Lenses and Optics

"This lens is soft" and other myths

Published December 22, 2008

One of the most common examples of anti-logic we see at LensRentals is the statement The lens is soft/front focuses/back focuses. Now don’t get me wrong, there are bad copies of lenses out there, as best we can tell ranging from 3% to 7% of lenses. And we know, despite our checkout procedures, that 1 of 400 lenses or so will be damaged in shipping and arrive not functioning. Sometimes there’s actual damage or misalignment of an element in the lens, although the vast majority of the time that’s not the case. Usually, the subject of the photograph is soft because the lens is not focusing precisely.

Three to 4 times a week we have the following conversation:

“The lens you sent me frontfocuses, it’s not good.”
“OK, we’ll overnight you a replacement.”

Then the first lens comes back and its perfectly fine when we check it out. But the customer is very happy with the replacement lens, it worked great even though the first one didn’t. So what has happened? Its rather simple, actually, and like most examples of anti-logic it stems from a wrong assumption: the customer knows his/her camera is ‘fine’ because it works with fine with their other lenses—none of them front focus or back focus.

The key to the puzzle is the definition of ‘fine’. Most people assume that ‘fine’ means ‘perfectly calibrated’. In reality, cameras are like any other manufactured item, calibration is within a given tolerance range. We don’t have privy to what the actual tolerance range Canon, Nikon, or the other manufacturers (except Zeiss and Leica) consider acceptable, so lets arbitrarily say the manufacturer will consider a camera or lens to be ‘in specifications’ if its + or – 3 ‘focus units’ from perfect. We can assume they reached this number because anything within + or – 3 focus units will be within the depth of field of a wide aperture (probably f/2.8) lens.

Let’s consider that I have a camera body that is -2 focus units from perfect, and a lens that is +2 focus units from perfect. Both are considered ‘fine’ according to the manufacturers definition, although they certainly aren’t perfect. However, the combination of a +2 lens on my -2 camera will be absolutely perfect, I’ll love the lens on my camera. After my experience with this one lens on one camera, I will write Sonnets on the various online forums about how great it is, and will tell anyone who doesn’t like it that they must be a bad photographer. I will have become the most dreaded online lifeform, a FLAO (Fanboy with Loss of All Objectivity).

But what if the lens was -2 focus units from the theoretical perfect and I put it on my -2 focus units from perfect camera? Well, it depends. If the lens is say an f/4 maximum aperture, probably not much: the depth of field from an f/4 aperture lens may well mask a bit of front focusing or back focusing. You might notice the lens frontocuses 3 feet in front of the subject at 20 feet if you pixel peep, but since the depth of field is 10 feet the subject is still in focus and the lens seems fine. I will probably describe the lens as very good, but not descend to complete FLAOdom.

But if its an f/1.4 lens with a very shallow depth of field, the front focusing will be noticeable: the subject will be out of focus and soft. If I know how to do a front/backfocus test I may have figured out the problem, but here’s the kicker: if I sent the lens in to the manufacturer to fix the problem they would check the lens out, say it was fine (because it is fine, its within specifications) and send it back. Only if I send the camera and lens together to be calibrated would the fact that the two together are out of focus be apparent, and then the manufacturer would be able to fix the calibration.

Ah, but there’s no free lunch. If the camera calibration was adjusted as part of the fix, I might find that another lens in my kit that used to be great, now backfocuses a bit. In the past, many full-time pros who were aware of these issues would send their entire collection of cameras and lenses to the manufacturer to be calibrated together. This was one of the original reasons Canon and Nikon formed their Professional Services groups. Most of the rest of us just made do or sent copy after copy of a given lens back until we got one that was sharp ON OUR CAMERA.

The bad thing is many, many people who did this then hopped on their online camera forum and made blanket statements like “I had to try 3 copies before I found one that was calibrated right”. In reality, what they should have said was “I had to try 3 copies before I found one that was calibrated right FOR MY CAMERA”. Those other two copies might well have been fine on someone else’s camera.

When you have a few dozen copies of each lens and each camera like we do, you quickly find out this is just a fact of camera reality. And the funny part of all this is the more expensive wide aperture lenses are the ones most likely to show the problem because their depth of field is so narrow and the in-focus portion of the picture is so sharp compared to the out of focus portion. That $200 f/5.6 zoom is not going to show a minor front focus problem because the depth of field is about half a mile. The $2,000 f/1.4 prime has a depth of field of a few inches and any problems are immediately evident (and the owner 10 times more invested in wanting a perfect lens).

The good news is newer cameras have taken all this into account and the fix is right at your fingertips. The following cameras all have a “lens microcalibration” feature: Canon 1D Mark III, 5DMkII, 50D; Nikon D3, D3x, D300, D700; the Pentax K20D, the Olympus E-30 and E-620, and the Sony A900. I’m surprised at how many people don’t take advantage of this feature – its a bit time consuming to do, but once done each of your lenses is locked in the camera’s memory and it will automatically compensate so that each lens is at a nearly perfect focusing plane whenever you mount it on the camera. I find the feature makes such a huge difference for most of my better lenses that I consider this feature alone makes the upgrade to one of the above bodies worthwhile.

Bad lenses (and cameras) will still exist, but the vast majority of front and back focus issues will be a thing of the past. And for those of you who don’t have this feature, we will continue, as we always have, to do our best to get you a lens that works great on your camera, even when it means sending a replacement.

Roger Cicala
More Lensrentals Articles

Addendum I recently saw the greatest real-life example of this ever, in an online forum where the poster states ’Canon’s New XX camera sucks’ (I’m eliminating names so the bots don’t pick this up and repeat it.) He goes on to say he had a body for several years, and a handpicked collection of lenses that he knew were perfect because he’d gone through several copies of each to get the sharpest one. Now he bought a new body and all his lenses sucked, and he’d now exchanged bodies twice and they still all sucked. So here is the perfect example of a person starting with a camera at the edge of tolerance, choosing through multiple selections a set of edge-of-tolerance lenses, and now generalizing that all the new bodies suck. The sad part is the new body has microfocus adjustment and he never even tried it. Just sent copy after copy back to the store.

Author: Roger Cicala

I’m Roger and I am the founder of Lensrentals.com. Hailed as one of the optic nerds here, I enjoy shooting collimated light through 30X microscope objectives in my spare time. When I do take real pictures I like using something different: a Medium format, or Pentax K1, or a Sony RX1R.

Posted in Lenses and Optics
  • I don’t ban people from this forum unless they engage in personal attacks, which you certainly have not.
    That being said, on-sensor PDAF that MILC uses does not require a separate PDAF sensor. On-sensor PDAF requires far less adjustment and on-sensor contrast detection AF virtually none. Remember, this is an article from 2008; in those days SLR with separate sensor PDAF was what existed. Things are better today, as they should be after a decade of advancement.

  • Dimy

    Of course you must be right because you are the specialist but its unnecessary complicating my above explained procedures. MILC also have PDAF btw but i’m not here to stroke ego. The result is the same. You may ban me from this forum if i am wrong.

  • You are correct about MILC. You are not correct about SLRs. SLRs don’t use contrast detection; they use phase detection which is a completely different thing requiring several levels of calibration.

  • Dimy

    I find all this talk about front and back focus very disturbing.
    The AF system of a DSLR is screwed down in the bottom half under the mirror and mechanically adjusted so the distance from the mirror is the same length of the light later to the sensor. The AF moves the focus of the the lens until the contrast is highest. No matter what the tolerance of the lens is.
    On MILCs the AF sensor is on exact the same plane as the sensor so it doesnt have to be adjusted at all.
    To me this all sound like maskara to serve the ego of bad fotographers and keep up service points?!

  • Mind’s Eye

    Read the manual?

  • Mind’s Eye

    It can be calibrated, but so you discovered, the micro+AF is not a linear linear. What most people do determine in their sons the today length or 2 to calibrate. Like everything he l in life, is a series of compromised and I’ll you can determine which focal lengths matter to you.

  • Mind’s Eye

    Correct, Micro-AF adjustments are not linear.

  • Mind’s Eye

    Manufacturing and physical realities dictate that it is impossible to make so bodies and lenses have a 0 offset. Hope is the purchase of real estate in the land of disappointment, it something just toA Roger’s quote

  • Mind’s Eye

    There’s no such thing as being about to find time every single device to the same exact perfect “10”. It’s just a matter of manufacturing and physical capability and all products are manufactured to with a certain tolerance of the ideal.

  • Mind’s Eye

    Rather than idiots, keys call them ignorant. While ignorant may have a negative connotation, the denotation is just unaware of information. Being unaware of a fairly technical soft & process isn’t a sign of idiocy. Better to be inclusive of those who need to learn this information than call them names.

  • Mind’s Eye

    No, best to reread the sticky and try to understand how different offsets from the ideal but with tolerances interact.

  • Mind’s Eye

    No, this isn’t how it works. You calibrate the lens on your cameras just once. The cameras saves the necessary information inside of the body so every subsequent to time you attach that lens it loads the calibration data. You do *not* need to recalibrate your body EVERYTIME you change lenses.

    So your doubt that Cabin and Nikon operate this way was accurate, just not your understanding of the micro-AF assistant process.

    I hope this clears up your misunderstanding and helps you move forward with making better photographs.

  • Eugene Powers

    The very first Nikon D800e, 16-35mm and 24-120mm I got all needed huge adjustment. On one of them even 20 would not help. And these are F4 lenses. So I personally delivered them (camera and lenses) to Nikon LA. After the fix tech support supervisor told me that camera and both lenses were way off. He also said that the body was adjusted within +/- 5. That is standard for Nikon bodies. I kept that particular body for a very long time and adjusted all my Sigma lenses to that particular body.

  • John, I don’t have any science behind it, but we rarely see a lens that needs more than a +/- 10 and I’m going to at least send all of those off to be checked. I see 6 and under very frequently just moving a lens from camera to camera, so I’m comfortable with that. But in between 6 and 10 is kind of gray, I think.

  • John Hernlund

    Hi Roger, I suppose that we are interpreting the statement slightly differently…that’s OK.

    In your blog post you’re talking about shift errors of just several units (+/- 2 or 3-ish in the AF Fine-Tune correction). Certainly such small shift errors would not be visible on film SLRs, we only see them on newer high resolution digital sensors. In such cases we can easily conclude that technology is out-pacing tolerances for phase detect AF systems.

    However, I’ve seen several lenses recently with much larger shift errors, requiring shift unit corrections greater than +/-10 (some almost as large as +/-20). Such a large error would most certainly be visible on film, and is not simply a matter of tolerances being outpaced by technology. In such a case we can say that tolerances of modern lenses are poorer than tolerances of lenses in the past (for which such issues were extremely rare).

    In any case, one has to decide when AF shift errors are too large to be acceptable. I could live with a lens that only requires +/- a couple units. But one that requires +/- ten or more units?

    Where do think we should draw the line?

  • OK, John, we know of that one for many years. Requiring a bit of AF fine tune is not the same as a focus problem. But anyone who wishes can, of course, send their lens and camera in and Nikon will adjust in the firmware. This is necessary when a zoom lens focuses differently at different zoom ranges.

    But having hundreds of cameras and thousands of lenses, all checked by factory service, there is enough variation that I can assure a given good lens may require a bit of AF adjustment on one given good camera but not another. Sending them in to the factory doesn’t change that.

    There is not now, nor has there ever been, a statement by any manufacturer that “any lens that requires AF fine tune should be sent in for servicing”.


  • John Hernlund

    There are many references on their multiple websites, this one is the most straightforward:

    “AF tuning is not recommended in most situations and may interfere with normal focus; use only when required…Service note: If your lens has a focus problem you should return it to Nikon Service as AF Fine-Tune is not intended to solve optical problems which will generally be outside of scope for this tool. If you have a focus problem which is always present with different lenses then this would indicate a camera setting issue or that the camera has received impact damage which is causing general defocus problems. Again return it to service.”

    See https://support.nikonusa.com/app/answers/detail/a_id/18105/~/how-to-use-the-af-fine-tune-function

    Photographers tend to differ on their response to this issue. Some refuse to keep a lens that requires any AF Fine-Tune. Others simply accept it and go on using flawed lenses. Those who choose acceptance do so even when they know Nikon says focus problems are generally indicative of more serious optical problems that cannot be fully corrected by AF Fine-Tune. And they are also aware that Nikon doesn’t recommend using AF Fine-Tune, saying that any lens requiring it should be returned for service by Nikon. But they think it is just the way things are, that manufacturers can’t make tolerances tight enough to satisfy modern digital sensor resolution. I disagree, of course, since companies like Nikon and Canon also make very high end industrial optics for tasks such as micro-chip fabrication, which have vastly more demanding tolerances than any DSLR. They can do it…they want to do it. But if customers don’t return faulty equipment and put pressure on them, then of course it is much less likely that they will fix these problems.

  • John, would you reference that please? That is not what Nikon has stated in the U. S., and not what Nikon Factory Service has stated to us ever.

Follow on Feedly