Saturday, April 15, 2023

Revisiting the Hebrew song "Birthday"

I wrote two and a half years ago about this song. I had occasion to listen to the recorded version a few months ago and did not like it at all: it seemed too fast and the arrangement was too 'busy'. I started work on a new arrangement that would be much slower, 80 bpm instead of 100 bpm. I'm not sure about the original key of this arrangement: the previous version started in B minor (mutating in the middle to A minor), whereas this new version seemed to start in C#m minor.

I tried recording the vocals to this in a few barren sessions without getting very far. The major problem is that the tessitura is very large, from the absolute bottom of my vocal range until the absolute top. Whilst I can normally reach the high notes (although without much power), the low notes are very much hit and miss. I even tried adapting my current arrangement such that I could place the original vocal on top, which is when I discovered that I had removed breathing noise from the original vocal in such a way to cause 'pumping'. This didn't sound very good and as I couldn't achieve reasonable sounding tone settings, I abandoned this idea too.

Considering the song one night in bed, I hit upon a rather radical solution: record the lower parts of the vocal in D minor, then record the high parts in A minor. By means of vocal tuning software, I could lower the first part and raise the second part, having the finished song in C minor. So today I set out to accomplish just that. It turns out that almost all of the song was sung in the first part, with only a few lines sung in the second part. After several takes and a great deal of technical manipulation, I had a completed version.

Listening to it with fresh ears, it seems obvious to me that certain lines were recorded at a separate time (the high lines); this is primarily because there is less of a pause between the end of a line sung by vocal #1 and the beginning of the line sung by vocal #2. Tonally there is also a difference, but this can be explained away by the different vocal register.  Also, there were a few lines towards the end of the first verse with which I had trouble, so I sang these again at the end (in C minor): it seems that my vocal technique had changed over the hours working on the song, as this part sounds like the pasted high part.

Cheating, but the end product is what counts, and quite probably no one else but me will know how I managed to record this song. A clear conclusion, though, is to try and find songs with a smaller range; this was an undocumented conclusion to my previous song, 'Wonderful days', from two weeks ago.



This day in history:

Blog #Date TitleTags
46915/04/2012[no title]MBA
102215/04/2017Return to KaizenDBA, Kaizen
121015/04/2019CommunicationsMobile phone
130915/04/2020Italian holiday filmHoliday, Home movies, Covid-19

No comments: