How useful is DxOMark's Perceptual MPix when comparing lenses on APS-C?
Asked 6/7/2016
6 views
2 answers
0
I'm comparing telephoto lenses for a Canon APS-C camera and noticed that DxOMark's Perceptual MPix scores can make surprising combinations look similar or even favor a non-L lens. For example, a Tamron 70-200mm f/2.8 VC scores higher than a Canon 70-200mm f/4L IS on a 550D, and some zooms with very different reputations can show similar perceptual MPix numbers. Does that mean the higher-scoring lens is actually better for image quality or autofocus? I'm mainly interested in daylight outdoor football and occasional portraits on APS-C.
Originally by Photography Stack Exchange contributor. Source · Licensed CC BY-SA 4.0
Photography Stack Exchange contributor
10y ago
2 Answers
4
Does this make the lens better than an L Lens?
It all depends on what way you mean when you use the word "better":
- Sharper at common apertures and focal lengths? At the center of the frame or over the entire field of view?
- Less chromatic aberration at a particular focal length and aperture?
- Less light falloff at a particular focal length and aperture?
- Wider maximum aperture? (hint: that's the biggest difference in the two lenses' DxO Mark overall score)
- Value for price?
- Ruggedness and ability to withstand harsh environmental conditions?
- Light and compact for hiking or long shooting session?
- Guaranteed compatibility with future Canon camera bodies?
- Autofocus accuracy? Autofocus speed? Autofocus frame-to-frame consistency?
As is often the case with comparisons between two zoom lenses, one might perform better at certain focal lengths and apertures and the other will perform better at other focal lengths and apertures. One might perform better in terms of chromatic aberration, the other may do better with regard to light falloff in the corners. One might give better image quality when carefully focused manually (as all tests at DxO Mark are conducted), the other may give better AF performance when tracking moving subjects.
There's no simple way to define which lens is "better" than another. On order to decide which lens is "better" for a particular use case, the requirements of that use case must be considered and applied to the performance of each lens.
In the case of the two lenses in question used for sports/action in daylight conditions the superior AF performance of the Canon 70-200mm f/4 IS is probably a larger consideration than the slightly better optical performance (as measured by DxO Mark) of the Tamron SP 70-200mm f/2.8 Di VC. The comparison of the same two lenses at The-Digital-Picture is a much more mixed bag, with the Canon sharper in the corners across the focal length range and even in the center at 200mm and f/4 than the Tamron. The Tamron has a reputation for less than stellar AF performance when use on moving subjects.
On the other hand, for portraits and concerts the larger f/2.8 aperture of the Tamron is probably the largest differentiator between the two. Especially when used with an APS-C camera that limits the low light performance of the smaller sensor.
P.S. - Someone please explain to me how that is a 13 P-MPix vs. 9 P-MPix score at DxO Mark based on their own measurements?

The same comparison (200mm @ f/4) at The-Digital-Picture.
Please keep in mind that most online testing sites test a single copy of a particular lens model. Copy-to-copy variation between one example of a specific lens model and another example of the same model can, and often does, vary as much or more than differences between comparing a single copy of lens X and a single copy of lens Y when both lenses have similar focal lengths, maximum apertures, and build quality.
Roger Cicala, founder and chief lens guru at lensrentals.com, rarely publishes test results of lenses unless he has tested and averaged the results from a minimum of ten samples of a particular lens. But Roger tends to only measure MTF, or "sharpness", at various points in the lens' field of view. He doesn't measure other things such as geometric distortion, peripheral light falloff, out-of-focus rendering (bokeh), etc. He's published more than a few blog entries regarding copy-to-copy variation between "identical" lenses.
Measuring Lens Variance
Fun with Field of Focus II: Copy-to-Copy Variation and Lens Testing
Things You Didn’t Want to Know About Zoom Lenses
Originally by user15871. Source · Licensed CC BY-SA 4.0
user15871
10y ago
0
Generated from our catalog & community — verify before relying on it.
Not by itself. DxOMark’s Perceptual MPix is only one summary metric, and it does not mean one lens is simply “better” in every way.
A lens can differ in many important areas: sharpness in the center vs edges, performance at different focal lengths and apertures, chromatic aberration, vignetting, maximum aperture, autofocus speed/accuracy/consistency, build quality, size/weight, and long-term compatibility. The DxO overall result can also be influenced heavily by aperture differences, so a 70-200mm f/2.8 and a 70-200mm f/4 are not directly reducible to one number.
Similar perceptual MPix scores between very different lenses are possible because that score is only summarizing resolved detail on a specific camera body. It does not capture all aspects of rendering or handling.
For your uses, look beyond the single score. Daylight football may favor fast, reliable autofocus and suitable focal length; portraits may favor the wider f/2.8 aperture for blur and subject isolation. Use detailed lens tests and sample images rather than relying on perceptual MPix alone.
Recommended products
UniqueBot
AI10y ago
Your Answer
Related Questions
Canon vs Sigma 70-200mm f/2.8 for weddings: how much does IS matter, and is Canon worth the extra cost?
How should I interpret DxOMark ‘perceptual megapixels’ for the Canon 100mm Macro on different Canon APS-C bodies?
What makes a 70-200mm f/2.8 more expensive than a 70-200mm f/4, and how do 70-300mm zooms compare to a 200mm prime?
Should I buy Canon EF full-frame lenses for a Canon 550D APS-C body?
Will a Canon EF 2x III extender work with Sigma or Tamron EF-mount lenses?