High Dynamic Range (HDR) content standards such as HDR10, Dolby Vision, HLG (Hybrid Log Gamma) have been presented over a last couple of years, and these standards commonly have both wider dynamic range and color gamut than that of the legacy content. However, it is practically hard to fully satisfy the HDR content standards with the current TV technology. Although various standards for display metrology regarding HDR content standards have been released to evaluate the performance of HDR TVs, it has been controversial whether the performance obtained from measurements using artificial test patterns is well consistent with the actual performance of HDR TVs from which viewers are expecting. For this reason, we examined how the performance of HDR TVs is varied in case measurements are made using actual HDR contents instead of artificial test patterns in terms of peak luminance, black level, and Electro-Optical Transfer Function (EOTF) accuracy. Our investigation was carried out using two different types of TV, an Organic Light Emitting Diode (OLED) and a Liquid Crystal Display (LCD). As a result, while the measurements of the OLED TV were quite coincided with the measurements made from artificial test patterns, those of the LCD TV were not.