Not really an issue, but one of the consequences of the noise being so much higher in the JK images in comparison to the i. there are faint but distinct and probably real i-band sources that get diluted into fuzzy green blobs in the colour PNG.
This is the primary motivation for pushing the QD harder in the SG and FAQ, I think.
It also means that a comparison between VICS82 and SDSS+CS82 (as Anu et al tried already) will be interesting.
Not really an issue, but one of the consequences of the noise being so much higher in the JK images in comparison to the i. there are faint but distinct and probably real i-band sources that get diluted into fuzzy green blobs in the colour PNG.