EOS RP front focus with EF lenses - Please CANON add MFA to RP!

My RP has front focuses with EF 50 1.4 and EF 85 1.8.

In real world i try to use this lenses in 3 to 6 meter range and beyond.

I cannot make a sharp picture at this range, even stopped down to f4. (Always front focus.)  With manual focus all two lens is sharp.

I do focus test on tripod with focus target at 3 meters.

All two lenses consistently front focusing regardless of focusing mode (spot, 1 point).

With EOS utility live view shooting is easily correctable with add three manual focus step after AF. But in RP body there is no option for micro focus adjustment. (On my old 50d is easily correctable with mfa in body.)

I have RF 35 f1.8. It is always focus perfectly, Eye focus, servo AF, always 100% focus hit rate.

Other manufacturers have a micro focus adjustment (for older lenses) in their MILC body (Sony A7 series, Nikon Z, Olympus  e-m1II has it.)

Please Canon enable it to RP too!

EF 50 1.4 at 3 meter wide open:

EF 85 1.8 at 3 meter wide open:

I thought mirrorless did auto 'MFA' every time it focuses. It's now redundant technology as far as I am aware.

I'm going to be following this thread. I keep reading that mirrorless doesn't need MFA because it doesn't have a separate focusing sensor. But what if the lens is the problem, not the camera? How do you account for that? Also, Nikon has MFA/Focus Fine Tune in their new Z bodies (not sure about Sony), is Canon doing something different so that it's not needed?

I quickly tried to replicate this issue. But (fortunately) my EF 85/1.8 is accurate with my RP.
Here's my first testshot. The focus point is about 2 stripes below the 0. Apologies for not being 100% accurate, but doing these kind of tests take up so much time if you want it to be 100% accurate and I wanted to do this quick and fast. Other than this, everything was fine: sturdy tripod, zoomed in at 10x magnification on the focus point, Canon mount adapter, newest firmware, level was 'on', 10 second timer, continuous led light, used a proper calibrator tool, etc.

image is SOOC. Shot at f/1.8 so RSA / focus shift is not a problem.

Looks like you have a rare problem, but maybe you could try again with better lighting and move a bit closer to your calibration chart. If it's still not good, I wish you all the best with getting it fixed.

gium wrote:

I quickly tried to replicate this issue. But (fortunately) my EF 85/1.8 is accurate with my RP.
Here's my first testshot. The focus point is about 2 stripes below the 0. Apologies for not being 100% accurate, but doing these kind of tests take up so much time if you want it to be 100% accurate and I wanted to do this quick and fast. Other than this, everything was fine: sturdy tripod, zoomed in at 10x magnification on the focus point, Canon mount adapter, newest firmware, level was 'on', 10 second timer, continuous led light, used a proper calibrator tool, etc.

image is SOOC. Shot at f/1.8 so RSA / focus shift is not a problem.

Looks like you have a rare problem, but maybe you could try again with better lighting and move a bit closer to your calibration chart. If it's still not good, I wish you all the best with getting it fixed.

It is weird that their 50 f/1.4 couldn't focus with an on sensor focus system.  My 50 f/1.4 focuses great on the R.  It used to front focus like in the OP's pictures when mounted on my T5i.

Some other problem is likely.  Fingers on the focusing ring?  Hand held?

Sony needs AFMA as it has lens adapters that use msemi-sivered mirrors for AF integration of older lenses.

Both Nikon and Olympus use PDAF and both have an inventory of legacy lenses that may need this help.

Fuji has PDAF but does not have AFMA ans it has no legacy lenses.

Panasonic uses CDAF, which, for all the criticism it gets, does not need AFMA as it uses direct measuring of the image.

Canon uses Dual Pixel AF which is a variant of PD.

It may turn out that all the lenses have an AF profile that is uniform for all .

We actually don't know what caused the issue and thus, like me, are speculating.

gium wrote:

I quickly tried to replicate this issue. But (fortunately) my EF 85/1.8 is accurate with my RP.
Here's my first testshot. The focus point is about 2 stripes below the 0. Apologies for not being 100% accurate, but doing these kind of tests take up so much time if you want it to be 100% accurate and I wanted to do this quick and fast. Other than this, everything was fine: sturdy tripod, zoomed in at 10x magnification on the focus point, Canon mount adapter, newest firmware, level was 'on', 10 second timer, continuous led light, used a proper calibrator tool, etc.

image is SOOC. Shot at f/1.8 so RSA / focus shift is not a problem.

Looks like you have a rare problem, but maybe you could try again with better lighting and move a bit closer to your calibration chart. If it's still not good, I wish you all the best with getting it fixed.

Thanks for the quick test!!!

My 85 1.8 in nfd (1 meter) distance almost good, but i want use it interiors with 3 meters or more distance.

Can you try a test shot from minimum 3 meters please?

kt_kt wrote:

I thought mirrorless did auto 'MFA' every time it focuses. It's now redundant technology as far as I am aware.

No all mirrorles uses CDAF cycle, CANON EOS R, RP and M is PDAF only.

Dual Pixel PDAF works well with new lenses (RF) that desinged to it. But older lenses sometimes not.

For example Panasonics-s are  CDAF only (and because this always 100% accurate in single shot mode with not moving subjects).

jalywol wrote:

Four suggestions:

1). Make sure the lens and camera firmware are up to date.

2) Try a different EF-RF adapter

3) Clean the electrical contacts on lenses, camera, and adapter.

4) If none of the above work, send in camera and adapter for repair.

-J

Thanks, of course possible that the body or adapter is buggy, but there is a confirmed front focus issue for the new RF 70-200 too.:

https://www.dpreview.com/forums/thread/4447512

J A C S wrote:

This is probably due to focus shift.

No focus shift. All test shot wide open. Focus shift occurs only stopping down the apperture.

realmadeira wrote:

I don't understand this. Sometimes I feel I am too old because I still remember that mirrorless camera is focussing right on the sensor and MFA is gone with SLR technology

Yes the focusing right on the sensor. This part is solved the half of the focusing problems from the dslr times.

But the second half of focusing problems comes from the focusing mechanism in lens built.

Canon dual pixel af is 100% PDAF in one shot focusing mode.

This means (simplified):

1. Body metering focus distance on sensor with phase detect points.
(It is 100% accurate, because metering on sensor.)

2. Tell to lens, drive the focus to metered distance.

3. The lens pull the focus tho desired distance, in one step.
(It is heavily depend on focus mechanism on lens, internal distance decoder, manufacturing tolerances etc. In older lenses it is not every time accurate.)
This is why other manufacturers enable MFA on theirs MILC.

4. The lens tell to body, that focus in the desired distance.
(And if lens focusing drive is not perefect, in real, the lens is not in focus.)
When you second time press the focus button the camera, the process comes again. But the lens "thinks", he is in good focus position, and report it back.

The only real solution is the full CDAF, or hybrid (PDAF after CDAF) cycle,

But the manufacturers rare using cdaf after pdaf, before its slowing down the focusing process. In this cases help the micro focus adjustment (if available...).

kt_kt wrote:

I thought mirrorless did auto 'MFA' every time it focuses. It's now redundant technology as far as I am aware.

Not auto MFA. It focuses on the sensor thus not requiring it. Just like LV on a DSLR. When using a DSLR normally is uses the mirror box assembly for AF and that has variation, thus the need for MFA.

R. Slaughter wrote:

This makes NO sense to me AND it worries me a bit as the primary reason I went mirrorless was due to focus frustrations with long lenses on my 7Dii and 5Div. I desparately need to take the time to make MFA to both of them on my Sig150-600 ; EF70-300L. (They’re both perfect with the 100-400ii) but the promise of the R never needing MFA (it’s effectively monitoring the final image adjustment during AF, it can’t miss, right?)

Originaly I was buy RP for this reason too.

There is lot articles how dual pixel af works. In here an interview with Canon engineers from this. They tell R focusing phase detect only, and "enough" accurate.

https://www.imaging-resource.com/news/2018/09/15/canon-eos-r-qa-with-the-canon-engineers

In first round i think, my lens(es) is simple a bad sample. (But all this is a speculation.)

On my old 50d my 50 1.4 and 85 1.8 needs AF micro adjustment in pdaf mode. In Live View it was 100% accurate (50d LV is cdaf only).

On the first canon milc (eos m - hybrid pdaf after cdaf) 100% accurate.,

Now on RP - both - bad luck.

Maybe it's an little incompatibility with my lenses, or my lenses simple buggy (or the ef adapter). This lenses are very old constructions, EF 50 1.4 is introduced 26 years ago...

My RP with RF 35 1.8 perfect focusing. I don't think its a body issue.