Why HD Coating? A Study of Lens Coatings

Started Sep 19, 2013 | Discussions thread
Flat view
MightyMike Forum Pro • Posts: 38,600
Why HD Coating? A Study of Lens Coatings

Lenstip occasionally does a test showing the light transmission through a lens at various wavelengths. They've done this 34 times in 252 reviews. Taking their data from those reviews we can see that not all lens coatings are created equal which is obvious, also that you don't always get the best coatings the company has to offer on every lens.

Pentax states that their light transmission with SMC coated glass is 99.8%, interestingly the closest they came in reality based on these tests is 99.4%. Now that isn't to say that Pentax was wrong, its just that its likely the highest possible transmission given the best combination of SMC coating and glass formula at the best wavelength and best angle of incedance of light. In otherwords right but not the whole story. This also brings up the very real possibility that not all glass formulas can accept all lens coatings and therefore each side of every lens group in an SMC lens may not have the full SMC coating on it. An example of this is the Nikon 24-70mm AF-S F2.8G ED, this lens has their nano crystal coating, however it only actually appears on 1 side of 1 element in the whole design of the lens.

Why do we need lens coatings a) For better light transmission which leads to more light hitting the film plane and likely better contrast and more accurate colours overall. b) Some would say that it helps against flare, I would agree to an extent but look at some of the reviews where the lenses have top quality coatings and some of them have nasty flare. So I would say that flare can be improved a little by lens coatings but its more the actual physical design of the lens that has a major affect on the way the flare shows.

So uncoated glass lets through 95% of the light per glass surface to air or air to glass surface connection. This means if you have a lens with 10 groups (20 surfaces that meet air) then the light transmission is just 35.8% (0.95^20)*100. Its not much light getting through after that many groups and this will cause an F2.8 designed lens to project F7.9 light onto the film plane. A loss of 3 stops over 10 groups of elements. The best lens tested had an overall rating of 0.995 across 7 groups of lenses (0.995^14)*100 = 93.2% light transmission overall. Therefore its F3.5 aperture projects F3.75 onto the film plane. About 1/6th of a stop of light loss. Its pretty clear glass coating is important.

In the following charts I got the lens name of course, the amount of element groups and the approx. transmission percentage at 7 different wavelengths from 400nm to 700nm. The min and max transmission and the average, i figured with 7 samples the average would be pretty realistic. The column designated "Avg Coating Efficency" has the (averaged across the lens) coatings transmission per 1 surface (remember 2 surfaces in a group). The "brightest" aperture of the lens and its effective t-stop. I have bolded the values of the more important columns.


Highest Efficiency to lowest Efficiency

Things of note!

1) Most lenses have a similar per group surface coating efficiency with anything over 99% being pretty good!

2) Most lenses lose between a third and half a stop of light

3) The Olympus 14-35mm F2.0 lens although having a decent efficiency of 99% it has 17 groups or a whopping 34 glass/air surface so its T-stop is actually T2.77 a loss of almost 1 stop of light

4) The Leica Summiccron-M 50mm F2.0 has the worst overall per surface coating but with just 4 groups and 8 glass/air surfaces it still only looses just a third of a stop of light.

5) Sigma appears to have put its best average coating on the 18-250 and it really helps as it has 14 groups (28 glass/air surfaces) but it only loses just shy of a half stop of light.

6) Its now reasonable to expect that every company has great coatings and not so great coatings and what is advertised isn't necessarily what is on every group.

So Why HD coating? I'd suggest i still can't answer that but I'd suspect that its got to be better in some way.

I was considering that one may get similar information and more samples is with the T-stop data on DXO. However I'm not sure how that would pan out as I couldn't say if that is an average and at what wavelength, it needs further investigation.

-- hide signature --

Mike from Canada
"I am not a great photographer! God is a great creator! All I do is capture His creation with the tools He has provided me."
'I like to think so far outside the box that it would require a telephoto lens just to see the box!' ~ 'My Quote :)'

Sony Cyber-shot DSC-T2
If you believe there are incorrect tags, please send us this post using our feedback form.
Flat view
Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow