Will I see any diffference if I fit a 24-70 f2.8L to my Canon 700D?

[ajohnw]

I allways wondered what that means. Optical there is a lens that creates an image and a sensor that registers that image. I don't mean the in-camera processing.

George

The main gain with some full frame cameras is noise and dynamic range on many of them at higher ISO's- especially the very top end models. There is a tendency for noise levels to increase as the pixel count goes up / size goes down so some of even the less expensive full frame cameras can offer an extra couple of stops for the same noise levels over crop cameras. In many ways this area is probably the main reason for buying one. Noise removal software improves eg the package I tend to use offers spatial control of the detail that isn't denoised away and a luminance curve to control what tone levels are denoised but many full frame cameras could still be used at higher iso's than the cameras I use.

Resolution and pixel densities is a strange subject - best voided but it's pretty clear that something like an 18mp Canon 1DX can produce very sharp images for all of the purposes that it's likely to be used for. You might say pixel counts go up because people like bigger numbers. A bit extremes maybe but as an example more recent bridge cameras are tending to use bigger sensors and less pixels. This really shows in some of the raw files these cameras produce and to some extent the iso settings that can realistically be used.

John
-

[george013]

John,
I just want to know what a lens makes it to be optimised for a certain camera. The mentioned camera's are all FF.

Quote Originally Posted by JR1 View Post
You need to remember that the top end lenses are optimised by Canon and Nikon for their top end cameras, on a D700 it will not be as good as on a D4s or D810 Canon similar camera.

George

[ajohnw]

I don't all together agree with the statement you have in quotes George.. The D810 in real terms is a high end consumer camera at around 1/2 the price of the D4s for instance. That doesn't stop pro's from buying them for the same reason as a Canon user might buy a 5D MkIII rather than a 1DX. Rumour has it, I'll never buy one to find out, that these very high end cameras are better in several respects as far as the cameras go but not the lenses for them. You might say they can make better use of them so optimised isn't really the right word.

What I find interesting about them is the pixel counts. Why are they on the low side? One reason is probably the use of higher ISO's. The other probably relates to the real resolution that can be achieved across the whole range of pro lenses. Those really do vary. The 24-105mm for instance isn't a particularly spectacular lens on full frame. One reason for that is that it's intended and priced to be sold with cameras at relatively low cost even though it has an L after it.

It is sort of optimised for full frame cameras as you can see from this - typical resolution on full frame 3200 lp/ph against 2,200 on crop. That's line pairs per picture height. The improvement is down to the size of the full frame sensor. The 5D II used for the tests is a 21mp sensor against the 1DX's 18mp. The chances are that the 5D will not capture any more detail than the 1DX even though it has more pixels. Same if the 5D III was used.

http://www.photozone.de/canon_eos_ff...5_4_5d?start=1

http://www.photozone.de/canon-eos/42..._4_50d?start=1

You can see from these though that as far as even performance and vignetting go that the lens is much better on a crop body. That is often the case but the results wont smack you in the face as I vaguely remember some one pointed out

This can't be a hard and fast rule as the lenses from all manufacturers vary and periodically they do update some - if sufficient number are sold so it's just a general trend. The much higher resolution figures from larger sensors is invariably true. This gives M 4/3 a bit of a problem so to part get round it the manufacturers have to produce better quality optics. They don't generally do this with aps crop lenses but do in some cases reach the same sort of level of performance from them as the full frame lenses have when fitted to crop bodies. It's a bit of a nightmare for buyers. However if some one intends to go full frame at some point it may make sense to buy FF lenses for a crop body.

As well as mentioning ISO aspects of fewer pixel full frame cameras I should also have mention higher resolution when the frame is filled with an image - in other words it will generally be possible to crop an aps view out of a full frame camera so the extra reach often mentioned on crop hasn't as much importance as people are generally inclined to think especially when they fit full frame lenses.

Like I said though these can only be general trends as lenses do vary but only within limits. I feel I have to repeat that. In capitols.

John
-

[rpcrowe]

I used a 24-70mm f/2.8L on a crop camera (at that time I was using a 40D) but, I also had a 30D which I usually carried with either a 70-200mm f/4L IS or a 12-24mm f/4 Tokina, depending on my projected needs. On a crop camera, the 24-70L was neither long enough at the 70mm side nor short enough at the 24mm side for me to use without a back-up lens. I did like the image quality but, did not like the heavy weight of the 24-70L "brick".

I replaced the 24-70L with a 17-55mm f/2.8 IS lens and use that lens in combination with the 70-200mm f/4L IS on a pair of 1.6x cameras. I use two Canon 7d DSLR's these days and it looks like I will be using this setup for a long-long while. It is just about perfect for me.

It may be a little on the heavy side and the extra body does cost money but, I look at my combination this way. If I were using a single 7D along with the 17-55mm f/2.8L and a 70-200mm f/2.8L, both the cost and the weight would be just about the same as my two camera and two lens set-up that I mentioned above. Additionally, I would have the PITA of switching lenses. BTW, I purchased both 7D cameras on sale as Canon refurbished models and paid well under $900 each for these bodies.

I get very satisfactory results from my 1.6x cameras. The 7D has an excellent auto focus system; better than most other crop cameras and most (non 1-D) full-frame cameras. I don't shoot video with my 7D and virtually never use my LCD for focus and framing. Therefore the lack of a articulating viewer doesn't bother me a whit! I shoot 95% of my images under ISO 400 so quantum ISO capability doesn't excite me...

I think that it is the ultimate in crop camera development. But there would be several tweaks to the 7D that would make it a better camera for my uses...

1. A/F capability at f/8 to enable me to use my 400mm f/5.6L lens with a 1.4x TC...
2. Two memory card slots (not a big thing but handy in some instances)...
3. ISO 50 or even ISO 25 capability to facilitate long exposures without the use of a ND filter...
4. Voice record capability to "take notes" regarding exposures...
5. Ability to communicate wirelessly with a computer so that I could shoot "tethered" without being physically wired to the computer...

I was sure that Canon would not incorporate these capabilities in the models following the 7D so I purchased my second 7D several years ago. I have never been sorry...

BTW: I love my 70-200mm f/4L on the crop camera and use it for possibly about 1/2 of my images (including all human and dog portraits).

[William W]

I’d still just like to know:

> to what lens(es) the OP will be comparing the 24 to 70 and the 24 to 105
> the reasons why the OP plans “to move up to” full frame.

WW

[BrianA61]

When I bought my first 'L' lens, I still had an old Canon EOS XTi with a 18-55 kit lens. It took OK pics but when I put the 24-105 lens on it and took my first shot, I immediately noticed a HUGE difference in clarity and sharpness. Yes, as mentioned, there are tradeoffs on kit lens vs the high end lens, BUT, IMO, the glass is where your best investment is. The camera body is just a matter of how many bells and whistles you want/need to capture the kind of photos you want. Spend as much as you can afford on good quality glass, then worry about the camera. I will stick with the L lenses and just suck it up when it comes to the inconvenience of such things as weight and cost. The end result is worth it, to me. Oh, and I've since upgraded to a FF (5D Mk2) to go with my L lenses. It's a heavy rig but takes some really good pics.

[ajohnw]

That's probably correct at the time you bought your 18-55 Brian. the period when kit lenses obtained a bad name. The equivalent one now is probably the ef-s 17-55 F2.8. Hardly any distortion and decent resolution. The ef-s 17-85mm offers similar performance on crop as the 24-105 does on full frame apart from full frames better resolution. In fact on crop the 17-85 tends to have higher central resolution.

What's really interesting is that the newer IS EF-S 18-55mm is pretty close to the 24-105mm on crop. Nikon's latest 18-55mm doesn't match up to it. Lots of fall of to the edges which is easily noticeable in some shots at full res - I made the mistake of buying one. Their 18-105mm is in some respects better.

Both of Canon's 24-70's are pretty good however my favourite test site reckons that the Tamron 24-70mm F2.8 is better in some respects, it holds up more at the longer end. Suggests that Canon need a MkIII now.

John
-

[george013]

I don't all together agree with the statement you have in quotes George..
John
-

I don't either. I just want to know the arguments.

What I find interesting about them is the pixel counts. Why are they on the low side? One reason is probably the use of higher ISO's. The other probably relates to the real resolution that can be achieved across the whole range of pro lenses. Those really do vary. The 24-105mm for instance isn't a particularly spectacular lens on full frame. One reason for that is that it's intended and priced to be sold with cameras at relatively low cost even though it has an L after it.

I think Nikon is making a studiocamera and a fieldcamera. In a studio you can regular the lights, mostly static subjects and a lot of time. In the field you have to do with existing light, so a higher ISO would be preferrable and time it takes to write the photo to the memorycard. And a senselsize which is half the size wil need a half shutterspeed to gain the same amount of movement.

I just wonder why there is a warning for the D800 to use only top-quality lenses due to the senselsize of the camera while there is no warning for the D7000, that has nearly the same senselsize.

To answer an often asked question, why going for a FF. To me it has been the better ISO of the D700. Coming from the D80, a very bad high ISO, via the D300, a bit better high ISO, to the D700, a for my purpose wonderful high ISO. And most of my shots are with a short focal length.

George

[Nicola]

Hi all,
collecting some test reports, I don't know if it's worth to spend a lot of money for 24-70 2.8 L II on crop sensors camera:
in addiction to some other issues, like a "not-standard" focal lenght (not wide angle and not tele) and lack of IS,
also the benefit in sharpness may not be present. See the comparison of sharpness between EF 24-70 f\2.8 L II and EF-s 17-55 f\2.8 on lenstip.com

Nicola

[ajohnw]

I wish that sites would standardise they way they report resolution Nicola. The trouble with lp/mm is that it doesn't tell people how much of that they can get on a sensor which will depend on it's size. I favour

http://www.photozone.de/all-tests

Going on lenses I have bought it's always given me a good idea of what to expect. There star rating at the end to summarise is very severe. Few zooms of any sort will get 4 for instance so I generally look at the analysis. Wish it had all lenses on the site but they are lenses some one has bought and taken to them. Many others get lenses supplied by the manufacturer.

As far as line pairs / mm are concerned 45 is reckoned to be a superb lens on crop and full frame but very few if any achieve it. On m 4/3 it seems to be reckoned to be about 55 and some do achieve it which is why in loose terms it can compete with crop cameras even with the smaller sensor but usually only when they are wide open. The reason for this is simple. The lenses are significantly smaller and optics has the problem that errors are scaled up as well when a lens is made larger so smaller lenses can be more precise but this doesn't make them cheap.

On cameras like the D800's and similar pixel density crop cameras it's rather difficult to get a clear idea if the extra pixels are actually doing anything useful. When an image is actually produce the information from the pixels is processed into a colour image and the software that is used for this has been improved over and over again over many years aimed at producing sharp images. Anti aliasing filters are disappearing from cameras as well. What these effectively do is "smear" images out at little so that aliasing can't occur. When they are removed the lens is doing this instead so in real terms the extra pixels can't be fully used otherwise aliasing could occur. On the other hand maybe they have come up with some software that part gets round this problem. The same effect might also cause manufacturers to limit the performance of their lenses on purpose. That's probably been happening on compacts and bridge cameras for years.

It is possible to look at the really raw data from a sensor but as it's black and white - just numbers with no colour it still doesn't tell the whole story. The easiest way to do it is to load a program called ImageJ and choose a documentation raw conversion. It does indicate how much light was captured by each pixel by producing a grey scale image that shows each pixel separately. In need of a life I compared 2 lenses some time ago using this method out of curiosity For even contrast levels the lower resolution lens won but the less even one could produce sharper normal images. I sold the more even contrast one.

John
-

[Nicola]

I wish that sites would standardise they way they report resolution Nicola. The trouble with lp/mm is that it doesn't tell people how much of that they can get on a sensor which will depend on it's size. I favour

http://www.photozone.de/all-tests

Going on lenses I have bought it's always given me a good idea of what to expect. There star rating at the end to summarise is very severe. Few zooms of any sort will get 4 for instance so I generally look at the analysis. Wish it had all lenses on the site but they are lenses some one has bought and taken to them. Many others get lenses supplied by the manufacturer.

As far as line pairs / mm are concerned 45 is reckoned to be a superb lens on crop and full frame but very few if any achieve it. On m 4/3 it seems to be reckoned to be about 55 and some do achieve it which is why in loose terms it can compete with crop cameras even with the smaller sensor but usually only when they are wide open. The reason for this is simple. The lenses are significantly smaller and optics has the problem that errors are scaled up as well when a lens is made larger so smaller lenses can be more precise but this doesn't make them cheap.

On cameras like the D800's and similar pixel density crop cameras it's rather difficult to get a clear idea if the extra pixels are actually doing anything useful. When an image is actually produce the information from the pixels is processed into a colour image and the software that is used for this has been improved over and over again over many years aimed at producing sharp images. Anti aliasing filters are disappearing from cameras as well. What these effectively do is "smear" images out at little so that aliasing can't occur. When they are removed the lens is doing this instead so in real terms the extra pixels can't be fully used otherwise aliasing could occur. On the other hand maybe they have come up with some software that part gets round this problem. The same effect might also cause manufacturers to limit the performance of their lenses on purpose. That's probably been happening on compacts and bridge cameras for years.

It is possible to look at the really raw data from a sensor but as it's black and white - just numbers with no colour it still doesn't tell the whole story. The easiest way to do it is to load a program called ImageJ and choose a documentation raw conversion. It does indicate how much light was captured by each pixel by producing a grey scale image that shows each pixel separately. In need of a life I compared 2 lenses some time ago using this method out of curiosity For even contrast levels the lower resolution lens won but the less even one could produce sharper normal images. I sold the more even contrast one.

John
-

John,
thanks for the feedback, I know photozone, and I usually compare results of several tests before buying a lens. In this case since the two lenses have been tested on the same sensor, LP\mm is good as a unit for the comparison. I agree with you that the sharpness is not the only parameter and the test numbers may not match the real life result on th customer camera.
Nevertheless, I think that usually, not only in photography, a very small improvement, even if actually measurable, may cost in term of money a lot, so must be kept in mind what is worth doing.
have a good day
nicola

[thegman]

Hi Bill, I have a canon 18-55mm kit lens and a tamron 70-300mm. The original post was a general question, not to compare my current gear. As I said, I have a friend that has given me some very good guidance, but I do know he does like like to buy the best gear no matter the cost.

[JR1]

John,
I just want to know what a lens makes it to be optimised for a certain camera. The mentioned camera's are all FF.


George

Some light reading

http://www.shutterbug.com/content/di...ake-difference

[William W]

. . . The original post was a general question, not to compare my current gear. As I said, I have a friend that has given me some very good guidance, but I do know he does like to buy the best gear no matter the cost.

Thanks for answering.
I did not realize that it was a general question: I thought you wanted a specific comparison to what you use now.

As a general answer here are a few points you might like to consider:
1. NO. An ‘L Series lens’ is not always the best gear for a particular task.

2a. In response to your friend’s comment ‘[my friend] says I'm wasting my money and that I should buy a full frame camera if I want to use this type of lens’: I would say that you are NOT wasting your money, but you should note that both the EF 24 to 70 and EF 24 to 105, were designed to be more generally useful/practical on a FF Camera: (i.e. in relation to the Zoom Range).

2b.There are quite a few Photographers who use a 24 to XXX on APS-C cameras. Some have a second zoom that is in the 10 to 24 range and they do not find that making a lens change-over at around 24mm being any inconvenience to them. I would. Changing a zoom lens at 24mm on APS-C is a real PITA for most of the Photography that I do

3a. In respect of your question about: ‘will I see any improvement in sharpness, contrast and colour’
That answer is: maybe, but sometimes, but not always. The kit 18 to 55 that you have, will likely be one of the latest iterations and (although you were asking a general question, obviously you still require a base for the comparison and your kit lens must be that base), so let’s take a few practical examples to explain.

3b. The ‘maybe’ part of my answer is in regard to your ability to note any differences. That is not a demeaning comment. It is often difficult to see any differences in two final images (i.e. IMAGE after POST PRODUCTION) when it that comparison is a double blind* A/B comparison.

3c. It is more likely that you will notice specific differences, when you begin converting raw files of images of a specific and similar nature, and or made under similar lighting conditions - for example if you were using the kit lens at F/5.6 at the 55mm end and you were taking Portraits in Low Level Available Light and you made the similar images with either of the 24 to XXX lenses and they also were set at F/5.6 and at about 55mm, then you would likely notice better contrast, better acutance especially at the edges and less Chromatic Aberration in the L Series Lenses.

3d. Also note that LACK OF Sharpness is often confused with Camera Blur. Note also that the TWO EF 24 to 70 F/2.8 L USM Lenses do NOT have Image Stabilization: and you might in some circumstances think these lenses are LESS sharp than the EF-S 18 to 55 (assumed that it does have IS).

3e. On the other hand, of you were using the Kit Lens at about F/8 and in the zoom range from 28mm to 40mm and if you were using Studio Flash, then it might be quite difficult to see much difference between the images that the kit lens made, compared to the images that the L Series Zooms made - you would need to look for CA at the edges; and acutance generally. I think you would be looking at 100% to begin to see any definitive issues – and if both images were POST PRODUCED carefully, there would be few eyes that could tell the difference in the final image in a double blind* A/B comparison, IMO.

*(note: ‘double blind’ is used here to mean that one would not know IF any one pair of images were actually making a comparison, between different lenses.)

4a. Consider that there are specific criteria to make L Series status. Note that one criterion for L Series status is that the lens must be able to mount to all cameras of that Series. The present series is “EOS” and therefore NO ‘EF-S Lens’ will ever be able to gain L Series status, no matter how good its optics might be.

4b.Whilst there might be excellent Lens's optics that made specifically to suit Canon APS-C Cameras; (for example EF-S 17 to 55 F/2.8 IS USM and EF-S 60 F/2.8 Macro USM) and whilst these lenses might be also the best suited lens for a particular task (the 17 to 55 a more suitable zoom compass and the 60 macro a more suitable FL macro on APS-C) and whilst both these EF-S lenses might render better results in various lab tests than some L Series Lenses: neither of the EF-S lenses will ever be able to gain and L Series Red Stripe.

5a. The status of L Series does not necessarily mean that all the characteristics of that particular lens which gains the L Series stripe are the best in that particular 'lens category' or that the lens itself is an improvement on all aspects of the previous lens, that it replaced.

5b. A recent example is one specific aspect of one of the lenses which you are considering. Comparing directly the EF 24 to 70 F/2.8L USM vs. EF 24 to 70 F/2.8 L MkII USM - both are labelled L Series. Obviously the MkII was released after the original version and it is considered to be Canon L Series quality and as such must meet the criterion to be ‘an improvement upon’ the lens which it replaced. But the original 24 to 70/2.8 has a more efficient lens hood by virtue of the zoom design of that lens, so in regard abating lens flare, the MkII is an inferior lens and a retrograde design release, but the MkII still has maintained an L Series badge.

***

As general ‘advice’, I think that you need to consider these elements and following these three steps:

a) LIST - WHAT you want to Photograph and HOW you want to make those photographs
b) LIST the LIMITATIONS of the gear that you have now, to do the above
c) LIST your IDEAL KIT (in today's gear and with respect to common-sense budget restraints) that you need to make the Photographs that you want to make in the way that you want to make them

Once you get a list of the LIMITATIONS that your gear has, then you can begin writing out your IDEAL KIT.

Once you have your IDEAL KIT written down, then you prioritize what is the MOST USEFUL bit of gear to get first and buy it.

As you have probably gleaned I do not read many LAB Tests and I think that your 'gereral question' goes way beyond and far deeper than 'sharpness, contrast and colour' comparisons.

WW

[george013]

Some light reading

http://www.shutterbug.com/content/di...ake-difference

To me it looks like a misty article.

A silicon sensor (and its protective glass cover) is far more reflective than photographic film.

I learned that the controlling of the flashlight in TTL-mode made use of the reflectance of the film and that the reflectance of a sensor/filter is to low to use. That's why the pre-flash was introduced to make use of the reflection of the shutter-curtain.

I think he is confusing the further "development of lenses" with "based on digital camera's", that are developing to. The article is from 2005.
George

[Manfred M]

Some light reading

http://www.shutterbug.com/content/di...ake-difference

Jeremy - I read the article and found that it was very much comparing film to digital cameras. The other issue is that the article is a good 10 years old (written in 2004) and while the physics have not changed, the ways of dealing with some of the issues have; as an example CCD sensors were still the "go to" design at the time, rather than the current sensors that are based on CMOS technology. A film emulsion, while certainly less reflective than a piece of glass, also reflected light and I do remember in those days enough comments regarding how reflections from the emulsion decreased image quality.

While I find some of the comments in the article a bit questionable, anything that is stated there would apply equally to a crop frame or full frame sensor camera, so your original premise that full-frame lenses are optimized for full-frame bodies does not hold water.