In the previous tutorial we entered into advanced territory, using MIDI to control the Volca Sample. This last tutorial takes care of one final question: using your own samples, instead of the default sample kit. As usual, this tutorial is the longer text version of the tutorial video on my YouTube channel (embedded bellow). If you have any questions or comments, using YouTube’s comment section is a great way of sharing thoughts and experiences with me and other Volca Sample users.
The Volca Sample has 4Mb of sample memory, which can save up to 100 samples (numbered 0 to 99) of audio. Current versions of the Development Kit released by Korg only allow 16-bit 33.0kHz mono samples (no stereo). Converting your samples to this format prior to using any uploader software is advisable.
The transfer protocol encodes the sample on a digital transfer format similar to what the older among us remember as the “Sinclair Spectrum Loading Sound”. This sound is sent to the volca Sample’s SYNC IN port. The strategy carries some important limitations: there is no way to automatically recover from a bad transmission, and if the Volca misses or misinterprets just one byte, it will abort with a “DCOD” error. Because there is always some probability for error during a transmission, this means that long transmissions (100-sample kits, for example) are very likely to fail. There is also little control on memory management, and all an uploader can do is to estimate how much memory is currently occupied, and will be consumed by the new samples.
A Few Words of Caution
This may be a question of personal preference, but I do like some of the samples that come with the Volca Sample. Korg offers a factory reset file (a 134Mb, 13 min long wav file) which will erase all of your samples and replace them with the original ones. This is far from ideal, so before messing with the original sample kit, I recorded every sample into a wav file (using audacity), which I then cut and saved on a folder. When I want a specific sound from that original kit, I just grab that one, and don’t worry about resetting the whole thing.
Tools for Uploading Samples
There are a number of tools that take your sample wave files (some also accept aiff and other formats) and “convert” them into the sound you have to send to your Volca. In this section I’m presenting the two tools I use the most: CEVS and the website Connect with Volca. This is not a review of either tool, just my impressions.
Caustic Editor for the Volca Sample (CEVS)
The Caustic Editor for the Volca Sample (CEVS)is an app developed by SingleCell Software that runs on IOS, Android, and Windows. Unfortunately, like most android phones, mine has some form os equalization on the audio output (in my case, Dobly sound) which has to be disabled ever time I want to upload samples. CEVS can run on linux, under wine, but does not upload the samples correctly, resulting in dcod errors. Nevertheless, I use it under linux for deleting positions in the sample kit (apparently deleting instructions are well sent). At the time of writing, I’m suspecting this has something to do with the audio codec used by wine to communicate with the soundboard (which should be solvable).
The main screen is a list of samples (the original ones are greyed out). Clicking on the wave opens the sample editor. Perhaps the greatest thing about CEVS is its sample editing capabilities. You can convert, cut, crop, filter and normalize your samples freely. You can also apply a number of interesting effects (distortion, phasing, reverb, etc) using the same FX engine as Caustic (if you don’t know Caustic, you should: this app is great both on the phone/tablet, as in the desktop). As if this wasn’t enough, you also have a small synth engine to generate your own sounds… it was using small samples generated from this engine and lopping that I started using the Volca Sample as granular synth. :) Oh and have I mentioned this app is free?
Despite being very powerful, CEVS is very centred around the paradigm: “I’m going to use this and only this phone to control this and only this machine”. Managing kits is very difficult. You can save multiple instances of the samplelist.txt file that appears in the CEVS directory/folder and replace the original with one of the copies to emulate different sample kits, but this is cumbersome, and does not allow sharing sample kits. The load dialogues could be a little bit more standardized. Indeed, even when running on the desktop, the program thinks it’s navigating o an Android file system.
Connect with Volca (website)
Pedro Vieira‘s website Connect with Volca is a web application that takes a number of samples and returns a wav file (zipped) that you can play to your Volca Sample whenever you need them, effectively making this a great way of storing ready-to-go sample kits. The site is also extremely simple: just click browse, select your sample (must be already a 16-bit mono file), fill in the position you wish, eventually hit the plus (+) button and add more samples (up to 15). When you’re satisfied, press “Click to create output file” and you will be prompted to save the resulting wav or zip file. Play it to your Volca Sample, and you’re good to go. And, of curse, you can keep the samplekit file, and re-load it at a later stage.
Korg’s SDK Tools
If you are keen on using the command line, Korg’s SDK for the Volca Sample (aka, Syro library) includes some pre-built example executables (for windows, mac and linux) that take their instructions from the command line and output. These are interesting tools, but a little difficult to use out-of-the-box. For the most part, I only use to send delete instructions.
There are a number of other tools for uploading samples to the Volca Sample. As the user base gets bigger and bigger, and more interest is being raised on this little machine. This is a small, non-comprehesive list of other software that can be used.
- Audio Pocket: This is Korg’s official uploader. It is great if you use Apple Software, but it hasn’t been ported to any other platform.
- Vosyr, by Frederikson Labs is another great app that runs on windows and mac. It is visually very appealing, and has the yet unique feature of allowing pattern editing (with automation!). It also allows projects (sample kits + patterns) to be saved, which is great. Unfortunately it does not have an open sample dialogue box, so you are forced to drag-and-drop every sample (not good when the app takes over 90% of your screen).
Dissecting the Upload Signal
As mentioned earlier, the way samples are transmitted to the Volca Sample resembles the way old Sinclair Spectrum computers used magnetic tape to save their data. Let’s dissect this a little bit. My sample kit contains only one sample (a TR707 bass drum taken from KB6.de) to be placed on position 0. The resulting wav file to be sent to the Volca is 680kb long (the sample itself is only 9kb). It’s spectrogram (using audacity) is shown bellow. As you can see, there is a hick frequency constant tone (the white line) that crosses the whole file. The data comes in 5 packages, and I’m guessing there is a starting transmission beep, then the instruction to clear slot 0, then the sample, and finally some form of clean up/end transmission.
The first 1.5s are always just the carrier signal, which runs quite hot (at about +2.8dB). The data itself is transmitted with in two frequency windows: one large window from about 100Hz to 10kHz, and another one from 10kHz to a little over 20kHz. And here lie all the problems: for once this signal is so hot that some consumer-grade sound cards may start to overdrive, and this will cause distortions, which will ruin the transmission. Also, any equalization will muffle the data transmission (specially the high-frequency region, since most equalization I know in consumer-grade products trend to emphasize the bass). Worst than that, any sound processing may bring some os these frequencies out of phase, and since the data is transmitted using very short (milisecond-long) beeps, it will kill your data. Perhaps this is why most errors appear only after 1.5s, which is the time the first data arrives at the Volca.
Tips, Tricks & Troubleshooting
Tip: This is more a personal preference, but I trend to make small sample kits using the Connecting with the Volca page (5 to 10 samples), and register them by type. For example samples 0-19 are usually a drum kit (or empty), PCM waves for melodires of granualar synthesis are usually stored between 20 and 30, bleeps and other noises are stored in the 40’s, volcal spinets in the 60’s and long loops (like the ones from the kaoscilator) are stored in the 90’s. Each sample kit only contains one type of samples, so I can change my drum kit without affecting any of the other categories. This also makes my wav files small and short.
Now, about the troubleshooting part. As it was mentioned, there are a number of things that can go wrong. This is my personal troubleshooting list. It covers most of the weaknesses I see in the protocol, but it is not 100% guarantied.
- Is the Volca firmware up to date? Check Korg’s website for updates. You can check your firmware version by turning the power on while pressing the [REC] key. At the time of writing, the latest firmware version is 1.30.
- Do you have any form of equalization on your pc/tablet/phone? This is the most common problem that goes noticed by users. Check for any equalization signs, such as Dobly, Dobly Surround, Beats by Dr.Dre, Bass Boost, Clear Sound, etc. Some consumer-grade cards may even have some form of hardware equalization, which cannot be disabled. In the latter case, you’ll need a usb sound card.
- Is the audio output connected to the Volca’s SYNC IN? Yes, I once was having trouble uploading samples because I plugged the jack into the SNC OUT by mistake.
- Are you using good quality stereo cables? They have to be stereo, and they have to be really good. Poor cables may sound ok, but may introduce some resistivity at certain ranges of frequencies, acting has filters. The only way to be sure about this is to test with a couple of cables of different brands and vintages and see if the problem goes away.
- Are you in air plane mode (for phones/tablets)? Sometimes sending or receiving something over wireless may interfere with your audio…
- Is any other software using the sound card? Notifications, alarms, ringtunes, etc… everything that may send a sound has the potential to ruin the transmission.
- Is the volume alright/is your sound card cliping? The Volca needs a hot signal, but if the signal is too hot, your sound card will start clipping. It is conceivable (although I never experienced it) that some sound cards will not provide a loud enough signal without distorting it. In that case, you have to use another sound card. However, for most soundcards, start by trying uploading with the volume at 100%. If there is an error, try 90%, then 80%, 75%, etc… eventually it should reach the sweet spot.
- Is the software you are using doing some form of eq/distortion? This came as a surprise to me: when I played a drumkit loader in audacity, the transmission stopped with the dcod error. However, the file worked fine when played using simpler programs such as mplayer or aplay. So, trying a different program may also help you solve your problems.
This ends the main series of Volca Sample Tutorials. For those interested, I also put together a guide on the Main Configuration Menu (Tutorial 7). However, I’ll keep posting new (smaller) videos on YouTube with Tips&Tricks for the Volca Sample.