D750 : focus adjustment needed?

Petruska said:

So let me get this right. You won't use Dot Tune, because the suggested target doesn't meet your standards.

Click to expand...

You are saying things I did not say, so you got it wrong.

I use another target part because I have it loaded on my computer ready for printing. Some other forum users have tested it and found nothing performs better. I am happy it does not fall into one of the categories that might cause an AF issue. In some circumstances it can fool contrast detect AF

If anybody comes up with what might be a more reliable test target I will try it and adopt it if it is better.

I consider the dot tune system as slightly inference as it relies on the reasonably accurate (as distinct from 100% accurate) focus confirmation light as a starting point. The system I use is based on the highest resolved image i.e. best focus.

So I put the camera on a tripod, with the 50mm/1.8. Tooks shots at F1.8, F2.8 and F4. Posted the F1.8 below. First one is PDAF, second one is liveview for each aperture. So does this need adjustments? Feel free to correct me if I'm doing this the wrong way (maybe need to get closer?)

(will do the same for the 24-120 later)


50mm1.8 - F1.8 - PDAF - center focus AF


50mm1.8 - Live View - small square center

So did some additional testing, both with the 50mm and 24-120. The 50mm seems to need about +6 and the 24-120 needs about -4 at the zoom end and -8 at the wide end, so I set it to -6.

Hans von der Crone said:

So I put the camera on a tripod, with the 50mm/1.8. Tooks shots at F1.8, F2.8 and F4. Posted the F1.8 below. First one is PDAF, second one is liveview for each aperture. So does this need adjustments? Feel free to correct me if I'm doing this the wrong way (maybe need to get closer?)

(will do the same for the 24-120 later)

Click to expand...

I don't subscribe to the "only one target will ever work" theory some people do, or else our cameras would never work in real life. However, Your AF target has to be flat and parallel to you camera (back). Your "ruler" can be angled as in your photos so you can see what's going on. But your target HAS TO BE PARALLEL to your camera.

Hans von der Crone said:

Feel free to correct me if I'm doing this the wrong way.

Click to expand...

You are not doing it right

Chuck Westfall is the Number 1 technical person at Canon USA. He does not think much of the type of test target you are using.

Chuck Westfall “I recommend using a flat, detailed target parallel to the focal plane. After reading through the PDF linked from your message, it appears that the author has missed a major point, i.e., any individual focusing point in a digital SLR is much longer than the simple line he is using on his chart. The nature of the AF sensors used by EOS digital SLR’s as well as those from other manufacturers is that they perform most reliably when the entire length of the focusing area sees readable detail. This condition is not satisfied by a thin line on a piece of paper. It's OK to include an angled chart in a test photo. In fact, Canon Factory Service Centers always do this. But the test target is always separate from the angled chart, and parallel to the camera's focal plane”.

The pdf link Chuck is referring to is similar to your target.

Whether your lenses need adjusting can only be determined if you test to a higher standard. In particular you need a type of target that does not sometimes cause a focus error that is not caused by a camera body or lens fault.

Got a link or some more info? Picking up a 70-200/4 today and if I find that is not as sharp as expeted, I will send all my stuff to Nikon to calibrate. Then I can be sure all is ok. Did some more af tests with the 50/1.8 and need to set it to +6 it seems. Getting pretty sharp shots with this lens a bit openened up. Still kinda annoying these adjustments are still needed in 2015 (not Nikon related. general slr stuff)

Hans vdC said:

Got a link or some more info

Click to expand...

The advice from Canon, which still stands as basic phase detect technology has not changed, was on a subscription forum about 8 years ago.

More information is in your camera instruction book. This says AF might not perform well with some types of subject, and when AF does not perform well consider switching to a different subject (at the same distance) or manual focus.

While I do not say no new equipment comes with a defect, over 99% of those who take the trouble to post an image when complaining about AF focus accuracy use a subject which Nikon (and Canon) caution might result focus error.

The vast majority of complaints are likely to be unfounded based on the expectation camera AF should be perfect in conditions where the manufacturers say AF might not work well.

This is not the same as saying some cameras that do not perform well with poor targets might not also fail to perform well with a good target. On the other hand posted photos supporting complaints using a manufactures guidance good target are rare. It seems most complaints would be withdrawn if equipment is retested to a good enough standard to confirm whether or not there is a fault.

Leonard Shepherd said:

While I do not say no new equipment comes with a defect, over 99% of those who take the trouble to post an image when complaining about AF focus accuracy use a subject which Nikon (and Canon) caution might result focus error.

Click to expand...

Small correction: which YOU caution may cause a focus error.

Up till now you have failed 100% of the time to either logically explain, or to back up your assertion with measurement data.

Marc Heijligers said:

Leonard Shepherd said:

While I do not say no new equipment comes with a defect, over 99% of those who take the trouble to post an image when complaining about AF focus accuracy use a subject which Nikon (and Canon) caution might result focus error.

Click to expand...

Small correction: which YOU caution may cause a focus error.

Click to expand...

A small but important correction - whatever Nikon DSLR are you own, Nikon say AF does not perform well with 6 types of AF subject in the instruction book.

I normally only have to comment on images that fall into one of the 6 Nikon categories - because it seems extremely rare to see an image that does not fall into one of the 6 categories.

I regularly make the point Nikon is a little harsh in implying AF never works well with one of the 6 categories and they might be better saying might not rather than does not.

Maybe the rarity of a posted image not of one of the six categories mean Nikon know more about AF performance than you do, unless you are saying Nikon are giving bad advice when giving guidance on getting good results using AF.

Leonard Shepherd said:

Marc Heijligers said:

Leonard Shepherd said:

While I do not say no new equipment comes with a defect, over 99% of those who take the trouble to post an image when complaining about AF focus accuracy use a subject which Nikon (and Canon) caution might result focus error.

Click to expand...

Small correction: which YOU caution may cause a focus error.

Click to expand...

I normally only have to comment on images that fall into one of the 6 Nikon categories - because it seems extremely rare to see an image that does not fall into one of the 6 categories.

Click to expand...

Whether an image falls into a category is a matter of interpretation. A symmetric subject does for instance not necessarily mean it is exposed symmetrically to the AF sensor. You need to understand the implementation before you can conclude from an image whether AF will fail.

THerefore, it is better to do some practical tests, e.g. like http://www.dpreview.com/forums/post/51284996

There you can see that the images that you propose as very good, good, or bad, all result in the same extremely good AF performance. I.e., in practice your concerns have no solid ground, and your interpretation whether an image falls in one of those 6 categories can be proven to be wrong by a test that every amateur can easily repeat.

Leonard Shepherd said:

Maybe the rarity of a posted image not of one of the six categories mean Nikon know more about AF performance than you do, unless you are saying Nikon are giving bad advice when giving guidance on getting good results using AF.

Click to expand...

Nikon gives perfect advice. The claim made is that your interpretation of what images fall in the 6 categories is flawed. It is shown by practical results. Nikon may sound harsh in your opinion (I don;t think so, it is rational stuff, no emotional quote), in practice their cameras are doing a tremendous good job regarding AF. It's a pity you are not recognizing the quality of their AF system in practical situations, and don't understand the extreme cases in which an image should be before one of the 6 examples apply. You talk in terms that 1 grain of salt can destroy a full pizza.

Focus perturbations are always there, whatever target you use. The small deviations resulting from most targets posted here, make them perfectly suitable to test AF performance. Your interpretation of the 6 cases reported by Nikon is not going to change that.

Again, it is your interpretation, not Nikon's truth as you always seem to claim.

Kaso said:

Kaso said:

Strange thing is my mirrorless seems sharp enough at 1/60sec for same subject with the 16-70/F4.

Click to expand...

We're you referring to your Sony a6000? It's an APS-C camera.

A friend of mine moved from a 1/1.5" 10MP PASM P&S camera (advanced in its class) to an APS-C (DX) 16MP D5100. He kept complaining during the first few weeks of using the D5100 that images used to be "so sharp" with that less expensive camera!

Click to expand...

I still say my Canon 70d is much sharper when you zoom in and pixel peep. I think it has to do with pixel density on the smaller sensor. I thought part of my problem was too slow shutter speed too. The other day I did headshots indoors at shutter speed of 160 on my 100mm lens and kept getting very soft images. I was a good 10 feet away shooting at f2.8 and still just ugh. 1 out of 5 were sharp. I was also shooting around 400ISO so its not noise.

Bottom line, I'm having a hard time with the term "tack sharp" and my D750 and I have Auto Tuned every one of my lenses.

dfwphoto said:

Kaso said:

dfwphoto said:

Strange thing is my mirrorless seems sharp enough at 1/60sec for same subject with the 16-70/F4.

Click to expand...

We're you referring to your Sony a6000? It's an APS-C camera.

A friend of mine moved from a 1/1.5" 10MP PASM P&S camera (advanced in its class) to an APS-C (DX) 16MP D5100. He kept complaining during the first few weeks of using the D5100 that images used to be "so sharp" with that less expensive camera!

Click to expand...

I still say my Canon 70d is much sharper when you zoom in and pixel peep. I think it has to do with pixel density on the smaller sensor. I thought part of my problem was too slow shutter speed too. The other day I did headshots indoors at shutter speed of 160 on my 100mm lens and kept getting very soft images. I was a good 10 feet away shooting at f2.8 and still just ugh. 1 out of 5 were sharp. I was also shooting around 400ISO so its not noise.

Bottom line, I'm having a hard time with the term "tack sharp" and my D750 and I have Auto Tuned every one of my lenses.

Click to expand...

My humble suggestion.

I believe that Canon and Nikon JPEG engines output different levels of perceived sharpness.

My own opinion, is that this comes out markedly in sensor processing tests. Where the JPEG and then the RAW's are compared. Canon's JPEG engine seems to give more punchy, whereas when the RAW's are examined, very often, the Nikon RAW is close or even better, in resolution and contrast.

I use a D300, I don't have a D750 and am now awaiting my newly acquired D610.

But, my own RAW workflow, yields substantially better output than the in camera JPEGs. Really, I think the out of camera JPEG rendering is crap. Although, C1 has had six years to get to C1 7 & 8 algorithms improved since Nikon issued the D300 with its then software.

This is without haloing or other severe artifacts. Sure, sometimes the contrats etc may be pumped or whatever, but it is not with artifacts and only a matter of interpretation.

But what I am saying is that a good RAW workflow will make all the difference, other things being equal (AF fine tune etc).

As an example, someone who shoots the same event regularly asked me, and surprised me, why my images were sharper than hers.

I was surprised a bit, but all I could say was that I use C1 Pro for my images. But truly, whatever software you are comfortable with, whether C1, Lightroom, ACDsee etc, should yield better results than the JPEG's.

By the way, I use a D300 and older Tokina 28-70 2.8. The lady uses a D600 + 24-70 2.8 AFS.

So, her equipment is at least comparable.

So, try a RAW workflow and give it a few months and see what you think.

If you already do use a RAW workflow, then I really do not know what to say.

Regards

--
Wishing You Good Light.

Thx for the advice and you're pretty much right. Even after tweaking it a bit, the jpg's from the camera look pretty bad to me. But I am getting used to the raw files of the D750, adjusted my sharpening quite a bit compared to what I did with my 5D3 and a6000 and I'm getting pretty decent results now.

It's good that you have the patience to keep correcting this, so others don't get misinformation.

A while ago I downloaded some of the target Leonard claimed to be poor and got nearly 100% excellent results with AF using center and side focus points, in both good and poor light.

Hans von der Crone said:

Thx for the advice and you're pretty much right. Even after tweaking it a bit, the jpg's from the camera look pretty bad to me. But I am getting used to the raw files of the D750, adjusted my sharpening quite a bit compared to what I did with my 5D3 and a6000 and I'm getting pretty decent results now.

Click to expand...

Glad you have somewhat resolved the issue.

One other suggestion. I believe that color profiles and contrast settings also impact perceived sharpness.

This possibly because good color profiles ensure no color bleed, or inappropriate balance in tones and saturation between colors, and contrast because that reflects the difference between edges as well.

So, I suggest, that if you have not yet done so, whichever software you use, see if it can make custom color profiles for your camera. Lightroom and Capture One can, though Capture One profiles I think are pretty good out of the box, but I tweaked mine for the D300.

For this you will likely need either a Colorchecker Passport or QP Card. Both of which may require a bit of learning the process to do so. (I don't use the 'proper' method, I use one for C1 Pro... works for me ;-) ).

http://www.amazon.com/X-Rite-MSCCPP...1166512&sr=8-1&keywords=colorchecker+passport

http://www.qpcard.com/

Also, contrast settings in your software conversions will have relevance, I believe.

Hope this makes some difference. Some may think it is a lot to do to get things right. But after the first time, it is pretty simple and also, once done, you don't have to go back to it a lot, unless say, you shoot weddings or food professionally , where you may be fairly particular on colors and tones and use the color cards for a few sample images at each event / venue, I guess (for those events a white balance card will probably be a must anyway).

Best Wishes.

--
Wishing You Good Light.