Friday, April 20, 2007


The rare times I write a scientific article, I spend hours looking for a non-silly starting sentence, and rarely I found one. An article written by Canet at al. (Anal. Chem. 1990, 62, 864-867) started with an apparently innocuous statement, whose implication is deep indeed:
Phasing an NMR spectrum [...] is still time-consuming and tedious with most existing software.

They hit the spot! Manual phasing is not difficult per se, it's difficult because those lazy programmers have not worked hard enough. If they make appealing interfaces, ergonomic controls, faster programs, simpler documentation, etc.. manual phase correction can become an easy task. The second sentence is not clear even if put into historical context:
This is generally accomplished by trial and error...

What they probably meant is that their computer was so slow and unresponsive that they renounced using the interactive adjustment. They probably used to correct the phase by directly typing the value of the parameters. If this was the case, however, the correct term would be "systematic search", or "iterative adjustment", not "trial and error". It would literally take hours in the latter case! The third sentence is mostly wrong:
The criterion to be met is left to the spectroscopist’s appreciation or experience.

Don't be mislead!

  1. The criterion is clearly stated: the phase is correct when all peaks are positive and completely symmetric.
  2. If properly taught, manual phase correction can be learned in one minute.
  3. It does not require a degree in physics.

The article then goes on describing yet another recipe for automatic phase correction, subsequently faded into oblivion. The most remarkable thing about the paper is that it was signed by four authors, which is the world record in the field.


Post a Comment

<< Home