IQ of Zooms vs Primes

[x-tc]

This might surprise some people:

 

Looks like on par zoom to prime IQ is no longer reserved to Fuji lens.  In fact, Sigma does one better on Nikon's own turf. 

[citral]

I have yet to see any review or test claiming the IQ of any fuji zoom is "on par" with a prime, except for the 18-55mm being better than the 18mm at every aperture (I think the 10-24 is better too, but at f/4 vs f/2 is not really the same whereas f/2.8 vs f/2 is really close) BUT having more barrel distortion which can be undesirable.

 

Can you please prove your point before making such an affirmation?

 

This could be misleading for people who are unsure what to chose. Thank you.

[citral]

By the way I've looked at the review, and if I understand correctly they are comparing FF lenses mounted on a cropped body, to a zoom specifically designed for crop bodies, and they say look, the corners of the zoom are sharper.

 

What did they expect?

[aceflibble]

It's more a case of glass being made for crop sensors vs glass being made for 35mm sensor. That Sigma 18-35mm is a helluva lens because it's made specifically for crop sensors. When you take a 'full frame' lens and put it on a crop sensor, every slight flaw becomes magnified. Same as if you go from a lower pixel count sensor to a much higher one.

For example, when Canon made the 24-105 L lens as the 'kit' lens for the 5D, that lens seemed great. On a 12mp full frame sensor, everything looked perfect. Below-average distortion for such a versatile zoom range and pin-sharp with no aberration.
Then the 5D Mark II came out with a 21mp sensor and the lens started to fall apart. Then people tried it on the 18mp crop sensor 7D and the lens looked even worse. The pixel pitch got so tight that every flaw became magnified by over four times. Huge colour fringing everywhere. Soft right across the frame. Out of focus areas rife with onion-ring artefacts. Now it's notorious as being one of Canon's worst lenses.

The Sigma 18-35mm f/1.8 is an incredible lens, there is no doubt about that. Put it against any full frame lens on the same crop body and it will win. I've come very close to buying it on many occasions, I've rented it for a couple of jobs and I would rate it as, for the price, the best lens any Canon or Nikon owner can buy for a crop body. But in the sharpness stakes it does get beaten by some other lenses made specifically for crop sensors.

And, of course, we can't get it on Fuji bodies anyway 

[Patrick FR]

some say the XF16-55 has Prime Quality... or at least gets pretty damn close to it: http://www.soundimageplus.com/soundimageplus/2015/3/3/fuji-xf-16-55mm-f28-r-lm-wr-fujinon-zoom-lens-review-as-good-as-a-prime-comparison-with-23mm-f14?rq=16-55

 

Soundimageplus compared it also to other primes. The conclusion is always the same: unbelievable IQ for a zoom lens.

 

I've never tested the 16-55 personally (very happy with the smaller 18-55 with OIS). But looking at the reviews, I'd say that you need a very, very good prime to beat the IQ of the 16-55.

[x-tc]

I agree that higher MP camera can exceed the resultion of a lens. However, the fact that the sigma zoom was design for crop is providing no advantage to the sigma . It's simply better glass than the primes. If there was any advantage here it would have been to the primes for using the mid part of the lens instead of the true corner. Again, it's just better glass. Anyway we spin it. Cropped center design is irrelevant as an advantage. The zoom simply resolves better everywhere on the glass compared to the primes. There is no need to make excuses for the primes.

 

If we want to say the zoom was designed for higher MP than the primes, perhaps. Keep in mind all these primes are used in 36MP FF nikons and the 35mm ff prime was released not too long ago.

[citral]

So in short let's just spin facts around until they fit our own biased opinion?

It's strange because I read only "almost", "quite" and "nearly" in the reviews. So let's not spin this : if absolute best possible IQ is what one wants, the primes still beat the zooms on sharpness, contrast, bokeh. Not to mention the IQ of the 10-24 at f/1.4 and f/2.8 is unknown.

Zooms have come a long way and are close, but not "on par".

This is ridiculous. Hoh hehhh I love so much my big zoom am just gonna declare it as good as primes and ignore everything including basic laws of physics and lens design lalala.

[x-tc]

Patrick,

 

Similar story with the XF 10-24 vs primes. Too close to call for 14, better than the 18. Got to hand it to Fuji and sigma. Zooms have come a long way.

http://www.fujivsfuji.com/10-24mm-f4-vs-primes/

[aceflibble]

So let's not spin this : if absolute best possible IQ is what one wants, the primes still beat the zooms on sharpness, contrast, bokeh.

Bokeh quality is, by its very definition, entirely subjective, as is contrast. Sharpness isn't, which is why I typically conclude with a comment on sharpness alone. I think it's important that everybody tries to remember that image quality is, mostly, down to interpretation.

Back in the day (and still, in some circles) you had loads of people arguing about Fuji film vs Kodak film. You'd often talk to some wedding photographer who would swear their Kodak film was categorically, technically better than a Fuji film they had tried, when really the Fuji film they had selected simply wasn't as suitable for that particular task as the Kodak film they were used to. Or maybe they just preferred the warmer colours of the Kodak film, or maybe they liked that there was more grain, or less grain, or whatever.

Point is, with any piece of equipment, everybody is going to have different definitions of what is good or bad. Sharpness—plain resolving power—really is the only thing you can actually compare, and even that can be confused by microcontrast. And of course, some people will complain that a lens is "too" sharp!

 

 

 

However, the fact that the sigma zoom was design for crop is providing no advantage to the sigma . It's simply better glass than the primes. If there was any advantage here it would have been to the primes for using the mid part of the lens instead of the true corner. Again, it's just better glass. Anyway we spin it. Cropped center design is irrelevant as an advantage. The zoom simply resolves better everywhere on the glass compared to the primes. There is no need to make excuses for the primes.

This is incorrect and a common mistake. You need to understand that pixel pitch and resolution are two different things. When you cram, for sake of argument, 18mp into a crop sensor, the pixel pitch is much higher than if you make an 18mp 35mm sensor. This greatly magnifies the qualities of the lens. You'll get more detail out of a lens which resolves well, but at the same time, any tiny flaws in that lens will become far more obvious.

 

Now, if you're making a lens specifically for crop sensors, you design that lens with this tighter pixel pitch in mind You don't just take a full frame lens and cut the corners off. You know you don't need to use as much glass, so you can afford to make the glass you do use as good as possible. You generally don't worry about achiving the highest resolving power, because the cropped 'magnification' is kind of zooming the lens in closer for you anyway, and instead you just make sure that every possible flaw is as well-controlled as you can make it. A moderately sharp lens with no aberration looks like it resolves far more detail with better microcontrast than a higher-resolving lens with fringing all over the place.

Conversely, if you're making a lens for 35mm sensors—as most primes and premium zooms are—you know you're dealing with a much more relaxed pixel pitch. You spend less time worrying about controlling flaws and more time on the overall quality. You're also trying to stretch the quality over a larger piece of glass.

Higher-resolution full frame sensors are bridging the gap a little, because they obviously get more out of a lens—both good and bad—than older 35mm sensors do, but they still don't have quite such a tight pixel pitch was the higher-reoslution crop sensors. So, to use my earlier example again, the 5D mark II (21mp 35mm) shows up some flaws, but it's something like the 760D (24mp APS-C) which will really magnify and exaggerate the clarity, or lack thereof.

 

Now, all that said, the fact of the matter is that very, very few companies bother to actually make really good lenses for crop sensors. The thought is that crop sensors go in cheap, low-end cameras and the people buying them won't want to invest in higher-quality and more expensive lenses. This is really the main reason that Sigma 18-35 zoom stands out so much. It's not that zooms overall are as good as or better than primes, it's more just that Sigma actually bothered to put some effort into making a premium crop sensor lens and that's a bit of a novelty.

 

This is where it becomes a very tricky subject to talk about with Fuji users, because Fuji are more or less the only company out there who are making multiple high-quality lenses specific to crop sensors. Look at the 60mm, for example, and the resolving power that has. Look at how sharp that is. It's a tiny piece of glass, 39mm at the filter thread, it's dirt cheap and it's a semi-macro, high-resolving portrait lens which is "too sharp" for most portaits! To get the same results from a full frame lens from another manufacturer, you need to be looking at the top-of-the-line stuff; I used to match it with the Canon 100mm f/2.8L Macro and that costs over twice what the Fuji 60mm does. Fuji's 18-55 zoom is actually sharper at the 18mm end than the 18mm prime is, but that's really because Fuji went out of their way to provide a 'kit' zoom which far exceeds the standards people expect from such lenses.

 

So this is really the wrong place and the wrong crowd to be comparing zooms vs primes. As Fuji users, we're used to zooms which are sharper than primes and cheap, light and small primes on crop sensors which are as sharp and as faultless as the very top red-band and gold-band lenses on the largest full frame sensors. We're spoilt because Fuji have designed all of their lenses to be as technically good as possible when paired with a crop sensor. There aren't any Fuji lenses which have soft haze, two stops of vignetting and six pixels of aberration in the middle of the frame. Us Fuji users are not the people who should be worrying about which type of lens is best. Fuji users don't need to worry about that. I used to use Canon and Mamiya and I used to have to weigh up primes vs zooms, quality vs flexibility, and I mainted a separate kit of lenses for my crop bodies (APS-C and 6x4.5 digital) as I did for my full frame bodies (35mm and 6x7 film). Then I ditched it all and picked up Fuji because I can buy whichever lenses and I know they're all equally good.

 

Maybe we'll come back to this subject when the X-Pro2 is out, if it indeed does have a 24mp sensor, or if that rumour of a 1.3x crop sensor Fuji is true. That will be the time to start measuring the primes vs the zooms. Even then, you're not going to see that much of a difference, I suspect. For now, just rest assure that no matter which lenses you buy for your Fuji, they're going to be fine.

[x-tc]

aceflibble,

 

I agree with everything you said in your prior post. except this part:  

This is incorrect and a common mistake. You need to understand that pixel pitch and resolution are two different things. When you cram, for sake of argument, 18mp into a crop sensor, the pixel pitch is much higher than if you make an 18mp 35mm sensor. This greatly magnifies the qualities of the lens. You'll get more detail out of a lens which resolves well, but at the same time, any tiny flaws in that lens will become far more obvious.

 

this was already taken into consideration. no mistake was made.   If it helps to simplify things if we are just talk about the IQ over the image circle of the cropped region on a FF sensor.  For example, we can mount both of these lens onto FF 36MP camera or 50MP or 1000 MP, whatever we like.   In both cases we have the same distance from exit pupil to sensor.  There should be no other design considerations here between full frame and cropped sensor other than making sure the image circle covers the required area of the seniors.

 

Having said that,  the image quality from the center of the frame out to the end of the image circle to cover the cropped area will be better with the Sigma zoom than the Nikkor primes in the range compared in the video.   At no point will the IQ be better with those Nikkor primes within the cropped area. This means the Sigma zoom's glass is better in this region from center all the way out to 2/3 from the center.   It pretty straight forward..  The Sigma zoom has better glass over this portion of the sensor, pixel density doesn't change the quality of glass.  Just the sample resolution.  We are aligned.

 

Let me know if you still think a mistake was made in the post that you responded to. Go back and reread it please.  Not sure why you thought something different.  We are in agreement.   

 

 

As for the rest, before this section in your reply and and everything after.  Yes. yes.. agreed.  We are aligned and glad we agree about the IQ of  Fuji zoom vs Fuji primes.  This is the very reason that I am mostly buying zooms on the Fuji platform (vs relatively heavy zooms on FF) and keeping fast primes for my FF Nikon for shallow DoF needs.

 

 

btw, funny to see you write some lens are too sharp. So true.   I have a FF Tamron SP 90mm/2.5 Macro and a FF Nikkor 85/1.8G.   Set them both to F2.5 and take a headshot. One will quickly discover that few female models will be happy with the result with the FF Macro. Just too darn sharp compared to the FF 85/1.8G.    So it's not just reserved for Fuji cropped sensors lens.

[citral]

Bokeh quality is, by its very definition, entirely subjective, as is contrast. Sharpness isn't, which is why I typically conclude with a comment on sharpness alone. I think it's important that everybody tries to remember that image quality is, mostly, down to interpretation.

Hmmm. Well I made a real life test between the 18-55 and the 35 both at f4 that got promptly ignored/discarded by x-tc.

 

Which I'm still not sure if he is a troll or not, because you seem to have some obvious good knowledge of how this works and he's discarding your comments as well, because they don't seem to concur with the opinion he has on his purchase, intended, as his bio describes, for recreational vacations souvenirs. So why would this even matter for this purpose is beyond me. You can't just beat the weight and versatilty of zooms for this use, but there is no need to claim they produce as good a picture as the primes, that would be kind of ridiculous for people that make art and/or big prints with their camera.

 

So I will repost the left top corners of this test shot I made of both and will let you decide which is which. After that we can still conclude there is no bad lens in the fuji range, it's a no-brainer. My 18-55 isn't going anywhere.

 

 

[milandro]

Apparently each era has a need for unresolvable questions , some of capital importance and some perhaps of a little more than a petty nature. 

 

Take Angels for example! For at least 1800 years, people have been debating all sorts of things about them, one of which was their gender which seemed to be a matter of some importance to many and which fueled many vehement discussion and was object of councils of learned men and women too.

 

 

All of this must have seemed to be very enticing, at least to the participants of such a debate judging by the efforts which they placed into it.

 

So, in this era of secularization of ours, it is pretty possible that such a learned but religious debate would gather less debaters around and that the deeply rooted need for a “ pilpul” of some sort could get fulfilled by a question such as: “ how many lines can a 10-24mm (or any other zoom lens of your choice)  resolve at any interval of its zooming range and aperture correspondent to all the “ prime” lenses available? "

 

I have no doubts of the merits of this analysis.

 

However the way I have solved it is a mere practical one.

 

Until now ( with the sensors being what they are and the lenses being optimized for those) Fuji zooms performance has always been rather good, equal or close to equal to the performance of fixed focal length lenses.

 

If one needs more that one of the fixed focal length lenses contained within the range of any given zoom, you spend more money and carry more weight than the zoom itself ( and this is even true of the so called gigantic 10-24mm) by acquiring two or more lenses contained in the range.

 

It might very well be the case that printing a photograph to the maximum of its achievable size I might notice an enhanced performance of a prime over the one of a zoom.

 

But I don’t print all that many of my photographs, let alone print them so big!

 

Should I ever need to do that, tough luck for me and any other zoom owner then! I suppose I will have to learn to live with the shame of such limitation.

 

Or, alternatively, not care too much about this extreme and improbable limits and just do my own thing, which I plan to do for a long time.

[citral]

I could not have said it any better.

[aceflibble]

x-tc: I wrote my second post at 1:20am, following a regrettable night of Eurovision drinking games, so yes, re-reading now, it does seem we're mostly agreed and there's just been some poor phrasing confusing things.

citral: See, I would guess the first image is a zoom and the second is a prime, but that's more due to the contrast. In terms of bokeh, to me they might as well be the same. Neither's smoother or rougher than the other in that regard. The second image does look sharper to me, but again, that looks like it's more due to contrast. I'm more of a prime user myself, but as you say, if you're looking for versatility and to keep the weight down, zooms—especially the Fuji zooms—do just as well at similar focal lengths and apertures. With other brands I do often seen a good argument to be made in favour of zooms or primes, but with the Fuji stuff... eh, it's all good. Give me the 56mm, the 60mm, the 90mm, the 18-55 or the 50-140 and I know I'll be able to get a good portrait out of any of them. (Maybe the 60 would need a weak diffusion filter, though!) I only really choose primes myself because it's just what I grew up using and when I use zooms I keep forgetting they can, er, zoom!

[citral]

To me the 18-55mm has a much more "nervous" out-of focus rendering, and that often distracts from the subject even when viewed at normal size.

 

Also, as it's on full manual the 18-55 is darker (due to the length of the zoom at 35mm the light has to travel longer to reach the sensor I suppose, tho I'm no expert so it might not be the decisive factor. In the Fuji vs Fuji series I think the reviewer is on auto speed so the camera makes up for the difference and they appear closer).

 

I'm not debating over personal preference, convenience etc. and there are many more factors to take in account than pure IQ for zoom vs prime.

 

What I maintain is that zooms still don't beat primes (and can't by design, unless one decides to ignore the rules of physics and optics) regarding pure IQ quality. That is just misleading and a test of a zoom designed for crop vs FF primes on crop don't prove anything at all.

 

Otherwise we'd have all Fuji studio portrait shooters use zooms.

 

End of the story.

[aceflibble]

Well, yeah, I think we're all agreed there. Sort of. The best zoom in the world categorically can not be as sharp as the best prime in the world, all other things being equal. And a lens made and for and being used with a specific sensor is going to do a better job than a lens designed for a different type of sensor than the one it is being used with.
But I still think Fuji shooters are in a weird position because the image quality from the current Fuji zooms is so close to the primes. (Or better, in the case of the 18-55mm and the 18mm prime.) So us Fuji users probably are not the best people to try to work out which is better! It's a much bigger issue on other systems where there is a wider range of lenses with more variation in quality. Like, that Sigma lens does out-resolve those Nikon primes, but there are a bunch of other Nikon primes and Canon primes which do beat the Sigma zoom. So you've really got to pay attention to exactly which model of lens you're talking about and how they're being used. Lab tests vs reality, different price ranges, different uses. There's a lot more going on than simply "Sigma made a nice zoom lens so I guess primes are dead."

[citral]

Yes. And as usual those last 2% improvement on performance cost a lot, like in many other topics.

 

I'm into ultralight backpacking and a 600g titanium pot sells for 60€ when the same volume in inox weighing 680g sells for 7€.

 

It may sound dumb to grab the titanium one, still it is worth it for people that either have enough income not to care, or for people who do the Appalachian trail. On the very long run it is the better investment because that litlle % you gain matters eventually. YMMV.

[x-tc]

Cheers aceflibble,

 

It seems that after almost 2 dozen posts we are back in agreement with my original statement. :-)

"Looks like on par zoom to prime IQ is no longer reserved to Fuji lens. In fact, Sigma does one better on Nikon's own turf. "

[citral]

I just can't cope with bad faith and drunk posting is a terrible idea. Sorry for these personal attacks, I find it irritating that you try to make people say what they have not quite said, instead of bringing some material to the discussion to prove your affirmation.

 

Put some effort into it, otherwise it's just pointless.

[x-tc]

There is a reason I ignore you. Try to figure out why that is.  Your grief is self inflicted due to your negativity and your disrespect for others which you feel are inferior;  qualities you are all too eager to demonstrate to everyone on this forum and I assume in your daily personal life also.  

 

Consider what you say and how others view it, impacts how they respond to you or if they ignore you completely going forward- like I have done.  I think when  and if you decide to become a thoughtful and civilized human being one day, your life will become a lot more satisfying.  Do with this advice whatever you wish.  Listen or discard it.  If you want to continue to attack me personally and laugh at other people, that is really up to you.   Whatever you think is going to make you ultimately pleased with yourself.   If you decide you want to make a change and turn your life around, you can start with your avatar that mocks others.  Frankly however, I doubt you will change but hopeful you will find your way none the less. Everyone desires to be happy. Good luck.

[olli]

I have no idea how my zooms and primes compare because in real life usage there is little difference and I see no point in setting up artificial shooting conditions to work it out. The only question that matters is whether the output is good enough for me. If it's good enough for me, it's good enough. It may in some technical or theoretical sense be the case that primes, on the whole, produce slightly better IQ than zooms, but in real life use this is largely irrelevant.

 

Of course there may be a handful of people for whom absolute IQ is critical but this is a tiny minority of photographers. The only time I have ever seen the difference in practice is when I visited an exhibition of wildlife photography. It was clear that, specifically, images shot on telephoto prime lenses (400-600mm) were visibly sharper and 'cleaner' than images shot on long tele zooms when the images are printed more that 48" or 120cm wide, but that's a very specific set of circumstances and you can't generalise from that with regard to practical usage.

 

What was also of interest from that exhibition was that when creating these very large (up to 72"/180cm) prints, both pixel count and sensor size had a greater effect than the type of lens used. In terms of sensor size there was little real difference between APS and FF, but there was a huge difference between both of these and the one image that had been shot with medium format (specifically a Phase One 65MP back.)

[milandro]

Nothing in ever really new under the sun.

 

The internet only gave a larger platform to this kind of discussions.

 

Years ago I was often asked to be one of the jury members (3) at a regional photo-club competition and of course part of the “ job” was to talk about the how’s and why’s we had chosen one picture over the other in the various categories.

 

Of course there were, at times, animated discussions on these matters and it was very interesting to see that the pro-jury was very much more interested in the end results than the participants who were very much more interested in the technical minutia.

 

Another thing.

 

About 15 years ago I was one of the photographers engaged into a lengthy and expensive digitization project.

 

We were shooting with a digital back ( scan back) for 4” x 5” large format cameras capable of shooting files of 720Mb.

 

The reason why we were shooting ( and keeping) such enormous files was not often clear to anyone. The people who had the direction of such project simply thought we are going to use the best possible file within the state of the art of the times so that no one could ever tell us that we didn’t do our maximum best.

 

The results way overshot the goal but that it didn’t matter.

[citral]

X-tc, you have to understand that what catches the eye of a photographer and he finds interesting or amusing does not necessarily mean he's mocking that. The woman on my avatar might very well be the nicest person on earth. It is a Martin Parr picture by the way. He's a very nice man who depicts what is absurd in our society and is extremely successful and well regarded.

 

Anyway, you only own zooms apart from the 27mm so I understand that you cannot make real tests to make your point if challenged. Yet you don't either provide a compilation of links for review, a sigma vs canon video does not lead to the conclusion "Fuji zooms have on par IQ with primes".

 

I could do that work, and it would not turn the way you want to hear it, so you could continue to pretend you ignore it because of who I am and not because of the results but it has little interest.

 

Rest assured I'm not judging you personally as I don't know you, only the way you act on the forum.

 

Anyway all in all it has little interest, if someone wants to claim on par IQ to make himself feel better with his purchase I don't know why I care so much in the end, so I will stop there and let you feel happy with your illusion

[milandro]

My fellow members.

 

I don’t want to patronize anyone.

 

The point of any discussion, anywhere, is not to get the upper hand or “ win” the argument. The point of a discussion is to exchange ideas and points of view.

 

Disagreement is the thing is the thing. 

 

I don’t necessarily want to “ convert” anybody to my thinking or diminish someone else’s position. Arguing one position is simply explaining one position to another person but ultimately we all have to resign to the fact that, thank G-D, we are all different!

 

There ’s a very long (!!!)  Italian poem called “ Il Cicerone” written by Gian Carlo Passeroni, a part of which is dedicated to the difference of taste and habits (culinary or otherwise) among humans none of whom are exactly like any other.

 

It is only the other day that two friends of mine were singing their praises of boiled rice, which, nice though it is, I always use as an example of although certainly healthy but uneventful and rather boring type of food!

 

So some like primes and some don’t.

 

Isn’t it wonderful that we can all buy what we like and let the others do what they prefer?

[MintMark]

There are lens review websites that take lots of measurements and I find these a useful resource for comparing lenses in a more objective way... two I often consult are

 

http://www.photozone.de/fuji_x

http://www.lenstip.com/index.html?test=obiektywu&producent=66&model=&typ=0&moc=0&przetest=1&szukaj=Search

 

They typically measure sharpness, vignetting, distortion and chromatic aberration, although I also appreciate lenstip's illustration of coma in the corners.

 

I recently tried to compare the 10-24 with the 14 and also with the 18-55. Each lens has pros and cons but I think the 10-24 at 14 compares well with the prime at f/4, although my important criteria might not be the same as someone elses...