I'm extremely happy with the performance of my SB-400 when taking photos of family and friends. The bounce works really well (even in portrait mode without a diffuser, in many cases). However, on occasion I'll see a photo where the light is coming a bit too much from above. Upward bounce shots are always going to have an overhead bias, and they usually look great in spite of it, but some times (every time for some people) you'll want a little extra light from the front, which will require some tweaking.
If you're like me, you have an appreciation for lightweight, low-cost (preferably free) alternatives for enhancing your flash output. I've tried many DIY techniques, but up until now I hadn't done a close side-by-side comparison of their output. So to help make things a bit more objective, I took dozens of comparison shots using different DIY diffusion techniques and different flash settings. I've compiled the most meaningful results into the image below. These shots were taken about 4 feet from the subject, with a wall about 5 feet behind the subject (a common social snapshot scenario I'd guess). All diffusers had the flash in the 90 degree position.
When looking at these photos I tend to examine the shadow under the bear's chin, since this is the type of thing we are hoping to meddle with. Also take note the texture of the fur and the amount of reflection it shows (this would be better if it were skin, but I don't think my wife wants to be seen on the internet, hah!).
Both white bounce cards disappointed me in this test. I was hoping to increase the surface area of the light in order to soften the shadows, but what happened is that both cards increased the intensity of the forward-facing light too much. You can see this in the photos; the brightness of the bear is overwhelming the ambient light of the wall behind him. The white cards look better than direct flash, but not by much.
To remedy the extensive light intensity I tried to find something less reflective for diffusion; in this case, some black construction paper. I think the switch to black actually worked. The forward output is lessened, and you end up with a nicely-lit bear face that balances well with the ambient light.
One thing I also noticed about the paper diffusers is that their output varied wildly depending on the precise angle the paper was at. At 90 degrees you get one look, but if it flops back 10 to 20 degrees you might as well not be using it at all! This inconsistency is definitely a concern.
What surprised me most was the results of the flash tilt settings. The 90 degree setting demonstrates what we are trying to improve: a generally attractive image with a few too many shadows underneath the bear's chin and tufts of fur. However, one tiny click of the flash position and things improve dramatically! The 75 and 60 degree settings both do an excellent job of illuminating the bear from more angles without drowning out the ambient light. In fact, I'd go so far as to say that the 60 and 75 degree settings are tied with the black card for the best looking of the bunch (the 60 degree looks almost identical to the black card!).
What I also found interesting in my tests is that the 60 and 75 degree settings (no diffuser) looked great regardless of whether I was only a few feet away or whether I was across the room. I had always assumed that the bounce angle was only good for projecting light further for cases where you were relatively far from the indoor subject. I was wrong! What is actually happening here (and you can see it if you look in the mirror) is that the angled settings are actually exposing a sliver of the primary light source to the subject. This works great!
It's worth noting is that, despite a much larger surface area, the large white card did not significantly improve the look of the shadows. If you look at the shadow on the wall, it is slightly better than the small white card, but if you compare it to the bare 60 degree setting you'll see that not only does the bear's chin shadow look just as good, but the better light balance makes the wall shadow look better too!
So one of my conclusions is that the surface area difference between the bare flash and the large white card is not significant enough to make a noticeable difference on a complex subject (a bear, at least) at this distance. Yes there is a difference, but unless your subject is standing right up against a wall it is difficult to see. Have a look for yourself:
The 60 degree setting has more defined edges to the shadow, but again, it still looks better than the large white card. The lesson here is to avoid taking portraits with people standing against a wall unless you have a fancy light setup. Or you can try a large black card for less forward intensity. That'll be next on my list.
If you're like me, you have an appreciation for lightweight, low-cost (preferably free) alternatives for enhancing your flash output. I've tried many DIY techniques, but up until now I hadn't done a close side-by-side comparison of their output. So to help make things a bit more objective, I took dozens of comparison shots using different DIY diffusion techniques and different flash settings. I've compiled the most meaningful results into the image below. These shots were taken about 4 feet from the subject, with a wall about 5 feet behind the subject (a common social snapshot scenario I'd guess). All diffusers had the flash in the 90 degree position.
When looking at these photos I tend to examine the shadow under the bear's chin, since this is the type of thing we are hoping to meddle with. Also take note the texture of the fur and the amount of reflection it shows (this would be better if it were skin, but I don't think my wife wants to be seen on the internet, hah!).
Both white bounce cards disappointed me in this test. I was hoping to increase the surface area of the light in order to soften the shadows, but what happened is that both cards increased the intensity of the forward-facing light too much. You can see this in the photos; the brightness of the bear is overwhelming the ambient light of the wall behind him. The white cards look better than direct flash, but not by much.
To remedy the extensive light intensity I tried to find something less reflective for diffusion; in this case, some black construction paper. I think the switch to black actually worked. The forward output is lessened, and you end up with a nicely-lit bear face that balances well with the ambient light.
One thing I also noticed about the paper diffusers is that their output varied wildly depending on the precise angle the paper was at. At 90 degrees you get one look, but if it flops back 10 to 20 degrees you might as well not be using it at all! This inconsistency is definitely a concern.
What surprised me most was the results of the flash tilt settings. The 90 degree setting demonstrates what we are trying to improve: a generally attractive image with a few too many shadows underneath the bear's chin and tufts of fur. However, one tiny click of the flash position and things improve dramatically! The 75 and 60 degree settings both do an excellent job of illuminating the bear from more angles without drowning out the ambient light. In fact, I'd go so far as to say that the 60 and 75 degree settings are tied with the black card for the best looking of the bunch (the 60 degree looks almost identical to the black card!).
What I also found interesting in my tests is that the 60 and 75 degree settings (no diffuser) looked great regardless of whether I was only a few feet away or whether I was across the room. I had always assumed that the bounce angle was only good for projecting light further for cases where you were relatively far from the indoor subject. I was wrong! What is actually happening here (and you can see it if you look in the mirror) is that the angled settings are actually exposing a sliver of the primary light source to the subject. This works great!
It's worth noting is that, despite a much larger surface area, the large white card did not significantly improve the look of the shadows. If you look at the shadow on the wall, it is slightly better than the small white card, but if you compare it to the bare 60 degree setting you'll see that not only does the bear's chin shadow look just as good, but the better light balance makes the wall shadow look better too!
So one of my conclusions is that the surface area difference between the bare flash and the large white card is not significant enough to make a noticeable difference on a complex subject (a bear, at least) at this distance. Yes there is a difference, but unless your subject is standing right up against a wall it is difficult to see. Have a look for yourself:
The 60 degree setting has more defined edges to the shadow, but again, it still looks better than the large white card. The lesson here is to avoid taking portraits with people standing against a wall unless you have a fancy light setup. Or you can try a large black card for less forward intensity. That'll be next on my list.
There are lots of variables: skin color and texture, room size, distance from subject. In a normal-sized house I've found that light from the bounce flash is enough to overpower the ambient light. If the the ambient light is unusually strong then of course the rules change. You'll be able to tell pretty quickly if you like the results.
