Sweet Spot for 24-105L

eichert12

For those 24-105L owners in the crowd, what have you found to be the sweet spot as far as sharpness goes? I still feel it's a little soft, and while I think it could be due to IS or perhaps camera shake and/or not a fast enough shutter, I wanted to get a feeling from other owners regarding what you feel is it's sweet spot. In what situations do you get the best results from the lens?

Cheers,
Steve

ziggy53

eichert12 wrote:

For those 24-105L owners in the crowd, what have you found to be the sweet spot as far as sharpness goes? I still feel it's a little soft, and while I think it could be due to IS or perhaps camera shake and/or not a fast enough shutter, I wanted to get a feeling from other owners regarding what you feel is it's sweet spot. In what situations do you get the best results from the lens?

Cheers,
Steve

Try the technique in the thread you linked and tell us how it affects your results.

pathfinder

I find a good tripod a big help, he says with a sly grin.

Try f8 ( two steps less than maximum - the usual spot for a lens best performance ) - in this shot you can see individual grains of sand falling in the light beam in the upper left portion of the image when printed at 20 x 30 inches. 24-105 f4 L at f8




This was shot with IS engaged. I never turn it off ( well, almost never)

I find this lens is on my camera 75-85% of the time, and if I do my part, it rarely fails me.

How large are the images you wish to print, Steve? And at what shutter speeds? I rarely shoot this lens at f4, it will be a bit soft at f4. For f4, I would choose an f2.8, or even an f1.4 lens, because they would be tack sharp at f4.

eichert12

That's a very nice example of how sharp the 24-105L can be! I'm rarely shooting at f8 with the lens due to the conditions I'm often shooting in (low light/indoors). I'm coming to realize that I'm expecting to much out of the lens at f4 and need to do a better job utilizing the lens in situations that it will "shine". Thanks for the thoughts!

jdryan3

eichert12 wrote:

That's a very nice example of how sharp the 24-105L can be! I'm rarely shooting at f8 with the lens due to the conditions I'm often shooting in (low light/indoors). I'm coming to realize that I'm expecting to much out of the lens at f4 and need to do a better job utilizing the lens in situations that it will "shine". Thanks for the thoughts!

I just posted a response to your other thread re: a low light WA, but I had to make some assumptions. Is it safe to ASSUME that in both posts, you mean no flash/low ambient indoor light? Tripod is not an option?

If so you may want to do a search on concert and/or no flash photography. You will quickly see most posters consider f/2.8 to be sloooww in those circumstances, and often shoot f/1.8 or faster with ISOs in the 1200 -3200 range, often at 1600.

Those are exactly the reasons I got my 35 L f/1.4, which is 3 stops faster than your 24-105 f/4. And I still bump up the ISO to 800-1600 on my 5D.

Both of these were shot 5D, 35L f/1.4, 1/100 ISO 1600 (and it wasn't all that dark on the stage).




Hope that helps (and in your lens decision on the other thread)!

pathfinder

The 35 f1.4 , the 50 f1.4, and the 85 f1.8 are all great for handheld in the dark. At f2.8 they should all be very, very sharp if there is no camera movement.

kini62

I find that my 24-105 is sharp at F4, at least sharp enough for people shots.

and by F5 it's really sharp.

Gene

gus

kini62 wrote:

I find that my 24-105 is sharp at F4, at least sharp enough for people shots.

and by F5 it's really sharp.

Gene

Im a bit of a sharpness freak at times & i used andys 24-105 at Glacier for 2 full days. I found it quite sharp at f/4 personally. Then again my first 10-22 was a pigs ear & my 2nd one is amazingly sharp

eichert12

kini62 wrote:

I find that my 24-105 is sharp at F4, at least sharp enough for people shots.

and by F5 it's really sharp.

Gene

Thanks Gene, I've heard that from others as well. I think the reason for the discrepency is at least in part due to the fact that the definition of "sharp" probably varies a bit from person to person. I've heard some people say the lens is sharp at f4, while others say it doesn't get sharp until f8. I've had somewhat inconsistent results, which I believe is in part due to camera shake, too slow of a shutter, and other "user errors". But I've heard enough people say the lens is soft at f4 to think that those things are not always the cause of the softness.

Cheers,
Steve

eichert12

jdryan3 wrote:

I just posted a response to your other thread re: a low light WA, but I had to make some assumptions. Is it safe to ASSUME that in both posts, you mean no flash/low ambient indoor light? Tripod is not an option?

That is a safe assumption. One of the things I need to get better at is taking photos at very low aperatures (1.8 and below). I always have a really hard time getting enough in focus, and/or the right thing in focus. When I shoot with my 50 1.4 at 1.4 I have a heck of a time getting good results. I guess I need more practice?

Going to check your reply in my other thread now....

Thanks,
Steve

eichert12

gus wrote:

Im a bit of a sharpness freak at times & i used andys 24-105 at Glacier for 2 full days. I found it quite sharp at f/4 personally. Then again my first 10-22 was a pigs ear & my 2nd one is amazingly sharp

I hope I don't have a pigs ear!

jdryan3

eichert12 wrote:

That is a safe assumption. One of the things I need to get better at is taking photos at very low aperatures (1.8 and below). I always have a really hard time getting enough in focus, and/or the right thing in focus. When I shoot with my 50 1.4 at 1.4 I have a heck of a time getting good results. I guess I need more practice?

Going to check your reply in my other thread now....

Thanks,
Steve

Wide Aperature = Shallow Depth of Field. That's why when you shoot wide open and up close, things like noses are in focus but both eyes aren't. Practice is the only way. I also try to set my focus point (usually center point), rather than let it hunt and decide which is the active point, in those low light, fast moving environments where the subject is close.

Also, always get the eyes in focus. If they are soft or OOF, then it doesn't matter how good an image it is. Sharp eyes allow for more forgiveness on the other aspects of the image being soft/OOF.

Dogdots

I've had my 24-105L for about 7 months now and its a real good lens. I've been happy with all my shots.

I did have a short period where I was really questioning my lens and its sharpness. But it was me not the lens that was at fault

I know that the sweet spot it where you set your lens, but really the lens shines on sunny days. I've also found that if I set my settings and then just shoot like a point and shoot it really hardly ever lets me down.

It does do well in low light to, but gotta use the tripod.

My lens very rarely is off my camera. Wouldn't trade it for anything else.

Dogdots/Mary

pathfinder

I did not mean to suggest that f4 was not a useful aperture for the 24-105, just that f8 is, generally, quite a bit sharper.

I just know from my own experience, that at f4 I find the images are not quite as crisp, as at f5 or f5.6, or f8. Some of that is DOF, some of it is inherent in the lens design.

I have lots of frames shot at f5 and f5.6 with a 5D and a 24-105 here

Check the exif, as a number of these images came from a G9 also.

This is at f5, handheld, sitting on the floor

LiquidAir

I find the 24-105 to be a bit soft at 105 and f/4 but not so much that I spend any time worrying about it. That said, for people shots in low light I prefer my primes (35/1.4, 50/1.4, 85/1.8 and 135/2) because IS doesn't stop subject motion.

eichert12

Thanks all! I'm going to continue to work with this lens to see if I can master it, as many of you have.

eichert12

jdryan3 wrote:

Wide Aperature = Shallow Depth of Field. That's why when you shoot wide open and up close, things like noses are in focus but both eyes aren't. Practice is the only way. I also try to set my focus point (usually center point), rather than let it hunt and decide which is the active point, in those low light, fast moving environments where the subject is close.

I'll have to try setting my focus point instead of letting it "hunt and decide". Often it takes a couple tries before I get the right focus point and even still I, at times, end up with the wrong thing in focus. I notice if mostly at apertures of 2.8 and under, which isn't surprising since those apertures result in a such a shallow DOF.

Cheers,
Steve

pathfinder

Steve,

I NEVER let my camera choose my AF point - I always ( unless speed requires giving up control) choose which specific AF point I plan to use, and place it precisely over the area where I desire focus. Letting the camera choose the AF points, for me at least, guarantees that the camera will choose precisely the wrong site to AF on.

I make a strong effort to avoid "focus-recompose" on shots done closer than 8 or 10 feet - it introduces focusing errors, and is not rec'd by Canon,either.

eichert12

Funny enough I think I've always let the camera choose cause I was worried about how long it would take me to select the AF point myself. However, now that I think about it I'm usually spending all day trying to get it to use the right focus point. I don't know why I never thought to stop letting the camera choose the wrong thing for me. I suppose it's becuase I got so used to it early on and learned to deal with the "suckiness" of it. I'm definitely going to take that control back from the camera, and hopefully start reaping the benefits!

Thanks for the advice!
Cheers,
Steve

eichert12

So I just did a search on "focus recompose" since that's what I've been doing. No wonder my results have sucked. It's because Focus Recompose Sucks! Thanks for pointing out something I feel stupid for not knowing!

pathfinder

Steve, as I said, I strongly try to avoid focus-recompose. But there are times, especially when shooting in low light, that it can be necessary, because there may not be enough light to focus quickly with a peripheral AF point. Just be aware of it and the risks it poses.

When we focused film SLRs manually ( back when the dinosaurs still roamed the earth ), every one did focus-recompose, because the manual focusing reticle was in the center of the viewfinder window. But there are errors introduced, as that web ink demonstrates. Beyond 15 -20 feet , the errors are pretty negligible.

Maybe, these points will help you find the sharpness in your images that you desire!

eichert12

I hope so!

Do you still do minor focus recompose when the thing you want to be in focus doesn't exactly line up with one of the camera's focus points? I'm assuming the way to go would be to pick the focus point that is closest to the thing you want in focus and then do a minor adjustment to get the composition you want? Or is there better way? Or do you usually find that the thing you want in focus lines up with one of the focus points?

Thanks,
Steve

LiquidAir

The error from focus and recompose is not nearly as large as many reports on the web would have you believe.

Here are the things to thing about when using focus and resompose:

How close is your focus point? If it is more than 6 feet away focus and recompose will rarely be problem.

Where do you want the place the focus point in the frame? If it is in the center third of the frame, focus and recompose will rarely be a problem?

How fast is my aperture? I have never seen focus and recompose issues when I stop down below f/2.8

When I have the 35/1.4 on my camera I often shoot in a way where focus and recompse is a problem. The 50/1.4 can also occasionally cause problems. If you have an 85/1.2 it might be an issue, but I have never seen a significant focus and recompose error with the 85/1.8. With the 24-105 you might be able to construct a case where it is a problem by placing the focus point in the corner of the frame at the very close end of the focusing range, f/4 and 24mm. However, in day to day use it shouldn't concern you.

Honestly with my 5D I get much more accurate focus using focus and recompose than I do with the peripheral focus points. I find that the focus and recompose error is smaller than the errors created by using the less accurate focus points. When I am shooting in a situation where focus and recompose is going to be a problem I either focus manually or use AF on a surrogate subject which is at the proper distance.

Here is a recent focus and recompose shot taken with the 24-105 at f/4 and 32mm (as well a ISO 1600). You can see in this shot that I have more than enough DoF at f/4.



Here is a sample focus and recompose shot with the 135L at f/2 (focus was on the cat ). This close up at f/2, focus and compose might cause a problem if I was putting the focus point in the corner of the frame, but for this shot it isn't an issue.



Both are sharp enough to comfortably print 16x24.

Glenn NK

The "sweet spot" for any lens isn't hard to find. Looking at the MTF charts in the link for the lens in question, it becomes apparent that the sweet spots vary with focal length:

http://www.photozone.de/8Reviews/lenses/canon_24105_4_is/index.htm


The MTF charts indicate which apertures/focal lengths provide the best from any given lens. What is relevant for someone with a 1.6 crop Canon, is that these tests were done with a 1.6 crop.

It's apparent that the IQ drops with decreasing aperture size (the exception being at the 105 mm focal length). The quality of the glass isn't getting worse at smaller apertures - the IQ is dropping because diffraction is overriding the gain from a smaller aperture.

For example, at 24 mm FL, I would prefer to use f/4 and f/5.6, but at 105 mm FL, I would prefer f/11. The IQ at 195 mm and f/11 isn't as good as it is at 24 mm and f/5.6, but at 105 mm, the "sweet" spot is at f/11.

For my 100 f/2.8 macro, I avoid f/2.8, and f/16 and smaller. If it wasn't for diffraction losses, it would be awesome at f/32.

http://www.photozone.de/8Reviews/lenses/canon_100_28/index.htm

In response to the post by LiquidAir, I would agree with him. It's simply a matter of understanding the geometry of a triangle. Focus/recompose uses the diagonal for focusing, and the (normally) long side for shooting the subject, so if the subtended angle between the diagonal and the sight line is small, the error will be small.

IMO, he is bang on in regard to the use of which focus points to use; and the technique of a "surrogate distance" is very practical and useful.

eichert12

Thanks for the opposing viewpoint regarding focus recompose. It sounds like the key is to understand when it can/will cause problems, and when it's a non-issue. I suspect it could be the cause of some problems I've had when shooting my 50 1.4, since at f1.4 there is such a shallow DOF and the slighest error in focusing can make or break the picture. As you guys mentioned with the 24-105 it's probably less of an issue.

LiquidAir, in the image you posted above that was shot with an ISO of 1600, did you run it through Noise Ninja (or some other noise reduction software)? I recently shot some pictures indoors at ISO 1250 and they were much noisier then I expected. That will likely be the topic of another post though.

Cheers,
Steve

eichert12

Glenn, thanks for that link as well! Very informative!

jdryan3

I agree with both Pathfinder & LiquidAir. In my earlier post, I noted that in low light large aperature situations where the subject is moving, I prefer not to let the camera choose my focus point. But there certainly are situations where I do use it - at greater distances, where DOF of the entire subject is OK; when the subject is moving in & around the frame and I can't select the point quickly enough (or do a focus-recompose). I also use AI Servo and not always One Shot :gasp

RE: Focus -Recompose. Not my preferred method, but again sometimes the only way to get the shot. Both of Liquid's are of situations where the subject is relatively stationary, where focus-recompose does work (love that shot of the cat!).

I think by really understanding what your camera & lenses can and can't do, (in hand with the basics of photography like DOF, shutter speed, aperature, etc), you will soon discover when to shift from your 'set methodology' to alternatives. Please just keep asking away - you'll get it.

Hey, if we all could always get clear, sharp shots of moving people in low lit spaces using ISO 100, f/11 @1/1000s, where would the fun be in that?

LiquidAir

eichert12 wrote:

LiquidAir, in the image you posted above that was shot with an ISO of 1600, did you run it through Noise Ninja (or some other noise reduction software)? I recently shot some pictures indoors at ISO 1250 and they were much noisier then I expected. That will likely be the topic of another post though.

That's a RAW conversion right out of Lightroom. I used Lightroom's noise reduction but nothing more exotic.