24-105 Resolution testing

[duydaniel]

I have a 19 inches LCD monitor.
Viewing this: http://www.graphics.cornell.edu/~westin/misc/ISO_12233-reschart.pdf
enlarging this file at 500%
and shoot this picture:

Handheld
5D3
iso 1600
105 mm
f4
1/160
IS on

exported straight from LR.
Any comments on resolution?

 

[Mt Spokane Photography]

I can't see any details in the image on my screen.

The chart should be well lighted and fill as much of the camera sensor as possible. That way, you can see problems with lens element alignment as well as center sharpness. Star charts are better yet for element alignment issues.

I'm might be missing the point here.

 

[duydaniel]

I can't see any details in the image on my screen.

The chart should be well lighted and fill as much of the camera sensor as possible. That way, you can see problems with lens element alignment as well as center sharpness. Star charts are better yet for element alignment issues.

I'm might be missing the point here.

this was very none scientific but if you download and zoom in, you will see soe detail since i didn't crop it.
my intention was making some reference or comparison where result somewhat reproducible.

 

[Sporgon]

You gotta be kidding.

All this shows is that your desk is more untidy than mine.

 

[Loswr]

...my intention was making some reference or comparison where result somewhat reproducible.

As Master Shifu said, "Excellent work…if you were trying to disappoint me.". Reproducible? A shot of a defined subject with a monitor of a defined size, fine. Handheld shot, unknown display resolution, contrast setting, distance, it's at an angle, etc.

 

[LetTheRightLensIn]

no

;D

honestly, it's hard to say, it's not at all reproducible without carrying out numerous complex calculations and carefully measuring out in many ways a replicated scenario, which would take a lot of effort and time; and then adding in hand-held with IS at only 1.6x over 1/FL makes it basically impossible to replicate, maybe some very advanced image analysis could try to measure potential micro-shake and remove it but....

the fact that it is wide open at 105mm for that lens and doesn't look totally soft might be a positive sign, but honestly it's hard to tell much of anything

 

[distant.star]

.
I can positively conclude the lens cap was removed for this shot.

That's a start.

 

[AlanF]

What a complete waste of time.

 

[Mt Spokane Photography]

Tough bunch tonight, don't give up quite yet, keep trying.

 

[Pi]

Looks good for that ISO but try ISO 100, tripod, no IS. Better yet - natural light, speed faster than 1/100.

 

[takesome1]

Crop the picture to show only the monitor.
Put the crop on full screen and switch back and forth between your crop and the 500x screen view.
This will show you the resolution drop from one the other.

Repeat the test;
Get your baggie out,
Roll up another big fat one and smoke it,
Then repeat the test again to see if you get similar results.

 

[duydaniel]

I can't see why you can't somewhat reproduce it (only if you want to compare your copy).
1) You have a pdf file and view at 500%
2) Since I didn't crop the image, you can compare your shot with mine at 1:1

Now speaking of screen resolution, it wasn't matter imo because when viewing the chart at 500%, even a super old 840x1024 turns into a 4200x512000 which is plenty of resolution.

Now if yours at 1:1 is smaller/bigger than mine meaning you were further/closer to the screen respectively.
I just hope it helps a potential buyer who may looking.

I checked the image at 300% and didn't feel a tripod would improve much.

 

[AlanF]

Looks good for that ISO but try ISO 100, tripod, no IS. Better yet - natural light, speed faster than 1/100.

It's a shot of an LCD screen (at an angle as well). So, how would natural light be better?

 

[duydaniel]

I only have a copy of 24-105 fortunately though the 5D3 is the highest resolution from Canon.
I hope it is helpful for someone looking for a comparison/ looking to buy one
Better yet if you have similar set up, you could tell me how was your copy compared to mine

 

[Sporgon]

I can't see why you can't somewhat reproduce it (only if you want to compare your copy).
1) You have a pdf file and view at 500%
2) Since I didn't crop the image, you can compare your shot with mine at 1:1

Now speaking of screen resolution, it wasn't matter imo because when viewing the chart at 500%, even a super old 840x1024 turns into a 4200x512000 which is plenty of resolution.

Now if yours at 1:1 is smaller/bigger than mine meaning you were further/closer to the screen respectively.
I just hope it helps a potential buyer who may looking.

I checked the image at 300% and didn't feel a tripod would improve much.

Two words you don't use in a test; hope and much !

Without going into great explanations your 'test' is likely to produce inaccurate information, so it is the worst type of test. The methodology may result in suggesting some inferior lenses are better than superior ones.

 

[Loswr]

I can't see why you can't somewhat reproduce it (only if you want to compare your copy).

What good is 'somewhat reproducing' the test conditions if those conditions aren't likely to produce reproducible results?

 

[Pi]

Looks good for that ISO but try ISO 100, tripod, no IS. Better yet - natural light, speed faster than 1/100.

It's a shot of an LCD screen (at an angle as well). So, how would natural light be better?

Natural light would be stronger. He can use ISO 100 and 1/100 s or faster.

 

[Loswr]

Looks good for that ISO but try ISO 100, tripod, no IS. Better yet - natural light, speed faster than 1/100.

It's a shot of an LCD screen (at an angle as well). So, how would natural light be better?

Natural light would be better if direct, natural light shining on a backlit LCD makes it brighter and increases contrast. In my experience, neither is true.