Canon 24-105mm f/4L USM II - Testing the Image Stabilization
Anna does not believe in subscription based software. I have been trying to convince her to get Lightroom CC. But she insist on getting the DVD ROM version of Adobe Photoshop Lightroom 6. So we went to Computer City, in the heart of the Wanchai, Hong Kong.
Contrary to popular belief, Hong Kong is not the shoppers paradise that most are led to believe. Although it is true that one can probably find whatever they need in Hong Kong, Internet shopping in the US has Hong Kong shopping beat - hands down! Needless to say, Internet shopping isn't a thing in this side of the bigger pond. So we had to do it the hard way - searching from shop to shop - in a mall dedicated to all things computers.
It was under this needle-in-the-haystack predicament that I decided to test the Canon 24-105mm f/4 L USM II... seeing that I have one with me... being the only reason why I'm reviewing this lens...
Don't sound so enthusiastic, would ya!
Okay! So I've never been particularly interested in the first version Canon 24-105mm f/4, especially with regards to its performance and rendering. At best, it was a handy lens to have, if you needed to go moderately wide to moderately tight, and if you had sufficient available light. A lot of ifs, if you ask me. Not exactly high praise or interesting too. But now that the updated version is here, I suppose I should write something about it. Only problem is, I don't know where my focus should be.
To make matters worst, I don't have my first version 24-105mm with me. I left it in New York. I figured I was going to get the new one in Hong Kong, so it made sense to leave the now previous version back in the US. Because of this oversight, I cannot compare the two.
So now what, genius? What am I going to write about? There had to be a story in this lens, somewhere.
So then it occurred to me that this lens is suppose to have improved Image Stabilization. I figured maybe I could test that? Unfortunately, I don't have the previous version's Image Stabilization to test, so how am I going to know if the Image Stabilization has improved?
In truth, there is no real way for me to know for certainty. However, if memory serves me right, I never ventured pass 1/15s hand held at 24mm with the previous version. But then again, I've never had to do any lens testing before, so that account is completely anecdotal. Also to consider is the increased resolution of the Canon 5D Mark IV. Again, I left my Canon 5D Mark III in New York, since I figure I would be getting the Mark IV in Hong Kong. The issue of more resolution is increased probability of visible lens shake, from the mirror slap. The Mark IV is only 6 megapixels less in resolution than the Nikon D800/810.
Of course I could have set the Canon 5D Mark IV at reduced resolution. But where would the fun be in that?
From what I've been told about this lens, the Image Stabilization is suppose to give us an additional 4 stops - or to say it another way, it is suppose to let us shoot an additional 4 stops slower without camera shake.
4 stops is quite a considerable cushion. What this means is if you're shooting at 105mm (with the rule of thumb to prevent camera shake being double the focal length for the shutter speed, which in this case is 1/200s to 1/250s), the improved Image Stabilization should let me shoot at 1/30s. And at 24mm, I should be able to shoot at 1/4s, using the same calculation methodology.
With that established, the objective of this shoot is to see if I could shoot at 1/30s at 105mm, and 1/4s at 24mm.
Now for the disclaimer. First, my shooting fundamental has never been all that steady. So if I can get at least a 50% hit rate, I would be satisfied. And second, I'm in Hong Kong, so I'm going to be shooting wide with this lens. With that being the case, I'm only going to show one example at 105mm. Although, if the truth be told, I just forgot to shoot at the other focal distances on the zoom range. Alright already... I goofed. But at least I did get one in at 105mm. It is the curse of being a Leica M shooter, with zooming in and out a completely counterintuitive habit.
Besides, how is it interesting to shoot up close at 105mm for portraits, or far away for reach at that focal length? From my perspective, the 24-105mm is a 24mm lens with the capability of reaching further - just in case you needed to... or if you happen to be shooting from the bleachers at the red carpet.
Also worth noting in this test is that image stabilization has nothing to do with keeping your subject still at lower shutter speeds. Crazy isn't it. As obvious as that may sound, it didn't exactly sink into my head, until after a couple of warm up shots, with Anna striking multiple poses. It was then I had to ask Anna to stay still, and not change poses, while I shot continuous bursts at lower shutter speeds. For Anna, the request was as counterintuitive as zooming for me.
For that reason, Anna had to strain her eyes wide open, in order not to blink. You may notice some sample images with Anna going the extra mile with my request. Understandably, she was not amused with my not-so-exemplary shooting fundamentals at lower shutter speeds. It meant she didn't move her eyes for extended periods of time.
In addition (as said earlier), the mirror slap of the Canon 5D Mark IV is probably contributing a certain amount of camera shake to offset the benefits of image stabilization, given the increase resolution, compared to the Canon 5D Mark III.
By the way, Computer City is an awful place to start the test. It is hideous inside, from an aesthetic perspective... although to be fair, it is very practical in its no-frills configuration. It is what it is - a place to buy all things computers.
We then decided to go down to the subway, since we've never shot there before. That too wasn't the best idea, given how fluid the crowd was, coursing in every direction. It made going tighter than 24mm difficult. Then I decided to go back to where I parked my car - which is more secluded and under unideal available light. That wasn't too bad, but I couldn't visually prove that I was shooting at slower shutter speeds.
Then I had a flash of inspiration. What if I shot at really slow shutter speed outside, where the available light is ample, in a crowded situation with people walking more predictably - albeit limited to 24mm. The idea was to capture motion blur behind a perfectly still Anna, with the motion blur being a visual demonstration of shooting at slower shutter speeds.
If I were to compare my hit rate to the sharpness of the Canon 35mm f/1.4L II, there is no comparison. The 35mm lens wins hands down. You can see the definition of a person's eye lash under extreme magnification. But with the new Canon 24-105mm, the rendering is somewhat less impressive - especially shot at lower shutter speeds. It doesn't look sharp at all.
Again, there could be many contributing factors to why the images from this lens look softer in that comparison, as already mentioned. But one has to remember why such a lens exists to begin with. It is that utility lens that you keep in your Canon lens bag, when you need to have a certain amount of flexibility, under decent available light. And if the available light isn't decent, the Canon 24-105 can still deliver, with the help of Image Stabilization. However, being able to deliver isn't exactly the same thing as being able to perform. This lens just does its job. Or at least, that's how I see it.
Besides, it wouldn't make sense for Canon to make this lens too good. Let us just say its a product differentiation issue, and leave it at that.
In retrospect, I think I got the previous version for video. Being less sharp was immaterial to me, when you're only shooting at lower resolution and conventional frame rates.
In the final analysis, I am of the opinion that the Canon 24-105mm f/4L USM II does what it is designed to do. My hit rate was almost 50%. But I think it would have been better, if I shot with less megapixels, given that the sample images look much better down sampled for web viewing.
The Canon 24-105mm f/4 is an extremely useful and capable documentary lens. It won't win your heart with its extent of sharpness or with the way it renders bokeh. It is not made to do that. I suspect that this lens will perform better under normal ambient light situations, at higher shutter speeds. I have seen evidence of improved sharpness, shooting 1/30s at 24mm, where the resolved detail isn't too awful under high magnification. However, it still didn't appear to be as sharp as the Canon 35mm f/1.4L II.
To be fair, when do we ever look that closely at an image. We don't, in the real world. It is true that lenses like the Canon 35mm f/1.4L II are sharper. But could those sharper lenses reach as far or go wider than the Canon 24-105mm f/4L II? I think not. With that being the case, there is a place for the Canon 24-105mm f/4L USM II in your bag. And believe me, there will be a time when you will need it. Besides, you would hardly notice it in your bag, when you're not using it. It's not exactly heavy for a 24-105mm lens.
It's a pretty nifty lens to have.
All images in this writeup have been optimize in Lightroom. One or two images may have been cropped, for the sake of framing... or rather, I forgot to level the camera... and I do hate unleveled images.