Please, help with D500 back-focusing - urgent :)

[Asanoth]

Hello,
I recently acquired a D500 to use with Tamron 100 - 400. My main motivation was the crop factor - previously I used D600 and it was either a DX body, or D750 AND 150 - 600 AND a lot of heavy lifting . I went for D500 over D7500 for obvious reasons.
It was hard to test in the field, as it has been fairly cloudy, but I cannot seem to get images that are sharp when viewed at 100 % in the back LCD. Sometimes in computer it is ok. I though it was because I was not used to handle a 600 mm equivalent, but after some DPreview discussion I tried to test it for front/back focusing.

Here are the photos of my focusing test, 5 cm mark is aligned with the box's wall:
https://drive.google.com/drive/folders/1B_bNAXq8y_DAsaUFtIlecxwwRVUPlNB2?usp=sharing

They are taken at 4,5 meters only. Do you see any back focusing? Should I setup the AF Fine tune to other value than 0? Plus values are obivously incorrect at the first sight, so I did not include them.

When I tried Auto AF Fine tuning, at 6 m, I kept getting 0, or sometimes plus or minus 1. So the AF fine tune thinks it is ok. I know I should use larger distance, but this is as far as I have two walls apart.

Most importantly, I am leaving for Thailand in two days: I planned to take the D500, as this will be mostly a wildlife trip, but I do not want to come home with unsharp images. I actually thought of taking D600 for its high ISO performance for forest photography, but like this I will get more pixels on the subject anyway, so it should be better (even though I was quite surprised with the large difference between D500 and D600). Taking two bodies is not an option, I also have some diving stuff with me and am already out of kilos.

Thank you for any tips.

 

[Patrick M]

The raw files certainly do not look tack sharp. There’s not enough depth to see back focusing. Not to my eyes anyway.
Q1 were these hand held or on a tripod?
Q2 did they look sharp through the viewfinder?

why is your metering “pattern”. Can you try matrix?

gets good reviews eg https://www.kenrockwell.com/tamron/100-400mm.htm

 

[gmart7717]

The Camera Info for DSC_0691-20.jpg shows the AF FIne Tune at -20 for that lens. You may have not saved your fine tune settings correctly. I would look at that again if I were you.

 

[Asanoth]

Thank you for your replies.
If it says -20 for *-20.nef file, it should be ok?

Patrick: I had the camera set on a sturdy table. I am not able to distinguish such a detail of sharpness in the viewfinder. I think I used spot metering, as I shot against a window.
What do you mean that there is not enough depth?

I had no issues with the lens on D600.

 

[Asanoth]

Yeah... So I am way more screwed than I though I could be.

I did new set of test shots, with remote flash, so they should be sharp without motion blur and all and with more depth. I also used a tripod, not a box on the table, but it actually seems to be less stable than the box (AF fine tune did not even work). Focal plane should be at 65 cm.

D500_new - Google Drive

drive.google.com

There seems to be quite some back focus at 400 mm at 5-6 meters distance. I think the best setup is -9.
But then I took test shots at 100 mm at around 3 meters, and there is actually front focusing! So once I move Fine tune into the minus region, all my close up 100 mm images will be out of focus. I would use 100 mm mostly for landscape or low aperture macro with close-up filter, but still.
I tested also 400 mm at 3 m and there it seems the best results are at around -15 (but I only did steps by 5).

I could not believe this, so I checked with D600. I used the same distance, so the ruler seems not so sharp, as it is 1,5x smaller. Anyway, I can still see some backfocusing at 400 mm. At 100 mm, I cannot see the ruler clearly on 15in screen, but it seems there is no front focusing, unlike on D500 - how is that even possible? Ät least close up at 400 mm, I would wind up with -10 to -15 again.

So, the issue is indeed with the lens. But how come I did not see it on my D600 photos before? Is that simply so that in real life these differences do not matter?
I quickly picked a few images from my selection from trip to Kenya this year: They do not seem to be out of focus...? (not sure what rescaling will do)

You can only see EXIF info for this image if you are logged in.

You can only see EXIF info for this image if you are logged in.

This one was at 112 mm with macro filter.
The cormoran's head seems a bit blurry, but I think that is because of the rescaling and it being shown in larger proportion compared to vertical photos.

You can only see EXIF info for this image if you are logged in.

But more importantly... What should I do now?
I ordered a Tap in console, local shop nearby to where I live luckily has a few for F mount in stock (not sure they will be able to process it tomorrow though), so I could, instead of packing, eh, be fiddling with the console tomorrow: But it will be already dark at 4 pm, so I will only be able to test the focus indoors, meaning I will be only able to go till 6 m. Maybe for longer distances the depth of field is thick enough so that it does not matter.
But once I set it up tomorrow, I will fly to Thailand the next morning and there is no unsetting it throughout the trip (I am not taking a laptop, or a tripot with me). So what if I find I messed it up? I cannot field test it anymore (even though it does not seem like this is something to be messed up, I mean the ruler is there for me). Should I just take a D500 when I know it works well in real life? Or will D500 work just as well and I should not worry about it, or rely on sharpening? I mean I could definitely use the extra reach.

Damn, I did not think I would come to this 30 hours before the trip...

 

[aolander]

Zoom lenses are problematic as far as focus tuning as different focal lengths may require different focus adjustments. And, focus tuning isn't just for the lens. It's for the lens/camera combination. I wouldn't worry about it. Set it somewhere in the middle, and I doubt you'll see any major issues in real life. You can get the Tamron Tap-In Console to set different settings for different focal lengths and distances.

TAMRON | TAP-in Console

An accessory that enables firmware updates and customized setups for selected Tamron lenses

www.tamron.com

 

[Asanoth]

And, why does auto af tune keep telling me zero?
edit: written before reading aolander's reply.

 

[DRwyoming]

So, the issue is indeed with the lens. But how come I did not see it on my D600 photos before? Is that simply so that in real life these differences do not matter?
I quickly picked a few images from my selection from trip to Kenya this year: They do not seem to be out of focus...? (not sure what rescaling will do)

Some of it may be pixel peeping and seeing small problems that won't necessarily show up in print but you may also have slightly bent the lens mount or otherwise developed a problem with the lens from a bump or scrape or something else.

The kind of AF Fine Tuning numbers discussed in this thread like -9 to -20 strongly suggest an actual lens or lens/camera mount problem. A bit of AFFT isn't unusual with DSLRs but big numbers like that aren't typical for properly functioning equipment.

Also, you really want to perform AFFT at representative working distances which are likely longer than a few meters for long focal length lenses. There are some online guides as to appropriate test distances based on focal length but the best is to figure out what distances you will typically use for this lens.

One possible confounding factor is that a previous owner may have entered large correction factors into the Tamron lens via the Tap-In console mentioned above. That shouldn't be a problem if you bought the Tamron new but if it's used then I'd definitely check to see what values were entered into the lens for different working distances and how that might impact any values you're editing in the camera itself.

 

[John Navitsky]

it may be an unpopular opinion, but i'd reset the fine tune back to zero and take a breath.

yes, fine tuning is a thing with dslrs.

however, i think usually other factors are bigger factors than fine tuning.

af misses, af speed issues, atmospheric issues, etc.

do you have any other lenses that would work that you are comfortable with?

trying to iron this out right before you go is going to be stressful and not a good way to set yourself up for success. and futzing with fine tune can be a big rabbit hole to go down.

 

[DRwyoming]

it may be an unpopular opinion, but i'd reset the fine tune back to zero and take a breath.

I second that, particularly when using a Tamron lens that accepts its own fine tuning values which can really tangle things up...

 

[aolander]

it may be an unpopular opinion, but i'd reset the fine tune back to zero and take a breath.

yes, fine tuning is a thing with dslrs.

however, i think usually other factors are bigger factors than fine tuning.

af misses, af speed issues, etc.

do you have any other lenses that would work that you are comfortable with?

trying to iron this out right before you go is going to be stressful and not a good way to set yourself up for success. and futzing with fine tune can be a big rabbit hole to go down.

 

[soundbyte]

Carefully clean the camera and lens contacts, that includes the large silver ring on each.
If these are dirty, sometimes invisible dirt, (eg finger prints) can cause focusing issues due to poor electrical connections.
Use a lint free cloth with a suitable fluid (eg: glass cleaner, Deoxit) do not use harsh cleaning fluids and put the fluid on the cloth not on the lens.
Buy some lens wipes (in your camera bag?) and use those as an alternate cleaning solution.

 

[apanella]

Fine Tuning... I spent a lot of time working with the D500 on a couple different long zooms.

I'll start with "perfect is the enemy of good". But what is "good" is a question I can only answer for myself.

I can relate with the "rabbit hole sentiment . Having been-there-done-that with a D500+long zoom rabbit hole, I would suggest that tuning with multiple focal lengths (available in the "console") is the "best it can get". But, it can take hours (days?) to get things just right for a given environmental condition. Once the lens tuning was done around 25C, I would use the in-camera to add any tweaks for ~35C and ~0C environments. I would consider all of this an advanced tuning technique and would never recommend this as a last-minute exercise as there are just too many things that can go wrong.


In this case, starting with the in-camera Fine Tune is probably as far as I would feel comfortable. I would recommend tuning the lens to the most used focal length (that is not the max focal length). I see you chose 400mm. Next, I would look at images in their base ISO to get the best view. I used ISO200 for the D500. I see ISO320 was used in that test images... that is probably OK. I would also choose a distance that fills the frame for the most common target size you will see in the field.

I have never looked at pushing the exposure as was done in the test images, but I don't think it will hurt.

As others have stated, technique can easily negate any gains done with Fine Tuning!!!
600mm x 1.5 = 900mm is not the same as shooting at native 400mm, especially with high pixel density cameras like the D500. I would recommend spending as much time as possible getting the feel of the longest distances when you are only filling a quarter of the frame BEFORE you go out for that first prize winning shot. Start with static subjects working up to slow moving subjects.


Oh, one more thing...
Focus can (does) vary across the image plane (function of the optics, especially those of long zooms).
I made the choice to tune to the center of the frame (the test images show the same choice**), thus accepting that when AF locks in at the edge of the frame (possible with the D500), it may result in the center of the frame not being perfect as hoped when pixel peeping. Nothing to fret as little post processing tends to solve these smaller things.
** I probably would have put the graticules of the tape measure in the center of the frame, too.


For the pics defined, I would pick -12 or -14 for the in-camera fine tune.

 

[Johan Schmidt]

Watch this green dot AF fine tune method for Nikon DSLRs

It's very accurate - I used this before I bought Focal automation software and the values achieved with green dot method were virtually identical.

 

[Asanoth]

I have to thank you all, this forum is very responsive!

For the pics defined, I would pick -12 or -14 for the in-camera fine tune.

And did you see what -10 does to 100 mm focal length at D500?

Johan: Thank you, I will try that method. I will skip the Tap in console, as I fear I would mess things up. Depending on how the results will be, I will try to do it in camera. In the "worst" case, I will take the D600 instead.

 

[Asanoth]

I did the dot method. Results were very consistent among multiple tests:

At 6 m distance:
400 mm +8 -> +1 or +7 -> -1
200 mm - 8 -> -2 or -9 -> -1
100 mm - +5 -> -3 or +4 -> -2

At 3m distance:
400 mm +8 -> +2
200 mm -8 -> 0
100 mm +6 -> -3

In the tests before it seemed like I should dial - 10 or so for 400 mm, not + 4.

I checked my photos from Costarica last year and I have (on full frame) over 2000 photos at 400 mm, 200 at 100 mm and the rest is sporadic. I did not use macro filter back then. I cropped a lot. The story was similar in Kenya, where I did a bit of macro, but the rest was a little less sporadic. I think most of 200-390 mm were either landscapes, or large animals where the focal plain is thick. I actually did this "analysis" when choosing a new body (D750 vs D500):

It is already nearly dark here, even though the day was sunny, so I cannot field test the lens.
I will try to take some test photos inside, if the value for 400 mm makes sense compared to what I did yesterday.

 

[EricBowles]

I'm with John - I would only consider AF Fine tuning outdoors with bright light and a very good test target. Your initial tests are so far from appropriate test conditions that they are likely to cause more harm than benefit. That lens is not going to be the sharpest lens, and it's got a relatively forgiving depth of field at f/6.3. The long end of that lens is going to be slightly soft but it's still okay as long as you avoid deep crops. It's going to be hard to get good values for fine tuning across the entire range of that lens.

Distance to test target does matter. At close up distances or macro distances, traditional testing does not work. It's just too hard to precisely focus at macro DOF levels so manual focus adjustments are typical.

 

[Asanoth]

Yeah... I cannot say I can wrap my head around these levels.
I do not know if macro with close up lens really follows similar laws, as my working distance is something between 33 and 36 cm. Shooting handheld, things like AF fine tuning are a few orders of magnitude finer than what my movement does, however still I want to stay.

I just would like to remind, that I started this (or, more precisely, that other DPreview) topic because I though my images were not focused properly even on high shutter speeds and low ISO.
So, what I did, is that I took a piece of sheet music, I taped it on that 6-meters-away wall and I took photos at 100, 200 and 400 mm at -8 and +8 compensation. +8 was always worse, by 200 mm it was very blurry, by 400 mm slightly soft. I tested -4 and results were similar. So I set it at -5. It is rather heavy-handed, but it seems it delivers.
I am not sure why the dot method did not work for me: even now, when I test with -5 at 400 mm, the focus indicator is off at -5. But the photos (of the sheet music on the wall, eh) are clearly better. If somebody wants to check, the photos are in the folder shared before.

 

[MartyD]

I learned pretty quickly that I could create a problem where one didn't exist because my focus fine tuning methods were poor.

I used Reiken Focal a lot to focus fine tune my lenses when I was shooting a D500. What I learned is that Focal requires a very specific process in order to get consistent results. It was very picky on lighting, target distance, and camera configuration. It also had the ability to check focus consistency and if it found a lens that didn't provide consistent focusing results at the same settings it would warn you that this was a problem. It was very noticeable when comparing the focus consistency of my Nikkor 70-200 f2.8 with the Nikkor 70-300 P series. The only time I would change fine tune values is when Focal would show repeatable results in testing.

I also never could make a good choice trying on my own to fine tune using a flat target. I found I needed to see the front or back focusing. I used to know a sports photographer who regularly tested his telephotos by throwing a ball out onto the grass, focus on the ball at maximum aperture, then checked the depth of field in the images. Seeing the distance in front and behind the focus point helps you see any front or back focusing. I did make some focusing testing charts that were mounted on a flat service that were long enough and I could present to the lens at an appropriate angle to span the complete depth of field at lens focal length and aperture. If you can see the entire DOF in a single image or even a series of images it was easier to see how the lens was behaving.

 

[ButlerKid]

After much time spent researching how to do focus tuning, and then time spent testing and adjusting various fine tuning values at various focal lengths, I finally reached some conclusions.

The results will vary based on just how perfectly precise one can be when repeating the test time and time again.

Trying to fine tune a super voom lens was an exercise in frustration. The amount of fine tuning needed usually varied throughout the focal range.

Better to test a zoom at the focal range one most often uses that lens. In long lenses, that is usually the longest focal length.

Thus, for zooms - I gave up trying to fine tune them. In fact, I no longer buy super zooms. For me, I'm much better off with a prime lens. I do still use a mid-range zoom like the 100-400, but I don't try to do focus fine tuning on it.

 

[apanella]

For a target.... the Velo LENS-2020 Lens Calibration Tool is a handy fixture.... it helps with setting the front and/or back focus.

A reasonable DIY version of a target is at the bottom of this page:

AF Microadjustment Tips

AF Microadjustment Tips

www.the-digital-picture.com

For best results, fill at least 75% of the frame with the test target after setting the focal length.
Indeed, a bright diffused light (cloudy day) is a better condition.
Don't forget, you are probably tuning the center of the lens)

From direct experience, in-camera Fine Tuning will yield the best results at around "user_selected_nominal_focal_length" +/- 100mm operating range for long telephoto zooms.
[The docks/consoles were invented tune multiple location ranges.]

Example:
If nominal tuning is 400mm, then the best focal range was 400mm-100mm to 400mm+100mm, or stated differently,
Using a nominal 400mm in-camera Fine Tune, the working range I found acceptable was 300mm to 600mm. YMMV


I would certainly recommend a bit of practice before going on a significant photoshoot, as a little trial-and-error promotes better understanding of the margins.


The one common thread of all comments, assuming the gear is operational to factory specs, Fine Tuning is a second order effect. [technique being the first]

 

[Johan Schmidt]

I did the dot method. Results were very consistent among multiple tests:

At 6 m distance:
400 mm +8 -> +1 or +7 -> -1
200 mm - 8 -> -2 or -9 -> -1
100 mm - +5 -> -3 or +4 -> -2

At 3m distance:
400 mm +8 -> +2
200 mm -8 -> 0
100 mm +6 -> -3

In the tests before it seemed like I should dial - 10 or so for 400 mm, not + 4.

I checked my photos from Costarica last year and I have (on full frame) over 2000 photos at 400 mm, 200 at 100 mm and the rest is sporadic. I did not use macro filter back then. I cropped a lot. The story was similar in Kenya, where I did a bit of macro, but the rest was a little less sporadic. I think most of 200-390 mm were either landscapes, or large animals where the focal plain is thick. I actually did this "analysis" when choosing a new body (D750 vs D500):

It is already nearly dark here, even though the day was sunny, so I cannot field test the lens.
I will try to take some test photos inside, if the value for 400 mm makes sense compared to what I did yesterday.

Imperative the flat target is set up correctly - distance between target and camera 50x focus length in mm, bright consistent light on target, camera sensor must be same height and parallel to target, using single point AF set to the centre point, set on a tripod, back button AF.

The D500 doesn't allow you to set AF fine tuning values at both ends of the zoom (as the Z9 does), so find a good average over the focal range unless you're only using one end predominantly.

 

[Lance B]

To test a lens for AF fine tune, I use this homemade set up. Works a treat. Focus on the bar code, then you can see if the lens is back or front focusing by looking at the rule that is on a 45 Deg angle.

This was my F mount 105 f1.4E on my D850. As you can see, it was spot on.

Close up

 

[carbon]

Patrick: I had the camera set on a sturdy table. I am not able to distinguish such a detail of sharpness in the viewfinder. I think I used spot metering, as I shot against a window.

Pardon, but are you shooting through a window?

Because if you are, with riflescopes I discovered that not all house windows are created equal, and all are likely to degrade your image. Some windows shockingly so.

 

[Maljo]

Agree