Pancake hood any good? A casual experiment

One hears a lot about how the pancake-sized hoods that Canon sells for e.g. RF 35mm, EF 40mm, and EF-M 22mm are useless because they are so tiny. But I've never seen anyone actually trying to empirically prove that they are indeed snake oil.

So today, I went to a park with my RF 35mm F1.8 STM to test the following configurations for flare:

1: A plain RF 35mm F1.8 STM. with a front filter. Reviews suggest this lens is relatively flare resistant, so a front filter is added to make the flare more visible.

2: RF 35mm + filter + the allegedly useless EW-52 pancake hood.

3: RF 35mm + filter + a "real" hood: Nikon HN-3, a steel screw-in hood for older F mount 35mm lenses, meaning it should be compatible with the RF's frame of view.Test shots follow. All were shot in RAW, and processed with default parameters in Lightroom without manual editing. The Sun is about half a frame above the top right corner of the photos.

No hood. The flaring is noticeable enough that maybe I didn't need that front filter to do this test.

Pancake hood. The flaring is not gone, but the contrast seems to have improved in the top right quarter.

"Real" hood. It's clearly worse than the pancake hood, and not much better than hoodless!

So, in this test, the pancake hood turns out to be not only not useless, but actually better than the "real" hood!

I'm frankly surprised by this outcome, and I only believe it because I ran several trials and they all look the same. I actually expected the pancake and real hoods to perform the same, as simple geometry dictates that the pancake hood should be just as good at blocking stray incident light as a larger hood situated farther away from the front element, but I do not understand why the former should outperform the latter.

Sorry for all those bad pics. This one is a bit better, I think.

Interesting ... thanks for testing it 

Its no surprise for me since I posted similar assessments several times here. The Light trap construction is quite effective and reduces the harming solid angle significantly.

Without the filter results would be even better ...

J A C S wrote:

That Nikon filter does not look very good, it is probably too reflective.

Well, it must be, based on this experiment. But you wouldn't be able to say that just by looking at the hood, which is painted black and serrated on the inside. Note that the Nikon hood photo is overexposed compared to the other two, making the hood look greyer than it actually is.

Incidentally, the build quality of the Nikon hood blew me away. I'm not being sarcastic. Literally a solid hunk of thick iron, it feels like it's built to do double duty as an army trench shovel. All for $6! Canon certainly doesn't build hoods like this, at any price. Which makes this result all the more disappointing.

The results could depend on the direction and on the character of the stray light.

True, which is why I called it a "casual" experiment rather than a carefully controlled one. This is the result from a single situation where the flare is highly aggravated. I'd love to read about similar comparisons in other situations, and with other hoods that are supposed to be better.

J A C S wrote:

quiquae wrote:

J A C S wrote:

That Nikon filter does not look very good, it is probably too reflective.

Well, it must be, based on this experiment. But you wouldn't be able to say that just by looking at the hood, which is painted black and serrated on the inside. Note that the Nikon hood photo is overexposed compared to the other two, making the hood look greyer than it actually is.

The hoods I use have black linen inside. This one looks a bit too shiny.

I have compared plastic hoods (no linen) and Canon ones. The difference is very clear.

While linen is a good light absorber, that isn't the whole answer here because EW-52 (the Canon pancake hood I used here) doesn't have linen either.

Canon only uses linen for some of its hoods. It isn't necessarily a case of Canon cheaping out on consumer grade lenses, by the way: EF-S 18-135 STM and EF 35 IS feature linen, while RF 24-105L and RF 50L do not.

https://www.dpreview.com/forums/post/62355126

Thanks for this test. I too have the pancake hood but I am curious about your rationale for putting the pancake hood on top of the filter. I was unsure about this and asked Canon customer service about it and they said filters should go on top of this hood. Not sure it makes a difference for one filter but I’m curious about tour thoughts on this.

A forward shift of the hood (by introducing a filter in between) can cause vignetting since it is narrowing the possible free angle ...  Its like putting more than one filter in fron of your lens especially harmfull in wideangles ...

Very good point, and the reason I put the pancake hood on top is because the Nikon hood had to go on top of the filter (being a screw-in), and I wanted to keep things consistent. As you can see, this pancake hood doesn’t cause noticeable vignetting when used with one thin filter.

I normally do deploy the hoods behind the filter, like Canon told you, mainly because that makes it easier to screw the filter off when required.