The final photometric accuracy depends on a proper background subtraction and color correction. Therefore, we give in Table 2.2 the observing mode used for verification of the calibration accuracy. In particular for faint sources, the simpler observing modes (off-on) are much less accurate than the ones optimised for faint source observations.

The calibration curves of the internal calibration sources (FCS) on celestial standards were established flux range by flux range using different sets of standards with appropriate brightness. This takes into account possible non-linearities of the ISOPHOT detector/read-out system. Therefore, we checked the calibration accuracy per flux range as far as possible.

For the checks measurements on known (calibration) sources were used, which are as far as possible independent of the data set used for establishment of the calibration curves of the internal calibrators. This was not always possible, but due to the reduncancy of calibration sources per flux decade, inclusion of a single measurement does not mean a circular argument in deriving the calibration accuracies. The names of the calibration sources used for the verification measurements are listed in Table 2.2, too. Due to limitations both in the number of test cases processed for scientific validation and in availability of suitable test cases not all observing modes could be addressed. In case of several examples covering the same flux range and showing somewhat different accuracy numbers the more conservative one should be used.

For PHT-S the quoted numbers are derived from a general assessment of the accuracy in deriving the spectral response function. Note that for PHT-S there is a direct conversion from the measured signal into flux density per wavelength range $F_{\rm \lambda}$ using this spectral response function established from a set of standard stars. There is no flux range dependent calibration using the internal calibration reference sources.

**Table 2.2:** Absolute and relative (filter-to-filter) calibration accuracies per detector achieved for various flux ranges in multi-filter mode. Hyphens indicate that no test case was available. Bold values indicate a significant improvement w.r.t. OLP V6.0, the version from which the time counter of the proprietary data period started. For the wavelength range of the detector system, please refer to Table 1.1.

detector	AOT mode	flux range	abs. accuracy	rel. accuracy	standard
		[Jy]	[%]	[%]

P1	P03 off & on	300 - 6000	10	15	$\alpha$ Boo
P1	P03, P05 off & on	70 - 1250	10	20	$\gamma$ Dra
P1	P03, P05 off & on	50 - 1000	10	10	Sirius
P1	P03 nodding	2 - 10	30	-	HR 6817
P1	P03 nodding	2	35	-	HR 5986
P1	P03 nodding	< 0.1	20	-	HD 172323
P2	P03 off & on	100 - 200	$\le$ 5	5	$\alpha$ Boo
P2	P03, P05 off & on	25 - 40	10	10	$\gamma$ Dra
P2	P05 off & on	$\sim$ 20	15	-	Sirius
P2	P03 nodding	3 - 6	$\le$ 25	-	HR 6132
P2	P03 nodding	3.5	$\le$ 5	-	HR 7310
P2	P03 nodding	0.5	$\le$ 10	-	HR 5986
P3	P03 off & on	10 - 25	10	20	$\alpha$ Boo
P3	P03 nodding	6 - 20	15	-	$\alpha$ Tau
P3	P03, P05 off & on	$\sim$ 5	10	30	$\gamma$ Dra
P3	P03 nodding	$\sim$ 3.5	15	-	Sirius
P3	P03 nodding	0.6	$\le$ 15	-	HR 7310
P3	P17/18/19	< 0.2	30	-	HR 5981
C100	P22 off & on	6 - 25	5	10	$\alpha$ Boo
C100	P22, P25 off & on	$\sim$ 5	10	-	$\gamma$ Dra
C100	P22 nod & minimap	$\sim$ 3.5	15	-	Sirius
C100	P22 minimap	0.2 - 0.6	5	-	HR 7310
C100	P37/38/39 4 times	< 0.2	30	-	HR 5981
C100	P37/38/39 2 times	< 0.1	30	-	SAO248381
C100	P37/38/39 2 times	< 0.1	30	-	IRC+60228
C200	P22 minimap	600 - 900	< 10	-	Uranus
C200	P22 minimap	50 - 120	25 (> 170 $\mu$ m)	-	Ceres
	P22 minimap, P25		15 (< 170 $\mu$ m)	-
C200	P22 minimap	$\sim$ 0.5	15	-	$\gamma$ Dra
C200	P25 off & on	$\sim$ 0.5	30	-	$\gamma$ Dra
C200	P22 minimap	0.2 - 0.4	30	-	HR 1654
SS	P40 off & on	2 - 5000	< 30	< 30
SL	P40 off & on	1 - 5000	< 30	< 30

$\fbox {\fbox{\parbox{15.5cm}{The goals for the improvement of the absolute and ... ...em[\bf $\cdot$] absolute accuracy chopped modes: $<$\space 30\%\end{itemize}}}}$

2.1 Absolute and Relative Filter-to-filter Calibration Accuracies for Point Sources