EF 100-400mm f/4.5-5.6L IS II USM Review WOW!

Edle wrote in post #17507039
So does anyone have indoor auditorium type shots to share?

Because of permissions, I don't have any shots to show, but I just finished a weekend indoor conference in which I used this lens. Many of the photos were nice and sharp with pleasant backgrounds. However, the slow SS's showed me that I should rent a 70-200 for our conference this autumn.
j-p

Thank you for the reply. I use my 70-200 2.8 on a 7D for all previous graduation shoots and have been very happy. I recently switched to the 5D MK3 which means I will lose some distance now so was considering the 100-400, but due to the 5.6 @ 400, not sure what kind of results to expect.

Edle wrote in post #17509064
Thank you for the reply. I use my 70-200 2.8 on a 7D for all previous graduation shoots and have been very happy. I recently switched to the 5D MK3 which means I will lose some distance now so was considering the 100-400, but due to the 5.6 @ 400, not sure what kind of results to expect.

It will depend on how much light you have to work with at the venue. The 100-400L will be two stops slower than the 70-200 f/2.8. If you were pushing the ISO of your 7D to its limit with the 70-200, the 100-400L may not be fast enough -- even on a 5D3. If your ISO was 1600 or lower on the 7D, though, you should be fine with the 100-400L + 5D3. I get very good results at ISO 6400 on the 5D3 as long as the exposure is correct.

I would suggest checking the EXIF data from previous outings to see where you were at with the 70-200 f/2.8 + 7D. That should provide you with the answer you seek.

Scott M wrote in post #17509153
It will depend on how much light you have to work with at the venue. The 100-400L will be two stops slower than the 70-200 f/2.8. If you were pushing the ISO of your 7D to its limit with the 70-200, the 100-400L may not be fast enough -- even on a 5D3. If your ISO was 1600 or lower on the 7D, though, you should be fine with the 100-400L + 5D3. I get very good results at ISO 6400 on the 5D3 as long as the exposure is correct.

I would suggest checking the EXIF data from previous outings to see where you were at with the 70-200 f/2.8 + 7D. That should provide you with the answer you seek.

Camera Canon EOS 7D
ISO 1000 (up to 3200 on high end)
Focal Length 200mm (320mm in 35mm)
Aperture f/2.8
Exposure Time 0.004s (1/250)

It's a fairly well lit place, Toyota Center in Houston, where the Rockets play.

Scott M wrote in post #17509153
It will depend on how much light you have to work with at the venue. The 100-400L will be two stops slower than the 70-200 f/2.8.

Yes and no. Yes, you need 4x the exposure time to get the same exposure of a blank wall as the 200/2.8 with the 400/5.6, but if you are shooting from the same spot, and have to crop the 200/2.8, you don't get any more subject light with the f/2.8. This is one of the greatest illusions that photographers face. When person A uses 200/2.8, and person B uses 400/5.6, and they are standing together and shooting the same thing at a distance, with the same shutter speed, the amount of light collected from each element in the scene is the same. f-#s only affect the light per unit of sensor area; they do not reflect the total light collected for any part of the scene. If A and B are using the same pixel density, the radius of the AA filter relative to the subject size is smaller, the coarseness of a Bayer CFA is less, and noise clumps and banding are finer with the longer focal length (bold italic text is an edit). What matters to the amount of light collected in such a scenario, with a given shutter speed is only one thing - aperture size (focal length divided by f-#). This is the same for 200/2.8 and 400/5.6, and 800/11. If you want less subject noise, get closer and/or get a larger aperture, and a longer focal length increases the detail, and makes the pixel/sensor artifacts smaller relative to the subject size.

Of course, a lower f-# can improve AF ability, and sometimes that is more important.

My model above breaks down when f-#s start to get too low. Because of the nature of the microlenses on sensors, for recent pixel sizes, an APS-C starts losing light when the f-# gets below 2.8, and a FF starts losing light below f/2 or so (the 5Ds will probably behave like the recent APS-Cs, unless Canon has come up with something new here with microlenses). So, 100mm f/1.4 would collect less subject light than 200mm f/2.8 with the same distance and shutter speed.

John Sheehy wrote in post #17509482
Yes and no. Yes, you need 4x the exposure time to get the same exposure of a blank wall as the 200/2.8 with the 400/5.6, but if you are shooting from the same spot, and have to crop the 200/2.8, you don't get any more subject light with the f/2.8. This is one of the greatest illusions that photographers face. When person A uses 200/2.8, and person B uses 400/5.6, and they are standing together and shooting the same thing at a distance, with the same shutter speed, the amount of light collected from each element in the scene is the same. f-#s only affect the light per unit of sensor area; they do not reflect the total light collected for any part of the scene. If A and B are using the same pixel density, the radius of the AA filter relative to the subject size is smaller, the coarseness of a Bayer CFA is less, and noise clumps and banding are finer. What matters to the amount of light collected in such a scenario, with a given shutter speed is only one thing - aperture size (focal length divided by f-#). This is the same for 200/2.8 and 400/5.6, and 800/11. If you want less subject noise, get closer and/or get a larger aperture, and a longer focal length increases the detail, and makes the pixel/sensor artifacts smaller relative to the subject size.

Of course, a lower f-# can improve AF ability, and sometimes that is more important.

My model above breaks down when f-#s start to get too low. Because of the nature of the microlenses on sensors, for recent pixel sizes, an APS-C starts losing light when the f-# gets below 2.8, and a FF starts losing light below f/2 or so (the 5Ds will probably behave like the recent APS-Cs, unless Canon has come up with something new here with microlenses). So, 100mm f/1.4 would collect less subject light than 200mm f/2.8 with the same distance and shutter speed.

I had to read this a few times.. I think I understand. Thanks for explaining!

Savethemoment wrote in post #17509631
I had to read this a few times.. I think I understand. Thanks for explaining!

Maybe I should have read it one more time, too. I left out part of a sentence, but I fixed it in an edit. The added text is bold italic.

Pondrader wrote in post #17508702
Just to reinforce the Hand held IS in this lens , shot after dinner tonight

1/40, f4.5, iso1250, 400mm Hand held 7DII + 100-400markII...,...,... I am not missing the push pull at all..,..,..not even for a second

IMAGE LINK: https://flic.kr/p/s1Ca​F4 Barred Owl_7587 by Jeff Manser, on Flickr

Lovely shot!

... but I would not miss the push pull for perching birds either. Or even panning shots of Birds in flight,. it's the other times I miss it.

John Sheehy wrote in post #17509680
Maybe I should have read it one more time, too. I left out part of a sentence, but I fixed it in an edit. The added text is bold italic.

I wasn't meaning to be critical of your writing, not at all. Only to say that the importance of actual aperture size, rather than just f stop, wasn't something I had understood with respect to the amount of light recorded for a particular object or person (shot from the same distance, with the same shutter speed) within the frame. Thanks again

John Sheehy wrote in post #17509482
Yes and no. Yes, you need 4x the exposure time to get the same exposure of a blank wall as the 200/2.8 with the 400/5.6, but if you are shooting from the same spot, and have to crop the 200/2.8, you don't get any more subject light with the f/2.8. This is one of the greatest illusions that photographers face. When person A uses 200/2.8, and person B uses 400/5.6, and they are standing together and shooting the same thing at a distance, with the same shutter speed, the amount of light collected from each element in the scene is the same. f-#s only affect the light per unit of sensor area; they do not reflect the total light collected for any part of the scene.


So, 100mm f/1.4 would collect less subject light than 200mm f/2.8 with the same distance and shutter speed.

It is very clear that you know what you are talking about but just so there is no confusion, it is important if anyone is now a tiny bit confused after reading this that a 400/5.6 is projecting an image that is 4 times greater than a 200/2.8 and thus while the same about of light is entering the entrance aperture it is being used to light an image 4x as large and therefore in unit terms is only being lit 1/4 as much as a 200/2.8 - it was however pointed out 100% correctly that if you are blowing up / cropping your 200/2.8 to get the same size image as a 400/5.6 then yes your are not getting any more light with the 200/2.8 than the 400/5.6 - and as was also pointed out, and as any sports or wildlife shooter knows, you can't always use a shutter speed that is 4x as long to get the shot.

I'm not sure I agree with your last statement (it is late however) - are they not the same after scaling per unit?

Great information by the way.

wallstreetoneil wrote in post #17510059
It is very clear that you know what you are talking about but just so there is no confusion, it is important if anyone is now a tiny bit confused after reading this that a 400/5.6 is projecting an image that is 4 times greater than a 200/2.8 and thus while the same about of light is entering the entrance aperture it is being used to light and image 4x as large and therefore in unit terms is only being lit 1/4 as much as a 200/2.8 - it was however pointed out 100% correctly that if you are blowing up / cropping your 200/2.8 to get the same size as and image as a 400/5.6 then yes your are not getting any more light with the 200/2.8 than the 400/5.6 - and as was also pointed out, and as any sports or wildlife shooter knows, you can't always use a shutter speed that is 4x as long to get the shot.

I'm not sure I agree with your last statement (it is late however) - are they not the same after scaling per unit?

Great information by the way.

Thanks for rounding that out in other terms. I appreciate everyone's input, slant and willingness to help here.

But I still want to see some indoor pictures with this thing!

Edle wrote in post #17510529
But I still want to see some indoor pictures with this thing!

These are from the week I got the lens during the winter - horrible hockey lighting.

5D3 / 6400 / F5 / 1/1000 / 183mm

IMAGE LINK: https://flic.kr/p/s2ZG​Ng

IMAGE LINK: https://flic.kr/p/r6fR​e6

IMAGE LINK: https://flic.kr/p/r645​dh

Thank you!!! At the risk of pushing my luck, any chance of indoor pics at 400mm?

wallstreetoneil wrote in post #17510059
I'm not sure I agree with your last statement (it is late however) - are they not the same after scaling per unit?

They're not the same because current sensors (including the microlenses) lose quantum efficiency when the f-# drops below a certain point.

With the 7D, as an example, f/4 gives twice as much light as 5.6, and f/2.8 gives twice as much as f/4, but f/2 fails to give twice as much light as f/2.8, and if you had an f/0.95 lens, you'd lose over a stop of light to this phenomenon. Canon hides this fact from you by scaling the RAW data to make the RAW numbers look the same as if this loss of light had not occurred. The factors are calculated into every metering the camera does, so when you shoot with an f/1.4 lens, the camera knows that it is metering with a loss, and compensates for it when the f-# of the actual shot is higher than 1.4.

With an f/0.95 lens, the scaling of the RAW data causes a loss of more than 1 stop of dynamic range, which is a tragedy, really, as Canon could just as easily changed the whitepoint in the RAW metadata, and maintained the full DR.

So, equivalencies break down when the f-# gets too low with current cameras. This gives an even greater benefit to big-pixel cameras when you want to shoot your f/1.4 or "faster" lens wide-open, as bigger pixels experience less of this loss with current microlenses.