SL2 Questions

[Tragg]

Advertisement (gone after registration)

I've skimmed a few reviews (will read in depth when I get a free moment) and find that they disagree on several fundamental points. So here are a few questions: To what extent are issues with the autofocus system inherent in the camera design or a product of the firmware? Secondly, what are pros/cons of using M lenses with this camera. Thanks.

[scott kirkpatrick]

AF opinions vary a lot, but the  center of what you will read is:

PDAF (Phase detect) is the best for continuous autofocus, because it uses special sensors on the chip (which replace image pixels) to predict the degree by which an object is out of focus and the direction in which to move he lens.  The cost is that those missing pixels sometimes are visible as bands or artifacts.

CDAF (contrast detect) just adjusts to get the sharpest focus, but may start out in the wrong direction and have to correct.  This fear is overblown, since there are software tricks for keeping track of what is in focus, and making tracking corrections outside of the time that the shutter is half-pressed.  So some impressive things (not everyone agrees) can be done with CDAF in firmware, and usually are.

The best place for an M lens is on an M camera.  The second best is on an SL or CL.  There is no third best.

Edited November 16, 2019 by scott kirkpatrick

[Jeff S]

3 minutes ago, scott kirkpatrick said:

The best place for an M lens is on an M camera.  The second best is on an SL or CL.  There is no third best.

Mostly true (wide angle lenses particularly) if user technique is adequate to realize potential IQ. But for someone with bad vision (especially using fast lenses wide open for shallow DOF) and/or unsteady hands, a camera with better EVF (with good focus aids and no blackout), or with IBIS, might be beneficial.  The SL2, with IBIS, might provide more benefit in this regard than even the SL, and potentially expanding lowlight handholding ability at slow shutter speeds.  Tools differ in more ways than one. Just because a sensor is optimized for a lens doesn’t mean the user is a similar match.

Jeff

[jrp]

Yes,. Apart from IBIS, it also depends on whether your subject is always central or not.

[Jk1002]

I am still curious about the viewfinder issues caused by the autofocus. Some reviews picked up on that, with continues AF I think, describing a wobble where the lens goes back and forth to achieve focus causing the image to blur.

DPreview mentioned but also PCmag. 

I get to try one out early December on one of those Leica store Tour stops afterwards I make up my mind.

[SrMi]

9 minutes ago, Jk1002 said:

I am still curious about the viewfinder issues caused by the autofocus. Some reviews picked up on that, with continues AF I think, describing a wobble where the lens goes back and forth to achieve focus causing the image to blur.

DPreview mentioned but also PCmag. 

I get to try one out early December on one of those Leica store Tour stops afterwards I make up my mind.

My understanding is that there are two different issues that some of the reviewers are complaining about:
- In AF-C, and especially with stationary subjects, the AF wobbles slightly and does not stabilize.
- The resolution of EVF drops while focusing (also mostly visible with AF-C, as focusing last longer)
I expect to use my SL2 in AF-S mostly.

 

Edited November 16, 2019 by SrMi

[Mr.Q]

Advertisement (gone after registration)

You need to determine if you track objects (in stills or video) using AF-C often. If not, the SL2 will be one of the faster AF-S cameras on the market.

I'd have to disagree with Scott regarding CDAF vs PDAF.  Panasonic has been working for years to perfect their DFD technology and it still sucks at tracking. Unreliable with a lot of wobbling. That's with MFT with a much deeper DOF.  It doesn't take a rocket scientist to realize it would likely be worse with a FF sensor.

As for the reviews, pay attention to the images taken with the camera. If the subject matter is mostly stationary, the reviewer will not complain about the AF performance. I trust Chris and Jordan over at dpreview on this subject. They shoot a wide variety of subjects and provide very thorough reviews.

[sillbeers15]

6 hours ago, Mr.Q said:

You need to determine if you track objects (in stills or video) using AF-C often. If not, the SL2 will be one of the faster AF-S cameras on the market.

I'd have to disagree with Scott regarding CDAF vs PDAF.  Panasonic has been working for years to perfect their DFD technology and it still sucks at tracking. Unreliable with a lot of wobbling. That's with MFT with a much deeper DOF.  It doesn't take a rocket scientist to realize it would likely be worse with a FF sensor.

As for the reviews, pay attention to the images taken with the camera. If the subject matter is mostly stationary, the reviewer will not complain about the AF performance. I trust Chris and Jordan over at dpreview on this subject. They shoot a wide variety of subjects and provide very thorough reviews.

Does PDAF equipped cameras provide 100% AF hit rate at all times?

[Mr.Q]

1 hour ago, sillbeers15 said:

Does PDAF equipped cameras provide 100% AF hit rate at all times?

No, night and day compared to DFD though.

[sillbeers15]

31 minutes ago, Mr.Q said:

No, night and day compared to DFD though.

'Day & Night' can be just 10 minutes difference in every evening in a 24 hour day.

You're a mature adult. Kindly be specific in arguments rather than throwing an emotional one.

Let's move on.

[Mr.Q]

Not emotional at all, you quoted me so I responded. I cannot fathom why you feel so strongly about this subject. I wonder, do you have experience with one of the latest hybrid AF cameras?

[sillbeers15]

2 hours ago, Mr.Q said:

Not emotional at all, you quoted me so I responded. I cannot fathom why you feel so strongly about this subject. I wonder, do you have experience with one of the latest hybrid AF cameras?

Yes I have. If I wanted it I would have bought it. Have you extensively handled the SL2 with at least 100 shots to evaluate the good / bad of the AF before making sweeping comments? I have handled it, but not long enough to make any conclusion. If I were like you, I would have reported to the world that I know everything and end up misleading folks. That’s where we differ. I only believe what I have seen and experienced. I only believe data and not hearsay. Please keep your shallow believe from YouTubers like Chris to yourself. In today’s internet information is abundant, be wise to pick up only the correct information instead of believing everything.

[thighslapper]

14 hours ago, Mr.Q said:

You need to determine if you track objects (in stills or video) using AF-C often. If not, the SL2 will be one of the faster AF-S cameras on the market.

I'd have to disagree with Scott regarding CDAF vs PDAF.  Panasonic has been working for years to perfect their DFD technology and it still sucks at tracking. Unreliable with a lot of wobbling. That's with MFT with a much deeper DOF.  It doesn't take a rocket scientist to realize it would likely be worse with a FF sensor.

As for the reviews, pay attention to the images taken with the camera. If the subject matter is mostly stationary, the reviewer will not complain about the AF performance. I trust Chris and Jordan over at dpreview on this subject. They shoot a wide variety of subjects and provide very thorough reviews.

 

2 hours ago, sillbeers15 said:

Yes I have. If I wanted it I would have bought it. Have you extensively handled the SL2 with at least 100 shots to evaluate the good / bad of the AF before making sweeping comments? I have handled it, but not long enough to make any conclusion. If I were like you, I would have reported to the world that I know everything and end up misleading folks. That’s where we differ. I only believe what I have seen and experienced. I only believe data and not hearsay. Please keep your shallow believe from YouTubers like Chris to yourself. In today’s internet information is abundant, be wise to pick up only the correct information instead of believing everything.

There is a lot of crap talked about the various merits of PDAF and CDAF and in particular Panasonics implementation (which presumably Leica has licensed).

To my knowledge Richard Wong is the only one who has published real world tests that compare the relevant cameras... and the results may surprise you. There is certainly not the chasm of performance and reliability that is claimed by many. 

https://www.youtube.com/watch?v=8k3a7LNe0Mw

[Mr.Q]

My apologies, I definitely seem to have hit a nerve. Mods, please delete post #7 onwards. Thank you.

[Donzo98]

As far as I have thought... the AF on the A9/II is the current state of the art, in terms of tracking, face/eye detect.

The S1/R, SL and presumably the SL2 are not in the same league.

I had an SL... and currently have an S1... no major complaints for what I shoot. I wouldn't mind having the current state of the art on the SL2, but I'm expecting the same AF to be the same as the S1.

The internet gurus seem to agree that the SL2 isn't any close when it comes to Sony AF. Are they wrong??

[sillbeers15]

8 hours ago, Donzo98 said:

As far as I have thought... the AF on the A9/II is the current state of the art, in terms of tracking, face/eye detect.

The S1/R, SL and presumably the SL2 are not in the same league.

I had an SL... and currently have an S1... no major complaints for what I shoot. I wouldn't mind having the current state of the art on the SL2, but I'm expecting the same AF to be the same as the S1.

The internet gurus seem to agree that the SL2 isn't any close when it comes to Sony AF. Are they wrong??

No. Not wrong. I’m sure we all acknowledge that the Sony A9II has a great product when it comes to AF competency as the produce is designed for AF priority. Neither are we trying to compare SL2 to what A9II is on AF capability as both cameras have different strengths and weaknesses. It is fair to share information regarding user experiences over each situation. However there is no value added insisting one camera or AF system is better than the other without supporting facts. Technology changes over time and it is rather fast. So what you and I know today can easily be irrelevant tomorrow if we are not kept in phase of the changes. CDAF alone AF system was lagging in AF focusing speed 10 years ago but not today. Do not mix that up with AF tracking accuracy. The drawback of CDAF is more energy and computing consumption. Causing a less accurate focusing on every individual frame on Continuous drive mode unless the processor is fast enough to calculate and ensure sharp images every frame ( even PDAF cannot achieve 100% hitrate). We yet to know how well the SL2 can track focus on motion despite knowing it uses a new Maestro III processor. 

It is simple experiment to know when there is no DFD support on CDAF. Just attach a non native L mount lens with a third party adapter to experience the slow and hunting of CDAF ( due to no lens contrast images data stored in camera to activate DFD). 

So kindly do not just,

PDAF= Perfect AF &

CDAF= Junk AF

Life is just not so simple.

[sillbeers15]

20 hours ago, thighslapper said:

 

There is a lot of crap talked about the various merits of PDAF and CDAF and in particular Panasonics implementation (which presumably Leica has licensed).

To my knowledge Richard Wong is the only one who has published real world tests that compare the relevant cameras... and the results may surprise you. There is certainly not the chasm of performance and reliability that is claimed by many. 

https://www.youtube.com/watch?v=8k3a7LNe0Mw

Richard’s video clearly demonstrated the AF calculation speed of the CDAF in the S1 surpassed that of the A7III when the video recording is routed to external device. It clearly shows the importance of sufficient processing power for AF makes the differences in AF tracking improvements.

So really not so much PDAF or CDAF that determines AF tracking speed (which will determine sharp images for tracking stills in motion) but rather the processor capacity is critical.

Another point that supports this argument is the Sony’s A9II improvements over A9 is only limited to firmware and more processor to support the 60 times per second calculation published in the product literature.

Sadly most camera users are believers of perception from hearsay rather than thinking through themselves on what makes sense. No wonder there are so many YouTubers out there passing on half truths.

[lx1713]

I mentioned in a post long ago that the AF speed of the SL is very much dependent on needing a much more powerful processor and that it was just a matter of time for future generations to catch up. The SL EVF was just slower than the AF capabilities, so often it's blurred in EVF but sharp in the capture. So the EVF plays a part on our perception.

I think having the additional pixels such as the current SL2 goes a long way towards better accuracy. Which is one of the reasons I decided to pay the deposit on the camera. The overall handling is just better.

One possible factor that may play an important part is the quality of your optics. The higher the clarity of your optics the easier it is for the processor to determine focus. One last factor comes from personal experience is that the more stable your platform, the better the performance of the AF. So a solidly built camera beats flimsy.

One anecdote, my 24-90 had all its mount screws loose in a recent job, I felt the lens wobble but couldn't figure out why. After the job and going thru nearly 2000 pictures, I had almost no out of focus images. I found out that the lens mount was quite loose. Each screw needed to be tighter by 1 to 2 revolutions. A CDAF system will beat a DSLR PDAF system in this situation. To be honest, I don't know if I should be pissed at Leica for having the lens mount become loose on me or awed at their  forethought of deciding on mirrorless and foregoing PDAF. In 35 years of owning cameras I've never had a lens mount come loose 

Edited November 18, 2019 by lx1713

[scroy]

 

On 11/18/2019 at 4:54 PM, lx1713 said:

I mentioned in a post long ago that the AF speed of the SL is very much dependent on needing a much more powerful processor and that it was just a matter of time for future generations to catch up. The SL EVF was just slower than the AF capabilities, so often it's blurred in EVF but sharp in the capture. So the EVF plays a part on our perception.

 

Then would AF on the SL2 (theoretically) be faster or more accurate using the TL lenses as the Maestro III processor is now working on a smaller( 20Mpix) image?

Has this been tested? It would certainly be interesting.

[Advertisement]

Hallo,

Look here: Leica L-Mount

[SrMi]

On 11/16/2019 at 10:24 AM, scott kirkpatrick said:

AF opinions vary a lot, but the  center of what you will read is:

PDAF (Phase detect) is the best for continuous autofocus, because it uses special sensors on the chip (which replace image pixels) to predict the degree by which an object is out of focus and the direction in which to move he lens.  The cost is that those missing pixels sometimes are visible as bands or artifacts.

CDAF (contrast detect) just adjusts to get the sharpest focus, but may start out in the wrong direction and have to correct.  This fear is overblown, since there are software tricks for keeping track of what is in focus, and making tracking corrections outside of the time that the shutter is half-pressed.  So some impressive things (not everyone agrees) can be done with CDAF in firmware, and usually are.

The best place for an M lens is on an M camera.  The second best is on an SL or CL.  There is no third best.

A PDAF phase is often followed by a brief CDAF phase, since CDAF is more precise than  PDAF.

The best place for an M lens is indeed an M camera. But the second best is apparently SL2, third best is SL. The rest may work fine only with some of M lenses.