400/4 vs 500/4 vs 100-400 vs 70-200/2.8

DavidP #28649

Forum Pro
Messages
29,088
Reaction score
1
Location
Conroe, TX, US
AKA,

The Great Dallas Lens Shootout!

On Saturday, August 17, 2003 the following DPR members met up in Dallas to do some serious lens testing (er, measurbating?):

DavidP
FredG
Girish Jadav
Jason Hutchinson

Here is the equipment used in the testing:



From left to right, the lenses are the 100-400 IS (at 400mm extension), 400/4 DO IS, 500/4 IS, and 70-200/2.8 IS. Also shown are a D60, 1D, and 10D, along with 1.4x and 2x TC's (both version II), an Angle Finder C, a remote release, and a Gitzo G1325 tripod with full Wimberly head.

Also shown is the test chart that we used. The test strips were downloaded from Norman Koren's site, http://www.normankoren.com

Note that we used 3 samples of the 100-400, 2 samples of the 400/4, and 2 samples of the 500/4. We also used two samples of the 10D.

We tried as many permutations and combinations as possible, but not every possibility was tried. If we'd done that, we'd still be there testing!

Here's a shot showing the test setup:



You can see the halogen lamp used for illumination, and the jerry-rigged reflector (a white board) used to even up the illumination across the chart.

Light levels were: ISO 400, f/4, and 1/320.

Here's a closeup of the test target:



Be sure to note the title and authors of the book on the bottom. ;)

Analysis will follow as we get all this data digested. Just on my 1D body, I took 108 images. We have a lot of stuff to sort through, and figure out what the conlcusions are, and how to present it.

--
The Lowest Paid Concert Photographer Around
http://www.neonlightsimaging.com/artshow/final.htm
Photography -- just another word for compromise

'Since we can't keep crime in check, why don't we legalize it and tax it out of business?' -- Will Rogers
 
the tests are all flawed!!! You didn't do it right, must be multiple user error!!!
AKA,

The Great Dallas Lens Shootout!

On Saturday, August 17, 2003 the following DPR members met up in
Dallas to do some serious lens testing (er, measurbating?):

DavidP
FredG
Girish Jadav
Jason Hutchinson

Here is the equipment used in the testing:



From left to right, the lenses are the 100-400 IS (at 400mm
extension), 400/4 DO IS, 500/4 IS, and 70-200/2.8 IS. Also shown
are a D60, 1D, and 10D, along with 1.4x and 2x TC's (both version
II), an Angle Finder C, a remote release, and a Gitzo G1325 tripod
with full Wimberly head.

Also shown is the test chart that we used. The test strips were
downloaded from Norman Koren's site, http://www.normankoren.com

Note that we used 3 samples of the 100-400, 2 samples of the 400/4,
and 2 samples of the 500/4. We also used two samples of the 10D.

We tried as many permutations and combinations as possible, but not
every possibility was tried. If we'd done that, we'd still be
there testing!

Here's a shot showing the test setup:



You can see the halogen lamp used for illumination, and the
jerry-rigged reflector (a white board) used to even up the
illumination across the chart.

Light levels were: ISO 400, f/4, and 1/320.

Here's a closeup of the test target:



Be sure to note the title and authors of the book on the bottom. ;)

Analysis will follow as we get all this data digested. Just on my
1D body, I took 108 images. We have a lot of stuff to sort
through, and figure out what the conlcusions are, and how to
present it.

--
The Lowest Paid Concert Photographer Around
http://www.neonlightsimaging.com/artshow/final.htm
Photography -- just another word for compromise

'Since we can't keep crime in check, why don't we legalize it and
tax it out of business?' -- Will Rogers
--
Troponin (Trop)
'No Limitation is Limitation'
and I still can't spell worth a dang!
 
From left to right, the lenses are the 100-400 IS (at 400mm
extension), 400/4 DO IS, 500/4 IS, and 70-200/2.8 IS. Also shown
are a D60, 1D, and 10D, along with 1.4x and 2x TC's (both version
II), an Angle Finder C, a remote release, and a Gitzo G1325 tripod
with full Wimberly head.
DavidP,

FYI - You flipped the position of the 500mm f/4 and 400 f/4DO in your caption. For the folks that don't know the 500mm is the tall one (second froml left) and the little guy (third from left) is the 400mm f/4 DO!.

Really looking forward to your results!!!

Thanks

Jim
 
This is my first attempt at analyzing these three lenses:

100-400 IS, 400/4 DO IS, 500/4 IS. For this analysis, I used only my copies of the lenses on my 1D body (just one of two bodies, actually). I also used only my teleconverters (I don't think we did any testing to see if different TC's made a difference).

The images I chose were with the 2x TC attached. Reason? I thought that the TC would highlight any differences in the lenses. The risk? That the 2x TC would so dominate the results, that I was essentially testing only it. I intend to do a similar test with no TC's to check for that.

I initially only looked at the wide-open (f/8 for the two primes, f/11 for the zoom) tests, but added the 400/4 stoped down 1-stop after the analysis.

Methodology:

1) Convert the images using YarcPlus with consistent parameters. No sharpening, white-balance taken from one image and applied to them all. Saved as 16-bit TIFs

2) Converted the images to JPG-12. Used this data to select the sharpest one (based on file-size).

3) Took the right-most Koren chart and used this -- completely arbitrary.

4) Posted the three wide-open Koren charts next to each other (100-400 on left, 400/4 in the middle, 500/4 on the right).

Here's the link to that image
http://www.neonlightsimaging.com/images/comparison_2x_wideopen.tif

5) Based on initial visual inspection, I concluded that these lenses looked remarkably consistent, though the 400/4 seemed to be less contrasty then the 500/4, and perhaps just a bit less contrasty than the zoom.

6) Decided to analyze contrast (taken as the difference between luminosity in the blacks vs. whites) at a few places along the chart.

7) Decided to look at the 2 lp/mm, 10 lp/mm, and 50 lp/mm markings on the Koren chart. Note that these values don't really mean anything in real lp/mm terms without some scaling. I'm not that interested in actualy lp/mm anyway, just comparisons.

8) Also note that due to problems getting exactly the same size, the 500 chart is slightly larger (in pixels) than the 400, which is slightly larger than the zoom. Hopefully this doesn't skew things too much. Getting the framing exactly right was far harder than we expected.

9) After looking at the values (which to me showed that the 400 is less contrasty than the 500, and almost the equal of the zoom, though perhaps a bit less contrasty) I wanted to compare the 400/4 stopped down one stop (which would make it the same aperture as the zoom).

10) I didn't included the chart of the 400 stopped down, but did include the analysis in my Excel table.

11) I used the rectangular marquee tool to select the blackest blacks and whitest whites at the three different lp/mm values for the 4 cases.

Those are listed, along with the difference between them in this table.
http://www.neonlightsimaging.com/images/comparison_table.tif

My conclusions:

I would say that the 400 stopped down is about the equal of the 500. At the same aperture as the zoom (zoom wide-open, 400 stopped down to match), the 400 is superior. Both wide-open, the zoom is just a bit more contrasty.

Essentially, when you get a 400/4 DO IS, you're buying an extra stop. When you use that extra stop, you pay the price of slightly reduced contrast. However, if you stop down to the same stop as you'd use with the zoom (applicable during most of the day, but probably not right after sunup and right before sundown), you'll gain contrast and you'll also gain in increased AF accuracy (IMO, at least on a 1D). Of course, with the 2x TC, you actually gain the ability to use AF on the 1D. With a 10D or D60, you can use AF with the 1.4x . . . unlike what you can do with the zoom with the TC added.

Compared to the 500, you're gaining the ability to carry it hand-held. If you're going to set-up on a tripod anyway, you're gonna prefer the 500. Both for its better contrast, but also for it's additional focal length.

I think the 400/4 does what its intended purpose is. But it does have a few tradeoffs.

Based on everything I've seen so far, I think I'm keeping this lens. And it's gonna be a lot easier now to consider selling the 100-400, though I may still have a problem doing that. But I think that a great lightweight combo to carry is the 400/4 DO IS and 70-200/2.8 IS, along with a 1.4x TC.

--
The Lowest Paid Concert Photographer Around
http://www.neonlightsimaging.com/artshow/final.htm
Photography -- just another word for compromise

'Since we can't keep crime in check, why don't we legalize it and tax it out of business?' -- Will Rogers
 
Oops. Thanks. I was working from memory. Should've been looking at the image.

It'd be kinda nuts to get the 400/4 DO if it were that much taller and heavier than the 500/4, wouldn't it? LOL.
FYI - You flipped the position of the 500mm f/4 and 400 f/4DO in
your caption. For the folks that don't know the 500mm is the tall
one (second froml left) and the little guy (third from left) is the
400mm f/4 DO!.

Really looking forward to your results!!!
--
The Lowest Paid Concert Photographer Around
http://www.neonlightsimaging.com/artshow/final.htm
Photography -- just another word for compromise

'Since we can't keep crime in check, why don't we legalize it and tax it out of business?' -- Will Rogers
 
As an aside, I took some "real world" shots today. Controlled ones, and not exactly great ones (or even wildlife). But it does address things that couldn't be addressed in the testing we did in Dallas yesterday.

I used a tripod, mirror-lockup, and a remote release. Turned off the IS on both lenses. And took a picture of a mailbox, with some nice things in the background to give various types of bokeh.

I tested the 100-400 and 400/4 DO at f/4 (400 only, of course), f/5.6, and f/8.

I used manual exposure. Everything seems to be pretty much the same exposure, except the 100-400 at f/5.6 -- I guess it was a bit brighter then.

I used Fred Miranda's CS-Pro on the LOW setting. White-balance was taken from one of the shots (picked it in YarcPlus) and applied to all shots. ISO was 200. YarcPlus was used with ARF=2.

Here are the shots. All saved as TIF 16-bits upon conversion, then saved as JPG-10 for display here.

http://www.neonlightsimaging.com/images/100-400_f56_mailbox.jpg

http://www.neonlightsimaging.com/images/100-400_f8_mailbox.jpg

http://www.neonlightsimaging.com/images/400_f4_mailbox.jpg

http://www.neonlightsimaging.com/images/400_56_mailbox.jpg

http://www.neonlightsimaging.com/images/400_f8_mailbox.jpg

The filenames should be self-explanatory.

--
The Lowest Paid Concert Photographer Around
http://www.neonlightsimaging.com/artshow/final.htm
Photography -- just another word for compromise

'Since we can't keep crime in check, why don't we legalize it and tax it out of business?' -- Will Rogers
 
Oops. Thanks. I was working from memory. Should've been looking
at the image.

It'd be kinda nuts to get the 400/4 DO if it were that much taller
and heavier than the 500/4, wouldn't it? LOL.
LOL - I was thinking "Ooh that 500 doesn't look as big as I'd have imagined - can I afford one?"
Unfortunately the answer's still "no" even if it isn't that small :-(

--
KRs
Chris

My meagre efforts are at http://www.dslr.co.uk
 
Actually, you can pretty much have your choice of the 400/4 DO IS and 500/4 IS for the same price.

Extra focal length + better images (however slight that advantage is or isn't), or the lighter weight?

Darnit, so many choices.
LOL - I was thinking "Ooh that 500 doesn't look as big as I'd have
imagined - can I afford one?"
Unfortunately the answer's still "no" even if it isn't that small :-(
--
The Lowest Paid Concert Photographer Around
http://www.neonlightsimaging.com/artshow/final.htm
Photography -- just another word for compromise

'Since we can't keep crime in check, why don't we legalize it and tax it out of business?' -- Will Rogers
 
Actually, you can pretty much have your choice of the 400/4 DO IS
and 500/4 IS for the same price.

Extra focal length + better images (however slight that advantage
is or isn't), or the lighter weight?

Darnit, so many choices.
Not for me there aren't - I can't afford either :-(
That said, I'd have great difficulty justifying it even if I could ;-)

--
KRs
Chris

My meagre efforts are at http://www.dslr.co.uk
 
As an aside, I took some "real world" shots today. Controlled
ones, and not exactly great ones (or even wildlife). But it does
address things that couldn't be addressed in the testing we did in
Dallas yesterday.

I used a tripod, mirror-lockup, and a remote release. Turned off
the IS on both lenses. And took a picture of a mailbox, with some
nice things in the background to give various types of bokeh.

I tested the 100-400 and 400/4 DO at f/4 (400 only, of course),
f/5.6, and f/8.

I used manual exposure. Everything seems to be pretty much the
same exposure, except the 100-400 at f/5.6 -- I guess it was a bit
brighter then.

I used Fred Miranda's CS-Pro on the LOW setting. White-balance
was taken from one of the shots (picked it in YarcPlus) and applied
to all shots. ISO was 200. YarcPlus was used with ARF=2.

Here are the shots. All saved as TIF 16-bits upon conversion, then
saved as JPG-10 for display here.

http://www.neonlightsimaging.com/images/100-400_f56_mailbox.jpg

http://www.neonlightsimaging.com/images/100-400_f8_mailbox.jpg

http://www.neonlightsimaging.com/images/400_f4_mailbox.jpg

http://www.neonlightsimaging.com/images/400_f56_mailbox.jpg

http://www.neonlightsimaging.com/images/400_f8_mailbox.jpg

The filenames should be self-explanatory.

--
The Lowest Paid Concert Photographer Around
http://www.neonlightsimaging.com/artshow/final.htm
Photography -- just another word for compromise

'Since we can't keep crime in check, why don't we legalize it and
tax it out of business?' -- Will Rogers
--
The Lowest Paid Concert Photographer Around
http://www.neonlightsimaging.com/artshow/final.htm
Photography -- just another word for compromise

'Since we can't keep crime in check, why don't we legalize it and tax it out of business?' -- Will Rogers
 
The 100-400 looks pretty good in your test even at 5.6. I would of expected the primes to be a bit more noticable in image quality.

I just got a sigma 50-500ex and im not real thrilled with it at 5.6 (6.3), it sharpens up nicely at f8 but Im thinking of exchanging it for the 100-400 (that is if delta international takes returns)

Ive read so much about the 100-400 not being that sharp at 5.6 and at least to my eyes it looks pretty good. Even more so compared to the DO prime which cost a good deal more

Any other shots and info you have in regards to the 100-400 I would love to see. I also am considering the "cheaper" primes such as the 300 f4 IS or the 400 5.6.
 
Yep. I had left off the semi-colons, and thought that all images had started loading, before going back and adding the semicolons.

I guess that one hadn't really started loading, and I missed it.

--
The Lowest Paid Concert Photographer Around
http://www.neonlightsimaging.com/artshow/final.htm
Photography -- just another word for compromise

'Since we can't keep crime in check, why don't we legalize it and tax it out of business?' -- Will Rogers
 
I think the 100-400 gets a bad rap. It's a much better lens than many seem to believe.

You have to remember that when people talk about the differences between some of these lenses, they are talking about things that most would never notice without meticulous attention to detail. And that would hardly be noticed by almost anybody without a side-by-side shot for comparison.

For example, the 100-400 beats the 70-200/2.8 IS + 2x TC in any kind of testing, when you compare side by side images. But if you needed the range, and didn't have the 100-400 with you (or couldn't afford both lenses), there's no reason to hestitate to use the 2X TC with that lens.
The 100-400 looks pretty good in your test even at 5.6. I would of
expected the primes to be a bit more noticable in image quality.

I just got a sigma 50-500ex and im not real thrilled with it at 5.6
(6.3), it sharpens up nicely at f8 but Im thinking of exchanging it
for the 100-400 (that is if delta international takes returns)

Ive read so much about the 100-400 not being that sharp at 5.6 and
at least to my eyes it looks pretty good. Even more so compared to
the DO prime which cost a good deal more

Any other shots and info you have in regards to the 100-400 I would
love to see. I also am considering the "cheaper" primes such as
the 300 f4 IS or the 400 5.6.
--
The Lowest Paid Concert Photographer Around
http://www.neonlightsimaging.com/artshow/final.htm
Photography -- just another word for compromise

'Since we can't keep crime in check, why don't we legalize it and tax it out of business?' -- Will Rogers
 
Somebody mentioned in a previous thread the "banding" he thought he saw. I don't know what image he was referring to, but it might've been one similar to this (same mailbox).

If so, what appears to be banding in some of the darker areas is really being caused by the brick veneer of houses. The detail is real, just very much out of focus.

--
The Lowest Paid Concert Photographer Around
http://www.neonlightsimaging.com/artshow/final.htm
Photography -- just another word for compromise

'Since we can't keep crime in check, why don't we legalize it and tax it out of business?' -- Will Rogers
 
IT did when I testerd it at close range - I still can't understand this "Infinity" Business with the Sigma 170-500 (and possibly the 100-400L as people seem to say it's soft) ....

IS it down to AF? . I guess so to a certain degree as I've been using the Sigma on the 1D and guess what - it's very good indeed at 500mm, still not too clever on the 10D / D60 unless stopped down to F11 - this is a pain as I lose the 500mm advantage on the 1D due to the FOV crop and only gain an effective 10mm over using the lens at 400mm on the 10D :( ..

As it happens, the TWO people championing the 100-400IS wide open at 400mm are Both 1D owners - hmmmmmmmmmmm..

--
Please ignore the Typos, I'm the world's worst Typist

The No1 Dedicated 1D forum in the UK -------->

http://www.1dforum.co.uk/php/phpBB2/

 
Going by the test chart download and your report (which didn't mention sharpness) , the expensive 400DO looks like incredibly poor value for money indeed (Over 7500 US Dollars in the UK) - the 500 looks incredible as expected, though it's one of canon's best ever lenses period!

--
Please ignore the Typos, I'm the world's worst Typist

The No1 Dedicated 1D forum in the UK -------->

http://www.1dforum.co.uk/php/phpBB2/

 
I think you'll find that "infinity issues" are really "AF issues", and not "lens quality issues".

Getting an Angle-Finder C should help answer that. Or just doing some "focus bracketing".
IT did when I testerd it at close range - I still can't understand
this "Infinity" Business with the Sigma 170-500 (and possibly the
100-400L as people seem to say it's soft) ....

IS it down to AF? . I guess so to a certain degree as I've been
using the Sigma on the 1D and guess what - it's very good indeed at
500mm, still not too clever on the 10D / D60 unless stopped down to
F11 - this is a pain as I lose the 500mm advantage on the 1D due to
the FOV crop and only gain an effective 10mm over using the lens at
400mm on the 10D :( ..

As it happens, the TWO people championing the 100-400IS wide open
at 400mm are Both 1D owners - hmmmmmmmmmmm..
--
The Lowest Paid Concert Photographer Around
http://www.neonlightsimaging.com/artshow/final.htm
Photography -- just another word for compromise

'Since we can't keep crime in check, why don't we legalize it and tax it out of business?' -- Will Rogers
 

Keyboard shortcuts

Back
Top