Angstrom method is a steady-state measurement for thermal diffusivity alpha using ac heating. Since thermal conductivity lambda is a better-known quantity, measured diffusivity a alpha is sometimes transformed into lambda based on relation lambda = C-v alpha using recorded or DSC measured C-v. However, Angstrom method itself is principally possible to extend to specific heat measurements, yet the accuracy is not promising due to the complexity of heat loss. Here we present a modified method for simultaneous measurements of thermal diffusivity and thermal conductivity with high accuracy by taking heat loss into account. A linear heat loss term m(2)T is introduced into the diffusion equation and the thermal conductivity lambda can be directly measured instead of specific heat. The measured thermal properties of commercial graphite sheets agree well with their nominal value. The origins of m(2) are also discussed, m(2) can be divided into amplitude independent and dependent part. From the basics of radiation and convection, the first-order radiation and convection comprise the amplitude independent part, while the dependent part includes higher order (dominated by second-order) radiation. Although the amplitude independent part agrees well with the extrapolated value of m(2) at zero amplitude, the second-order radiation cannot fully cover the measured amplitude dependent part. This discrepancy is further explained by floating temperature baseline variation due to residual heat during heat oscillation. (C) 2015 Elsevier Ltd. All rights reserved.