Facinating lens-to-lens image quality comparison tool.

[Scott Stephen]

http://www.the-digital-picture.com/R...ple-Crops.aspx

This link is for a really good lens-to-lens image quality comparison tool. I have not seen anything else quite like it. You can see and compare image quality of 2 lenses side-by-side from pretty much the entire lines of Nikon, Canon, Sigma, Tamron and ZEISS. Plus a few others. You can even see each lens at multiple aperatures and at various zoom lengths.
Shots are of test pattern charts, so you can easily see blur, distortion, chromatic abberations, center sharpness, edge softness, etc..
They include pretty much the entire line of Nikon, Canon, Sigma, Tamron and Zeiss lenses. You can compare various Nikon lenses to each other, or compare Nikon to Canon or Zeiss to Sigma or whatever.

You can even compare the longer lenses with teleconverters mounted, though it is not readily apparent how. Just push the lens past its maximum native focal length and you get to see it with a teleconverter attached.

I see a lot of lens questions posted. I hope this is helpful to someone. Cheers!

[John Morton]

Well that's about the most interesting thing I've seen this week - thanks for posting this link!

[Scott Stephen]

Great! You are more than welcome! Do be careful though. If you have a tendency to get wrapped up in this stuff, this can transfix you for long periods!

[terrib]

Thank you for posting this. I've bookmarked it for future reference. I thought it was interesting also, besides comparing different lenses, to be able to compare different apertures of the same lens to see how choosing "the sweet spot" can affect your final image rather than choosing an aperture at random when you have a choice.

[Kris V]

Thanks for posting this very informative link!

[ajohnw]

It's interesting to compare canon's long zooms against primes, I've always thought that the zooms were good. Thanks for posting.

I wish it offered other than 1D cameras to see the effects of diffraction with aperture against pixel size. The 1D range has rather large pixels.

I sometimes test my own. Mainly curiosity. Old lenses against new. I generate this for a lens

http://www.flickr.com/photos/4899461...57630787239818

Shutter shake blur vertically but about 1.5 pixels otherwise. ImageJ is the easiest thing to use to obtain an undemosaiced raw and blow it up. The crosses in the shot mark pixel boundaries. There is a shadow across the black GB and the splotchy black in the letters is down to light scatter in the lens. The full shot is in the same set as well. The test might interest other masochists. It's probably best not to know.

John

[The Blue Boy]

That's a cracking little find Scott, thanks for sticking it up.

[DocE]

Hi,
I am new to photography and this site, having just retired. I post as DocE on NatureShare.com and Nikonians.com, and have posted 180 or so bird photos on NatureShare. I am building a quiver of equipment for my primary interest in wildlife photography with an emphasis on birds. I have been struggling with which lenses/cameras to acquire and have no commercial bias or interest. I am an amateur totally. I tried to set up my own test system at home for testing back focus and image clarity, and found it challenging and difficult. I used the ISO 12233 chart suggested by one Nikonian which seems to be a good standard for comparison of image quality. I found the digital picture site and was delighted to be able to compare lenses side by side for image quality and other parameters.

I wonder if others have used this site and find the comparisons transfer to use of lenses/cameras in the field. Their lens comparisons are done in manual focus mode (to take back focus issue out of the comparison), yet I doubt many use these lens in this manner. So what does it mean? Should these tests be used to make lens purchase decisions?

If anyone has found another site that does these kind of comparisons I would love the link.

Thanks

[ajohnw]

I use several sites and do find that lenses follow the results shown. There is always the chance of a poorly constructed lens cropping up though.

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

http://photographylife.com/

http://www.photographyblog.com/

http://www.lenstip.com/lenses_reviews.html

http://www.cameralabs.com/

http://www.ephotozine.com/

A mixed lot but the 1st is best. I don't think resolution crops tell the whole story. I recently picked a lens on the basis of even resolution putting up with more than ideal levels of chromatic aberration. Lenses are swings and round abouts. Often there has to be a compromise some where.

For checking focus I find this best

Anybody tried this? AF / Focus checking

It's very sensitive. The image I posted was taken via a 16mm F2.8 180 degree field of view on full frame cameras lens that has a massive depth of field. I use it on APS.

John
-

[Manfred M]

Unless you are shooting a very high resolution camera from a heavy-duty (rigid), sandbagged tripod at base ISO you are never going to get anywhere near the theoretical resolution out of the lens. In my view, too many people spend far too much time worrying about this one single performance characteristic when they spend most of their time handholding shots and making shooting decisions that are based on other factors; shutter speed, DoF etc.

If you are looking at pure resolution, stick to primes as these (generally) offer superior theoretical performance to zooms, and shoot in their "sweet spot" when selecting your working aperture.

I don't think that I've ever looked at resolution charts when buying a lens; and I have quite a few ranging from cheap kit lenses and those turned out buy lower cost third party lens makers, through to expensive standard pro glass and I even have a couple of highly specialized, special purpose lenses.

I guess I spend my time taking pictures rather than worrying about lens test results. In my view, these test sites often test things that are easy to test rather than things that are meaningful to the photographer...

[NorthernFocus]

...I wonder if others have used this site and find the comparisons transfer to use of lenses/cameras in the field. Their lens comparisons are done in manual focus mode (to take back focus issue out of the comparison), yet I doubt many use these lens in this manner. So what does it mean? Should these tests be used to make lens purchase decisions?...

DocE, where did you find the explanation as to how they do the testing on the digitalpicture.com site? I would be interested in reading it. Before using any of this type of information I like to understand the basis for it. Photozone as recommended by John is another good site.

Regarding your question, if they do use manual focus the testing is still valid. When comparing lenses what one should be interested in is the image quality that the lens is capable of. Front/back focus issues with camera AF are camera issues, not lens issues. Modern DSLRs have tuning functions in them to correct AF accuracy issues. It is up to us as photographers to understand our gear and to ensure that it works properly/optimally.

By the way, adding a name that you prefer to be addressed as is helpful so we can sound friendly when addressing one another when posting here on CIC.

[ajohnw]

Unless you are shooting a very high resolution camera from a heavy-duty (rigid), sandbagged tripod at base ISO you are never going to get anywhere near the theoretical resolution out of the lens. In my view, too many people spend far too much time worrying about this one single performance characteristic when they spend most of their time handholding shots and making shooting decisions that are based on other factors; shutter speed, DoF etc.

If you are looking at pure resolution, stick to primes as these (generally) offer superior theoretical performance to zooms, and shoot in their "sweet spot" when selecting your working aperture.

I don't think that I've ever looked at resolution charts when buying a lens; and I have quite a few ranging from cheap kit lenses and those turned out buy lower cost third party lens makers, through to expensive standard pro glass and I even have a couple of highly specialized, special purpose lenses.

I guess I spend my time taking pictures rather than worrying about lens test results. In my view, these test sites often test things that are easy to test rather than things that are meaningful to the photographer...

I can't all together agree with that Manfred but I do accept that taking the results given by the dot de site I linked to that these need not relate to actual resolution in the camera. In fact the test uses a mathematical extrapolation of something that is only really accurate if the lens was optically perfect which in real terms they can't be. They do give an indication of resolution though but it wont beat actually measuring it.

Lenses do vary and in the past I have owned lenses with limited resolution and that mainly shows up as poor contrast. The tests also show how much vignetting I can expect, probable best aperture, likely level of chromatic problems and distortion. As I mentioned I do tend to go for even resolution across the frame as drop off in the corners can be seen.

Tripods aren't my thing. I shoot hand held and have taken lots of pixel sharp shots. I tend to use lighter gear which helps. Basically I rely on IS and and use some sense on selecting exposure times. In fact the only shake problem I have had was on a tripod - testing a telephoto that didn't have a separate mount to balance it up so the camera had to be clamped down. Pixel density on that camera are higher than your D800. The only tripods I own are relatively light. In fact the one I have used most and am prepared to carry around is really light. On top of that I have a manfrotto 190B and a Velbron CX mini.

If I find I need more support with say the none IS Nikon 300mm and the D7000 out will come the monopod. Probably if I use it on M 4/3 as well. Which monopod though. I have an early manfrotto, a carbon manfrotto which in real terms is no lighter or usefully stiffer than the old one I wonder why I bought it actually and also a Jessops one with some years on it. I have an Olivon pan and tilt 2 way head. Super smooth and easy to guide but weighs nearly the same as the monopods. The Jessops thing comes with a built in tilt head that is light and serviceable. I suspect I know which one I will use.

Shake. I have posted hand held shots of the moon using 300mm and 2x digital convert on m 4/3. Also of a Robin 300mm. I fact the only tripod shots I have posted were 10 sec star shots taken on the light weight with a 50mm lens. It could be argued on exposures that long shake doesn't really matter if it's a well damped tripod. Also hand held D7000 plus 300mm but the exposure times were unrealistically short. i also have 2x and 1.4x converters for that should prove interesting. I believe in trying rather than assuming I need 1/2 cwt of sand bags and a tripod that won't flex when that is hung on it. Basically do something about a problem when the fact that it is a problem is known.

Actually I sometimes wonder if a shooting stick might help. There are a couple of near antique sit on types kicking about but the other types with a V at the top could also help. I also know some one who carries his camera plus a rather large lens on a home made wooden pole.

I should add that I am not the most stable of people especially if a strain to keep things steady. it's best not to think about it. Raise camera focus and shoot.

John
-

[DanK]

Thanks for posting this.

I don't go quite as far as Manfred. I do think this site can be useful, and I have often used it myself. I find it particularly useful for looking at the aspects that are most likely to be a problem, e.g., corners and shots wide open.

However, I agree that under many conditions, small differences in lens quality have no practical impact. For example, if you are going to display on the web and are not going to crop dramatically, you will be displaying at such a low resolution that lens resolution will rarely matter. On the other hand, if you are printing large or cropping severely, it may matter. I have never done anything systematic with this, but I suspect that in many cases, other aspects of lens quality may matter more.

I wish it offered other than 1D cameras to see the effects of diffraction with aperture against pixel size. The 1D range has rather large pixels.

They do. It depends on the lens. In many cases, they offer two bodies.

[Black Pearl]

I have a few issues with test sites like the one posted - though I will break my advice and go have a play.

1* Some people will search for their lenses, hunt for the optimum setting then stick with them regardless of the subject or scene in a blind belief that it will make their pictures better. They won't shoot wide open as it isn't quite as sharp or they won't stop right down because of diffraction when doing just that might make their photograph a million times better in an overall way.

2* Many will find that (exactly as Manfred said) they can never achieve the test results in real world shooting (due to poor technique) but will then spend incredible amounts of time and energy searching for this mythical Shangri-La without ever looking within.

3* In real world shooting, with very good technique, sympathetic processing and correct sharpening the vast majority of minute differences are cancelled out and (within reason) it doesn't matter how many lines per millimetre your combination can record.


...now them I'm off to see how good my glass is

[Geoff F]

That site sounds like a nice idea but it just locks up my computer everytime I try to select anything.

[ajohnw]

If the truth was known most people look at review sites. My view is simple there are 2 basic types. This is a great camera or lens and others that demonstrate just how good it is. Shades of in between as well but most are pretty polarised one way or the other.

I just don't see the point in this is great sites as they serve no useful function at all other than to tell people that this and that equipment is available.

Expecting lenses to meet resolution tests is basically a bit daft. Do we all shoot black and white? All pixels record black and white. Colour takes 4 and interpolation of one sort or another. All the figures do is give people an indication. As to black and white have they been taken from raw or from camera jpg's. In my view the lens line type resolution tests in the initial post are the worst of the lot even though they are in some ways the most accurate - in black and white.

The list I posted is mixed. One for instance can be good place to go for full sized images from bridge cameras along with raw's to develop Another measures a cameras colour accuracy when they test bodies. Some present the info in different ways. It's a case of if I fancy a particular lens I do my best to find some useful information about it.

I missed out dpreview who test some lenses as well. i have found some of their tests a little curious.

On the whole as far as lenses go people get what they pay for but there are variations at any price level and unfortunately exceptions as well.

John
-

[Manfred M]

John - my real issue with all of these sites is that they go out of their way to praise the lenses and perhaps in the lab environments, where they were used for one day, they indeed perform well for this short period..

The one thing that I have noticed is that most modern lenses are optically very good, given either the range the work under (I'm thinking of some of the ludicrous ranges; 18-300mm, etc.) or super high speed (some of the fantastic f 1.4 primes or the f/2.8 pro zooms). On the other hand, most of these are not built to anywhere near the mechanical standards and precision of those produced just a generation or two ago (Zeiss and Leica products are really the main exception).

I don't have a problem with modern engineered resins, that are used on external parts as these take banging a lot better than metal (metals dent, resins rebound), but on the other hand they don't wear as well. The rubberized grips really don't wear all that well, and frankly plastic filter threads are always a risk for cross-threading when an aluminum or brass filter is screwed in. Plastic lens mounts, plastic components on wear parts, etc, hardly ever get mentioned.

I don't like seeing part lines on moulded parts; especially on lenses that run into the four figures cost wise; These should be cleaned up (it can be done for pennies).

I have some lenses that I'm sure will last 20 or 30 years, but I would be suprised if some of the modern pro (expensive) lenses will last even half that long.. I've seen how well the internal autofocus motors are made, and I'm frankly unimpressed...

[Adrian]

An interesting debate. My own amateur perspective is that using charts like this is an excellent way of comparing lenses....if you like taking pictures of charts. If however you like to go outside the lab and take pictures of other stuff, these comparisons become insignificant as long as you are using quality equipment in the first place and adopt good technique.

Frequenting internet sites about photography tends to indicate that we photographers can get a bit overly obsessed about the tools.

[ajohnw]

I don't get obsessed about tests just use them to pick lenses. As a for instance I recently bought a DX 18-105mm. I looked at all of Nikon's 18 to this and that and concluded it was the right compromise in the right areas so bought one. I could have gone 18 to something longer but from what I could see it will crop to something reasonably longer if needed. It's not a brilliant lens and neither are the other similar lenses that go even longer. Could have bought the 16-85mm but is it significantly better, no not really. Just probably costs more and I have something wider than 16mm anyway.

I think the jury is out on how long current AF will last. Never been into it but suspect the actual drive has no moving parts and it actually tightens up the fit of one lens I have. No moving parts is generally a good idea. Faults I here about are generally wiring related. Plastics entered fishing tackle a long time ago. Carbon fibre in them and actually it's the metal bits that cause problems. I'm used to that sort of change. Who knows what they put in lens plastics. I don't. but one thing for sure the fact that something is made of plastic doesn't mean it will wear out any quicker. Parts are lighter so loads are lower. Out of interest I took a metal manual prime telephoto apart that was kicking about and no longer of any use to me.and guess what everything ran on plastic including threading for focusing. Also high tech grease. Before that they used metal on metal and crap grease that gums up in no time. The lens with plastic dates back to the 70's. I've also dismantled a metal on metal lens date maybe pre wwII.

Can't say I am keen on Nikons rubber focus grips. People do have them work loose. Suggests junk materials or even incomplete curing. Others use soft grip plastics or mould in as per Olympus.

I would also suggest that the longest lasting mount would be the right plastic into metal. Plastic compresses so if it doesn't flex due to being overloaded no problem really. Plastic cameras usually have a metal chassis carrying the sensor and lens mount. Again no problem and probably easier to water proof too.

I suppose I have a different view about "advances" as I am a design engineer and did a lot of work with plastics early in my career. I don't see any real problems with it providing the job is done properly. Sometimes it isn't.

Anyway now some might see that I don't look at test results the way some might think I do. I do like to look at 100% res shots though and if I see fuzz at a pixel level I am likely to sell the lens. Maybe not the 18-105mm though which may well be ok anyway. A lot depends on what the lens is to be used for.

John
-

[Brev00]

I like reading lens testing website results but take them with a grain of salt. They appear 'scientific' yet one does not have to go far to find disqualifying aspects. With this particular website, different cameras are used to reach the results posted. Once variables outside of the actual lenses are left uncontrolled, the results are interesting but not what I would consider valid. Plus, who knows what sort of lens quality variability is affecting the results. The results they got with the Tamron 70-300 vc are nowhere near as sharp and contrasty as I get. Do they af fine tune and if so do they do that equally well with all lenses? I don't know. So, it is nice if their results back up mine but,in the end, the beauty of digital photography is that we are all our own lens testers.