How sharp is sharp?

[EOS mE]

sorry if this is a newbie question. but i've seen some photos on here that are super sharp, tack sharp even. now, by sharp do you guys define it as sharp when viewed at 100% or 100% crop? or when after the photo has been finished with PP?

cuz i just picked up the 70-200L f2.8 IS. i took some photos and at 100% crop, they are not as sharp as i had imagine. not even tack sharp without any PP. but i'd think that an L lense should be sharp no matter what right? or is that wishful thinking in a perfect world?

[malla1962]

I don't think you will find any lens tack sharp at 100% without PP,when viewed at 100% the image is about 4ft long.;)

[liza]

Most DSLR images require PP to counter the effect of the anti-aliasing filters. It has little to do with the lens.
Just the nature of the beast. :)

[EOS mE]

oh.. ok. and all this time, i thought it needs to be viewed at 100% to be sharp. silly me. i've done some indoor shots today and yesterday when i got the lense. i've been too sick to even shoot a stable photo. can you believe that even with the IS on.. i can still shoot a blurred photo at 1/60s?

i hope to God i get better soon tomorrow.. i need to go out and take some photos and put this baby to the test.

[malla1962]

I don't think you will have any problems with a 70-200f2.8Lis,its a superb lens
just mack sure you set the distance switch on the side of the lens for better focus speed.:D

[ed rader]

sorry if this is a newbie question. but i've seen some photos on here that are super sharp, tack sharp even. now, by sharp do you guys define it as sharp when viewed at 100% or 100% crop? or when after the photo has been finished with PP?

cuz i just picked up the 70-200L f2.8 IS. i took some photos and at 100% crop, they are not as sharp as i had imagine. not even tack sharp without any PP. but i'd think that an L lense should be sharp no matter what right? or is that wishful thinking in a perfect world?

here are a couple 100% crops with no PP shot with the 70-200L f4 IS.

ed rader

http://www.fototime.com/9D0C9E29F5B5EB0/orig.jpg

http://www.fototime.com/D96CD4755D94825/orig.jpg

[EOS mE]

I don't think you will have any problems with a 70-200f2.8Lis,its a superb lens
just mack sure you set the distance switch on the side of the lens for better focus speed.:D

thx. i've currently set it at 1.4mm-infinity. i'd figure unless if my object is 8ft or more, then i'd have to set it at 2.5m-infinity.

here is one photo i took just outside of my balcony. did some light PP in ACDSee PRO. center focus on the red leaf in the center.

[mcminty]

Take it outside, in the sun, and shoot some stuff at F5.6 or F8. Shutter speed around 1/500 or higher. It should be a little sharper. :)

Andrew.

[EOS mE]

here are a couple 100% crops with no PP shot with the 70-200L f4 IS.

ed rader

http://www.fototime.com/9D0C9E29F5B5EB0/orig.jpg

http://www.fototime.com/D96CD4755D94825/orig.jpg

Ed, those are supurb shots! even at 100% those are still sharp! looks like i need to test this lens out in real life and see how it performs.

[malla1962]

thx. i've currently set it at 1.4mm-infinity. i'd figure unless if my object is 8ft or more, then i'd have to set it at 2.5m-infinity.

here is one photo i took just outside of my balcony. did some light PP in ACDSee PRO. center focus on the red leaf in the center.Nice start.:D

[EOS mE]

Take it outside, in the sun, and shoot some stuff at F5.6 or F8. Shutter speed around 1/500 or higher. It should be a little sharper. :)

Andrew.

thanks, i've heard that the IS @ 2.8 should be pretty sharp too. i haven't tried shooting it at F4.. but i'm sure at F4 it should be very very sharp! it just has to be. i'm going to blame it on my sick self right now for not be able to produce sharp photos. it's not the lense.. it's me. :cry:

[EOS mE]

Nice start.:D

thanks malla.. i'm pretty much grounded to my bed. wify don't like me walking around and playing with my camera stuff while i'm sick. heck.. i'm not even suppose to be on POTN right now.. but she's showering.. haha.

[Hellashot]

thx. i've currently set it at 1.4mm-infinity. i'd figure unless if my object is 8ft or more, then i'd have to set it at 2.5m-infinity.

here is one photo i took just outside of my balcony. did some light PP in ACDSee PRO. center focus on the red leaf in the center.

Try an object that is easier to focus on for testing. From my point of view, the focus point is not easy to identify. And use autofocus. It looks like you took a big jump up in lenses from what you had.. maybe too big a jump? :)

[GyRob]

oh.. ok. and all this time, i thought it needs to be viewed at 100% to be sharp. silly me. i've done some indoor shots today and yesterday when i got the lense. i've been too sick to even shoot a stable photo. can you believe that even with the IS on.. i can still shoot a blurred photo at 1/60s?

i hope to God i get better soon tomorrow.. i need to go out and take some photos and put this baby to the test.
although you can get sharp shots at 1/60s and lower if your at the 200mm end great care still needs to be taken IS is good but not a failsafe.
Rob.

[rklepper]

here are a couple 100% crops with no PP shot with the 70-200L f4 IS.

ed rader



Very nice Ed. These really look good.

[SYS]

You'll love the lens more you get to know it. I had the same reaction when I first got my 24-70L. I'd dare say here that the image problem is at least 90% due to user errors, not the lens.

[Filmus]

cuz i just picked up the 70-200L f2.8 IS. i took some photos and at 100% crop, they are not as sharp as i had imagine. not even tack sharp without any PP. but i'd think that an L lense should be sharp no matter what right? or is that wishful thinking in a perfect world?

Some 100% unprocessed crops all using the 70-200 F4 IS.

http://k47.pbase.com/o6/29/239629/1/70417049.0bZs2LXD.pbaseIMG_2_320sec.jpg

Max. dimension of an image allowed on POTN is 800 pixels. Larger images may be linked to, but not attached or embedded.

Jon

Full size imgaes (Examples of slow speed hand held, not the sharpest)
http://www.pbase.com/carmel_palti/image/70663445
http://www.pbase.com/carmel_palti/image/70663448
http://www.pbase.com/carmel_palti/image/70663504
Using Canon TC 1.4
[/URL]http://www.pbase.com/carmel_palti/image/70509430 (http://www.pbase.com/carmel_palti/image/70663440)
ISO 800 Sample
[url]http://www.pbase.com/carmel_palti/image/72111831&exif=Y

[The Hardcard]

Sharp is a somewhat relative term also. it is common for people - when describing that lens as sharp @ 2.8 - to mean compared to most other lenses at 2.8.

[EOS mE]

Some 100% unprocessed crops all using the 70-200 F4 IS.

http://k47.pbase.com/o6/29/239629/1/70417049.0bZs2LXD.pbaseIMG_2_320sec.jpg

Max. dimension of an image allowed on POTN is 800 pixels. Larger images may be linked to, but not attached or embedded.

Jon

Full size imgaes (Examples of slow speed hand held, not the sharpest)
http://www.pbase.com/carmel_palti/image/70663445
http://www.pbase.com/carmel_palti/image/70663448
http://www.pbase.com/carmel_palti/image/70663504
Using Canon TC 1.4
http://www.pbase.com/carmel_palti/image/70509430
ISO 800 Sample
http://www.pbase.com/carmel_palti/image/72111831&exif=Y

jon.. now those are some scary pics! super sharp!

[EOS mE]

Try an object that is easier to focus on for testing. From my point of view, the focus point is not easy to identify. And use autofocus. It looks like you took a big jump up in lenses from what you had.. maybe too big a jump? :)

maybe so. the only other lense i had before this is the 17-55, which is my walk around lense for everything. that is sharp.

although you can get sharp shots at 1/60s and lower if your at the 200mm end great care still needs to be taken IS is good but not a failsafe.
Rob.

thanks Rob. having the IS is a plus.. but i don't bank on it. stable hands and good body position is more than a must. but i do find 200mm is a bit harder for me to shoot compared to the photos at 70mm.

[EOS mE]

anyone have any photos shot with 70-200L 2.8IS @ wide open? i shot some at F4 with my IS and it's niceeeee..... but at 2.8, it's so-so...

[Bob_A]

here are a couple 100% crops with no PP shot with the 70-200L f4 IS.

ed rader

http://www.fototime.com/9D0C9E29F5B5EB0/orig.jpg

http://www.fototime.com/D96CD4755D94825/orig.jpg

Ed, are these no PP as in shot in RAW and converted to jpeg with sharpening turned "Off" or shot in jpeg with the sharpening set to "Standard" (i.e., some in camera processing) but no additional sharpening in post?

It's just that they don't look like images that are possible with no PP unless they have been sharpened in-camera.

Great images by the way :)

[ed rader]

Bob -- they were shot large jpeg and i've never changed my camera default so there is in-camera sharpening.

ed rader

[EOS mE]

Bob -- they were shot large jpeg and i've never changed my camera default so there is in-camera sharpening.

ed rader

Ed... i like your linky on your Canon Gear... that's the same classical song my wife and I had played at out wedding.

btw.. is your 70-200 a F4 version or the 2.8IS version?

[ed rader]

Ed... i like your linky on your Canon Gear... that's the same classical song my wife and I had played at out wedding.

btw.. is your 70-200 a F4 version or the 2.8IS version?

Rui -- thanx. canon was played at our wedding too :D .

i have the f4 IS.

ed rader

[EOS mE]

i'm seeing that the photos at 100% crop at 2.8 is a bit softer compared to the photos shot at F4. is that normal?

i'm comparing the shots straight out of camera in RAW and DPP2.2 without sharpening into JPEG.

The first one is at wide open 2.8 and the second one is at F4. both are at 100% crop.

[pcasciola]

Comparing f/4 to f/2.8 is like comparing apples to oranges. Stop your f/2.8L IS down to f/4 and it will be (should be) tack sharp. With the f/4 lens the light is passing through a lot less glass, same as when things get sharp when you squint your eyes. Well, the squinting works for those of us that don't have 20-20 vision anyway. :)

My 70-200/2.8L IS with no PP (and I mean RAW with no in camera JPEG sharpening) are slightly soft at 100% when I shoot wide open, but sharpen up quite nicely with a simple unsharp mask of 300/0.3/0 as a starting point. Those are the values Canon suggests using to compensate for the softness added by the anti-aliasing filter on the 1-series cameras, and I believe the 20D is in the same ballpark.

EDIT: FYI, shooting JPEG with the default sharpness setting is sharpened, RAW is not.

[EOS mE]

Comparing f/4 to f/2.8 is like comparing apples to oranges. Stop your f/2.8L IS down to f/4 and it will be (should be) tack sharp. With the f/4 lens the light is passing through a lot less glass, same as when things get sharp when you squint your eyes. Well, the squinting works for those of us that don't have 20-20 vision anyway. :)

My 70-200/2.8L IS with no PP (and I mean RAW with no in camera JPEG sharpening) are slightly soft at 100% when I shoot wide open, but sharpen up quite nicely with a simple unsharp mask of 300/0.3/0 as a starting point. Those are the values Canon suggests using to compensate for the softness added by the anti-aliasing filter on the 1-series cameras, and I believe the 20D is in the same ballpark.

EDIT: FYI, shooting JPEG with the default sharpness setting is sharpened, RAW is not.

thanks philip. now that makes sense. when i stopped it down to F4.. it's sharp without any PP. good thing you mentioned it being soft at 2.8... cuz i thought i was seeing things.

[pcasciola]

Don't forget that at f/2.8 you are also dealing with a much more shallow depth of field, so for sharpness tests I try to stick with flat objects, like printed material.

[EOS mE]

oh... thanks philip.. let me find something else to shoot. so flat like a magazine flat?

[pcasciola]

Yeah, magazines and newspaper work. My favorite test subject is the lens box. :D

Of course, you should use a tripod and remote shutter release too if you can.

[abat]

here are a couple 100% crops with no PP shot with the 70-200L f4 IS.


Ed, out of interest, would you care to post the same shots with PP and sharpening. I would find that instructive.
Andrew

[ed rader]

Ed, out of interest, would you care to post the same shots with PP and sharpening. I would find that instructive.
Andrew

Andrew -- here is one and a link to my zoo album...all pics are with the f4 IS. some with 1.4 TC.

i gotta eat dinner, walk the dog and play with my 5d ;) .

ed rader

http://www.fototime.com/inv/4E0FC3BF736F478

http://www.fototime.com/D96CD4755D94825/orig.jpg

http://www.fototime.com/6E9AA8E45858236/orig.jpg

[Pinto]

I know you are excited, but take your time. Make sure you put this lens on a stable tripod and spend some time with it before you come to any conclusions. This lens is relatively heavy and has a definite learning curve.
I do not understand why anyone would give you images taken with the f4.0 as examples for you when that is a totally different lens. ???
If you just purchased a 70-200 f4, why would you want to see images from a 2.8? It doesn't make any sense, so don't pay any attention to them. Only take advice and compare images from posters with the same lens.
This is a great lens. Take your time. Good luck.

[Bob_A]

Bob -- they were shot large jpeg and i've never changed my camera default so there is in-camera sharpening.

ed rader

Thanks Ed, make sense. Looks like you have a nice copy of the f4L :) I'm really happy with mine also ... even though it's the non-IS version :)

[EOS mE]

Yeah, magazines and newspaper work. My favorite test subject is the lens box. :D

Of course, you should use a tripod and remote shutter release too if you can.

thanks philip. unfortunately i don't own a tripod yet. :( but i have notice this lense is a bit heavier compared to my other two lens (17-55 & 50 f1.4) :lol:

should i take the photos at 200mm or at 70mm or both for comparision?

[pcasciola]

If it were me, I would shoot 70mm, 135mm and 200mm at f/2.8 and each 1/3rd stop increment up to f/5.6 to find your sweet spot at each focal length. Use the center focus point (this will enable high accuracy focus since you have the f/2.8 version), and make sure it's on a high contrast edge. Shoot RAW and then compare the crops. If I weren't so busy I would test mine too just for comparisons sake.

Another thing I like to do when testing a lens is take a photo of a ruler or measuring tape at an angle to the lens at f/2.8. You will see the acceptable focus window, which should be centered on the point your focused on. This will help you find out if your lens is possibly front or back focusing.

[abat]

i gotta eat dinner, walk the dog and play with my 5d ;) .

Thanks Ed, wonderful shots. Enjoy your dinner ..... oh, and your 5D too.
Andrew

[EOS mE]

I tried out this baby today out in the field. here's a shot of my daughter.. 2yr old mini doxie, Chloe. handheld and shot at f4.. this is super sharp lense!! i'm in love!

straight out of camera with very very slight usm applied.

[asabet]

I would define tack sharp as an image which yields a processed 100% crop that looks like a processed resized image. For example, take a look at Gene Oleynik's 500/4L and 400/5.6L sampler galleries here: http://www.pbase.com/gaocus

In my experience, mainly with my dad's copy, the 70-200/2.8L IS is not going to give tack sharp images at f2.8 and 200mm. Maybe there are some better copies out there. Either way it is a phenomenal zoom and plenty sharp at all focal lengths and apertures.

[Pinto]

Very nice image, EOS mE.
Looks like you are getting a handle on that lens. All you need is some practice.
Your next purchase needs to be a good tripod.

[cjm]

I have gone partially blind in my Right eye, but hopefully its only a tempory thing but how sharp is Tack sharp? This was taken with my 24-70 L of my Dialated Eye after seeing the eye doctor. It is a 100% crop straight from the camera! Tell me that isnt Tack sharp?

[EOS mE]

Very nice image, EOS mE.
Looks like you are getting a handle on that lens. All you need is some practice.
Your next purchase needs to be a good tripod.

thanks pinto. yes.. tripod is next on my list. i'll be researching for them next year. gotta pay off these lens first... hehe :lol:


I have gone partially blind in my Right eye, but hopefully its only a tempory thing but how sharp is Tack sharp? This was taken with my 24-70 L of my Dialated Eye after seeing the eye doctor. It is a 100% crop straight from the camera! Tell me that isnt Tack sharp?

christopher.. glad to hear your eye is better.. but man.. that's just too close of a call. looks like i can almost poke your eyes out.. haha :lol:

[cjm]

The more I use this lens, the pickier I think I become with Pixel Peeping. Before you know it I will be expecting every lens I ever own to be this sharp. Hope not. :lol:

[EOS mE]

haha.. it's ok chris.. i'm already addicted to pixel peeping.. guess that brings a new term to PP (pixel peep) instead of post process.. haha :lol:

[RgB]

This bird is a Rainbow Lorikeet.

I'd call this sharp especially at f/2.8 using my 70-200mm f/2.8L IS shot in Standard large JPG added slight USM though not needed.

Here is the full photo
EXIF Summary: 1/250s f/2.8 ISO160 195mm
http://homepage.mac.com/danielsperanza/other2/sharp1.jpg
100% Crop
http://homepage.mac.com/danielsperanza/other2/sharp2.jpg

I got 36 pics in my backyard of about the same quality.

[EOS mE]

nice Daniel. mine is sharp also.. just found out that i need to work on camera posture with this baby on the 30D.

[trailstar]

30D + Sigma 17-70

100% crop, no PP

[RgB]

nice Daniel. mine is sharp also.. just found out that i need to work on camera posture with this baby on the 30D.

I have to think more about posture too.

You dog pic is very good. I just took my dog to the vet to remove a cancer on his ear flap:( first test look benign.

[EOS mE]

I have to think more about posture too.

You dog pic is very good. I just took my dog to the vet to remove a cancer on his ear flap:( first test look benign.

daniel, thank God it's on on his ear flap. mostly, good thing you guys found it first! hope he's recovering.

[RgB]

He's back home and the vet removed it all and thinks it shouldn't come back:) I dropped him off this morning at 9 AM and picked him up at 6 PM and he even wanted to go for a small walk when he got home:D

[EOS mE]

He's back home and the vet removed it all and thinks it shouldn't come back:) I dropped him off this morning at 9 AM and picked him up at 6 PM and he even wanted to go for a small walk when he got home:D

wow.. that's good. most doggie would be so drugged out to go for a walk. when our doxie got spayed, she was too drugged out to even wanna eat. so she just slept even after she came home. such a mommy's girl...

[RgB]

wow.. that's good. most doggie would be so drugged out to go for a walk. when our doxie got spayed, she was too drugged out to even wanna eat. so she just slept even after she came home. such a mommy's girl...

It was hits ear so it wasn't major operation where they opened him up. He did apparently sleep in the recovery room straight after the going under:confused:

[DrPablo]

Sharpness is only a meaningful adjective when applied to the desired output of your photo.

If you were able to do the digital equivalent of contact printing, i.e. print all your images at capture size, your print dimensions would be 15x24mm (the size of a postage stamp) and you couldn't tell the difference between a Canon and a Quantaray lens even if viewed up close.

On the other hand, if your output is an 16x24 inch enlargement, then you need to enlarge your captured image 25 times. The bigger you enlarge, the more intermediate fine detail your image will require in order to look sharp from a given viewing distance. So the ability of a ridiculously sharp Zeiss lens to resolve and transmit fine detail will matter much more at big enlargements, where that resolved detail becomes visible.

The problem with a 100% crop, is you're looking at the image at an output size far beyond what you're ever likely to print. So why would you take a lens and enlarge its image beyond its resolving power, just to prove a point about how it looks at some different size? If you want to know how an image will look when enlarged to 11x14, then make its dimensions 11x14 and do a detail-crop at print size -- don't enlarge to 4x6 feet and expect to draw some inference from that.

Furthermore, absolute pixel counts do NOT determine sensor resolution. Pixel density does. By definition, the 5D can resolve a theoretical maximum of 61 line pairs per mm (assuming line pairs that are 1 pixel wide). That is the lowest resolution of any DSLR ever made -- the 400D can resolve 81 line pairs per mm, because there are simply more photosites per mm. And yet the 5D has more total pixels. So if you enlarge to 100% from the 5D the image will be larger than that of the 400D; and yet you'll resolve more fine detail from identical crops from the 400D image because of its pixel density. But then you make a print and you'll find that the 400D image will suffer more because of its increased enlargement factor.

These factors render the 100% crop a fun but nearly useless way of testing lens performance, especially when the camera-to-camera variables are taken into account.