J-band photometry
=================

SNe  Instrument+filter  Standard star system 
--------------------------------------------
99Q  SOFI+J,Keck+J      Persson              
99ff NIRC+J             Persson - see #1     
00fr ISAAC+Js           Persson              
--------------------------------------------

#1 We calibrated the J-band data from VLT and UKIRT using observations
of 10 standard stars from Persson et al. The Keck data were calibrated
by transferring the photometric ZPs to a galaxy from the UKIRT
observations to a galaxy in the field of SN 1999fn.

A larger sample of A0 stars to check the photometry
===================================================

The sample was selected from the Hipparcos catalogue and cross matched
with the 2MASS catalogue.

I've chosen those which are more that 30 degrees from the plane and those
which have B-V < 0.05.

The median colours of the remaining 184 stars are:

B-V =  0.010
V-J =  0.051
J-K = -0.004

If the median B-V colour is due to extinction, the extinction corrected
colours are:

B-V =  0.000 by definition
V-J =  0.029
J-K = -0.009

=> An offset of 0.029 magnitudes between the IR (Persson) and Optical
(Johnson) systems.

Transformations to the Bessell system
=====================================

Following the transformation given in

http://www.astro.caltech.edu/~jmc/2mass/v3/transformations

one can derive the following relation J_LCO (Persson) and J_BB (Bessell and
Brett) is

J_BB=J_LCO + 0.032*(J-K)_BB +0.036,

which, for an A0 star, simplifies to

J_BB = J_LCO + 0.036

This is very close the offset of 0.029 magnitudes inferred above.

Conclusions
===========

Since the photometry of all the SNe are tied to the Persson system, it
may be justified to add 0.036 magnitudes to the J band magnitudes of
all SNe. It is not clear if Tonry, Reiss and other collaborators have
done this, but I doubt it.

If we choose not to apply the correction, then a suitable
systematic error for IR-optical colors should be 0.05 magnitudes and
this should be added to the systematic error which is listed in table 8.

If we choose not to apply this correction, then the systematic error
is 0.02 magnitudes.

Transformation to the natural system
====================================

However, there is an additional offset because the filters
will not exactly match those used in the above equations and
standards stars are typically not A0V stars.

We make this additional transformation by adjusting the
ZPs and this done by integrating theoretical spectra through
experimental filter curves.

As an example, the ISAAC observations were done in the Js filter
and the standard star system was the LCO system. The standards in
this system are G0V to G4V stars, so there will be a non-zero 
correction to the ZP. The corollary being that if the standards were
A0V stars, there would be no correction.

For the observations of Beethoven, the correction to the ZP is

-0.012 magnitudes if we want to work with LCO system and
-0.012 magnitudes if we want to work with BB system and

This is a correction which is already applied to the data. See
section 4.1 of Serena's paper.