Cropped Lens vs Full Frame Lens on APS-C Sensor

jasonZed · May 27, 2017

I have a theory that I'm working on and I wanted to see if it had any merit to it. I have a Canon Rebel T5 (non-i) and I had decided to buy what I thought would be a better lens than the kit lens. Now I like to shoot landscapes and so I thought I should start building up the gear for a full frame camera. After a bit of research I decided to start with a Sigma 17-50mm 2.8 EX DC OS HSM lens. Got some good reviews and the glass seemed to be pretty good. However, after using the lens and T5 together for a year now, I've noticed that the focus isn't spot on. Even when I manually focus, it's not as sharp as the kit lens (Canon EFS 18-55mm 3.5-5.6 IS II).

Here's a comparison, the left is the kit lens, right the Sigma (keep in mind this is just to show my point not for quality reasons):

ec60054fc2f346feb703d6a06a34c232.jpg.png

If you notice in these two pictures, the edges on the stitching and in the shadows are a tad bit crisper on the left (kit lens). And it's more exaggerated when I'm shooting landscapes.

So here's my theory; could it be the fact that I'm using a full frame lens on a cropped-sensor camera and since the rear element is a different size and distance from the sensor, it's not ever going to be as sharp?

I also know that the higher end cameras have the ability to do auto-focus micro adjustments but in the case, the focus even when done manually is never as sharp.

And I know that I need to upgrade my body, I'm already exceeding it's limitations, but it was a Christmas gift from my parents and I wanted to use it as long as I could stand to. I'm even looking at maybe the Sone Alpha line... the A7's are priced pretty good and they are full frame. But right now I'm not tied to any brand. I've shot with Canon, Nikon, and Pentax and I've gotten pretty good results with any one of those.

Anyhow, just let me know if I'm way off base on the full frame lens on an APS-C sensor theory.

KCook · May 27, 2017

That Sigma is a crop sensor design, not full frame!
The f-stop for the sample shot is not given. f/2.8 could explain softness, or was this stopped down?

Kelly

Tom_N · May 28, 2017

jasonZed said:
So here's my theory; could it be the fact that I'm using a full frame lens on a cropped-sensor camera and since the rear element is a different size and distance from the sensor, it's not ever going to be as sharp?

The distance from the sensor shouldn't be an issue. In Canon's system, the EF-S lenses are allowed to protrude further into the body, but the mount was designed specifically with full-frame EF lenses in mind, too. Each type of lens is responsible for bending the light in the way appropriate for its own rear element to sensor distance.

It's more likely that you are seeing differences in the quality of the particular lenses. Differences in settings (aperture, shutter speed, ISO), lighting, or post-processing (e.g., in-camera sharpening) could affect things, but I assume you've controlled for those.

Gerry Winterbourne · May 28, 2017

jasonZed said:
I have a theory that I'm working on and I wanted to see if it had any merit to it.

It doesn't (more below).

jasonZed said:
I have a Canon Rebel T5 (non-i) and I had decided to buy what I thought would be a better lens than the kit lens. Now I like to shoot landscapes and so I thought I should start building up the gear for a full frame camera. After a bit of research I decided to start with a Sigma 17-50mm 2.8 EX DC OS HSM lens.

The first problem with your theory is that this isn't a FF lens so you are basing your theory about FF v crop on a comparison of crop v crop.

jasonZed said:
Got some good reviews and the glass seemed to be pretty good. However, after using the lens and T5 together for a year now, I've noticed that the focus isn't spot on.

If you really mean focus - whether the lens is properly focused at the point you want in perfect focus - that's a matter of either your technique or the interaction of the body's and the lens's calibration regardless of the size of lens.

I suspect you don't mean focus as such but resolution ...

jasonZed said:
Even when I manually focus, it's not as sharp as the kit lens (Canon EFS 18-55mm 3.5-5.6 IS II).

... which this tends to confirm.

Have you done your comparison with all settings equal for each pair: you obviously can't compare them at f/2.8. You'd need to compare at (say) f/3.5, f/5.6, f/8.

jasonZed said:
Here's a comparison, the left is the kit lens, right the Sigma (keep in mind this is just to show my point not for quality reasons):

Does this comparison do as I suggested? Which f-stop is it at

jasonZed said:
So here's my theory; could it be the fact that I'm using a full frame lens on a cropped-sensor camera

No, because you aren't.

jasonZed said:
and since the rear element is a different size and distance from the sensor, it's not ever going to be as sharp?

No. The design of a lens includes every element, not just the rear one. Once the lens is designed it must be mounted on the lens: the lens mount distance is what keeps the lens the right distance from the sensor. For Canon DSLRs the mount distance is 44.00mm regardless of sensor size.

--
---
Gerry
___________________________________________
First camera 1953, first Pentax 1985, first DSLR 2006

Galleries by Gerry Winterbourne

www.pbase.com

[email protected]

Mystery member · May 28, 2017

Your theory is wide of the mark, as has been mentioned. Being f/2.8 doesn't make it a full-frame lens, if that is your misunderstanding.

In general, excellent FF lenses will perform equally well on APS-C and FF, but a so-so FF lens may be found wanting on APS-C. I have a number of FF lenses for my APS-C Nikon, and all of my Sony lenses are FF and eminently suitable for both a6000 and a7.

It's worth considering that very few long lenses are made specifically for APS-C. About the only one that I can think of is the Sony 55-210mm zoom. So, provided that the lens will actually fit the APS-C camera, there's no barrier to good performance because it happens to be FF.

Your comparison is not definitive because the main feature of both images is the noise, and it's impossible to gauge the sharpness.

Here's a fairly sharp image, again it has no photographic merit, being a test shot from a FF prime lens on my APS-C Sony a6000.

Michael Fryd · May 28, 2017

jasonZed said:
I have a theory that I'm working on and I wanted to see if it had any merit to it. I have a Canon Rebel T5 (non-i) and I had decided to buy what I thought would be a better lens than the kit lens. Now I like to shoot landscapes and so I thought I should start building up the gear for a full frame camera. After a bit of research I decided to start with a Sigma 17-50mm 2.8 EX DC OS HSM lens. Got some good reviews and the glass seemed to be pretty good. However, after using the lens and T5 together for a year now, I've noticed that the focus isn't spot on. Even when I manually focus, it's not as sharp as the kit lens (Canon EFS 18-55mm 3.5-5.6 IS II).

Here's a comparison, the left is the kit lens, right the Sigma (keep in mind this is just to show my point not for quality reasons):

If you notice in these two pictures, the edges on the stitching and in the shadows are a tad bit crisper on the left (kit lens). And it's more exaggerated when I'm shooting landscapes.

So here's my theory; could it be the fact that I'm using a full frame lens on a cropped-sensor camera and since the rear element is a different size and distance from the sensor, it's not ever going to be as sharp?

I also know that the higher end cameras have the ability to do auto-focus micro adjustments but in the case, the focus even when done manually is never as sharp.

And I know that I need to upgrade my body, I'm already exceeding it's limitations, but it was a Christmas gift from my parents and I wanted to use it as long as I could stand to. I'm even looking at maybe the Sone Alpha line... the A7's are priced pretty good and they are full frame. But right now I'm not tied to any brand. I've shot with Canon, Nikon, and Pentax and I've gotten pretty good results with any one of those.

Anyhow, just let me know if I'm way off base on the full frame lens on an APS-C sensor theory.

It's possible to produce very sharp images on a crop body with a full frame lens.

Keep in mind that the sharpness of a lens can vary with aperture. Most lenses tend to be their sharpest when stopped down 2 or 3 stops from their maximum aperture. An f/2.8 likely will be sharper at f/8 than at f/2.8.

A lens that opens very wide may not be as sharp as you expect when wide open.

This also might be a focus issue. Your lens might have a front or back focus issue. You may see an issue where the auto-focus is consistently focusing slightly in front of, or behind where it should.

Obviously, these are generalities. I don't have enough information for specific speculation on your issue.

I would suggest you try a few tests. Put the camera on a tripod and take a few test shots at a variety of apertures. See if some apertures are producing noticeably sharper images. You should also try taking a picture of a ruler angle away from you. See if the part of the ruler most in focus is closer or further from you than you expect.

You should also try manual focus with the camera in live view (use 10x magnification) (camera still on tripod). This bypasses the default phase detect auto focus, and should bypass front/back focus issues.

photosen · May 28, 2017

jasonZed said:
So here's my theory; could it be the fact that I'm using a full frame lens on a cropped-sensor camera and since the rear element is a different size and distance from the sensor, it's not ever going to be as sharp?

Nope, it's because sigma has to reverse engineer Canon's autofocus... My EF lenses focus perfectly on both my 30D and 70D: 35mm f2, 50mm f1.4, 70-200 f4; just as well as my EFS 24mm 2.8 and 10-22. My Tamron 90mm 2.8 hunts a lot more, but once it focuses it's spot on.

jasonZed · May 28, 2017

Okay, first off, thank you everyone for the corrections. For some reason I thought the Sigma was a FF lens.... Like I said, I bought it over a year ago and I know I was looking at both FF and Crop sensor lenses. I just forgot which one I ended up going with... So, that totally debunks my theory.

And thank you everyone for the insight and patience... I appreciate the constructive responses.

KCook said:
That Sigma is a crop sensor design, not full frame!

Got it! My mistake!

photosen said:
The f-stop for the sample shot is not given. f/2.8 could explain softness, or was this stopped down?

Both lenses at f/3.5, which is Canon's max.

Kelly === [QUOTE="photosen said:
Tom_N, post: 59623232, member: 672906"]
The distance from the sensor shouldn't be an issue. In Canon's system, the EF-S lenses are allowed to protrude further into the body, but the mount was designed specifically with full-frame EF lenses in mind, too. Each type of lens is responsible for bending the light in the way appropriate for its own rear element to sensor distance.

It's more likely that you are seeing differences in the quality of the particular lenses. Differences in settings (aperture, shutter speed, ISO), lighting, or post-processing (e.g., in-camera sharpening) could affect things, but I assume you've controlled for those.

Thanks for the info, I'm starting to think that this lens isn't just as sharp as the Canon kit lens. No post processing was done... just straight raw image and converted to jpg in LR. All settings (f-stop,shutter speed, ISO) were the same.

===

Gerry Winterbourne wrote:> If you really mean focus - whether the lens is properly focused at the point you want in perfect focus - that's a matter of either your technique or the interaction of the body's and the lens's calibration regardless of the size of lens. I suspect you don't mean focus as such but resolution ...[QUOTE="photosen said:
Even when I manually focus, it's not as sharp as the kit lens (Canon EFS 18-55mm 3.5-5.6 IS II).

... which this tends to confirm.
[/QUOTE]
Not sure I'm following on the resolution part; I'm using the same camera body... so the resolution is the same. I agree sharpness between lenses doesn't necessarily mean focus issues, just could be the lens isn't as crisp as the other.

photosen said:
Have you done your comparison with all settings equal for each pair: you obviously can't compare them at f/2.8. You'd need to compare at (say) f/3.5, f/5.6, f/8.

Does this comparison do as I suggested? Which f-stop is it at

f/3.5

photosen said:
No. The design of a lens includes every element, not just the rear one. Once the lens is designed it must be mounted on the lens: the lens mount distance is what keeps the lens the right distance from the sensor. For Canon DSLRs the mount distance is 44.00mm regardless of sensor size.

Thanks for the info! Learning everyday!

===

WryCuda, post: 59623232, member: 672906"]
Your theory is wide of the mark, as has been mentioned. Being f/2.8 doesn't make it a full-frame lens, if that is your misunderstanding.

In general, excellent FF lenses will perform equally well on APS-C and FF, but a so-so FF lens may be found wanting on APS-C. I have a number of FF lenses for my APS-C Nikon, and all of my Sony lenses are FF and eminently suitable for both a6000 and a7.

It's worth considering that very few long lenses are made specifically for APS-C. About the only one that I can think of is the Sony 55-210mm zoom. So, provided that the lens will actually fit the APS-C camera, there's no barrier to good performance because it happens to be FF.

Your comparison is not definitive because the main feature of both images is the noise, and it's impossible to gauge the sharpness.

Yeah, my mistake... I should have double-checked the lens was actually FF before making the assumption. Thanks for the info!

===
[/QUOTE]

Moon · May 28, 2017

jasonZed said:
Not sure I'm following on the resolution part; I'm using the same camera body... so the resolution is the same. I agree sharpness between lenses doesn't necessarily mean focus issues, just could be the lens isn't as crisp as the other.
===

All lenses are not created equal. Higher quality and/or better designed lenses are literally able to resolve more/finer detail. Regardless of the 'resolution' of your sensor - different lenses will allow for the visible 'resolution' of more or less detail. These 'better' lenses also tend to be sharper (or give the appearance of greater sharpness), better contrast, higher saturation and greater dynamic range.

jasonZed · May 28, 2017

Moon said:
jasonZed said:

Not sure I'm following on the resolution part; I'm using the same camera body... so the resolution is the same. I agree sharpness between lenses doesn't necessarily mean focus issues, just could be the lens isn't as crisp as the other.

===

Click to expand...

All lenses are not created equal. Higher quality and/or better designed lenses are literally able to resolve more/finer detail. Regardless of the 'resolution' of your sensor - different lenses will allow for the visible 'resolution' of more or less detail. These 'better' lenses also tend to be sharper (or give the appearance of greater sharpness), better contrast, higher saturation and greater dynamic range.

Mystery member · May 28, 2017

jasonZed said:
I'm starting to think that this lens isn't just as sharp as the Canon kit lens. No post processing was done... just straight raw image and converted to jpg in LR. All settings (f-stop,shutter speed, ISO) were the same.

Not sure I'm following on the resolution part; I'm using the same camera body... so the resolution is the same. I agree sharpness between lenses doesn't necessarily mean focus issues, just could be the lens isn't as crisp as the other.

I'd be surprised if the Sigma wasn't better than the Canon kit lens, particularly if you are comparing them at f/3.5. I use the closely related Sigma 17-70mm on my Nikon and it's quite a good lens.

It's rare for that the sensor resolution to be matched by that of the lens, and lens performance is quite often expressed in terms of effective resolution.

For example, the test image that I posted was from a 24MPix camera, but I can only make a case for the lens actually producing 13MPix of detail*. However, a commercial testing site (DxOMark) only rates the same combination at 11MPix "Perceptual Megapixels". Note that many people have issues with the testing methods used.

As a matter of interest, the same site rates the lens at 32MPix on a 42MPix FF camera, which relates to your original question. (On my FF 24MPix camera, I find a resolution of about 15MPix).

Have you tested your lenses using "Live-View"? This should eliminate issues with lens calibration.

* 4500x3000 = 13MPix

KCook · May 28, 2017

jasonZed said:
Both lenses at f/3.5, which is Canon's max.

That should have cleared up the softness in the f/2.8 zoom. So we are left with a defective lens, or missed focus, or shutter shock. No filter was on the Sigma, right?

Kelly

jasonZed · May 28, 2017

KCook said:
KCook said:

Both lenses at f/3.5, which is Canon's max.

Click to expand...

That should have cleared up the softness in the f/2.8 zoom. So we are left with a defective lens, or missed focus, or shutter shock. No filter was on the Sigma, right?

Kelly

Correct, no filters... I plan on doing some more "scientific" tests tomorrow with the Sigma, just to see if I can rule out defects in the lens. I'll do a focus test at different apertures and get the lighting better suited for testing. That's another gripe of mine on this T5, even 6400 ISO is very noisy, even when not zoomed in. Granted, that's a debate I've seen a lot going on...

jasonZed · May 28, 2017

So here are some examples of photos that I've taken with the T5 and the Sigma...

This one really isn't too bad, but I had to do a lot of post processing to get the sharpness there...

No amount of work in the post processing could really sharpen this one up. I took several of this one and went back and forth on manual and auto-focus and this is best I got out of the bunch. So, I'm starting to think that this lens doesn't have good f/2.8 quality... but....

Here's one of the many I took and not really any better at f/4.5.

Not bad here...

Isn't too bad here either...

Again, a lot of post processing time to get it this crisp. Would have been a better picture if it weren't for the car...

Granted, I was pushing this at 1/50sec at 17mm, but I would think that I could have a sharper image...

Lastly, a pano just because I love the mountains!

Tom_N · May 29, 2017

jasonZed said:
So here are some examples of photos that I've taken with the T5 and the Sigma...

I wonder if the issue is that you have a few very bright spots (the lights), and a lot of very dark ones (everything else). The camera may have exposed the picture correctly overall, but the sensor wouldn't have collected very much light for the dark areas.

So is the "lack of sharpness" a lens issue, or a simple case of "insufficient light" noise?

jasonZed said:
Isn't too bad here either...

If you use the DPR 'loupe' to look in this cat's eye, you'll see a reflection of the grass and sky that he's staring at.

Although, from his expression, and the body language of his ears, I was half-expecting to see the reflection of the image of a songbird that he was planning on having for dinner!

Gerry Winterbourne · May 29, 2017

Extract from your first post "... using the lens and T5 together for a year now, I've noticed that the focus isn't spot on. Even when I manually focus, it's not as sharp ,,,"

jasonZed said:
jasonZed said:

If you really mean focus - whether the lens is properly focused at the point you want in perfect focus - that's a matter of either your technique or the interaction of the body's and the lens's calibration regardless of the size of lens.

I suspect you don't mean focus as such but resolution ...

jasonZed said:

Even when I manually focus, it's not as sharp as the kit lens (Canon EFS 18-55mm 3.5-5.6 IS II).

Click to expand...

... which this tends to confirm.

Click to expand...

Not sure I'm following on the resolution part;

You started by saying the focus isn't spot on even when you focus manually, which means one of two things - either you aren't focusing properly even in manual, which can't be a fault in the lens; or you don't really mean focus isn't spot on but that the sharpness isn't what you'd expect.

Sharpness is a perception: it is governed by a combination of resolution and contrast. Lens tests, if conditions are consistent, measure the resolution part of sharpness. A combination of camera and lens isn't perfect: the lens reduces the ability of the combination to show fine detail and so does the camera (its sensor when the camera is digital).

There are ways of testing the lens without a camera but that's not what you're doing; with the lens on the camera you are testing the combined effect.

jasonZed said:
I'm using the same camera body... so the resolution is the same.

The sensor part of resolution is the same; the lens part probably isn't. Here's a greatly simplified explanation: wrong in details but a fair outline.

The best resolution your camera can give is about 3500 horizontal lines (this is roughly the same as the rows of pixels but not exactly so). The best resolution a lens can give can vary dramatically between lenses and at different apertures for a given lens. If the bare lens could give over 3500 lines then the combined output would be limited to 3500 by the camera. If the bare lens could deliver only 3000 lines that's what the combined output would be.

I've looked at test reports of your two lenses. Unfortunately can't find them both at the same site so the results aren't directly comparable. However, it is clear that most of the time neither of them gives 3500 lines so in this case what you say ("so the resolution is the same") isn't true.

jasonZed said:
I agree sharpness between lenses doesn't necessarily mean focus issues, just could be the lens isn't as crisp as the other.

"... as crisp" is a colloquial term. You've actually said what I've said - some lenses give less resolution than others.

jasonZed said:
jasonZed said:

Have you done your comparison with all settings equal for each pair: you obviously can't compare them at f/2.8. You'd need to compare at (say) f/3.5, f/5.6, f/8.

Does this comparison do as I suggested? Which f-stop is it at

Click to expand...

f/3.5

With a very few exceptions all lenses start relatively soft at their maximum aperture, improve on first stopping down and then gradually soften again. Here are links to the tests I mentioned (ignore the numbers in the Canon test - it uses a camera with less pixels than yours; what matters is the shape of the charts) http://www.photozone.de/canon-eos/831-canon_1855_3556stmis?start=1 https://www.ephotozine.com/article/...x-dc-os-hsm-interchangeable-lens-review-15056

As you see, both lenses are about equally soft at f/3.5 so the comparison you show just confirms that. The Sigma, though, improves faster before starting to soften again.

You really need to do your tests at all apertures in good light - or use long exposures - to eliminate the effects of noise.

jasonZed · May 29, 2017

Tom_N said:
jasonZed said:

So here are some examples of photos that I've taken with the T5 and the Sigma...

Click to expand...

I wonder if the issue is that you have a few very bright spots (the lights), and a lot of very dark ones (everything else). The camera may have exposed the picture correctly overall, but the sensor wouldn't have collected very much light for the dark areas.

So is the "lack of sharpness" a lens issue, or a simple case of "insufficient light" noise?

I'm not sure, I'm thinking this just isn't a good low light lens.

Tom_N said:
Tom_N said:

Isn't too bad here either...

Click to expand...

If you use the DPR 'loupe' to look in this cat's eye, you'll see a reflection of the grass and sky that he's staring at.

Although, from his expression, and the body language of his ears, I was half-expecting to see the reflection of the image of a songbird that he was planning on having for dinner!

I loved this comment. Although, this guys favorite snacks are socks!

KCook · May 29, 2017

jasonZed wrote:

jasonZed said:
So here are some examples of photos that I've taken with the T5 and the Sigma...

This one really isn't too bad, but I had to do a lot of post processing to get the sharpness there...

Darn! I wish you had shown these shots prior to PP, instead of after. Sharpness here is Ok in view of the low light and very high ISO. If you want really sharp SOOC results in these conditions switch to a full frame camera.

jasonZed said:
No amount of work in the post processing could really sharpen this one up. I took several of this one and went back and forth on manual and auto-focus and this is best I got out of the bunch. So, I'm starting to think that this lens doesn't have good f/2.8 quality... but....

Here's one of the many I took and not really any better at f/4.5.

Tripod shake. Alas the really solid tripods cost a small fortune. Budgets suck!

jasonZed said:
Not bad here...

Isn't too bad here either...

Furry critters are a challenge for any AF system. For real sharpness add flash.

jasonZed said:
Again, a lot of post processing time to get it this crisp. Would have been a better picture if it weren't for the car...

Granted, I was pushing this at 1/50sec at 17mm, but I would think that I could have a sharper image...

Missed focus. You are shooting with a Rebel after all. This is one of those rare instances where spray-and--pray can help. A little.

jasonZed said:

Panos are weird. I would never use one for lens diagnostics.

In short, you are pushing the T5 more than the Sigma.

Kelly

Bob · May 29, 2017

Some lenses are just not that great.

It could also be that each lens has a different sweet spot in terms of aperture setting.

Cropped Lens vs Full Frame Lens on APS-C Sensor

Member

Forum Pro

Forum Pro

Forum Pro

Forum Pro

Forum Pro

Veteran Member

Member

Senior Member

Member

Forum Pro

Forum Pro

Member

Member

Forum Pro

Forum Pro

Member

Forum Pro

Veteran Member

Keyboard shortcuts