Friday, August 2, 2013

Unusual conversations – analysis methodology

I have presented the results of various analysis throughout this blog without going into any details about the analysis methodology. However, this time, (with no intentions of bragging) I will elaborate the analysis method as we will try to drive some analysis conclusions at the end of this section. Knowing how the analysis was performed will be a useful context to relate to the results presented at the end.
  1. Though Raja’s compositions run to over 4,500, a database of 1.600 songs were considered for this analysis (a 35% sample of the population). The sample set has all the languages set to music by Raja other than Marathi
  2. The 1,600 songs covered in this analysis span all the 5 decades – 1970s, 1980s, 1990s, 2000s, 2010s.
  3. A scan of the 1,600 interludes is done to isolate ‘usual’ and ‘unusual’ conversations
  4. The focus is more on the ‘unusual’ conversations. Every ‘unusual conversation’ is checked for the rules of C&R, and a segment is isolated and recorded in a spreadsheet. A single interlude can have more than one C&R segment
  5. For every identified segment, the two instruments involved in the conversation is recorded in the spreadsheet, along with the name of the film and the release year
  6. The lead column in the spreadsheet is ensured not to be violin, guitar or flute
  7. A pivot analysis is done on the lead instrument to get a list of unique combinations and those become the 14 categories I plan to present. A sorted list based on the lead instrument is kept for reference in creating the clips
  8. The collection of data was done over 6 months and the assumption is that the sample database is a close representation of Raja’s output and is good enough for predicting/analyzing trends related to this technique
  9. Based on the result set, clips are isolated from the songs and assembled for presentation with the appropriate naming convention for easy cataloging
  10. Analysis conclusions are derived from the analysis spreadsheet to have data support to the conclusions

I hope the analysis helps reset some perspectives about Raja and his use of C&R technique in his music, such as:
  1. Not everything he has done with this technique revolves around the violin, guitar or flute
  2. The breadth and depth of use of this technique can be matched by no other composer, even if we consider his ‘unusual’ conversations
  3. Some of the instrument combinations are ‘unthinkable’ before his time, and few composers have tried these even after he has demonstrated them in his interludes

No comments: