We present number counts , luminosity functions ( LFs ) and luminosity densities of galaxies obtained using the Sloan Digital Sky Survey Sixth Data Release in all SDSS photometric bands .
Thanks to the SDSS DR6 , galaxy statistics have increased by a factor of \sim 9 in the u -band and by a factor of \sim 4 - 5 in the rest of the SDSS bands with respect to the previous work of ( ) .
In addition , we have achieved a high redshift completeness in our galaxy samples .
Firstly , by making use of the survey masks , provided by the NYU-VAGC DR6 , we have been able to define an area on the sky of high angular redshift completeness .
Secondly , we guarantee that brightness-dependent redshift incompleteness is small within the magnitude ranges that define our galaxy samples .
With these advances , we have estimated very accurate SDSS DR6 LFs in both the bright and the faint end .
In the ^ { 0.1 } r -band , our SDSS DR6 luminosity function is well fitted by a Schechter LF with parameters \Phi _ { * } = 0.90 \pm 0.07 , M _ { * } -5 log _ { 10 } h = -20.73 \pm 0.04 and \alpha = -1.23 \pm 0.02 .
As compared with previous results , we find some notable differences .
In the bright end of the ^ { 0.1 } u -band luminosity function we find a remarkable excess , of \sim 1.7 dex at M _ { { } ^ { 0.1 } u } \simeq - 20.5 , with respect to the best-fit Schechter LF .
This excess weakens in the ^ { 0.1 } g -band , fading away towards the very red ^ { 0.1 } z -band .
A preliminary analysis on the nature of this bright-end bump reveals that it is mostly comprised of active galaxies and QSOs .
It seems , therefore , that an important fraction of this exceeding luminosity may come from nuclear activity .
In the faint end of the SDSS DR6 luminosity functions , where we can reach 1 - 1.5 magnitudes deeper than the previous SDSS LF estimation , we obtain a steeper slope , that increases from the ^ { 0.1 } u -band , with \alpha = -1.01 \pm 0.03 , to the very red ^ { 0.1 } z -band , with \alpha = -1.26 \pm 0.03 .
These state-of-the-art results may be used to constrain a variety of aspects of star formation histories and/or feed-back processes in galaxy formation models .