Tamron SP 70-200mm f/2.8 Di VC USD G2, calibrating with the Tap-In Console


I remember when I got my first 70-200 f/2.8 lens, it was the first version of the Nikkor, and it was a fantastic lens on the 12 Mpx cameras I had back then (D300s and D700). The feel, the look, the balance and about everything was perfect. I still really really like the look of a 70-200 f/2.8 lens on a DSLR, I think it looks simply amazing. Obviously with the increase of megapixels, the older lenses didn’t keep up and needed to be replaced. In the beginning of last year, I got the Tamron SP 70-200mm f/2.8 Di VC USD G2.

To be honest, a 70-200 f/2.8 lens has its limited use because of its focal length, but when it fits, it truly shines. It’s a bit heavy, but well worth it by the quality it produces.

Equipment-0485

This post should be read together, or preferably after these posts:

Reikan Focal, focus calibration software HERE.
Nikon AF fine tune, how and why HERE.
Tamron Tap-In Console, hardware and software to alter plane of focus among other things HERE.
Tamron 24-70 f/2.8 VC G2, calibrating the lens with Tap-In Console HERE.

A short version of the above articles is something like this. Nikon has had the ability to alter the plane of focus for more than a decade in the menu system in camera. Tamron has had the ability to alter the plane of focus with their Tap-In Console for a couple of years, making changes to the lens directly. The plane of focus is an imaginary line at 90 degrees from the camera where the focus should be at its sharpest, if it’s not you need to compensate with AF fine tune or in-lens compensation. Reikan Focal is a computer software that helps you find the best value to add either in-camera or in-lens. If this doesn’t make sense to you, I advise you to read the full articles linked to above.

70-200 2

When I first got the 70-200 G2, I noticed some photos that were not in completely good focus while others were very good. I know now that it all had to do with aperture and focal lengths used in the different photos, and probably also the focus distances. As mentioned in the 24-70 G2 article, it was totally unexpected that the lens should behave like this, so after consultation with the seller, the lens was returned. The new sample behaved differently but at the same time quite similar to the first copy, as the same was for the 24-70 G2 I also returned.

As we can see from the calibration, the focus is quite all over the place at closest focus distance with massive swing from plus to minus dependent on the aperture. Not terrible bad at 100 mm, but at both ends and 135 mm it is a considerable variation.

70-200 0.95 meter

Luckily this calms down a bit at medium focus distance, but still I think the variation in both ends are pretty large. In the middle though, it’s fairly stable with little variation only.

70-200 5 meter

And finally, here are the numbers at infinity.

70-200 infinity

The reason I didn’t do more at smaller apertures, is that the curve in the Focal program is getting increasingly flatter and thereby more and more unreliable at longer focal lengths. The star (*) behind the numbers at 5 meters is that the curve that is so flat that Reikan Focal won’t give a final number, so made my best guess from the curve (should maybe have not written anything like at infinity).

A short explanation of the letters after the values. I’m using Reikan Focal software to help me determine the correct value, and Focal are using four words to tell how good or the quality of the final value is. The best is Excellent, then Good, then Acceptable, and finally Poor. If I get an Acceptable or Poor result, I normally do another run to confirm the first one, and sometimes even a couple of more runs if there is a big difference in the values and in the end calculate the average value.

Another update: I’ve been playing with the Tamron SP 15-30 f/2.8 VC lens last days, and I think I’ve come across a rule of thumb on how to calibrate at infinity, finally. Some may say it’s quite obvious, and I think I’ve been partly fooled by the name “infinity”. In this context it’s not “endless far away”, and when I finally accepted that fact, the pieces started to fall into place more easily. A common thread with all Tamron SP f/2.8 lenses I own, is that the longest reported focus distance correlate somewhat with the focal length. Take the focal length in half and replace millimeters with meters, like 15mm will be 7.5 meters, 70mm will be 35 meters and 200mm will be 100 meters. This is more or less exactly what the longest focus distance reported from the lens and camera, that are embedded onto the image file. In practical terms it means how far away is needed to get to the longest focus distance. It’s just trial and error, as the real measured distance is hardly only a guideline. Take a photo of the focus target and see what is the reported distance in the file and move accordingly if needed. I also think it’s a good idea to not move too far away, only enough and then some. Means if you exactly reach longest focus distance at a certain distance, add a few % so you are sure it is actually enough (I’ve seen reported focus distance move quite a bit between calibration runs without the actual distance have been change). Can confirm that this worked out to be the best solution (so far), get to the closest point from the target that still gives the longest focus distance at that particular focal length (just remember that it will change from focal length to focal length).

Don’t have any before and after photos, but here is one I took today that shows to focus is pretty good. Focus point in the middle of the photo.

Test-0312

The big question is obvious how to calibrate a lens like this when the values are all over the place and we can only add one value for all apertures? That’s up to the owner of the lens obviously, and for me I buy f/2.8 zooms to be used at large apertures. So, I would aim for maybe f/4 at zero with bias towards f/2.8 if the values vary too much. For the future one can hope that Tamron takes the next step and make it possible to calibrate at several apertures too.

70-200

This article is a bit shorter than the one about the 24-70 G2, but I feel it would have been too long if it was added into the 24-70 G2 article, so I recommend you to read that one too to get the full picture, and don’t forget that this is about calibrating process and not about the final performance.

Copied from the 24-70 G2 article with some minor adjustments, as I expect not all of you will read that one:

I guess when you’ve read so far you have a feeling the whole story is somewhat negative, and I agree. But remember what this is, a post about calibrating the lens, not about how it performs in daily life. So, I think this slight negative angle is justified, I think Tamron could have provided better user manual on how to do the calibration. In their tech pages online, they could also have explained the behavior of the lens at different lens designs without giving away any design secrets.

As mentioned this is not a post to talk about the performance of the lens, but I will say I really love Tamron lenses and how they make the photos look, so don’t let this post scare you away from any of the fantastic SP lenses from Tamron. I will not hesitate to buy another Tamron lens in the future.

70-200 3

Equipment in use for the calibration:

Reikan FoCal 2 calibration software on my PC.
Reikan focus chart printed in highest quality on 120g non-glossy paper for close and medium distance, printed in A2 size for infinity.
Tamron Tap-In Console and Utility.
Nikon D750.
Tamron SP 70-200 f/2.8 Di VC USD G2.
Manfrotto 475B tripod.
Kirk BH-3 ball head.
2x Dynaphos soft-boxes with total about 450 watts of low energy lights for illumination of the focus chart indoors and daylight outdoors.

NB, the values mentioned throughout this post is only valid for my lenses together with my camera, do not expect they will fit your lens and camera, they will most likely do it worse! So, do your own calibration and find your own values. I posted those values only to prove the point, not to give the final answer, and that is also why the Tap-In utility is shown empty.

70-200 1

If you have some comments or questions, so, come on, let me hear from you in the comments below!

Thanks for visiting!

Categories: GearTags: , , , , ,

10 comments

  1. I too am trying to figure out the best way to calibrate via the tap-in-console. For infinity distance, I am wondering if multiplying the longest focal length by 50, then dividing by 1000, and multiplying by 2 would work. On the 70-200 that works out to the 20 meters Tamron quoted you, and on a 24-70 it works out to 7m. Thoughts?

    Liked by 1 person

    • Hi Nathan, your thoughts are much appreciated!

      First off, I’m writing this at work, so I don’t have my camera and lenses with me to test and confirm.

      As you might have seen, I’ve just finalized the calibration of the 24-70 G2 and at infinity at 70mm I ended up with a distance of approximately 40-45 meters (did not measure at the time), which is slightly more than the rule of 50 times the focal length.
      I’m just thinking out loud here, but I think your theory will bring infinity too close. Also remember despite calibrating closest and medium focus distance is the same for all focal lengths, at infinity it will change with the focal lengths too, so we can’t use the same distance at 70, 100, 135 and 200mm. With the 24-70 G2, it looks to be quite straight forward using 50 times the focal length, but at longer focal lengths, I haven’t found a solid solution yet. I’ve tried to aim at something approximately 90 meters away at 200mm, and then without refocus, aim at something 300 meters away and scene is blurry in the viewfinder, which means in my mind that infinity at 200mm is more than 100 meters.

      I’m using, as mentioned, Reikan Focal to do the calibration, and it seems like the program is not designed to calibrate at longer focal lengths at infinity. I’ve tried using a house at 300 meters, mobile phone antenna at 800 meters, a power line mast at 2 km, but the result is all over the place. It could also bee that those are too far away? Maybe infinity isn’t so “infinity” after all? Maybe the target should be between 100 and 150 meters? Hm, maybe 60 times the focal length is a better choice? Just by writing this gives me new ideas!

      I’m also unsure how reliable the so called “Dot tune” is at such distances?

      In a way I’m glad I don’t have the Tamron 150-600 VC, although I’ve got the Nikkor 200-500 VR, which focus very good thought.

      I will crack this nut, this has haunted me for months now.

      How have you done the calibration so far? Did you measure distances? Have you repeated the calibration and is it consistent? Did you use largest aperture only?

      Like

      • I feel like I have had so much trial and error in this process it is impossible to cover it all in this message. Maybe we can have a skype call sometime to compare notes? If you’d like to, let me know via email nathanquint@gmail.com

        On 50 times focal length. I think I may be doing something wrong. When you say 50 times the focal length being 40-55 meters, I am also multiplying by 50 but getting a much smaller number. If 70mm * 50 = 3500mm, then that is only 3.5m. I always thought this sounded too close, which is partly the reason I wanted to double it. That being said, what am I doing wrong?

        As far as distance to target, I measure the distance because I couldn’t find in my EXIF data a distance to target measurement (I shoot Canon). Now the way I see it, ideally you would begin focus at a camera reported distance. The reason being, I don’t think the camera has anything else to go by. Actual distance is arbitrary if the camera doesn’t know what it is. All it has access to is what it measures the distance to be.

        As far as aperture, on the 70-200 I use f2.8 and on the 24-70 I use f3.2 since the lens appears much sharper at 3.2. I think this will give focal better data to work with, even if the depth of field is slightly wider at 3.2 than 2.8.

        It’s hard to say how consistent things have been because I keep changing variables as I refine the process.

        Liked by 1 person

  2. Thank you Nathan! This proves that two brains works better than one! Once again a solution that looks to be working, turns out to be faulty. I only multiplied with 50 and got a big number that looked awful similar to the reported maximum focus distance for the 70-200 G2 at both 70 and 200mm, where the reported distance is 35m at 70mm and 100m at 200mm, the problem was that I’ve overlooked a “0”. The quick answer would then be to multiply the answer by 10, but if memory serves me right, the maximum reported distance on the 24-70 G2, is the same at 24 and 35mm and again the same at 50 and 70mm, still it could be guideline. I really wish that Tamron could have been a little bit forthcoming with information regarding the calibration process, and I know your feeling about lots of hours of trial and errors. Unfortunately the weather will not cooperate here the next days, so I can’t go out and do more tests for a while.

    There is a lot of small programs that can read exif data, just do a search and I’m sure you will find some. To my knowledge Canon also reports focus distance.

    That the 24-70 is sharper at f/3.2 is also my experience.

    I think it would be better that we continue to write here, as I’m convinced that other struggle with this as well, and by this we are helping them too.

    How do you do you calibration? Program/method?

    Like

  3. Hi Isis & Ove,

    I have had a bit of a breakthrough, but it does not include Reikan Focal. I am not sure I even trust Reikan anymore. What I did was purchase the Datacolor SpyderLENSCAL tool, and used it as followed. I started by running Reikan Focal auto calibration at 70mm, and 1m distance from the target, over two calibration runs. The Reikan auto calibrate tool suggested a microfocus adjustment of -6 on the first run, and -8 on the next. I decided to go with a final value of -7. After that, I started a tether capture session in Lightroom, and pointed the camera at the SpyderLENSCAL target. I took an image using liveview focusing as a reference, then took an image using viewfinder focusing with the Reikan adjustment values applied. The Reikan adjustment values did not match the focusing achieved with liveview.

    I then set all adjustments to zero and began testing manually changing the in-camera micro focus adjustments. I would take an image, compare it to the live view reference image, and adjust the micro adjustment in-camera accordingly until I had a visually indistinguishable match. Once I had a perfect match, I would then repeat the focus test a couple of times to verify it was solid, and after that set the value in the lens via the Tamron tap-in tool. I would then check one more time, with the in-camera adjustments set to zero, and the adjustment applied to the lens via the tap-in tool. Via this method, I had a nearly perfect match between the image results achieved via live view focusing and the viewfinder. Reikan suggested a MFA of -7, and the datacolor spyder lenscal gave me a result of 0.

    I repeated this process on the following lenses for all calibration distances documented in the Tamron tap-in console, with the exception of infinity.

    90mm Macro G2
    35MM f1.8
    24-70 G2
    70-200 G2

    In all cases, at nearly every focal length, and focus distance, the results were better manually using the Datacolor Spyder LENSCAL tool. This was backed up by walk around tests, comparing live view shots, to viewfinder shots.

    I have invested HOURS, I am guessing 30-40 hours trying to use Reikan Focal to get good calibration results, but every time I thought I had it right, I would get some unexpected results, or the results would be inconsistent. I still want to do some more real-world testing, and figure out how to calibrate at infinity, but it finally feels like the lenses are working as they should be.

    You asked about my setup, this is my setup for using Reikan.

    Canon 5D Mark IV
    Aperture Priority mode.
    Two YN-600 Air LED lights powered by batteries.
    Either the small or large hard targets purchased from Reikan
    I measure the distance to the target with a measuring tape, the point I measure on camera is the sensor marker on the upper left of the viewfinder.

    A few notes about Reikan I want to document here.
    If you’re trying to use it to calculate for infinity at a far distance, the ambient light needs to be close to the target light. The problem is that the target takes up a small amount of the image, and since you cannot do the test in manual mode, Reikan uses spot metering to determine its exposure. Since the target is smaller than the spot, the dim background makes it into its metering calculation, and the target end up getting very overexposed.

    In general, Reikan seems very finicky, requiring perfect lighting, the viewfinder to be covered, only shooting in AV mode, you can’t print on a laser jet, it should be the right paper, hard targets preferred… Using Spyder LENSCAL was MUCH MUCH easier, and I have a lot higher confidence as I can easily see the results with my own eyes as they are happening.

    Liked by 1 person

    • That is fantastic Nathan!

      I will take a closer look at this device from Datacolor, using their calibration tool for our computer screens with good results.

      I understand your points about Reikan, but this program is not made for calibrating Tamron or Sigma lenses. At medium focus distance, like its meant to be done with first party lenses, I have a feeling it works quite well. The problem seems to be at shortest focus distance partly, and infinity especially. I’ve had some conversation with Reikan, and to my understanding they have very little experience with calibrating third party lenses.

      Regarding over exposing the target outdoors, I’ve found that dialing in some exposure compensation worked fine with me. And by the way, how does Canon adjustment steps correspond with Tamron steps, are they one to one? For Nikon you need to multiply with 1.5 or more to get it right, depends on focal length and focus distance.

      Nice set of Tamron lenses you’ve got there, how do you like the 35mm? I’m getting the Tamron SP 15-30 f/2.8 VC tomorrow, and it will be interesting if it follows its longer siblings in need of a calibration.

      Awesome that you would share you thoughts and experience here Nathan!
      Thanks!

      Like

      • When I tried to dial in exposure compensation, if I remember correctly, the program replaced my setting with its own which was close to 1 stop + exposure compensation.

        The MFA steps appear to be one to one. As I said I would make the change on the lens, and then test again to make sure the viewfinder image matched the live view image and it was good.

        I still need to figure out exactly how to calibrate for infinity. I was looking into the hyperfocal distances, but the distances are so long it seems that it would be quite hard. Have you found out anything else on this? I thought about finding a friend that speaks Japanese and having them reach directly out to Tamron Japan.

        The 35mm has made some lovely images, but at times I had focusing issues. I hope they will be resolved now.

        Liked by 1 person

  4. Hi
    I just purchased Tamron 90mm, 24-70mm G2, 70-200mm G2 and Tap-in. About the infinite adjustment I just got the answer from Tamron EU: “For adjustment for infinite you need minimum 25meter, after adjustment via tap in check it on longer distance.” Do you have found any usable solutions?
    Best
    Kimmo

    Like

    • Hi Kimmo!

      Sorry for the late reply, I saw you comment yesterday, but had no time to reply.
      To calibrate at infinity has been a real headache, and I’ve spent countless hours to try to decipher the enigma. I’ve tried several solution that looks promising, but after some test runs it doesn’t work after all.
      One working theory I’ve been trying lately is the one of reported distance in the image file. I don’t remember the exact distances now, I have have it written down at home, but its close to half the focal length “changed” into meters. Like at 70mm focal length, half is 35mm and then change it to 35 meters. Then I put up the target and start testing in the 35 meter range. I start then at about 27 meters and take a photo of the target, at the same time I drop a piece of wood to indicate exactly where, take a couple of small steps back and repeat the process until I’ve reached 38 meters or something. Then I take a look at each photo in a exif viewer to see where does the embedded reported distance reach maximum focus distance. As I’ve marked each place where I took a photo with a piece of wood, I can find the exact spot easily. If that happen to be at the third spot, I choose the forth to be sure it’s far enough from the target (but not too far away, as that seem to not work at all). If not any of the photos has the longest focus distance, I move accordingly and try again, the same if all has the longest distance, I move closer and do it all over. When this focal distance is done calibrated, move to the next and repeat…
      I hope this somehow makes sense, I feel the description is a bit poor….

      Another thing is that I use Reikan Focal as my program to determine to correct focus, and the program is made to calibrate at medium focus distance and not at infinity, so I feel it gets more and more unreliable the further away the target gets. Have a sense that the last page about calibrating lenses and using Reikan Focal is not written yet.

      Please write down your ideas, thoughts and other findings, these posts about calibration are hugely popular and the more accurate I can make them the better. Also mention your camera and how you are doing the calibration, program or method.

      Take care
      Ove

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: