Cutting and Labeling

The cutting process is long, tedious, and frustrating because even after noise reduction the noise floor is much higher than I’d like.

For each note I have to find the onset, and then cut it at the closest zero crossing.

Screen Shot 2014-04-04 at 9.41.28 PM

I also have to find the end of the decay, which including the reverb, is subtle and can be difficult to ascertain. Again once I find the end point I have to cut it at the closest zero crossing.

Screen Shot 2014-04-04 at 9.40.58 PM

Once I have all the notes cut, I listen to them for any clicks and pops, meaning that I missed the zero crossings, and to make sure that I have all of the notes in order, with no doubling, or messed up notes. After that’s all said and done I can start labeling. Despite the samples being in different keys, I am labeling them in the spirit of the organ, where each stop, no matter the octave, is centered on C4. With the 61 Key manual that means that samples range from C1 to C6, in the bastardization of scientific pitch notation I am using to label my files. This system equates the physical location of the key on the manual to the label, as opposed to the pitch.

Screen Shot 2014-04-04 at 10.43.11 PM


I’m not going to make any excuses, there is one reason why I am not going to upload my files tonight, I don’t feel it’s good enough for submission. This was caused by a combination of things, particularly my poor scheduling, in regards to balancing this, other projects, and work, but also because I made some early mistakes in the process (nosie floor, and sample length) which have been giving me trouble for a while now, but in my stubborness, I’ve not admitted it.

The noise floor is making it so that there is a noticeable dropout between notes, when played through headphones. This could possibly be fixed if I were to create a VST from scratch which filled the in-between notes with the room tone of the church. But that’s something that’s over my head.

Also, because I recorded long enough samples to make looping easy, I am having trouble making notes which are not held for at least the recorded duration of the sample use the natural decay recorded in the sample. I may have to edit shorter versions of the note and rely more on looping.

The point is, my files won’t be uploaded until Friday evening. I take full responsibility for my lateness, including any penalties that come with that. There will also be two more blog posts which chronicle my process, including the editing of the samples and the patch creation, which should go up around the same time as my files.

I apologize for this, and regret not being more open with these problems.

Recording Process

The recording process was both easier and more difficult than I thought it would be. In my infinite foresight, I failed to confirm that I had assistance during my session, requiring me to transport, set up, and recorded solo.


I used an X/Y stereo mic setup, as previously described, running it into ProTools through the MOTU 896 at 96k. I used the Earthworks SR30 small condenser microphones, which are cardioid microphones with exceptional response.



I placed my microphones in the pews, where I started with facing them towards the front of the congregation, but due to the (suprisingly) low level of the organ, I re-positioned them facing the pipes. I’ve already spoken about the awful room noise I experienced. I knew it was going to be a problem while I was in the space, but after 45 minutes of trying to compromise a position between accurate sound of the room and lower noise, I resigned myself to the fact that I couldn’t position out the hiss, hum, and rumble of the room.


Because the recording setup was so far from the manuals, before each stop I had to start recording and run upstairs to the manuals, leaving about 40 seconds of pre-roll on each file. Once I got to the manuals, I played each note for approximately 8 beats, or 2 bars, at somewhere around 120bpm.


There were some oddities to the organ which I noticed as I was working, one being that certain stops had leakage, where the stop being open inherently produced a low level tone from air leaking into certain pipes.This, like the noise floor, wasn’t something I was able to fix with my limited time and inexperience with organs. I also noticed that certain notes had slight variations in the attack, which are hard to describe other than to say that they were dissimilar to the other notes. After discussing my session with Gordon, who is the caretaker of the Organ, he told me that the difference in attack was caused by dust, and that the organ was over-due for a cleaning. Again, there wasn’t too much I could do about it at the time.


In conclusion, the problems with the noise floor, the air leakage, and the inconsistencies of the attack, leave me to characterize my sample gathering as somewhat of a disappointment, if not an outright failure. The only saving grace being that the recordings are incredibly true to the space, and the organ, even if that means that they don’t sound as crisp and clean as I had hoped.


Choosing a Sampler 2: Choosing a Sampler Harder

Despite my uncertainty, I’ve settled on the sampler I will be using to build the patches.

Kontakt, while expensive, is feature packed, and highly regarded as one of the top software samplers available. The amount of material available to guide me through the process of creating my patches will hopefully make this process as painless as possible, and was in large part the reason I chose to move forward with Kontakt.

This means that I will be looking into the possibility of trans-coding the kontakt files for use in other samplers, although my plate is pretty full right now and it’s not my biggest priority.