C100: Canon 24-105 is a Horrible Documentary Lens in Very Low Light

nedcam

nedcam

Well-known member
I have been using it on the C100 for over a year but I usually am not in the dark situation I recently found myself in lately. I loved the lens for the most part although 90% of the time I am not doing the editing of it's footage. However...

I am a documentary shooter at heart, even if it's a corporate or TV gig they usually hire me for a doc look. So I am always trying to stay on the good side of the editor giving wide/tight/wide/tight, repeat. Always. So in this situation what I noticed, having to crank up to ISO 1600-2000, this lens is a Pig in Low Light at the tele end. You can't cut the tight shot with the wide unless you really played with brightness in post. Ugh. I haven't had a lens with so much transmission loss at the tight end. I guess I never really noticed it before because I was never in the soup on tight and then having to do the edit myself.

I was doing a doc for a charity in a dark ballroom of children doing activities. So naturally I shoot the kid wide but then, most importantly, I need the smiling face. This lens has such a drastic drop off there's no way to use the tights. And there's no time to change lenses.

So a note to myself for the future is just stay wide when in very low light. I'm wondering what's in this focal range that doesn't drop off so much?
 
I use this lens for both stills and and a little video. It has never been my fav lens, yet when I look at my favorite images, many were taken with it. It isn't as sharp or fast or as wonderful as a prime. But if I can only have one lens for the day, the 24-105 would likely be the choice; what else would you use? The twice as expensive 24-70 is sharper and faster but loses some reach in the bargain. The 24-105 isn't a terribly expensive lens, the OIS is pretty decent and the focal range is really useful. It was never even thought of as a video lens until Canon introduced EOS cams with video. As the high iso abilities of the newer dslr cams improved, fast lenses aren't as important as they used to be except where very shallow DOF is desired.

In post rant mode, what else would you use for similar function and similar dollars? I don't know of any so I guess that's why I still have mine. Not too much love for this lens in places, but it gets it done.

Grant
 
Keep your f-stop wide open at 4 and you won't experience any loss of transmission through the zoom range. Whenever I have that lens mounted, I do my best to expose for f/4 via my ND's and/or ISO range rather than close the aperture down. Once you get used to that kind of lateral thinking in exposure, it becomes quite a useful lens. Not my favorite, but definitely one of my most used for documentary shooting.

Even if you're stopped down just a 1/3 of a stop, you'll get pretty significant loss of T when zooming that isn't present when the lens is wide oen.
 
Cowpunk 52: I was wide open at f4, and when I zoom to tele there is a definite transmission loss at the tele end. You can not cut the far tight shot with the wide shot unless you do work in post to brighten the tight shot. You mention NDs, I was in a dark interior, I had to crank the ISO up.

The focal length is great for doc shooting but when shooting doc I need to snap in and get that tight shot, I can't have it go down what looks like an entire stop, regardless of what the manufacturer's literature states. There is a definite loss. Unless I got a lemon.

I must say that the transmission loss is much more noticeable while editing the footage than it is on the field monitors. When you cut the wide to the tight at f4 under low light conditions I would say it's unprofessional looking, thus needing to stop and play with brightness and bring up even more noise. I am not happy and don't mind paying more if I could find something covering that range and has no trans loss at tele.
 
Last edited:
You might want to check your copy, it should not lose any transmission when wide open. I've tested multiple 24-105's with gray cards on waveform monitors at full wide and full tell, and none have exhibited any transmission loss at wide open. When I get home I'll upload some of the stills from one test.

If yours is losing T when zooming in wide open, I'd look at getting it serviced or replacing it.
 
One thing to keep in mind, if your iris function is set on "fine," you may have actually been stopped down 1/8 or 1/4 of a stop without knowing it. The camera's f-stop readout would still say f4.0 until you actually hit f4.5, and you would be getting a loss of T while thinking you were wide open. That's happened to me before.
 
nedcam, it IS possible you got a "less than the best" copy of this lens. I have noticed over the years that when a noticable drop in I.Q. is observed suddenly at any of the farthest ends of a lens' function, and it's not a lens where everyone already talks about that, then you probably got something other than the best copy.

This always gives me the cold-sweats when I buy new lenses. I'm left wondering: "how will I know if I got a poor one?" HUGE flaws are usually obvious and the seller will swap you out in most cases, but what about when we get one that's only "half-flawed"? Or a "quarter-flawed"? How will I know? (I am not surrounded by people who have mass amounts of lenses on hand for me to always compare to.)

Cowpunk52's test is probably the best.

Ok, I'm a dark cloud of gloom today....sorry.
 
Thanks I will look into that once I get the C100 back from service. I believe it is set to fine, that would be my nature when I originally set things up. When I pick it up perhaps I'll drop off the lens.
 
No it is a fact that it does get darker at the end of the lens, at least a 1/2 stop

pick one up and see
 
I've used two 24-105s in the past two years, the one I own and a borrowed lens I needed when mine wasn't available (from a still photographer). I wasn't able to do side-by-side comparisons, and I wasn't shooting in low light, but the borrowed lens never looked right to me., especially at the long end, and I was happy to get mine back. It was older by about a year, but otherwise the same lens.

I guess there are variations in these mass-produced lenses that aren't immediately obvious unless you have a couple to try - I was lucky to get a good one, although it also loses a third to half a stop at 105 - but that shouldn't be enough to cause the problem described.
 
Here's a couple stills from a basic test I did a while ago on a rented 24-105. The display info on screen speaks for itself, but pay attention to the waveform spot meter.

24mm f4
105mm f4
 
Mine definitely loses about half a stop wide open, and far more stopped down. Aperture cable recently got ripped (a common issue) and I'm debating whether to even bother repairing it.

(RETESTED - It does not!)
 
Last edited:
It's pretty well documented on photo sites that the Canon 24-105 DOES ramp near/at the top end even though it is supposed to be a constant aperture f/4.
 
People's experience is real. But so is Cowpunk52's test. Look at his body of work - that guy knows what he's talking about. Seems like he got one of the better copies. That will happen from time to time.
See what I mean about the cold-sweats when buying? unless you have several to compare to - you may simply never know if that's all the lens can do, or if you got a poor copy.
 
Sorry guys, but the loss in T just doesn't happen when wide open. I've tested nearly a dozen copies from various rental houses and ones that I've owned. They all perform exactly as the test I posted above, which shows no loss of transmission in the center of frame (vignetting is another story). If your lens is losing T when zooming in with a wide open aperture, then something is wrong with the lens.

Interestingly, you'll find nearly every EF zoom lens suffers from this same phenomena with few exceptions (notably 70-200 flavors). The 16-35, both 2.8 and 4, as well as versions 1 and 2 of the 24-70 L have the same characteristic of losing T through the zoom range when stopped down, but not wide open. I've spent a lot of time testing many copies of these lenses because I've used them on dozens, if not hundreds, of various television series episodes.

I encourage everyone to test lenses with scopes, because you cannot trust your eye when it comes to these things. I promise you'll find identical results to mine.
 
I hate to sound like a cheerleader - I don't even know the guy - but Cowpunk52 really is right about running his bare minimum test. That test won't lie.
Really, look at his body of work - the links are in his Sig. He's one of the people we are lucky to have around this forum. The fact that we can get free advice on so many things from folks like him here still blows my mind sometimes.
 
Thanks DPStewart *internet high five*

Honestly, though, I've got no skin in this game, I just like to share the things I've learned. I have been very fortunate to learn from very talented people that didn't have to teach me anything, and I like to pay that forward to others.
 
Thanks for doing the test but mine loses at least a half stop, looks like more, when wide open on tele. So if you all are saying it shouldn't be doing that then I will bring it into Canon although it is out of warranty. Presently this is an unusable documentary lens in low light because I need to do a crash zoom tight for the editors and it's just creating extra work (and noise) since the tele end is darker. Very disappointed. But hey, exteriors it's beautiful!
 
You must log in or register to reply here.
Back
Top