MTF charts - FF & MFT

Discussion in 'Open Discussion' started by Timur, Feb 21, 2016.

  1. Timur

    Timur Mu-43 Regular

    124
    Nov 10, 2015
    Timur
    Researching into my next lens purchase I started looking at the Tokina 11-20 (as was suggested by some fellow users here). So once I actually opened two reviews from lenstrip side by side to check their galleries for Tokina and for the Oly 7-14 (wanted to compare at 100% for safe measure), I can clearly see that Oly is miles and beyond in terms of sharpness and detail. In fact, even the excellent Nikkor 24-70 2.8 has lower numbers in their charts. I was wondering - why that is? I've used the Nikkor zoom, it was excellent and pin sharp. For comparison, the Nikkor @24mm has almost 50lpmm, Tokina @20mm reaches 54lpm, and Oly is stated to have peaked at 80lpmm. I'm not trying to start a battle here, just want to understand if mft glass is actually sharper, or there are some technical aspects that skew this data. Just going from gallery shots oly 7-14 actually looks like a way sharper lens than both of tokina and nikon.

    Links with the MTF data:
    Nikon Nikkor AF-S 24-70 mm f/2.8G ED review - Image resolution - LensTip.com
    Tokina AT-X PRO DX 11-20 mm f/2.8 review - Image resolution - LensTip.com
    Olympus M.Zuiko Digital 7-14 mm f/2.8 ED PRO review - Image resolution - LensTip.com
     
  2. pellicle

    pellicle Mu-43 All-Pro

    Feb 10, 2010
    Southport, OzTrailEYa
    pellicle
    People become quickly confused about transferability of information from one format to another. Lens testing also needs another factor (which is magnification) to understand it.

    Magnification means "how big is that thing you're capturing on the sensor"

    We look at 36mm wide captures of (pick your pixel count, but lets say 5000 pixels wide) and compare that to a 17.5mm wide capture of essentially the same framing. So its clear that we are enlarging one image more than the other.

    Perhaps this old blog post of mine will help you understand this point.

    in my view ...: lens tests: looking for better pictures, what's missing?

    Its not light going, so take your time to try to follow the points.

    In short this is why DxO's P-MPix system is useful as it shows that same effect. IE that the same MTF results in a different image quality on different formats.

    Essentially the best thing to do in evaluation of a lens is look at shots captured on the same format and see if you're happy enough with those results.

    I had a Tokina 12-24 on a 20D and was quite satisfied with it.
     
    • Agree Agree x 1
  3. Timur

    Timur Mu-43 Regular

    124
    Nov 10, 2015
    Timur
    Thanks for the link!

    unfortunately I wasn't able to find photos taken with Tokina 11-20 (or 11-16) with adapter with a focal reducer to a MFT camera. So I just looked at the photos in the galleries, and compared the images at 100%.
    Since the resolution of the photo is very close on the EM1 and the canon 50d they used (4608×3456 vs 4752×3168), the oly lens just seemed way sharper and detailed to me.

    http://pliki.optyczne.pl/tok11-20/tok11-20_fot05.JPG
    http://pliki.optyczne.pl/oly7-14/oly7-14_fot01.JPG
     
  4. pdk42

    pdk42 One of the "Eh?" team

    Jan 11, 2013
    Leamington Spa, UK
    Magnification and pixel pitch are the issues here. Back in my Canon days I had a 50d (APSC) and used a 100-400L with it. I never got good results and almost sold it. However, when I "upgraded" to a 5dii (FF), I found that the lens performed much better. The pixel pitch of the 5dii was pushing half of the 50d's and because the image needed less magnification to make a print or on-screen view, any lens imperfections were less brutally exposed.
     
  5. Timur

    Timur Mu-43 Regular

    124
    Nov 10, 2015
    Timur
    So, am I understanding this correctly assuming that the larger sensor (FF) has less magnification into the image, and so it produces sharper images than same lens on a crop body? And it's worse the further into crop you go? Meaning that apsc will perform better than mft in that regard? I did notice that on my rokinon 85 1.4 lens, the photos were much sharper with the same lens on d800.

    But then why mft lenses show higher results in the MTF chats, and not lower?

    I'm gonna have to read the article first I guess.
     
  6. pellicle

    pellicle Mu-43 All-Pro

    Feb 10, 2010
    Southport, OzTrailEYa
    pellicle
    Hi
    welcome :)

    yes, that's probably true ... which sort of happend to me when I got into m43 in 2009, so picking was a craps shoot.

    Myself I've come to the conclusion (after trying lots of stuff including the Oly 9-18 on adapter, which is a great lens btw) that image quality is really only one part of the story, the other is weight.

    I did a shoot for a conference back in 2009 with my G1 and set out packing my stuff to do the gig. I was looking forward to not having to tote around large heavy gear and was quite pleased with my m43 outfit that I'd built up (including the 9-18) . When I picked up my bag I laughed aloud and declared to the room "what happened to my light weight camera outfit!"

    If I'm going to make my choices in m43 lenses now I pick (unless there is no alternative) lighter native lenses. These are designed specifically with the system in mind and I'm quite sure there is nothing left to wring out of the lenses in terms of optimised quality (and indeed some software tweakes needed in post) for these system lenses.

    For a start even APS-C lenses (let alone "full frame") are optimised for a larger sensor, and so less magnificatio as well as coverage.

    The important thing to me is less glass between incoming light and sensor (known as transmission).

    Its possible that a Tokina will perform as well as the native one, but ask these questions:
    • will it be bigger and more unweildy?
    • will it be that much cheaper?
    • will it be observably better on print or www distribution?

    Here is some images I've put together on the 9-18 when I was considering ditching it and replacing it (which I now have) with the 14mm + the GWC1 for when I wanted wider.

    9-18-14-45Morph.

    Then a few lenses together (such as the Pany 7-14)

    widesCompared.

    and some tangential thoughts but which may be similar to your thinking and may feed information into your process, specifically size and handling (with a bit of IQ tossed in)

    in my view ...: Panasonic 0.79 wide adaptor on the 14mm (image quality)

    Best Wishes
     
  7. pellicle

    pellicle Mu-43 All-Pro

    Feb 10, 2010
    Southport, OzTrailEYa
    pellicle
    correct ... the examples in my blog post make that clear and why ...
     
  8. Timur

    Timur Mu-43 Regular

    124
    Nov 10, 2015
    Timur
    I read that the 4/3 lenses were optically corrected (as opposed to software in-body correction on the m4/3 lenses), so they were basically same size as the full frame glass.
    I'm quite happy with the size of my system in general, I took my EM1 + Panaleica 24 1.4 everywhere, and I had the oly 45mm in my pocket or bag most of the time without even noticing it most of the time)

    Currently I use my Samyang 12mm f2, which is a great lens in my opinion, and it's plenty sharp and small in size.
    Now I wanted to try an even wider lens, and need to make a choice. I like the panasonic 7-14, but I'm on oly body, and I don't want to deal with the purple artifacts. The olympus 7-14 f4 looks like a nice and solid lens, but it's also pretty big, and goes for about 750-800$ used. New olympus pro zoom is amazing, and in a perfect scenario it's a lens i'd get, I just need to think if such a large investment into new-ish landscape photography territory for me is justifiable. Tokina + lens turbo II would end up around 550$, the cheapest out of the bunch. Maybe Oly 9-18 is what I need for now, to see if I'm going get into shooting super wide)

    Anyway, just my train of thought.
     
  9. pellicle

    pellicle Mu-43 All-Pro

    Feb 10, 2010
    Southport, OzTrailEYa
    pellicle
    I think you should research that carefully before scrubbing the 7-14 based on some web site showing backlit branches against a bright cloud.
     
  10. Timur

    Timur Mu-43 Regular

    124
    Nov 10, 2015
    Timur
    I pulled out few forum discussions about this, and most of the people seem to agree that it prominent. According to them it's not related to CA or flare, but coating on the lens. People mention that problem comes up mostly in the darker situations (night cityscapes or interior shots), but also when there are challenging light situations in day time. Like here for example:
    Panasonic 7-14mm Flare - Case Study for a solution
     
  11. pellicle

    pellicle Mu-43 All-Pro

    Feb 10, 2010
    Southport, OzTrailEYa
    pellicle
    like all tools none is perfect

    I've seen many great images made with the 7-14 ... I've seen many pixel peepers who provoke issues by pushing their gear, but somehow don't really make many "great images".

    Its only you who can decide what tool suits you. I'm only able to answer the original question asked about lenses for FF being used on 43 ... the points about MTF are (I hope) clarified in my blog post. I know I spoke of using 4x5 and 120 film in there, but the point is "size of sensor" and explained as to WHY that is.

    Best Wishes
     
  12. Timur

    Timur Mu-43 Regular

    124
    Nov 10, 2015
    Timur
    Well I think the point of forums such as this one and internet in general is that we can communicate and research, and find the best possible solution for one's needs.
    Thank you for your input.
     
  13. Turbofrog

    Turbofrog Mu-43 Hall of Famer

    Mar 21, 2014
    Yes, M4/3 lenses are sharper and resolve more than lenses on larger formats.

    But (there's always a but), that advantage is usually eliminated by the additional sensor area.

    If the site still reports direct MTF figures, most usually report it in lw/ph which is line widths per picture height - this is used to normalize the values against the sensor size. On Lenstip, however, the value is reported in terms of lp/mm - line pairs per mm. So in this case, since the M4/3 sensor is half the diagonal of an FF sensor, you need to multiply the FF value by 2 to match the M4/3 value. That will give you true resolving power of the sensor + lens combination. You can do the same with the APS-C lenses, but use 1.3x (2x crop divided by 1.5x crop) instead.

    So on an APS-C body, the 52-55 lp/mm of the Tokina 11-20 probably gives you a little lens resolution than the Olympus 7-14 (52-55*1.3 = 68-72 lp/mm). The 24-70/2.8 there peaks around 47 lp/mm, so will give perhaps 15-20% better resolution than the M4/3 lens when used on a full-frame body.

    However, keep in mind also that these measurements are very sensitive to the specific body that they are measured on. The bodies used on LensTip are typically old bodies with low MP counts and anti-aliasing (AA) filters, both of which mean that the comparisons are not always easy to make. Even within a format, while you can assume that all lenses will perform better on a new 16MP AA-less body like the E-M1 compared to the 12MP E-PL1 that's used for testing (and even better on a 20MP body like the PEN-F or GX8), how much better the lens will do is not necessarily obvious.

    The APS-C and FF bodies used are also older, usually in the 10-15MP range, though in some cases they are the latest bodies (new Canon reviews seem to be done with a 5D Mark III, and the Nikon reviews use a 25MP D3X).

    ...basically, the numbers can only tell you so much. They'll give you a rough order of magnitude. But the other thing they won't tell you is the effect that adapting a lens designed for one format to another, or the influence of the focal reducer, or any of those other factors. Generally adapting to M4/3 will tend to cause some theoretical image quality penalty due to the thickness of the sensor cover glass, but well-designed focal reducers will take this into account and correct for it, and will improve the MTF values so that you're getting a similar 1.4x multiplication factor...

    It gets complicated! Best not to get too obsessed with the numbers and just see if the lenses give you the results you like.

    EDIT: I had included the wrong multiplication factor for APS-C. I was thinking of the 1.5x crop factor relative to the FF, not the ~1.3x factor of APS-C relative to M4/3. Oops! So the Olympus should indeed be a bit sharper than the Tokina.
     
    Last edited: Feb 22, 2016
  14. bikerhiker

    bikerhiker Mu-43 All-Pro

    Dec 24, 2013
    Canada
    David
    Sharpness is a subjective measure not a quantifiable measure. What's sharp to you may not be sharp to me image wise. Also, lens MFT data are tested with a flat field test chart at a specific distance and may perform differently in real life. Case in point is the Nikkor 24-70 which performed poorly in many MFT test charts, but do well in real life shooting as you and myself had observed. The reason the Nikon 24-70 performed not so well with MFT charts is due to the design of the lens itself. In order to achieve smoother out of focus transition (known as BOKEH), certain design choices had to be made. Here's the video link for "The Philosophy of Nikkors" -- New "Philosophy of Nikkor" video with some technical details on the 24-70mm f/2.8E ED VR lens | Nikon Rumors

    Clearly, every lens out there are designed towards specific brand standards. The key is HOW the lens would render the image. And that's important because ultimately we are ALL photographers. We photographers use lenses to take pictures. We are not using lenses to measure numbers, dynamic range, noise dbm and lines per inch per height. That's only what a camera operator does -- a breed of photographers who often been confused to true photographers. There are also things we human eyes have limitation that these MTF numbers would do no service to the additional acuity. Basically, don't assume your ability to discern more detail and sharpness with your eyes is the same as others in this world. Many earlier studies conducted by the Kodak labs by Granger and the likes had proven image acuity is a very subjective measure. It's based on what you see. The numbers, MTF or SQF (Subjective Quality Factor) are just a guide to allow us to get closer to the conclusion.

    My advise to you is this. Look at some real samples made by all 3 lenses, the budget you can afford and the needs fulfilled for your photography needs. Then buy the lens that fulfill your needs and ignore the rest. You buy the tools to address your photography needs. If you see Olympus clearly to be better than the other 2 and you can afford the price, then buy it. Ignore what others are saying. Ultimately, you are buying for you! You are not buying a lens just so that, you can make other people happy. That's the youth generation where a young person is still lacking on his or her identity security. Adults had gone past that..
     
    Last edited: Feb 22, 2016
  15. Timur

    Timur Mu-43 Regular

    124
    Nov 10, 2015
    Timur
    Ok. The reason I was concerned in the first place, is because the sample photo from the review looked so soft (unusable soft to me), and I wanted to make sure I wouldn't make a purchase that I would regret later.
    I think if you opened the image I'm referring to, you'll see for yourself. It's not theoretical, or psychological, it's a clearly observable quality of the image. In the later posts the discussion went into educational territory, which was interesting to me personally.

    Now, I don't see asking for opinions on the forum intended for that as being insecure or immature, this is what this forum is for in my opinion, being in the community tied together in some way by owning the same camera system, sharing a hobby (or profession). Plus this particular combination (a UWA Tokina on a mft body) is not a common one, and I could not find many photos taken with it.

    I personally don't see how uneducated action like buying a lens without researching about it can be considered as an "adult decision". I appreciate the input, but I'm not sure your position of higher ground is appropriate as you do not know me and are a complete stranger on the internet, so I don't see why you think judging me is appropriate. Besides, if I follow your own advice, I should ignore what you are saying, and keep making educated purchasing decisions at my own pace.
     
  16. Timur

    Timur Mu-43 Regular

    124
    Nov 10, 2015
    Timur
    It makes sense, thank you for the excellent and to the point info. From what I gathered so far, neither option is a bad one, so no loss for me)
     
  17. Turbofrog

    Turbofrog Mu-43 Hall of Famer

    Mar 21, 2014
    You're welcome. I just noticed that I made a bit of a mistake in my math: I had included the wrong multiplication factor for APS-C. I was thinking of the 1.5x crop factor for APS-C relative to the FF, not the ~1.3x factor of M4/3 relative to APS-C. Oops! So the Olympus should indeed be a bit sharper than the Tokina, at least according to those lens tests.

    52-55*1.3 = 68-72 lp/mm, not the ~80 lp/mm values I had originally used with the incorrect 1.5x multiplication factor.
     
  18. bikerhiker

    bikerhiker Mu-43 All-Pro

    Dec 24, 2013
    Canada
    David
    Timur,

    The video I linked was meant to give you the reasons why Nikkor lens performance, while not as good on paper against others, actually perform quite well. Nikon actually provided sample tests of Company A and Company B zoom lenses. It just so happened what sometimes look good on paper may not necessarily perform well in real life as sample tests shown. Company A and B zoom lenses are actually 2 of the top performing well known lenses comparable to 24-70 Nikkor. In fact, they scored a bit better than the Nikkor in MTF tests too!

    Secondly, my comment on "adult decisions" wasn't meant for you but for the general buying public. Sometimes, internet posting can be misunderstanding without having face to face communication. Anyhow, because the general public buying decisions are heavily influenced by test charts and bloggers who had a very short time testing the lenses, they could not have discovered the benefits of the lens in a very short time nor could they tell anything about a lens simply by looking at numbers. It requires more than just a few days of testing or hours on the MTF chart to really tell what a lens could do. Which was why Nikkor put out this video. To explain that, despite the rather lackluster performance on the test bench and the unfairness of others in the blogging community putting this lens down, that the Nikkor 24-70 is a great performing lens under a wide variety of conditions. If these conditions meet yours, then you buy the lens. If company A meets your conditions better, than buy company A or B. That's the whole spiel of the video.

    Personally, I don't understand why people are so fixated in worrying about test charts. Are people buying lenses these days to take test charts? I don't think people spend a lot of money just to take test charts no?!? More than likely, you are buying a lens to take photographs of something more interesting than test charts.
    So find real samples of photographs taken by capable photographers on the lenses you want to buy and make decisions based on those. That's how I based on buying decisions on all my lens purchase. I look at real photographs taken by people and see how they look to me. If they look sharp to me, then I'll buy. Resources available here on the forum provide ample samples of the Oly 7-14 f/2.8 by some of the finest photographers I've seen and then some. To me, the samples speak for themselves.
     
  19. Timur

    Timur Mu-43 Regular

    124
    Nov 10, 2015
    Timur
    However you read it, your last paragraph seems to be aimed at criticizing me, or the "youth generation" as you put it. If you did not intend it, you certainly have not made it easy to see it that way. As I already stated above (I guess you did not read the entire thread), I've used the Nikkor 24-70 on a d800 body, and I found it exceptionally sharp, so I'm not disagreeing with you on that point. I do not see a reason to go in depth with that particular lens as it isn't even my choice of purchase, it was more of a side point tied into a general question - why MFT lenses have higher MTF numbers in charts, which was answered by a person just above. My question had nothing to do with why another FF lens would perform better to 24-70, or worse. I have a feeling that you are arguing a separate point from the ones discussed here (again, m4/3 numbers being higher than FF/apsc ones), there's a specific answer to why that is given by a person above. I understand the concept of products being designed for a specific purpose, it's just like in any other industry - IT, computer hardware, photographic equipment, fashion, even food industry.

    Reading the charts does have value to me personally, not a largest influence on the overall decision, but as an addition to the rest of the information that I'd gather about a product (in this case lens) that I'm thinking of buying. As i do not posses years of experience in the industry, and I do not know each and every lens that exists (like the lens in question - I did not know about the Tokina), reading feedback (user reports and reviews) and more technical info (charts etc) has value for me. You can easily assess if the lens is worth investing into if you research enough. Is there a reason you think I buy lenses going from MTF charts alone? Isn't that a lot to assume about a random person you've never met in person?
     
  20. Timur

    Timur Mu-43 Regular

    124
    Nov 10, 2015
    Timur
    A person in another thread promised to post some photos taken with a combo for me, so hopefully I'll be able to see the photos and make a final decision. If it's a bit softer than tripple the price olympus, it's not a deal breaker to me. It's just hopefully (and now I'm almost sure of it) its not as soft as that damn photo I posted earlier in the thread, that looked worse than any kit lens you can get for under 100$ on mft :)