Speedlight SB

If you don't mind manually setting the exposure information (ISO, aperture) from the camera into the flash, the SB-28 will take a fine picture with the D50. I have an SB-26, and it works very well in A mode. In the end, a flash is a flash is a flash, but advanced capabilities like iTTL make it a lot easier to get better flash exposures under more circumstances more reliably. With expertise or trial and error, you don't even need automated flash; you could always do everything with the SB-28's manual mode. But, for real life use, the SB-600 or 800 are certainly a lot easier to use and make it more likely to not miss shots.

I just bought an SB-800 because:

1) Having to set all of the information into the SB-26/8 is inconvenient and error-prone.

2) iTTL offers the ability to get more accurate auto-metering under more circumstances (such as off-camera macro shots) a lot more easily.

3) The SB-800 offers a number of other benefits through the included diffuser and color balance filters.

I have since discovered that the SB-800 works very well along with the SB-26 using the SB-26's "D" remote triggering mode (triggered off the main flash). It is very easy to use the SB-800's iTTL to meter the main subject very accurately and use the SB-26 for auto (or manual) fill (e.g., to light the background or to provide secondary lightling to the main subject). As I understand it, the SB-28 does not have the "D" remote triggering mode.

As far as your photos are concerned, I think they are generally pretty over-exposed. You would have to use proper exposure compensation on the SB-800 in order to get a good flash exposure under the challenging conditions that you faced. However, I believe that the SB-800 would have produced a more balanced image in auto-everything matrix mode (if that is your objective). Exposure compensation of about -2/3 EV on the SB-28 would have helped your picture, too. It looks like you used direct flash as opposed to bounce flash. Assuming there was a white or light ceiling in the room, bouncing at about 60 degrees with some diffusion or a reflector would have produced more natural-looking results from either the SB-28 or SB-800 (with -1/3 to -1 EV exposure compensation for matrix metering) Best results might have been SB-800 TTL mode, spot metering on the dog with 0 or slightly positive exposure compensation since the dog is very light colored.

Doug