the original audio file, and then decides how much of each frequency band to boost or attenuate. And it uses a lot of frequency bands instead of just a typical smoothened out curve with 3 to 10 peak regions. What I’ve done is choose a reference recording, have AAMS analyze its spectrum, and make my MIDI orchestation output have the same spectrum. To be on the safe side, I used a reference file with an Open Audio License from Wikimedia:

http://en.wikipedia.org/wiki/File:Pietro_Mascagni_-_Cavalleria_Rusticana_-_Intermezzo_Sinfonico.ogg

It doesn’t have to be the same song, but it’s reasonable to assume that reference based automatic mastering will give best results if the reference is playing the same notes as your project. In most cases, it may be enough that the reference has the same style or genre or instruments, i.e. it is enough that it sounds similar to what you want. What I’ve done is closer to exactly what I want.

AAMS Intermezzo Spectra

AAMS Intermezzo Spectra. The Red and Blue plot shows the spectrum of my project while the yellow and green plot shows the spectrum of the reference obtained from Wikipedia. Overall, they look similar, even rising and falling at the same frequency regions.

AAMS Intermezzo Suggestions

AAMS Intermezzo Suggestions. How my project audio file should be equalized. Notice how there are a lot of frequency bands. I wish I understood what this plot means beyond knowing that it looks cool.

Interestingly, the spectra didn’t differ so much meaning that the SSO + Freeverbtoo combo already sounds realistically mastered enough to begin with. That may also be the virtue of classical orchestra music being more reproducible, keeping the recording setup as simple as possible and having no special effects whatsoever. But I can still imagine many cases wherein AAMS could be a lifesaver, especially if your speakers/headphones are not the best you could get, then you could still be safe if your song sounds like a professionally mastered reference song through quantitative spectral merits. It may also help if you feel that your sample libraries are not sounding the way you imagine it should.

Conclusions

Although digital MIDI orchestration  motivates us to create our own orchestra compositions, it is  a good learning experience to work on an existing that tune you know well. This raises your standards as you try to best replicate the known song. It is like the difference of drawing a fictional face you imagined, and drawing a portrait of someone you know well. The former does not impose a definite right and wrong. Making covers also makes you aware of what the tools are capable of, and what are their limitations.

Happy music making!

Farewell (case study)

April 20, 2013 Leave a comment

After a long break from making music due to my thesis, here I am again! This time I’ve decided to share my workflow on a recent project. You can listen to the final result below:

The song

Album art from the Escaflowne Original Soundtrack 3 from which the original music can be found.

The song is part of the Vision of Escalowne‘s original soundtrack (third CD album). I’ve seen this anime more than a decade ago, made cassette copies of my sister’s CD, and listened to it regularly back in highschool. For a cartoon, it has a soundtrack that goes beyond what you would normally expect. It’s one of the reasons I got hooked into orchestra music (as opposed to the serious and profound classics that is less accessible to my less mature mind back then). It’s composed by Yoko Kanno and Hajime Mizoguchi and (probably) performed by Warsaw Philharmonic Orchestra.  The  score was based on this transcription by ThePochaccos and all thanks to him/her for doing a great job.

Unfortunately it is not a public domain score, so I won’t be able to share it that easily. But basically, I transcribed the video using MuseScore. It is not that difficult, there are parts in the piano that need “voices”, but the whole piece is mostly strings (for me., that’s actually easier and more fun than emailing the youtube user). (I hope to share something public domain next time).

I’ve been itching to get Reaper for a long time already, but so far, my projects are not too complex, so  this setup is still fine.

Sonatina Symphonic Orchestra

For this song, I decided to do a proper demo of SSO. I’ve always used multiple orchestra samples and layer the different results in Audacity. Using SSO by itself has also revealed some of its shortcomings. Sometimes the release does not sound well, giving an unnatural sound at the end of the note (i.e. in long contrabass notes or in violas). I work around this by shortening the note until the odd sounding part is no longer audible. The slow attacks also made it less favorable for fast short notes, making them sound mushy, and making the melody less defined. I try to remedy this by increasing the note velocities or decreasing the note velocities of the background instruments. I wonder whether using a VST SFZ player, instead of Syntfont’s native sfz support, might give better results, but I have not really explored this option.

Manual looping*

Update: As it turns out, this manual looping workaround was unnecessary ans I should apologize for the misinformation. My mistake was to directly load the sfz file to Synthfont (an older version back then) instead of using an sfz playing VST. SSO strings loop nicely with Plogue sforzando. The text is maintained for historical purposes and as it may still be helpful for other libraries that do not have looping in them or, for another application, to hide the distinct repetitive sound of looped samples.

Another shortcoming of SSO is its lack of looped samples for some instruments, which has forced me come up with a tricky work around. The first note of the first violins is 9 bars long (31 seconds at 70 BPM!). At first I thought of editing the SFZ samples, but that seemed overkill for a single note. So next, I imgained how a real orchestra would actually play a half minute note. If it were a single violin, the maximum amount of time you could slide the bow over a string would be limited by the bow’s length and the minimum bow velocity needed to produce an acceptable sound, maybe five or 10 seconds (my imagination’s approximation). Restarting the bow slide would have made a new note. But an ensemble of more than 10 violins doesn’t have to simultaneously restart their individual bowing. So while one violinist restarts there are about 10 others who are still bowing midway, hiding the restarting guy and creating an illusion of continuity. That’s my guess.

Back to the MIDI editor, I implemented “manual looping” by making an extra first violins track (not to be confused with the second violins in the score). I broke the whole 9 bar note into shorter segments that can be played by SSO. The extra violin track continues the note when it’s about to end in the original violin track. Then the original violin track continues the note when the extra violin track’s note is about to end. Hence, by alternating and overlapping these two violin tracks for the same note, I get a manually looped violin note. To mask the attacks of this repeating violin, I align them with the attacks (note start) of the other instruments in the score. Of course, these alternating violin tracks must have the same volume and panning and go through to the same effects chain.

It may also be worth noting that the SSO updated sustain violins worked better for this trick.

Guitar = Guitar Pro

Since I can’t compete with a real orchestra, I generally avoid making inferior copies of something that it is already great (except for personal studies or demos). Who would listen to that? At the very least, I would change an instrument to give a different feel that is worth listening to. Hence, I changed the piano part into guitar. Being a more common and accessible instrument, and being a long time guitar player myself, I could relate more with the sad sound of a guitar.

I’m also known among my friends for advocating Guitar Pro (GP) as a virtual guitar addition to their DAWs. Even though I play guitar well (used to?), and own many guitars (too many to remember), recording guitars with my limited laptop studio setup has never given me satisfying results. GP actually started as a tablature study program (coincidentally at the age when I was crazy about studying guitar tabs). When it started out, there was a free alternative that can do as much, Tux Guitar. But since Guitar Pro’s introduction of RSE (realistic sound engine), it has, in my opinion, left Tux Guitar far behind. GP would not integrate with a DAW like a VST or soundfont, but it’s notation based interface, optimized for guitar articulations, makes it far more intuitive than any VST I know. GP can simulate vibrato, hammer on/pull offs, ringing, chord “brushing”, harmonics and many more with a few mouse clicks and without having to tweak MIDI parameters.  And the demos sound realistic enough for me (listen or download here). It’s probably the guitar equivalent of Finale + Garritan combo, but at a price below a hundred dollars. Software that unify sophisticated music notation and virtual sound production is really something we should be thankful for (although I would also hope for piano roll integration).

Conclusion

There’s nothing really new here, but I would be happy if a newcomer in digital music production/midi orchestration would learn something from this. Note that the only tool that costs money is Guitar Pro, although I’ve also donated a small amount to Synthfont as it is very useful to me and it was the first thing to exactly match what I was looking for before, a simple tool that applies soundfonts and VSTi’s to an existing MIDI file that is not as overkill as a full blown DAW. With diligence, passion and knowledge of what great tools are available out there, making quality music, one that you can mix in to your iPod or MP3 player, is no longer a thing that can only be done with professional thousand dollar studios.

Happy music making!

If buying can wait, wait until December

January 1, 2013 Leave a comment

When I told my land lady before that I wanted a piano in my room, she advised me to buy one in December as prices would go down.

The same rule applies for virtual instruments and other music making software. If you can wait, wait until December. I had seen many discounts, and hence, was able to purchase Miroslav Philharmonik Orchestra and Kontakt 5 at 1/3 and 1/2 their regular prices respectively. In fact, there are too many such promos, that I decided to only discuss the products I actually bought and just give a more comprehensive link in KVR Audio for the rest. Elsewhere in the interwebs, audible (UK) has given a 30% off on all audiobooks (I’m a lazy reader) and Actual Tools sold their multiple monitor taskbar with a 50% discount. For sure there are more discounts for other products during the December holidays.

Christmas has some good effects. I would like to believe that deep in their hearts, vendors just simply want to give, when they are possessed by Christmas spirits. But of course, marketing wise, Christmas promos get them more publicity, attract customers who would otherwise never buy their products, and for those selling tangible stuff, get rid of the year’s unsold products. In any case, if you are not tricked into impulsive buying, it is a win-win situation IMHO.

So, if you think a product is great, but have no immediate use for it, it may be good idea to be patient and wait. Also, if it software delivered via download, you don’t have to worry that it will run out of stock. If the software is famous/reputable, its website would most probably still be around within the year.

You might even realize that you can actually live without it as time goes by and you discover alternatives. :)

Risks and considerations

What you could worry about is that the price may go up as you wait. There is the possibility of inflation, and prices may rise with continuing software development expenses. But so far, I’ve seen price increases are moderate when they do happen. Another consideration is major software upgrades. There are a lot of complaints when software undergo extreme make-overs like when MS Office 2007 was introduced. As the most affected are those who are used with the previous version, this may not be a problem for new customers. I haven’t used any previous versions of Guitar Pro before, so unlike what others may claim, I find Guitar Pro 6 to be just fine. Whether or not an upgraded version will be discounted or for free if you have a previous version is another concern. It could be sad to pay extra a few months after buying a previous version.

Coda

What a great way to start a year by waiting for it to end. I actually don’t feel very comfortable talking about all that buying stuff and whatnot as I prefer talking about free stuff which all of us can appreciate. But as consumers or prosumers, buying things is just part of getting things done.

Happy New Year! :)

Beyond MIDI with Synthfont

December 19, 2012 Leave a comment

So what do you do with a MIDI file that sounds1 like a cheap toy when played with your computer? Load it on Synthfont!

There are a lot of suspicious websites and software out there when you search for “MIDI to MP3”. I doubt that most of them would improve the sound quality. They just convert the format, and you have no control over how the resulting MP3 file sounds. Although they may serve the purpose of making your arranged music more compatible to more players and devices, it is still painful to listen to. In many cases, you simply have a much bigger file.

This is why it is better to go beyond a passive format conversion and take control of the process. Give your hard work the quality it deserves. It is also a fun learning experience that can open doors to new hobbies or even new career opportunities. Knowing that you can create high quality music tracks, with just your computer, is quite empowering.

Even if you can tweak the MIDI further through automation and control changes, it is not very motivating if the instruments still sound bad (just like how having a bad sounding real musical instrument discourages you from playing it). Also, there are many instruments or instrument ensembles that are not included in the General MIDI specification. What concerns me most, is that you don’t have ensembles for violins, violas, cellos and contrabasses. Instead you have generic “strings” (instruments 48 to 51).

A Beginner’s DAW

After being initially discouraged by the complexity and bloat of other free digital audio work stations (DAWs), I eventually found something that simply does what I want to do. Synthfont makes it straightforward to apply soundfonts and VSTs to an existing MIDI file. It is “MIDI-centric” not containing audio editing/recording functions like those found in audio editors like Audacity or Wavosaur. It is also not bundled with annoying loops and samples that you’ll never use. (I find it both ridiculous and discouraging how some DAWs boast that they are bundled with gigabytes of samples).

Getting better sounds using soundfonts and VSTs

To demonstrate a simple comparison, the MIDI resulting from the previous Sekaiju tutorial was modified. Notes were made shorter than notated to emulate the player’pauses. The chords were voiced out and split to different string instruments. Slightly higher velocities were also assigned to shorter notes. Below, you can listen to two versions of the same MIDI file. One is rendered with Window’s default synth while the other is rendered using free soundfonts and VSTs from the web.

Silent Night in Synthfont

Silent Night in Synthfont using custom soundfonts, VST instuments and a reverb VST effect.

  1. Rendered2 using GM.DLS3 found in “WINDOWS\system32\drivers\”. This is how it would have normally sounded if played directly as MIDI file with the default Windows sounds.

  2. Now, this version uses custom soundfonts, VSTis, and VST effects.

    The following were used:

Although the second example is far from the best possible result, it is much more pleasing than the first one. Normally, I’d spend more hours until I become happy with the results, but I have a self imposed Christmas deadline. Like any artwork, a good musical arrangement should take its time. There are also a lot of free soundfonts and VSTs out there, aside from what I’ve used. So there is plenty of room for experimentation and improvement.

Similarly, there are also other good DAWs out there, both free and commercial. But back then, when I was starting, the other free alternatives were not the easiest to learn, and were a bit overkill. Hence, Synthfont, because of it’s “plug and play” paradigm and ease of use with existing MIDI files made it a winner for a beginner “DAW-ist” like me (wow, that sounds like some kind of religion). Its website actually has a tutorial, but I never really read it as things are quite straight forward to figure out. On the funny side, I always discover something new because I never really read the tutorial. :P This just shows that I don’t have to ingest tons of information before I can get satisfactory results.

Other tools of the trade:

These are little tools frequently used for other practical stuff when creating MP3 files. Their functions can also probably be done through Synthfont, but it is just far more intuitive to use dedicated tools. Furthermore, they can be used independently of what other software you use to create MP3 files.

  • Mp3 Gain: Normalize the volume of your Mp3 file, so you don’t have to adjust the volume when your player goes through different files. 92 dB is standard while softer pieces such as piano solos seem to go well with 89dB.
  • Mp3 Tag: Add more meaningful extra information in your MP3 file. Let people know where it came from.

Foot notes:

  1. The sounds are NOT in the MIDI file. But we commonly associate cheesy sounds with MIDI files because of the Roland GS Sound Set used by the Windows SW Synth (which dates back to 1996).
  2. This defeats the purpose of using Synthfont. But I think it is more preferable to do it this way compared to installing one of the many other competing not-so-well-known converters you may find online.
  3. It is possible to replace GM.DLS and get better default MIDI sounds.

Happy holidays/music making!

Sekaiju tutorial: A simple example (Part III)

December 16, 2012 Leave a comment

Introduction

It’s been a month since I had written Part II of this tutorial back in Hungary, and Christmas is now a few days ahead. By now, Sekaiju 3.4 is already released with a few added features and bug fixes. So now, I finally had enough pressure to finish the Silent Night midi project I’m working on. One of the reasons for this great delay is that I would rather use other free music notation tools such as MuseScore, Finale Notepad or Musink which I just recently discovered. Another is I would guess that many DAW oriented users do most of their work through the Piano Roll, which was already discussed. Anyway, enough of these excuses. Let’s just finish this tutorial series and enjoy the holidays!

If you haven’t seen them yet, I would suggest you go through the following tutorials:

Again, here is the basis of the MIDI file we’re building from scratch. Since this part would be about using the Musical score editor, we just copy the score directly as we see it. Perhaps, one motivation for using a MIDI editor instead of a dedicated notation editor is that you can further tweak the notes in the Piano roll, deviating a bit from notation to make it sound more human and expressive, since as I had mentioned before, MIDI and music notation are not equivalent. As we chose the key of C, there is no key signature indicated on the score.

Silent Night

Silent Night by Franz Gruber, from Wikifonia. [Published by benoit on December 14, 2006 © reserved by Musicopy]

Creating notes through the Musical Score

Showing the Musical Score window

If you have the file (from the previous tutorials) opened, a Musical Score representation can be shown by selecting the “Show new Musical Score Window” button (♫) in the main toolbar or in the “View” menu.

Basic editing tools

Minus the Line tool, the editing tools in the Musical Score window are mostly identical to those found on the Piano Roll. The major difference is that you can not resize a note by dragging. Instead, you select the notation symbols that correspond to the note length you want to use.

Sekaiju: Pen, Erase, Select, Preview

Toolbar Buttons found in both Musical Score and Piano Roll windows. From left to right:Pen, Erase, Select and Preview.

The buttons are the Pen, Line, Erase, Select and Preview. When selected, these buttons do the following:

  • Pen:
    • Draws a new note with a pitch determined by its position and duration by the the selected note symbol
  • Eraser:
    • Erases an existing note.
  • Select:
    • Selects multiple notes and allows multiple notes to be simultaneously moved.
    • Also works for a single note.
    • Allows keyboard or menu actions like Copy, Cut, Delete to the selected notes.
  • Preview
    • Plays existing notes when the cursor is dragged on top of them (you will see a vertical line).

Towards the right of the toolbar, length indicating music symbols can be seen:

Sekaiju: Musical Note Symbols

Buttons for selecting the note length: Whole Note, Half Note,Quarter Note, Quaver Note, Semi-Quaver Note, Demi-Semi-Quaver Note and Dotted/ Triplet modifiers

These buttons determine the length of the note that is added when using the Pen Tool. The first 6 buttons, are the (common?) note lengths.While the last two, the “dot” and “triplet” modifies whichever note you have selected among the first 6 buttons.

Entering notes through the Musical Score interface.

Continuing the flute (melody) track

1. From our previous work, look at the staff corresponding to the Flute track. It should be easy with the tracks labeled. There is no need to explicitly select the track as notes are entered on the staff that you click, and the staves are already separated for each instrument (unlike in the Piano Roll where all the track’s notes are using the same workspace). If you add a note on a staff of a different instrument, that instrument’s track gets selected.

2. Zoom to a comfortable level by clicking the plus or minus (“+” or “-“) buttons at the lower right corner along the scroll bars or by using “Ctrl +” or “Ctrl -” shortcut keys.

3. Set the position quantization. Since the smallest note is an eight note or quaver (♪) and the notes on the song are multiples of a semiquaver, we could choose 60-Quaver from Snap dropdown (third). We have to do this because, unlike other score editors, note entry is position dependent. For example, if you create a whole note on an empty bar and click somewhere in the middle, the whole note will start roughly where you clicked and go beyond the bar, causing tied notes instead.

4. Use the beat markers as guides. Note that in each bar/measure (bounded by blue vertical lines), there are 3 vertical gray lines. This indicates the beats in our 3/4 time signatured-song. Also use the measure numbers as a guide.

5. Choose the Pen tool and select a note symbol corresponding to the note length you want to enter. Since we are using a music sheet as a guide, we simply copy the note symbols as we edit. Enter the note with the correct length by clicking on the appropriate location on the staff. Remember that note entry is position dependent (like in the Piano Roll).

Dotted notes and triplets are obtained by pressing down one note length then also pressing down either the dot (“.”) or triplet (“3”) buttons to modify that note length.

Sekaiju: Musical Score Note Entry

Note entry through Sekaiju’s Musical Score window.

Continuing the strings (harmony) track

1. The same procedures as with the Flute track, except that you will be putting notes into the Strings staff. There is no need to explicitly select the Strings track as it already has a separate staff. Notice that it is using a bass staff which makes more sense with the notes we previously entered with the Piano Roll.

As in the previous tutorial, for simplicity, I only used the root of the indicated chords in the transcription and play them an octave below the melody.

Silent Night: Measures 10-21

Silent Night: Measures 10-21. Flute and strings. The whole song is available as a MIDI file.

Quirks

As the musical score editing behaves very similarly to piano roll editing some unusual things can happen.

  • Notes can overlap. You place a note that starts before the previous note of the same pitch ends.
  • There is no “page view” or “wrapping”. You have to scroll horizontally to see the rest of the song.
  • There is no concept of “voices” so note entry is dependent on position. You can easily create notes that don’t have to be vertically aligned.
  • Similarly, there is no concept of rests. For example, if you place a quarter note on an empty bar, it doesn’t necessarily have to be in the start of that bar.
  • Notes can not be dragged-resized. Just erase and replace with the right size, or use the Piano Roll instead.
  • I haven’t figured out how to tie notes. The last “Peace” note in Silent Night is longer than a whole note, two tied dotted half notes. Of course, this is easy to remedy in the Piano Roll. But this is one of the expected shortcomings of Musical Score editing not a main purpose of Sekaiju.
  • Likewise, I also don’t see any quick way to flatten or sharpen a note. This would be used if there are accidentals in the song. This can be done of course, but not like how you normally do it in a full blown score editor.

Moving forward

This is a very simplified version of Silent Night, as the emphasis is really getting familiar with Sekaiju. The MIDI file is available for further study. As an exercise, I would suggest completing the chords by adding thirds, fifths and sevenths instead of just the root notes (refer to a chord chart if necessary). Also the built in MIDI synth of Windows sounds too cheesy to be enjoyable. Hence, a big step (big in improvement, not so much in effort), is to use the MIDI file with Synthfont or other DAWs together with good soundfonts, VST instruments and VST effects.

Also, for what it’s worth: Merry Christmas! :)

Categories: Music making Tags: , ,

MIDI on Windows 8 (Acer W500)

November 27, 2012 Leave a comment

Basically, MIDI input works. But there are several Windows installation related problems (not MIDI related though). If MIDI worked before (in Windows 7), I think it should be fine after an upgrade.

Despite all that hype about the iPad and Android, I never really jumped in since that meant leaving a lot of things behind. I should admit, more than 60% of my work, or my life for that matter, is done through a Microsoft Windows machine. Being an occassional Windows programmer, there’s just so much to miss from a classic good old fashioned all purpose operating system. iOS and Android are rather specialized to phones and tablets. And doubts start to arise when USB peripheral devices come to mind. I also dislike how the iPad or Kindle devices prevent expansion through the lack of SD card slots or USB ports, which also shortens the lifetime of the built in solid state storage device.

Windows 8 on the other hand, is meant to be a successor of the Windows series. Hence, it is designed for a broad range of computing devices. Though I suspect that they might trim features here and there depending on what kind of device it is being installed on. MIDI input worked on my tablet back when it was running Windows 7 Pro. Now, I need to find out if that would be the case on Windows 8.

Although there have been reports of earlier releases of Windows 8 not supporting MIDI devices, this seems to work smoothly on my W500 that had been upgraded to a recent official release (for roughly 20 USD). My USB MIDI keyboard (Korg microKey) is recognized by Synthesia (which is the MIDI application most likely to be used with the device). As can be seen in the screenshot below my Korg microKey is recognized and used (A3 note is pressed). The Microsoft GS Synth is still also present.

Synthesia on Windows 8 with MIDI input

Synthesia on Windows 8 with MIDI input (also my first Windows 8 screenshot). USB MIDI input device recognized and used. Microsoft GS Synth still exists.

It’s not completely Metro (thankfully)

Windows 8 introduced the Metro GUI which is best used for touch screen devices. But it did not remove traditional “windows” applications (those boxy things with borders and buttons on the top right corner that we easily take for granted).  There is now a distinction between “Desktop” and “Screen” apps. Desktop apps are the ones we are most familiar with from previous versions of Windows while Screen apps are designed for a more recent touch interface (big buttons and fonts, support for gestures, simplified UIs, etc). If you successfully install and run a non touchscreen good old fashioned app, a familiar looking desktop and taskbar will appear behind the window of that app (the start menu is gone though, but there are alternatives). That also hints that there is a good possibility that your previous DAW/MIDI programs will still run. It’s supposed to be the next Windows after all. In fact, Sekaiju which is made with an old version of Microsoft Visual C++ (Win 9x era) works on Windows 8 (after going through some paranoid security questions).

Sekaiju on Windows 8

Sekaiju on Windows 8 “desktop mode” with some recorded notes. The Korg microKey MIDI controller is also seen in the device manager.

Windows 8 installation annoyances

Something always goes wrong. Especially when trying new things.

1. My SD Card was formatted!

Both the local drive and the removable SD card in my device are 32 GB. At first I was having trouble that the installation does not want to use the disk I’ve selected and formatted. As it turns out the solution was to remove the removable SD card from the tablet. And as it further turns out, it appears that installation was selecting and formatting my SD card!!! Farewll backups. Although I had backed up the primary drive with Macrium Reflect, files that were only in the SD card will have to be restored with special means (Photorec seems to be helpful. EaseUs Data Recovery may do a better job, preserving file names and folders but it is much more expensive than Windows 8!). Moral lesson, remove SD card or any other unnecessary media before installing Windows 8.

2. Product key won’t work if you clean installed over a previous Windows.

Formatting before installing a fresh OS had always been the preferred way as “upgrading” always had issues. The problem when activating, a clean install is not seen as an upgrade, and hence the upgrade product key is thought to be invalid. “Fortunately”, many had the same problem and someone came up with a work around. Quoting from a brilliant, wonderful human being named BinaryInk as I found in the answers.microsoft.com forum

“The work around for this, while probably not officially supported for obvious reasons (they want more money), is to change a registry key. This was posted for Windows 7 update keys doing the same thing on a forum (though I had an update version AND did a clean install MULTIPLE times without having to do this) but worked without an issue on my laptop running Windows 8 regardless.

1. Run the registry editor (regedit)
2. Find the following key: HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\Setup\OOBE
3. Change the value for ‘MediaBootInstall’ from 1 to 0
4. Open an elevated command prompt (run as admin)
5. Run the following command: slmgr -rearm
6. Reboot

If you already entered your key, check the activation: for me it was already activated and I needed to do nothing more. If not, type in activate windows and type in the key; it should work. Also, do yourself a favor and export this key from regedit and save it somewhere if you ever are required to do another clean install. I know I did.”

Categories: Music making Tags: , , , ,

Hungary, where I learned about SFZ internals

October 31, 2012 Leave a comment

It’s quite a coincidence how the first meaningful SFZ tutorial I found in youtube is made by a Hungarian. I’m visiting Hungary at the moment for other business matters, and I just happened to be killing time during the night, when I found this video tutorial for using Garritan’s Aria player to sample instruments.

(He’s just using Garritan to play the samples, so don’t let that “scare” you)

After seeing the video, I learn that sfz is like a markup language like html or ini files. Prior to this, I have never even thought about loading sfz files in a text editor (just loaded one now in SciTE). So unlike sf2, sfz files themselves do not contain audio, but instructions or mappings on how to handle audio files located elsewhere. I had always though that the wav files in the “samples” folder in the Sonatina Symphonic Orchstra was just there for curious people who want to tweak it. They’re actually the files loaded in the DAW. I even did not notice that the sfz files themselves are too small to contain audio data. Thanks to synthfont working smoothly.

Back to Hungarians, I also appreciate how, in general, they seem knowledgeable about music or musicians. I was having dinner with a group of biologists and the conversations eventually lead to wine, cars and then to artists like Edvin Marton and local groups, how many independent musicians are great in doing covers but bad with their own compositions (sorry to the indies out there), and how you can enjoy even a “bad” song once you play it with an instrument.

Elsewhere in Europe mealtime conversations lead to sports, gym, the weather or outdoor activities … hmm …

Anyway, for those interested in building sfz files, here is the reference from Cakewalk:

The sfz Format: Basics

If your DAW does not support sfz natively, you can get Cakewalk’s sfz player VST. Prologue also recently released a free sfz player called sfrorzando.

Happy sampling!

The Linux Sampler, something interesting (other formats also)

October 29, 2012 Leave a comment

It seems that the determining factor in choosing a sample format is not the format itself, but the availability of samples in that format.  SF2 seems to date back in the era of computers with 32MB RAM, thus being rather old, widely used and compatible to many DAWs. It is quite popular since there are a lot of free samples in this format (just look at hammer sound which dates back to 1997). I happened to learn about SFZ due to the Sonatina Symphonic Orchestra. Being more recent than SF2, it offers some advantages, such as sample round robin support and a more convenient way to build sample libraries (supposedly, I have not tried yet). Recently, I’ve been trying out the Kontakt Player as Kontakt (NKI) seems to be preferred by financially well-off musicians, and is therefore associated with good/professional quality (the free samples found on bedroom producers blog tend to have this format).

Linux Sampler

Nice dark theme too (Fantasia frontend hotlinked from wikimedia)

Another sample format that I recently wanted to try is the GIG (Gigasampler) format which is currently preferred by the Linux Sampler. Although I’ve been familiar with this before, the “Gig” made me think gigabytes, which scared me a bit. That is not necessarily true though. Furthermore, opensource/free software like the Linux Sampler have the mysterious power of motivating “prosumers” on sharing works they have done with it. In fact the reason I got interested in the Gigasampler format is the available samples in the Linux Sampler website which include a Yamaha Concert Grand and London Philharmonia Orchestra Instruments (licenses still pending though). Below is an impressive demo of the grand piano:

Update: This seems to be the same sampled piano on this site: http://sonimusicae.free.fr/matshelgesson-maestro-en.html

What is the best choice then?

I usually find this question irrelevant. Although there is certainly a best format in terms of programmability, flexibility, convenience when working with MIDI and other technical aspects, in the end it is your ear that decides which sample to choose. Being able to use any of the formats allows you to pick the best of all those worlds. My favorite beginner DAW, synthfont, supports SF2, SFZ and GIG (*most), though I’m still missing out in the NKI (Kontakt) world. This question will matter though if the same sound sample is offered in several formats. In which seldom case, I’ll personally go for SFZ being more advanced than SF2, more compatible to my DAW than GIG and not tied up to a proprietary sampler like Kontakt.

Other (free) software for creating instrument samples

  • Viena (SF2) (not to be confused with Creative’s Vienna)
  • SFZed (SFZ)


There’s also the SWAMI Project although I’m reluctant to recommend it as I’m not totally into Linux when it comes to Audio (reliance of its available DAWs on Jack is to blame in my case…) and it still needs donations before Windows (or Mac) versions would be released.

Categories: Music making Tags: ,

Sekaiju tutorial: A simple example (Part II)

October 23, 2012 Leave a comment

Creating notes through the Piano Roll

Having set the song’s general properties in Part I of this tutorial. We’re now ready for the most involving part, editing notes. If you haven’t seen the previous parts of this tutorial series, please go to the following links:

With a sheet music for reference, we can now see the value of Sekaiju’s Musical Score interface. But for now, we will discuss another commonly used note entry method in DAWs and other MIDI editing programs, the Piano roll. We will be “manually” converting a sheet music of Silent Night into MIDI.

Silent Night

Silent Night by Franz Gruber, from Wikifonia. [Published by benoit on December 14, 2006 © reserved by Musicopy]

If you can read music, then you can interpret the notation and convert it to piano roll (I also realized that this is a great way of practicing sight reading without a musical instrument in hand). Each major division is a measure/bar and the minor divisions are quarter notes. Since we are using 3/4 time signature, we see 3 divisions per measure. For reference middle C is MIDI note number 60 which can be located in the sideways keyboard in the left.

As I mentioned previously in the introduction to Sekaiju’s user interface, zooming is done by pressing the little plus and minus (+ -) buttons along the scroll bars, or by using “Ctrl +” and “Ctrl –” keyboard shortcuts. (I have to rewrite it here as it was not obvious to me back then). Also, the Piano Roll window is shown by selecting “View -> Show new Piano roll window” or clicking the button with little purple rectangles looking like a piano roll.

Basic Piano Roll editing tools

I find it difficult to tell someone how to draw without telling him what the drawing tools do. Hence, I have to go through a few basics before we can create a simple song. Feel free to skip this section if you want finding things out through your intuition :).

Sekaiju: Pen, Line, Erase, Select and Preview

Toolbar Buttons in the Piano Roll. From left to right:Pen, Line, Erase, Select and Preview.

The buttons are the Pen, Line, Erase, Select and Preview. When selected, these buttons do the following:

  • Pen:
    • Draws a new note with a pitch determined by its position and duration by the set or previous note length.
    • Extends an existing note by dragging a note if it is close enough in the left or right edge of that note.
    • Moves an existing note by dragging when it near the middle of that note.
      • The note is played in the process, even if it is not moved to a new place.
  • Line:
    • Draws a series of connected notes with equal lengths connecting a starting point and an end point. As it goes through the chromatic scale, I seldom use this for note drawing.
    • Does the same in the automation part below the piano roll (where it is most probably intended to be used). Useful for making linear volume ramps.
  • Eraser:
    • Erases an existing note.
  • Select:
    • Selects multiple notes and allows multiple notes to be simultaneously extended to the left or right, or to be moved.
    • Also works for a single note.
    • Allows keyboard or menu actions like Copy, Cut, Delete to the selected notes.
  • Preview
    • Plays existing notes when the cursor is dragged on top of them (you will see a vertical line).

Entering notes through the Piano roll interface.

The flute (melody) track

1. We start by selecting the flute track (second) in the right panel. This means that notes we will be drawing will belong to the flute track. We may also select the flute track on the Track dropdown just beside the Preview button.

2. Set the snap length. Since the smallest note is an eight note or quaver (♪) and the notes on the song are multiples of a semiquaver, we could choose 60-Quaver from Snap dropdown (third). (I normally ignore those tick numbers when working as there is enough visual information in the piano roll).

3. Choose the Pen tool and start drawing on the piano roll grid area. Create notes by clicking on the grid. Each new note’s size will be the same as the previous note’s size (or the default size for the first note). Resize the notes by dragging their edges with the pen tool.

Creating notes on the Piano roll.

You may play your work at any point using the Playback button in the main toolbar or hitting the spacebar in your computer keyboard.

Adding the strings (harmony) track

1. Select the strings track from the right panel or the Track dropdown list.

2. Since there is only one chord per measure in the song, we can use a larger snap like 120-Quarter note.

3. Proceed with the Pen tool, just like when editing the flute track.

For simplicity, I only used the root of the indicated chords in the transcription and play them an octave below the melody. This may look quite boring as the first few measures use C chords. But there you have it, just repeat the steps to finish the song and you will get MIDI song file from scratch!

Silent Night in Flute and Strings.

The next and final part of this tutorial will deal with using the musical score window which may be easier if we are copying directly from a score as in this example.

Categories: Music making Tags: , ,

Sekaiju tutorial: A simple example (Part I)

October 22, 2012 Leave a comment

Introduction

Perhaps the best way to learn any software is to start with a simple step by step example that users may replicate.

I will be dividing this tutorial into several parts to avoid cluttering a single web page (and to publish the whole tutorial in smaller installments).

Since the cold lonely season of gratuitous money spending is approaching, I decided to use Silent Night for this tutorial. I will use score found in Wikifonia as the basis for the MIDI file. I’ll also use the friendly key of C version so we don’t have to worry about key signatures. For the instruments (not specified in the score),  I chose a flute for the melody and strings for the harmony. For those new to Sekaiju, please check my overview about its user interface. And for those who do not have it yet, download the latest version from the Sekaiju website.

Silent Night

Silent Night by Franz Gruber, from Wikifonia. [Published by benoit on December 14, 2006 © reserved by Musicopy]

Part I: Setting up basic song properties

Before creating notes, there are some basic things you have to define in a song. Hence, in this first part you will learn how to:

  • Give descriptive names to the tracks
  • Choose instruments for the tracks
  • Set the tempo
  • Set the time signature
  • Set the key signature (if you want to)

If you’re the type who doesn’t like reading, you can zoom on the pictures below. I’ve made them as informative as possible. Now, here we go.

1. Creating a new track. Sekaiju starts with a MIDI file with 17 tracks without by default (16 used for storing notes). You can also create a new track from the File menu or pressing “Ctrl N“or clicking the “New File” button in the toolbar.

2. Name the tracks (optional). On the “Track list” window, enter names under the “Name” column. I just wrote the title of the song in the first track. I then named the second and third tracks “Flute” and “Strings” respectively (they can be any other valid name). Naming tracks is not strictly required, but it is a good practice and it will make your work easier in the long run.

3. Choose instruments. Either click the little arrow buttons Under the “Program Number” column or enter the instrument number directly. The flute and strings are 73 and 48 respectively.  The first track can not contain notes and can not be assigned to an instrument since it has a special purpose for containing general properties like tempo, time signature and key signature and other author specific information.

Sekaiju Tack List

Track names and instruments are modified through the track list window. (For steps 2 and 3)

4. Now may be a good time to save your file. Saving is standard, just like in most Windows programs. If you wish you can use Sekaiju’s skj file format instead of mid. It has some extra features on top the standard midi file format like keeping the colors of your tracks if you decided to change them (double click the boxes in Color column). You’ll eventually need to save as MID though for interoperability with most audio/music programs.

The next steps are done through the Event list window. To open this window, go to the menu and choose “View -> Show new Event list window” or click the button on the toolbar that looks like a table/spreadsheet.

5. Tempo. This is modified through the Event list. The tempo in this example sheet is not specified meaning that it is the default 120 BPM [Citation needed?]. By default, Sekaiju also uses 120 BPM, but lets use a different tempo for the sake of learning. A bit slower, like 100 BPM. Within the first few rows, you will find an “Event kind” called “Tempo“. In the Value (1 2 3) column on the same row, change the 120 to 100. Note that the Microsec / Quarter note part automatically changes from 500000 to 600000, so you don’t have to worry about this.

6. Time signature. For this you have to look for the Event Kind called (you’ve guessed it already) “Time Signature”. Silent night is in 3/4 or Waltz as the score also indicates. Most songs (that I know of) are in 4/4 and Sakaiju also uses this by default. As with the tempo, you only need change the 4/4 part to 3/4 and the other numbers are taken care of.

7. Key signature (optional). This is specified by the number of sharps (#) in the “Key Signature” Event Kind. The default has zero sharps being the key of C Major (as with anyone beginning to read sheet music). You have to replace the word “major” with “minor” if you want the relative A minor. I decided that key signature is optional as it will not affect the actual pitches of the notes in the song. However, this will affect how the song will look (with accidentals) in the Musical Score window or if you eventually import the midi file into score editors.

Event List window

Global song properties like tempo, time signature and key signature are modified in the Event list window. Note that these parameters are stored in the first track.

These general properties, tempo, time signature and key signature are stored in the first track. They are also seen in the first few MIDI events as this information are necessary before the notes can be interpreted. People who have a fair amount of MIDI know how would notice that Sekaiju is tightly developed around the MIDI file format.

The next part of this tutorial series will discuss the creation of notes through the piano roll.

My apologies if you’re reading this and later parts are not yet available online. Nevertheless, I hope that the information available here and the previous tutorial is enough to help you learn the rest on your own. (These apologetic sentences will be deleted soon, hopefully).

Categories: Music making Tags: , ,