Primes always have better IQ than Zooms.... Not

[springbock]

This is easier to see, The 70-200 @ 70 is even a bit better IQ wise than the Canon 60mm macro. As well as better than the Pentax 200mm macro at 200. So even though I wasn't claiming zooms could be sharper than primes it appears it can happen. And please don't tell me the Pentax 200 mm macro and the Canon 60mm macro are not good primes.



--
'The truth is rarely pure and never simple' Oscar Wilde

 

[spt_gb]

Trot over to SLRGEAR.COM and compare the blur plots for the 70-200 f/2.8 with the 200 f/2.0, the prime is sharper at all equivalent apertures.

 

[springbock]

What... no rebuttal to the lens comparisons I showed? I took the effort to make my comparison easy and accurate for you and others. I'm not going to follow your link to a web site and then dig around for 2 lenses you chose!

You said NO good zoom is better than a good prime! I gave proof that you are wrong. You seem to be one of these "photographers so set in their outdated attitudes when it comes to image quality with primes and zooms"

Trot over to SLRGEAR.COM and compare the blur plots for the 70-200
f/2.8 with the 200 f/2.0, the prime is sharper at all equivalent
apertures.

--
'The truth is rarely pure and never simple' Oscar Wilde

 

[rsn48]

I did glance at the Pop photo article and I did notice that some of the primes were faster than the zooms; for me this is important - available light, better out of focus background, etc. As has been reported more than once here, IQ isn't the only factor that can be important in a lens. For example, I'd rather us the Canon 50 f1.8 (or to be truthful, the Sigma 50 f1.4) over any zoom out there that has the 50mm incorporated into it - a much faster lens wide open AND IQ is better wide open.

Some primes are getting long in the tooth, Canon isn't paying much attention to primes except at the long tele end. At the long tele end, there are no zooms that match the IQ of a prime e.g. Canon 100 - 400 IS versus Canon 400 f5.6

There are some superb zoom lenses out there, the Canon 70 - 200 f2.8 and the Sigma 100 - 300 f2.8 come to mind with great IQ. But even sharper is the Sigma 150 macro lens, for example.

But not all primes are excellent and not all zooms are "so-so."

--
Rationally I have no hope, irrationally I believe in miracles.
Joni Mitchell

 

[springbock]

I did glance at the Pop photo article and I did notice that some of
the primes were faster than the zooms; for me this is important -
available light, better out of focus background, etc. As has been
reported more than once here, IQ isn't the only factor that can be
important in a lens. For example, I'd rather us the Canon 50 f1.8
(or to be truthful, the Sigma 50 f1.4) over any zoom out there that
has the 50mm incorporated into it - a much faster lens wide open AND
IQ is better wide open.

Some primes are getting long in the tooth, Canon isn't paying much
attention to primes except at the long tele end. At the long tele
end, there are no zooms that match the IQ of a prime e.g. Canon 100 -
400 IS versus Canon 400 f5.6

There are some superb zoom lenses out there, the Canon 70 - 200 f2.8
and the Sigma 100 - 300 f2.8 come to mind with great IQ. But even
sharper is the Sigma 150 macro lens, for example.

Once again not true!...

The Sigma 150 f2.8 macro.



The Sigma 150 macro is a very good lens. For Macro it is superb but it it not necessarily a sharper lens in that focal length for things other than macro.

But not all primes are excellent and not all zooms are "so-so."

--
Rationally I have no hope, irrationally I believe in miracles.
Joni Mitchell

--
'The truth is rarely pure and never simple' Oscar Wilde

 

[Chato]

I did glance at the Pop photo article and I did notice that some of
the primes were faster than the zooms; for me this is important -
available light, better out of focus background, etc. As has been
reported more than once here, IQ isn't the only factor that can be
important in a lens. For example, I'd rather us the Canon 50 f1.8
(or to be truthful, the Sigma 50 f1.4) over any zoom out there that
has the 50mm incorporated into it - a much faster lens wide open AND
IQ is better wide open.

Some primes are getting long in the tooth, Canon isn't paying much
attention to primes except at the long tele end. At the long tele
end, there are no zooms that match the IQ of a prime e.g. Canon 100 -
400 IS versus Canon 400 f5.6

There are some superb zoom lenses out there, the Canon 70 - 200 f2.8
and the Sigma 100 - 300 f2.8 come to mind with great IQ. But even
sharper is the Sigma 150 macro lens, for example.

Once again not true!...

The Sigma 150 f2.8 macro.



The Sigma 150 macro is a very good lens. For Macro it is superb but
it it not necessarily a sharper lens in that focal length for things
other than macro.

But not all primes are excellent and not all zooms are "so-so."

All these tests you post have the word "subjective" in the heading. What does that mean?

Dave

 

[rsn48]

Let me instead of the Sigma 150 macro throw in the Sigma 50 f1.4, look this one up at Pop photo. I just read the review there, find me a zoom with better characteristics, and fast.

Again in general what I say is a no brainer: some zooms are excellent and some primes are sharper, but the development of lenses has gone towards zooms which explains why you can't get a decent prime from Canon below a 35mm. This is why many were surprised that Sigma came out with its new 50, it seems to go against the grain of lens development - towards zooms. Perhaps you can post the Pop Photo IQ chart of it.

--
Rationally I have no hope, irrationally I believe in miracles.
Joni Mitchell

 

[springbock]

Hi Dave...

"SQF was developed by Edward Granger, a professor at the Rochester Institute of Technology and senior scientist at the Eastman Kodak Company during the 1970s, as a better way of measuring lens sharpness than the prevailing system at the time, known as Modulation Transfer Function (MTF). MTF measures a lens’ ability to reproduce discrete bars (or lines) in increasingly smaller arrays.

Sharp lenses can reproduce more “line pairs per millimeter” (lpm) than can unsharp lenses. MTF numbers, however, are meaningful only to optical engineers, so Granger set about designing tests that could have real-world meaning to most photographers.

The actual SQF tests are performed on an optical bench: The tested lens is mounted on a movable, computer-controlled stage and focused at infinity. Light from a laser is used to illuminate a metal plate into which very fine crosshairs have been etched—2.5 microns wide for a normal lens, or much smaller than the diameter of a human hair. (Larger crosshairs are used for short focal length lenses; smaller crosshairs, for long lenses.)

The crosshairs are optically altered by a device called a collimator to appear to be at infinity. The collimated light passes through the crosshairs, through the center of the test lens, ultimately striking a high-resolution CCD array where the “light spread” (blur) of the image is measured.

The greater the light spread, the less sharp the lens. We measure the spread from the full range of lens apertures.

After testing the center of the lens, we continue with off-center and edge measurements. The results are compiled, compared, and weighted: Image center readings comprise 50 percent of a lens’ SQF score; off-center readings, 30 percent; and image edge readings, 20 percent. (The percentages are slightly different for digital lenses.)

"By running the readings through a series of computerized calculations, SQF can quantify lens sharpness and interpolate the rate at which sharpness will deteriorate as image size increases".



http://www.popphoto.com/cameralenses/2564/optic-nerves-afraid-to-buy-the-wrong-lens-sqf-subjective-quality-factor-page4.html

--
'The truth is rarely pure and never simple' Oscar Wilde

 

[springbock]

Again in general what I say is a no brainer: some zooms are excellent

No... not just excellent, some zooms are sharper than some primes!

and some primes are sharper,

Some yes, but not always and they are decreasing every year.

but the development of lenses has gone
towards zooms which explains why you can't get a decent prime from
Canon below a 35mm. This is why many were surprised that Sigma came
out with its new 50, it seems to go against the grain of lens
development - towards zooms.
Perhaps you can post the Pop Photo IQ chart of it.

I don't need to... you need to read more carefully... I never said every zoom is sharper than every prime.

--
Rationally I have no hope, irrationally I believe in miracles.
Joni Mitchell

--
'The truth is rarely pure and never simple' Oscar Wilde

 

[springbock]

Determining a lens sharpness and contrast by using Subjective Quality Factor, once again (developed by Edward Granger, a professor at the Rochester Institute of Technology and senior scientist at the Eastman Kodak Company) has to be light years ahead of forum members opinions.

To me the no brainer is getting your information from experts and not back yard self proclaimed lens experts ;-)

One thing I find on this forum is there are lots of people eager to state their opinion with no evidence to back it up!

--
'The truth is rarely pure and never simple' Oscar Wilde

 

[ebsilon]

Hi Dave...

"SQF was developed by Edward Granger, a professor at the Rochester
Institute of Technology and senior scientist at the Eastman Kodak
Company during the 1970s, as a better way of measuring lens sharpness
than the prevailing system at the time, known as Modulation Transfer
Function (MTF). MTF measures a lens’ ability to reproduce discrete
bars (or lines) in increasingly smaller arrays.

Sharp lenses can reproduce more “line pairs per millimeter” (lpm)
than can unsharp lenses. MTF numbers, however, are meaningful only to
optical engineers, so Granger set about designing tests that could
have real-world meaning to most photographers.

The actual SQF tests are performed on an optical bench: The tested
lens is mounted on a movable, computer-controlled stage and focused
at infinity. Light from a laser is used to illuminate a metal plate
into which very fine crosshairs have been etched—2.5 microns wide for
a normal lens, or much smaller than the diameter of a human hair.
(Larger crosshairs are used for short focal length lenses; smaller
crosshairs, for long lenses.)

The crosshairs are optically altered by a device called a collimator
to appear to be at infinity. The collimated light passes through the
crosshairs, through the center of the test lens, ultimately striking
a high-resolution CCD array where the “light spread” (blur) of the
image is measured.

The greater the light spread, the less sharp the lens. We measure the
spread from the full range of lens apertures.

After testing the center of the lens, we continue with off-center and
edge measurements. The results are compiled, compared, and weighted:
Image center readings comprise 50 percent of a lens’ SQF score;
off-center readings, 30 percent; and image edge readings, 20 percent.
(The percentages are slightly different for digital lenses.)

"By running the readings through a series of computerized
calculations, SQF can quantify lens sharpness and interpolate the
rate at which sharpness will deteriorate as image size increases".



http://www.popphoto.com/cameralenses/2564/optic-nerves-afraid-to-buy-the-wrong-lens-sqf-subjective-quality-factor-page4.html

--
'The truth is rarely pure and never simple' Oscar Wilde

Great explanation, I've wondered what this SQF concept in the PopPhoto test actually meant. Strange that it should be termed Subjective - from this elaborate explanation "Objectve" would certainly be a more suitable term, wouldn't it?

------- Eirik ----------
n
[]O]

Visit my gallery at http://eirikbs.smugmug.com/

 

[JohnnyReno]

To me the no brainer is getting your information from experts and not
back yard self proclaimed lens experts ;-)

I sure hope I don't come across as such ^ ^

I hate those guys...

• S.G.

 

[tko]

First of all, PP is probably the worst tester in the business. Riddle me this - why are they the only ones who bother to give funny color ratings at different print sizes? Did they test wide open? Their testing is so old fashioned and just plain crippled it's not even funny.

Then they picked a really funny set of primes. You see, a decent prime is F1.4, and F2.0 is almost slow for a prime, while F2.8 is fast for a zoom. So normally you'd be comparing say, F1.4 to F2.8, a whole other difference in performance, DOF, and light gathering. You'd almost think they cherry picked the lenses to make the zooms look good.

It costs big bucks to make a fast prime. Sure, at F5.6 most zooms look pretty good, even the kit lenses. Go to F2.8 that's twice the lens diameter, 4 times the glass needed, maybe 8 times the cost or more. Lens speed costs big time.

Tell you what. Compare a 50MM F1.4, a 85MM F1.8, or a 135MM F2.0 to the nearest zoom and we'll see what happens.

So, I wouldn't read PP and proclaim myself an expert and all the old pros obsolete. It takes more than one quick read to equal experience.

 

[spt_gb]

posting Popular Photographer's SQF plots you ignored what I posted, and as far as I can see for the obvious reason that pop photo's test database is somewhat inadequate to the task. Where is the SQF plot for the Canon 200 f/2.0? What about the old 200 f/1.8 even (reckoned to be one of the sharpest lenses ever made)? You've not got enough data to work with, and such that there is seems to be inconsistent (why do some plots start at 4x5 and others at 5x7?).

Despite your posted description's statement of how wonderful the SQF methodology is, the world in general doesn't seem to be buying it. It's too simplistic, it doesn't tell us about chromatic aberrations (shining a monochromatic laser through the lens can't possibly do this), it doesn't tell us about corner performance vs center, vignetting or image distortion. Web review sites like SLRGEAR and PhotoZone give you better data sets and more lenses. Find me a good prime and then a better zoom from these data sets.

You said NO good zoom is better than a good prime! I gave proof that
you are wrong. You seem to be one of these "photographers so set in
their outdated attitudes when it comes to image quality with primes
and zooms"

Trot over to SLRGEAR.COM and compare the blur plots for the 70-200
f/2.8 with the 200 f/2.0, the prime is sharper at all equivalent
apertures.

--
'The truth is rarely pure and never simple' Oscar Wilde

 

[tabloid]

I have a Sigma 70-200 f2.8, and a Canon 100mm f2.

At f2 the Canon knocks the spots off the Sigma at 2.8, (at any focal length).

At f8 i cant tell the difference between the two lenses.

Moral of the story: if u want real quality at full aperture, get a prime.

Camera used: Canon 1Ds 3

http://tabloid.zenfolio.com/p542383149/

Regards

George Richardson.

 

[TaoMaas]

Here's why the macros are performing worse than the zoom. From the explanation article: "The actual SQF tests are performed on an optical bench: The tested lens is mounted on a movable, computer-controlled stage and focused at infinity. "

They're testing the lenses at infinity. Probably the worst range for a macro. So we should probably toss out the two macros or else ask to see a test of the Canon zoom at 200mm, focused at it's closest focusing distance.

 

[Chato]

Thanks for the explanation.

After reading your post I checked around the Web, and it would appear that using SQF is sort of an accurate "poor mans" guide to lens quality. Not as accurate as MTF, but a good ball park way of determining quality.

But even though I am a fan of high quality zooms, all things being equal, I still believe that primes give a higher quality. Often enough in todays world this "difference" is academic.

Dave

Hi Dave...

"SQF was developed by Edward Granger, a professor at the Rochester
Institute of Technology and senior scientist at the Eastman Kodak
Company during the 1970s, as a better way of measuring lens sharpness
than the prevailing system at the time, known as Modulation Transfer
Function (MTF). MTF measures a lens’ ability to reproduce discrete
bars (or lines) in increasingly smaller arrays.

Sharp lenses can reproduce more “line pairs per millimeter” (lpm)
than can unsharp lenses. MTF numbers, however, are meaningful only to
optical engineers, so Granger set about designing tests that could
have real-world meaning to most photographers.

The actual SQF tests are performed on an optical bench: The tested
lens is mounted on a movable, computer-controlled stage and focused
at infinity. Light from a laser is used to illuminate a metal plate
into which very fine crosshairs have been etched—2.5 microns wide for
a normal lens, or much smaller than the diameter of a human hair.
(Larger crosshairs are used for short focal length lenses; smaller
crosshairs, for long lenses.)

The crosshairs are optically altered by a device called a collimator
to appear to be at infinity. The collimated light passes through the
crosshairs, through the center of the test lens, ultimately striking
a high-resolution CCD array where the “light spread” (blur) of the
image is measured.

The greater the light spread, the less sharp the lens. We measure the
spread from the full range of lens apertures.

After testing the center of the lens, we continue with off-center and
edge measurements. The results are compiled, compared, and weighted:
Image center readings comprise 50 percent of a lens’ SQF score;
off-center readings, 30 percent; and image edge readings, 20 percent.
(The percentages are slightly different for digital lenses.)

"By running the readings through a series of computerized
calculations, SQF can quantify lens sharpness and interpolate the
rate at which sharpness will deteriorate as image size increases".



http://www.popphoto.com/cameralenses/2564/optic-nerves-afraid-to-buy-the-wrong-lens-sqf-subjective-quality-factor-page4.html

--
'The truth is rarely pure and never simple' Oscar Wilde

 

[jrtrent]

Web review sites like SLRGEAR and
PhotoZone give you better data sets and more lenses. Find me a good
prime and then a better zoom from these data sets.

Photozone convinced that for most of my needs (outdoor shooting in good light at a normal focal length), the DA 35mm f/2.8 was, on balance, worse than my Samsung variant of the the SMC-DA 18-55mm f/3.5-5.6 AL kit zoom. In order of importance, I'm looking for low geometric distortion, low chromatic aberrations, even sharpness across the frame, and decently low vignetting (though I agree with dpreview's somewhat less critical standard, that corner fall-off less than 1 stop below the center really isn't a problem).

I had high hopes for the 35mm macro because macros generally have vanishingly low geometric distortion and very even sharpness across the field. However, the kit zoom at 35mm beats the prime in every area except vignetting, and even there shows less than one stop fall-off wide open. To keep it simple and reduce my work, here are some numbers from Photozone at my most commonly used aperture of F/8:

macro
0.4% barrel distortion
chromatic aberration 1.16 pixel width
center/border resolution 2183/1971 LW/PW
vignetting 0.24EV

zoom
0.00574% barrel distortion
chromatic aberration 0.6 pixel width
center/border resolution 2160/2040 LW/PW
vignetting 0.23EV

I still want a prime for its smaller size, possibly lower weight, wider aperture (would sure be nice for indoor museum shots) and (fingers crossed) decent focus and depth of field scales. But my kit zoom will get me through til Pentax sees fit to release the 30mm prime currently on its lens map, and hopefully the new prime will test well.

 

[rsn48]

Primes have over the past decade been put on the back burner as more and more photogs scramble for zooms so the companies have decided - the money is in the zooms. Fair enough, but I think Sigma with the new 50 has tapped into a market that is rebounding from zooms and wanting decent primes.

What is needed is some decent FF primes below 50 that are contemporary, not recycled mid-90's models or earlier. And these primes should be fast, between f1.8 and f1.4 - any lower and they become unaffordable for many. I was just looking at an advert in Shutterbug for the leica prime that is below f1 and I think it was costing around $3500 or higher, shaky memory here.

Utrawides are like long tele zooms, most discover on the long zoom tele's that the majority of their pics are taken at the longest focal length. With ultra wides most discover that the majority of their pics are taken at the widest focal length; SO IF THEY CAN find a decent prime that is sharper, why not use it; this is why so many Canon 400 L f5.6's are sold - sharper than the 100 - 400 and reasonably priced because it is an f5.6

But for a good idea of what to buy, look in the bag of well known wedding photographers, usually they have a series of fast primes as well as a series of faster good IQ zooms.

--
Rationally I have no hope, irrationally I believe in miracles.
Joni Mitchell

 

[Leonard Migliore]

Thanks for the explanation.

After reading your post I checked around the Web, and it would appear
that using SQF is sort of an accurate "poor mans" guide to lens
quality. Not as accurate as MTF, but a good ball park way of
determining quality.
But even though I am a fan of high quality zooms, all things being
equal, I still believe that primes give a higher quality. Often
enough in todays world this "difference" is academic.

Yeah, I believe that too. But, like the Flying Spaghetti Monster, proof is hard to find. I set up my tripod in my back yard and shot the following (no redeeming artistic value):



with a 16-85VR set to 55mm and a 55mm f.2.8 Micro Nikkor, both at f/5.6. Now we know who'll win that one, but what I got was

16-85VR



55 Micro



I can't see any useful difference.

Full files are at

http://www.laserk.com/media/Lenstest/zoomfull.jpg
http://www.laserk.com/media/Lenstest/microfull.jpg

Of course, the sun decided to come out while I was changing lenses but I don't think that affects the results.

--
Leonard Migliore