Help in Understanding DXOMark's Score: Okay, I wanted to buy a new lens for my Canon 70D, now I have the Canon EF 50mm f/1.8 STM lens. I was trying to decide between the EF 85mm f/1.8...

baygolf Loc: DMV

Okay, I wanted to buy a new lens for my Canon 70D, now I have the Canon EF 50mm f/1.8 STM lens. I was trying to decide between the EF 85mm f/1.8 USM or the EF 100mm f/2 USM. When I looked these lenses up on DXOMark, I was surprised to see that the 50mm had a better DXOMark score [22] than the 85mm [19], (the 100mm f/2 score was 23). From all the reviews I have read the 85mm is an excellent lens and some would consider equal to the 85mm L lens. I was also concerned about the focal length since I wanted this lens for portraits: on a 70D the 85mm = 136 and the 100mm = 160. Now based on focal length I'm leaning towards the 85mm, but base on the DXOMark score I should just use my 50mm and save some $$$. WOW! If I had a full frame then the 100mm would be my selection. When I looked at the DXOMark score for another lens I have the EF-S 40mm f/2.8 STM it has a score 19, which is the same as the 85. What I'm I missing? Is the 85mm not high-quality portrait lens on an APC camera?

BHC Loc: Strawberry Valley, JF, USA

baygolf wrote:

Okay, I wanted to buy a new lens for my Canon 70D,... (show quote)

Perhaps one error is putting too much faith in the actual rating, rather than in the range. On a 1-10 scale, 2.4 versus 2.6 will show as a 2 to 3 range, not a huge difference, indeed rather insignificant. On the other, a difference between ratings of 3 and 7 should indicate an obvious difference in quality. On a scale of 1-100, the same proportional differences apply.

travisdeland Loc: deland, FL

I own the 85mm, F/1.8 lens-screw the scores-it's an absolutely great lens. I've used it with my T3i, 70D, and now my 7DmkII. Buy it and you'll never look back.

MT Shooter Loc: Montana

baygolf wrote:

Okay, I wanted to buy a new lens for my Canon 70D,... (show quote)

When comparing lenses on DXO you need to make sure the comparison was done on the same body, it's selectable. Different bodies can yield widely different lens scores due to differences in sensors and processors.

baygolf Loc: DMV

MT Shooter wrote:

When comparing lenses on DXO you need to make sure the comparison was done on the same body, it's selectable. Different bodies can yield widely different lens scores due to differences in sensors and processors.

Thanks, I was comparing the same body: 70D.

baygolf Loc: DMV

travisdeland wrote:

I own the 85mm, F/1.8 lens-screw the scores-it's an absolutely great lens. I've used it with my T3i, 70D, and now my 7DmkII. Buy it and you'll never look back.

Thanks, I guess I was really questioning DXOMark scoring.

joer Loc: Colorado/Illinois

baygolf wrote:

Thanks, I guess I was really questioning DXOMark scoring.

Much of it is splitting hairs. How you PP and image may have a bigger affect.

jerryc41 Loc: Catskill Mts of NY

I believe their testing is accurate, and their results are significant. You have to decide if what they are testing is important to you. Before I buy a lens, I check all the comparisons I can.

https://www.imaging-resource.com/IMCOMP/COMPS01.HTM
http://www.dpreview.com/products/compare/lenses
http://lenshero.com/lens-comparison
http://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx
http://www.lenstip.com/lenses.html
http://www.dxomark.com/Lenses/Compare
http://www.lenscore.org/

CO

LensTip.com tests lenses in all of these categories. There're listed at the bottom of each page on the website. They do very though testing. They tested the first two lenses you listed. Click on the links below to read their reviews.
1. Introduction
2. Pictures and parameters
3. Build quality
4. Image resolution
5. Chromatic and spherical aberration
6. Distortion
7. Coma, astigmatism and bokeh
8. Vignetting
9. Ghosting and flares
10. Autofocus
11. Summary
12. Sample shots
http://www.lenstip.com/444.1-Lens_review-Canon_EF_50_mm_f_1.8_STM_Introduction.html
http://www.lenstip.com/291.1-Lens_review-Canon_EF_85_mm_f_1.8_USM-Introduction.html

GENorkus Loc: Washington Twp, Michigan

Mogul wrote:

Perhaps one error is putting too much faith in the actual rating, rather than in the range. On a 1-10 scale, 2.4 versus 2.6 will show as a 2 to 3 range, not a huge difference, indeed rather insignificant. On the other, a difference between ratings of 3 and 7 should indicate an obvious difference in quality. On a scale of 1-100, the same proportional differences apply.

You look for the range needed first!

Think about it. When you want a car to haul around a large family, I doubt you'd be looking in a motorcycle shop, or even a City Car dealer.

Mark7829 Loc: Calfornia

There are so many techno-freaks out there. Beginners think that the camera and lens is what makes the picture and those 2 pts on DxO mark will make a huge difference. Well, your all wrong. The photographer makes the picture. The camera and lens only takes the picture. The other critical and all important attributes in photography is composition and light. In good composition and light, you can take an award winning image with an iPhone. When you master composition and light it does not matter what camera you have and what lens you use. You will take a great shot anywhere and everywhere.

davidrb Loc: Half way there on the 45th Parallel

baygolf wrote:

Thanks, I guess I was really questioning DXOMark scoring.

DXOMark's scoring isn't suspect. Have you ever seen a photograph taken by DXO? NO, and you won't. The lenses are lab rats. There is nothing about their testing that compares lenses or bodies used in actual photography. Tell a professional photographer that his/her lens is inferior because DXO says so. That's a real conversation ender. If your photography takes you to a laboratory for shooting then DXO is the place for you to go for ratings. The only rating that counts is the one given by someone who shoots in the real world. DXO does not fit that category.

sirlensalot Loc: Arizona

I think DxOMarks testing and scores are quite relevant if you are using only that criteria. The way a lens is used by a photographer can be pretty subjective. A bench test is the same criteria and standard for all lenses. Cameras and lenses absolutely make a difference if you are specializing in a particular field, such as sports as an example.

Architect1776 Loc: In my mind

baygolf wrote:

Okay, I wanted to buy a new lens for my Canon 70D,... (show quote)

Remember DXO is just their opinion using their parameters not general scientific standards. If the 85 is for portraits it is perfectly good.
I do not own it but have rented it for a specific purpose and it worked absolutely great, no regrets.

GENorkus Loc: Washington Twp, Michigan

Most seem to be talking about or against test scores, but when you figure in how most product is made...

Each component is made to exacting specs. Each hunk of glass, or metal, or plastic, or ??? used has a percentage of perfection.

When you put two really close to perfect things together they either become more perfect or farther away from it. After you do that a hundred time, it could be the best thing ever, or just barely within the tolerances given for the item.

Such is the case with lenses, your car, products, anything made. So guess what? Test scores are often just an "educated guess" of everything combined.