When working with digital photograhpy in SL, I find it helpful when planning workflow, to start at the end and decide how the image will be presented to the audience. From there, then in what way it might need editing/modifcation/adjustment between the moment of capture it till it reaches that audience.
I will give some examples to help illustrate how it is possible to work with the challenge these "black lines," for two diffrenent, and common, workflow situations.
1. The image will be caputered at the same resolution it will be presented and/or litlle or no image editing is required.
If this is the case, then an very usul approach is to utilize a screen capture program. Set the anti-aliasing to the highest levels possible along with all the other graphics settings preferred, set shot, close the UI (control + atl + F1) and capture the screen. This is also very useful in situations where the computer is able to generate Depth of Field and/or Shadows and yet the Snapshot feature inside the viewer will not produce/record them (common with AMD/ATI graphics cards).
Ths is an excellent article in freeware Screen Capture programs:
2. The image will be shown at high resolutions (Full HD for example) and it will require image editing/process work.
Here what needed is the cleanest possible base image to work with, and one that is also hopefully LARGER than the final output.
First let's get the larger image by enabling High-res Snapshots. This will allow you to take images that are up to 6016 x 6016. This is an incredibly large image and very helpful for even print work. This link shows how to use this feature:
Now, the issue of jaggy lines in final output. If the final output is LESS than the orginal image capture size, then when you lower the resolution of the image for final output your are performing anti-aliasing. Your graphics program will resize the image and preform a smoothing operation (resampling) in the process. In Photoshop, the options for Resampling the image are found in the image size menu (control + alt + I). Most image processing applications have similar features. Having this very large base file to wrok with, will make is very easy for the program to produce excellent results.
To ENABLE anti-ailaing AND High-res Snapshots will likely produce images that are completely indisiquishable in this situation.
In all, most SL photography is often for used in world or on photosharing sites like Flickr. For any inworld work, it is unlikely that users will notice if anti-aliasing is used or not. In the case of Flickr, images are most often veiwed 'in-stream' at about 500 pixels per side (max - see: http://www.flickr.com/help/photos/). Again, in this case, even if you upload higher resolution images to your Flickr stream, the website will resize (and more than likely - resample) the images for your stream page. Users can only see higher resolutions if they specifically request so.
I hope that helps.