The refractive index profiles in Fig. 2 were incorrect. The proper designs are presented here with additional comments.
© 2008 Optical Society of America
It was recently pointed out to us that the refractive index profiles given in Fig. 2 were incorrect ; they correspond to a different spectral target. The proper refractive index profiles of the “house-like” filter are given here. The total optical thickness is increased to 32µm. This was necessary to get a flat reflectance across the long-wavelength tower. However, the design substitutions do not affect the main conclusions of the paper.
We also would like to clarify the SWIFT-phase operation and the iterative process for the design of the “house-like” filter. A portion of the long-wavelength tower was not included in the SWIFT-phase calculation. This improved the refractive index distribution and the spectral fit in Fig. 2 (b), and potentially made the final refractive index profile more predictable since the required corrections were smaller. Moreover, the refractive index sampling interval was relatively crude (91.84 nm in optical thickness) in Fig. 2 (b). During the subsequent iterations, the refractive index sampling interval was reduced to improve spectral fit.
More information about the refractive index profiles can be obtained by contacting the authors.
References and links
1. X. Cheng, B Fan, J. A. Dobrowolski, L. Wang, and Z. Wang, “Gradient-index optical filter synthesis with controllable and predictable refractive index profiles,” Opt. Express 16, 2315–2321 (2008). [CrossRef] [PubMed]
2. The authors are grateful to Dr. P. Verly for spotting the mistake and checking the final designs.