Report: New Canon Super Telephoto Lenses Coming in May

[neuroanatomist]

They already have a Zero extender for all lenses.

Yep, and it’s one of the cheapest Canon products!

 

[ThomasTH]

Yep, and it’s one of the cheapest Canon products!

But still more expensive than the Sony and Nikon version.

 

[justaCanonuser]

Just a brief update for those who followed my Nikon Z8 & Z 600mm f/6.3 PF (w and w/o 1.4x TC) versus R5 II & EF 600mm f/4.0 III (w and w/o 1.4x TC) AF issues and are interested in my reports. My wife and I found, that the Z8 still offers in their AF menu an option to micro adjust its AF to a lens, in contrast to Canon's ML cameras (as far as I know). So that was a strong hint, that Nikon's AF system even in the age of ML cameras may need AFMA. I dusted off my old Spyder Lenscal and we found indeed, that the Z8 & Z 600mm f/6.3 PF had a massive back-focus, with 1.4x TC even worse. So we AFMA'd the system for both combos. Unfortunately, since then there was no chance to really test this combo with birds, due to bad weather and a lack of time, so we still have to wait for real world results. Of course, we do hope we'll see a substantial improvement.

So, what do we learn from that? Obviously, Nikon's AF system seems to work much more like the old DSLRs, in contrast to Canon's DPAF based system, since there is still an option - and sometimes a strong need - for AFMA. When I have more time, I will try to do a little research to find more in-depth information about Nikon's Z AF system. What I have in mind is that the AF system of Nikon's first Z cameras got a lot of bad reviews, just with about the arrival of the Z9 and later Z8 the test results improved substantially.

 

[john1970]

Historically, Canon tends to announce professional grade big white telephoto lenses in the spring so a May announcement would be consistent with past behavior. For example, the Canon RF 100-300 mm f2.8 was announced on April 20, 2023.

 

[justaCanonuser]

Historically, Canon tends to announce professional grade big white telephoto lenses in the spring so a May announcement would be consistent with past behavior. For example, the Canon RF 100-300 mm f2.8 was announced on April 20, 2023.

I'd only be skeptical if the announcement would be April, 1st

 

[justaCanonuser]

Due to a lack of time, I asked Perplexity (what I prefer since several years now over ChatGPT) why Nikon's Z system still offers AFMA and Canon's not, and explained in my prompt our specific AF problems with our Z8 & Z 600mm f/6.3 PF combo (with and w/o 1.4x TC). Perplexity's answer might be interesting for some nerds (like me) here:

Nikon Z cameras still offer AF fine-tuning because mirrorless autofocus is not automatically immune to optical tolerances. The camera can focus on the sensor plane very accurately, but a given camera/lens/teleconverter combination can still show small mechanical or optical offsets, especially when a teleconverter is added and the effective aperture gets smaller.[1][2]

Why the adjustment exists

The main reason is that focus accuracy is not just a software problem; it is also affected by lens design, manufacturing tolerances, adapter/teleconverter alignment, and how the camera interprets phase-detect data. Even in mirrorless, the system can be “close but not perfect” for a specific lens combination, so Nikon keeps fine-tune as a correction layer for edge cases and for users who need maximum precision.[3][2][1]

With long lenses, wide apertures, and teleconverters, the depth of field becomes very thin, so even a tiny offset looks like obvious backfocus or frontfocus. That is why a birding setup can seem fine in general use but still miss consistently on small distant subjects.[4][3]

Nikon vs Canon approach

Canon’s RF mirrorless system is generally more self-correcting in practice because Canon has leaned heavily on on-sensor phase detection and lens communication designed as a tightly integrated mirrorless platform. Nikon’s Z system also uses on-sensor phase detection, but Nikon still exposes AF fine-tune because it allows compensating for residual per-lens/per-body/per-converter variation rather than assuming every combination is perfect out of the box.[5][1]

The key difference is not simply “Canon has dual pixel, Nikon does not.” It is more about how each manufacturer implements lens-body calibration, predictive AF behavior, and tolerance management across the whole optical chain. Canon’s system tends to hide more of that from the user; Nikon gives the user a manual correction tool when the combination is off.[2][1]

Your specific setup

A Z8 plus Z 600mm f/6.3 PF plus TC 1.4x should normally focus well, so a “massive backfocus” points to an out-of-spec combination, a teleconverter/lens alignment issue, a body/lens calibration mismatch, or a focus-shift behavior that becomes visible only at the longer effective focal length and smaller maximum aperture. The fact that the issue appears both with and without the TC suggests the base lens/body calibration is already off, and the TC may be exaggerating it rather than causing it alone.[6][7][4]

Practical interpretation

So the short answer is: Nikon keeps micro-adjustment because mirrorless AF is not mathematically perfect for every optical combination, and Nikon chooses to let users compensate for residual system error. Canon’s RF system often feels more automatic because Canon’s implementation relies more on integrated on-sensor focusing and hides calibration differences better, but it is not because mirrorless AF fundamentally cannot need calibration.[1][5][2]

A useful rule of thumb is that if one Nikon Z lens or lens-plus-TC combo needs a large correction, that is usually a sign to test the lens, TC, and body separately, because the camera is compensating for a real mismatch rather than a normal tuning quirk.[8][6]

My personal conclusion: looks like we'd have to send this combo to Nikon's service because of too big production tolerances - typical Nikon's lousy quality, I have to say after many years of using Canon and Nikon gear side by side

 

[AlanF]

Due to a lack of time, I asked Perplexity (what I prefer since several years now over ChatGPT) why Nikon's Z system still offers AFMA and Canon's not, and explained in my prompt our specific AF problems with our Z8 & Z 600mm f/6.3 PF combo (with and w/o 1.4x TC). Perplexity's answer might be interesting for some nerds (like me) here:

Nikon Z cameras still offer AF fine-tuning because mirrorless autofocus is not automatically immune to optical tolerances. The camera can focus on the sensor plane very accurately, but a given camera/lens/teleconverter combination can still show small mechanical or optical offsets, especially when a teleconverter is added and the effective aperture gets smaller.[1][2]

Why the adjustment exists

The main reason is that focus accuracy is not just a software problem; it is also affected by lens design, manufacturing tolerances, adapter/teleconverter alignment, and how the camera interprets phase-detect data. Even in mirrorless, the system can be “close but not perfect” for a specific lens combination, so Nikon keeps fine-tune as a correction layer for edge cases and for users who need maximum precision.[3][2][1]

With long lenses, wide apertures, and teleconverters, the depth of field becomes very thin, so even a tiny offset looks like obvious backfocus or frontfocus. That is why a birding setup can seem fine in general use but still miss consistently on small distant subjects.[4][3]

Nikon vs Canon approach

Canon’s RF mirrorless system is generally more self-correcting in practice because Canon has leaned heavily on on-sensor phase detection and lens communication designed as a tightly integrated mirrorless platform. Nikon’s Z system also uses on-sensor phase detection, but Nikon still exposes AF fine-tune because it allows compensating for residual per-lens/per-body/per-converter variation rather than assuming every combination is perfect out of the box.[5][1]

The key difference is not simply “Canon has dual pixel, Nikon does not.” It is more about how each manufacturer implements lens-body calibration, predictive AF behavior, and tolerance management across the whole optical chain. Canon’s system tends to hide more of that from the user; Nikon gives the user a manual correction tool when the combination is off.[2][1]

Your specific setup

A Z8 plus Z 600mm f/6.3 PF plus TC 1.4x should normally focus well, so a “massive backfocus” points to an out-of-spec combination, a teleconverter/lens alignment issue, a body/lens calibration mismatch, or a focus-shift behavior that becomes visible only at the longer effective focal length and smaller maximum aperture. The fact that the issue appears both with and without the TC suggests the base lens/body calibration is already off, and the TC may be exaggerating it rather than causing it alone.[6][7][4]

Practical interpretation

So the short answer is: Nikon keeps micro-adjustment because mirrorless AF is not mathematically perfect for every optical combination, and Nikon chooses to let users compensate for residual system error. Canon’s RF system often feels more automatic because Canon’s implementation relies more on integrated on-sensor focusing and hides calibration differences better, but it is not because mirrorless AF fundamentally cannot need calibration.[1][5][2]

A useful rule of thumb is that if one Nikon Z lens or lens-plus-TC combo needs a large correction, that is usually a sign to test the lens, TC, and body separately, because the camera is compensating for a real mismatch rather than a normal tuning quirk.[8][6]

My personal conclusion: looks like we'd have to send this combo to Nikon's service because of too big production tolerances - typical Nikon's lousy quality, I have to say after many years of using Canon and Nikon gear side by side

I used Reiken Focal for may years to AFMA DSLRs. They still produce their software for mirrorless, which is basically a waste of time for Canon as it just gives some limited information, but it does work with Nikon AFMA, and they explain why and the improvements. https://www.reikanfocal.com/why-nikon-mirrorless.html

It was very good indeed for DSLRs and I found it invaluable and the quickest and most accurate AFMA tool.

 

[Marc73]

Since, the current 400 & 600 primes are EF lenses adapted to the RF mount; their due for a refresh. Before the world cup would make sense, but that doesn't mean they will do it.

As for a switchable teleconverter; I would prefer a completely separate unit. That way it can be used on other lenses and when it's needed. A built in teleconverter adds weight, length, and cost to the lens. I only need 1.4, not a fan of a 2.0 teleconverters ( my personal opinion). They have a working switchable teleconverter from the EF 200-400mm, that has shown reliability for years. Why haven't they adapted this into a separate teleconverter; there is no doubt in my mind it would sell. I would buy it. SEE DISCUSSION BELOW. I've learned you can't make a generic separate switchable teleconverter.

In my opinion, since the introduction of the RF mount Canon has put the long glass needs on the back burner. During this time, Sony & Nikon have developed better options and it's time for Canon to step up. The classic example is the gap between the 100-500mm and 400/600mm primes in the "L" series lenses. If you want upgrade to 600mm or faster glass from your 100-500, your ONLY option is a $13 to 14k prime. Then there is the mythical 300-600mm that has been teased for years. This would be a great wildlife lens, if their smart enough to make it a variable aperture of f4 to 5.6. For those of us that shoot in low light and want better bokeh; a fixed f5.6 aperture is unacceptable at the 300 to 400mm range for the anticipated price of at least 7K.

What about the 200-800mm?

 

[justaCanonuser]

I used Reiken Focal for may years to AFMA DSLRs. They still produce their software for mirrorless, which is basically a waste of time for Canon as it just gives some limited information, but it does work with Nikon AFMA, and they explain why and the improvements. https://www.reikanfocal.com/why-nikon-mirrorless.html

It was very good indeed for DSLRs and I found it invaluable and the quickest and most accurate AFMA tool.

Thanks for this valuable tip, Alan. I called today the local Nikon service, they will check the combo, I have just to find a time slot to visit them. They promised they can do this service while I am waiting. Let's see what Nikon's experts tell us.

 

[justaCanonuser]

A lightweight RF 500 mm f4 would be great as well and a nice surprise.

I bet that we never see again any new 500 mm f/4.0 prime from Canon! It wouldn't make sense since the 600mm f/4.0 lenses are so light now, and keeping two production lines that would cannibalize each other of such prime lenses wouldn't make sense economically.

Looking at Canon's recent strategy it is more probable that may see a new tele zoom covering the 500mm range in future what is faster than f/7.1 at the long end...

 

[Dragon]

Due to a lack of time, I asked Perplexity (what I prefer since several years now over ChatGPT) why Nikon's Z system still offers AFMA and Canon's not, and explained in my prompt our specific AF problems with our Z8 & Z 600mm f/6.3 PF combo (with and w/o 1.4x TC). Perplexity's answer might be interesting for some nerds (like me) here:

Nikon Z cameras still offer AF fine-tuning because mirrorless autofocus is not automatically immune to optical tolerances. The camera can focus on the sensor plane very accurately, but a given camera/lens/teleconverter combination can still show small mechanical or optical offsets, especially when a teleconverter is added and the effective aperture gets smaller.[1][2]

Why the adjustment exists

The main reason is that focus accuracy is not just a software problem; it is also affected by lens design, manufacturing tolerances, adapter/teleconverter alignment, and how the camera interprets phase-detect data. Even in mirrorless, the system can be “close but not perfect” for a specific lens combination, so Nikon keeps fine-tune as a correction layer for edge cases and for users who need maximum precision.[3][2][1]

With long lenses, wide apertures, and teleconverters, the depth of field becomes very thin, so even a tiny offset looks like obvious backfocus or frontfocus. That is why a birding setup can seem fine in general use but still miss consistently on small distant subjects.[4][3]

Nikon vs Canon approach

Canon’s RF mirrorless system is generally more self-correcting in practice because Canon has leaned heavily on on-sensor phase detection and lens communication designed as a tightly integrated mirrorless platform. Nikon’s Z system also uses on-sensor phase detection, but Nikon still exposes AF fine-tune because it allows compensating for residual per-lens/per-body/per-converter variation rather than assuming every combination is perfect out of the box.[5][1]

The key difference is not simply “Canon has dual pixel, Nikon does not.” It is more about how each manufacturer implements lens-body calibration, predictive AF behavior, and tolerance management across the whole optical chain. Canon’s system tends to hide more of that from the user; Nikon gives the user a manual correction tool when the combination is off.[2][1]

Your specific setup

A Z8 plus Z 600mm f/6.3 PF plus TC 1.4x should normally focus well, so a “massive backfocus” points to an out-of-spec combination, a teleconverter/lens alignment issue, a body/lens calibration mismatch, or a focus-shift behavior that becomes visible only at the longer effective focal length and smaller maximum aperture. The fact that the issue appears both with and without the TC suggests the base lens/body calibration is already off, and the TC may be exaggerating it rather than causing it alone.[6][7][4]

Practical interpretation

So the short answer is: Nikon keeps micro-adjustment because mirrorless AF is not mathematically perfect for every optical combination, and Nikon chooses to let users compensate for residual system error. Canon’s RF system often feels more automatic because Canon’s implementation relies more on integrated on-sensor focusing and hides calibration differences better, but it is not because mirrorless AF fundamentally cannot need calibration.[1][5][2]

A useful rule of thumb is that if one Nikon Z lens or lens-plus-TC combo needs a large correction, that is usually a sign to test the lens, TC, and body separately, because the camera is compensating for a real mismatch rather than a normal tuning quirk.[8][6]

My personal conclusion: looks like we'd have to send this combo to Nikon's service because of too big production tolerances - typical Nikon's lousy quality, I have to say after many years of using Canon and Nikon gear side by side

The calibration for individual differences really doesn't make sense. Sensor PDAF is measuring at the point of focus so there should be no errors. If there is some residual error in the PD algorithm when concatenated with a particular lens, then fine tuning via CDAF on first use should clean that up and that may be what Canon is doing. An algorithm that starts from far OOF and tries to make it all the way in one shot clearly could have some issues, but with modern fast sensors, multiple focus checks as the lens motor is homing in seem logical. Contrary to the AI response, I would say Nikon has a software problem.

 

[AlanF]

The calibration for individual differences really doesn't make sense. Sensor PDAF is measuring at the point of focus so there should be no errors. If there is some residual error in the PD algorithm when concatenated with a particular lens, then fine tuning via CDAF on first use should clean that up and that may be what Canon is doing. An algorithm that starts from far OOF and tries to make it all the way in one shot clearly could have some issues, but with modern fast sensors, multiple focus checks as the lens motor is homing in seem logical. Contrary to the AI response, I would say Nikon has a software problem.

Here is what a another AI engine writes:

Why the Feature Still Exists​

If the path error is gone, why did Nikon keep the menu option in the Z6, Z7, Z8, and Z9? It comes down to Phase Detection (PDAF) vs. Contrast Detection (CDAF).

• PDAF is a "Predictor": Most Nikon Z cameras use on-sensor Phase Detection. PDAF works by looking at two different views of the subject and calculating exactly how far the lens needs to move. It’s incredibly fast, but it is technically an open-loop calculation.
  
  
  
• Manufacturing Tolerances: Even with on-sensor PDAF, slight variances in how a specific lens's motor reacts to a command can exist. While the camera should see that it's in focus, the PDAF algorithm might consistently stop the lens a tiny bit short or long due to the lens's specific mechanical behavior.
• Spherical Aberration (Focus Shift): Some lenses shift their focus point slightly as you close the aperture (stop down). Since cameras usually focus with the lens wide open, the "perfect" focus point at f/1.8 might be slightly different than at f/2.8. AF Fine-Tune allows a pro to "offset" this shift for their most-used aperture.
  

---

3. Using Adapted F-Mount Lenses​

The most practical reason Nikon kept AF Fine-Tune is the FTZ Adapter. Many photographers use older DSLR (F-mount) lenses on their mirrorless bodies.

• These older lenses were designed for the mechanical "tug-of-war" of DSLR systems.
• While the mirrorless sensor is more accurate, some older lens motors may not "settle" perfectly on the target when driven by the new mirrorless algorithms.
• If you find your favorite 85mm f/1.4 G lens is always just a hair soft on your Z8, the AF Fine-Tune lets you bridge that gap.

 

[justaCanonuser]

The calibration for individual differences really doesn't make sense. Sensor PDAF is measuring at the point of focus so there should be no errors. If there is some residual error in the PD algorithm when concatenated with a particular lens, then fine tuning via CDAF on first use should clean that up and that may be what Canon is doing. An algorithm that starts from far OOF and tries to make it all the way in one shot clearly could have some issues, but with modern fast sensors, multiple focus checks as the lens motor is homing in seem logical. Contrary to the AI response, I would say Nikon has a software problem.

We'll hopefully find out via Nikon service what's the problem. But, based on our experience with a Nikon gear side by side by a comparable Canon gear I can only wrap up: Nikon has severe quality problems in the past 15 years anyway. Based on all the various failures that happened with our Nikon gear I wouldn't recommend to change from Canon to Nikon.

 

[justaCanonuser]

Here is what a another AI engine writes:

​

• Manufacturing Tolerances: Even with on-sensor PDAF, slight variances in how a specific lens's motor reacts to a command can exist. While the camera should see that it's in focus, the PDAF algorithm might consistently stop the lens a tiny bit short or long due to the lens's specific mechanical behavior.

This describes what we observe so far most exactly. We did not yet adapt any older F mount lens, and I could not yet find any reliable information about a possible focus shift of this Z 600mm f/6.3 PF lens. Thanks for your kind help, Alan!

 

[Michael Clark]

The symptoms with the Z8 and the Z 600/6.3 are: the camera focuses fast (with no TC attached) and gets in particular very close to in-focus images with the first frame(s), but then starts to struggle with following frames, and the AF sort of micro-pumps around the precise focus position.

That sounds like the original 7D in AI Servo mode.

 

[Dragon]

Here is what a another AI engine writes:

Why the Feature Still Exists​

If the path error is gone, why did Nikon keep the menu option in the Z6, Z7, Z8, and Z9? It comes down to Phase Detection (PDAF) vs. Contrast Detection (CDAF).

• PDAF is a "Predictor": Most Nikon Z cameras use on-sensor Phase Detection. PDAF works by looking at two different views of the subject and calculating exactly how far the lens needs to move. It’s incredibly fast, but it is technically an open-loop calculation.
  
  
  
• Manufacturing Tolerances: Even with on-sensor PDAF, slight variances in how a specific lens's motor reacts to a command can exist. While the camera should see that it's in focus, the PDAF algorithm might consistently stop the lens a tiny bit short or long due to the lens's specific mechanical behavior.
• Spherical Aberration (Focus Shift): Some lenses shift their focus point slightly as you close the aperture (stop down). Since cameras usually focus with the lens wide open, the "perfect" focus point at f/1.8 might be slightly different than at f/2.8. AF Fine-Tune allows a pro to "offset" this shift for their most-used aperture.

---

3. Using Adapted F-Mount Lenses​

The most practical reason Nikon kept AF Fine-Tune is the FTZ Adapter. Many photographers use older DSLR (F-mount) lenses on their mirrorless bodies.

• These older lenses were designed for the mechanical "tug-of-war" of DSLR systems.
• While the mirrorless sensor is more accurate, some older lens motors may not "settle" perfectly on the target when driven by the new mirrorless algorithms.
• If you find your favorite 85mm f/1.4 G lens is always just a hair soft on your Z8, the AF Fine-Tune lets you bridge that gap.

That all makes perfect sense but doesn't address the idea that the system should be able to tweak the goal on the fly since it has access to many samples as the lens is focusing. I see no reason why mirrorless PDAF can't work similarly to CDAF with the exception that it should be able to avoid overshoot and hunting because it inherently knows whether the instant focus is front or back and CDAF does not. If the lens motor does overshoot, the system should still be able to back it up in no more than one extra try. I still think Nikon has some software optimization to do if the previous comments are correct. That said, I have a Tamron 18-400 that won't focus precisely at the long end on any body (SLR or Mirrorless). The Tap-In console has no effect when uses on a DPAF mirrorless body, and manual focus notably better than AF. Interestingly it comes very close to correct focus when used on a 5DSR in live view mode. That camera uses CDAF in live view, which would suggest that even Canon does not iterate focus attempts as much with DPAF as is does with the CDAF in the 5DSR. It also suggests that Tamron didn't get even the EF focus algorithm quite right. Their response when questioned was a request to send the lens and the camera body to them for fine tuning (but ironically, the Tap In fine tuning has no effect) but that doesn't make the lens universal as it would only match it to one body if successful. Since I don't have that much use for the lens in RF land, I demurred. As an aside, you would think if the problem is overshoot, hitting the half press on the shutter button a second time would improve the focus, but it does not. So maybe the Nikon tweak will be more compatible with 3rd party lenses that don't quite follow the instructions they are given. It would be an interesting test if someone has a copy of that lens in Nikon format.

 

[Michael Clark]

Due to a lack of time, I asked Perplexity (what I prefer since several years now over ChatGPT) why Nikon's Z system still offers AFMA and Canon's not, and explained in my prompt our specific AF problems with our Z8 & Z 600mm f/6.3 PF combo (with and w/o 1.4x TC). Perplexity's answer might be interesting for some nerds (like me) here:

Nikon Z cameras still offer AF fine-tuning because mirrorless autofocus is not automatically immune to optical tolerances. The camera can focus on the sensor plane very accurately, but a given camera/lens/teleconverter combination can still show small mechanical or optical offsets, especially when a teleconverter is added and the effective aperture gets smaller.[1][2]

Why the adjustment exists

The main reason is that focus accuracy is not just a software problem; it is also affected by lens design, manufacturing tolerances, adapter/teleconverter alignment, and how the camera interprets phase-detect data. Even in mirrorless, the system can be “close but not perfect” for a specific lens combination, so Nikon keeps fine-tune as a correction layer for edge cases and for users who need maximum precision.[3][2][1]

With long lenses, wide apertures, and teleconverters, the depth of field becomes very thin, so even a tiny offset looks like obvious backfocus or frontfocus. That is why a birding setup can seem fine in general use but still miss consistently on small distant subjects.[4][3]

Nikon vs Canon approach

Canon’s RF mirrorless system is generally more self-correcting in practice because Canon has leaned heavily on on-sensor phase detection and lens communication designed as a tightly integrated mirrorless platform. Nikon’s Z system also uses on-sensor phase detection, but Nikon still exposes AF fine-tune because it allows compensating for residual per-lens/per-body/per-converter variation rather than assuming every combination is perfect out of the box.[5][1]

The key difference is not simply “Canon has dual pixel, Nikon does not.” It is more about how each manufacturer implements lens-body calibration, predictive AF behavior, and tolerance management across the whole optical chain. Canon’s system tends to hide more of that from the user; Nikon gives the user a manual correction tool when the combination is off.[2][1]

Your specific setup

A Z8 plus Z 600mm f/6.3 PF plus TC 1.4x should normally focus well, so a “massive backfocus” points to an out-of-spec combination, a teleconverter/lens alignment issue, a body/lens calibration mismatch, or a focus-shift behavior that becomes visible only at the longer effective focal length and smaller maximum aperture. The fact that the issue appears both with and without the TC suggests the base lens/body calibration is already off, and the TC may be exaggerating it rather than causing it alone.[6][7][4]

Practical interpretation

So the short answer is: Nikon keeps micro-adjustment because mirrorless AF is not mathematically perfect for every optical combination, and Nikon chooses to let users compensate for residual system error. Canon’s RF system often feels more automatic because Canon’s implementation relies more on integrated on-sensor focusing and hides calibration differences better, but it is not because mirrorless AF fundamentally cannot need calibration.[1][5][2]

A useful rule of thumb is that if one Nikon Z lens or lens-plus-TC combo needs a large correction, that is usually a sign to test the lens, TC, and body separately, because the camera is compensating for a real mismatch rather than a normal tuning quirk.[8][6]

My personal conclusion: looks like we'd have to send this combo to Nikon's service because of too big production tolerances - typical Nikon's lousy quality, I have to say after many years of using Canon and Nikon gear side by side

It seems to me that with MILCs the issue is usually how accurately the lens can move compared to how much the camera told the lens to move, and how well the camera can know how much the lens actually moved without continuing to focus during the movement when processing cycles might be better spent for predictive analysis of where the target will be when the shutter (mechanical or electronic) is actuated. Move and refocus is much slower than measure focus, tell the lens which way to move and how far, then confirm the lens has moved that far without measuring focus again, and take the picture.

As others have also pointed out above, not all lenses hold an exact focus position once the AF motor stops moving it, either. Wear and tear leaves room for "play", just like an older car has more steering wheel play than when it was new. If you stop turning the wheel to the right and start turning it back to the left, you have to move it an inch or a few before the movement of the steering wheel starts turning the tires in the opposite direction.

Doing AFMA calibration for friends, I've seen cases where the correction need to be slightly different depending on if the lens started out towards MFD or towards infinity before AF was actuated.

 

[Michael Clark]

That all makes perfect sense but doesn't address the idea that the system should be able to tweak the goal on the fly since it has access to many samples as the lens is focusing. I see no reason why mirrorless PDAF can't work similarly to CDAF with the exception that it should be able to avoid overshoot and hunting because it inherently knows whether the instant focus is front or back and CDAF does not. If the lens motor does overshoot, the system should still be able to back it up in no more than one extra try. I still think Nikon has some software optimization to do if the previous comments are correct. That said, I have a Tamron 18-400 that won't focus precisely at the long end on any body (SLR or Mirrorless). The Tap-In console has no effect when uses on a DPAF mirrorless body, and manual focus notably better than AF. Interestingly it comes very close to correct focus when used on a 5DSR in live view mode. That camera uses CDAF in live view, which would suggest that even Canon does not iterate focus attempts as much with DPAF as is does with the CDAF in the 5DSR. It also suggests that Tamron didn't get even the EF focus algorithm quite right. Their response when questioned was a request to send the lens and the camera body to them for fine tuning (but ironically, the Tap In fine tuning has no effect) but that doesn't make the lens universal as it would only match it to one body if successful. Since I don't have that much use for the lens in RF land, I demurred. As an aside, you would think if the problem is overshoot, hitting the half press on the shutter button a second time would improve the focus, but it does not. So maybe the Nikon tweak will be more compatible with 3rd party lenses that don't quite follow the instructions they are given. It would be an interesting test if someone has a copy of that lens in Nikon format.

Except while all of this focusing and refocusing is going on the subject is often moving erratically and changing the distance between subject and camera. There aren't many use cases for still life photography with Super Telephoto lenses. And the camera's processor(s) needs significantly more processing cycles to predict where that moving subject will be when the shutter actuates than it needs to perform a focus measurement.

With DSLRs Canon placed a higher priority on speed when using the optical VF with dedicated PDAF array, and a higher priority on accuracy when using Live View with main sensor based DPAF.

 

[mimbu]

I'll reply for the benefit of others who may read this thread, in the hope that your cluelessness doesn't engender false hope in others. Spreading misinformation seems to be a pattern for you, among other distasteful habits that you exhibit.


Yes, it does have to be different. A flip-out TC works when the TC is added within the lens. Notice how in the block diagram of the EF 200-400/4 that I posted above, there is a (weak) converging group behind the TC optics. The same is true for any lens with a flip-in TC. Here's the Nikon 600/4 + 1.4x, which has even more optics (but still net weak convergence) behind the TC group.

View attachment 228801

The only way in which your statement would be correct would be for a lens to be designed to take a drop-in TC, like a really fat drop-in filter holder. There would need to be optics behind that big hole in the lens barrel. Moreover, in these expensive lenses the TCs are designed for optimal performance with each lens' optics. For example, the TC group in the Nikkor 600/4 + 1.4x has an SR element (their equivalent of Canon's BR elements aka 'blue goo'), whereas the TC group in the Nikkor 400/2.8 + 1.4x does not.

In other words, each supertele lens would need it's own specific drop-in TC, rather than having a generic drop-in TC for a series of lenses that would have more of a negative optical impact than tailor-made TCs for each lens. So even the kludgy idea you are suggesting (well, that you would have suggested if you actually understood optical design) would not happen. A dedicated TC for each lens...why make a drop-in version at all, then? That would be foolish, something that manufacturers' lens designers are not. That's why both Canon and Nikon have made lenses with flip-in TCs, not completely removable drop-in TCs.

And not separate flip-in TCs...because physics. The nice thing about physics is that it remains true even when people like you don't understand it.


No. Clearly, you don't understand what I posted, or the underlying Canon patent application. More importantly, you don't understand the relevant optical concepts. A lens cannot maintain infinity focus with an extension tube behind the lens, and if all the optical elements of a TC mounted behind the lens were to move out of the optical path, you would have an empty tube behind the lens...i.e., an extension tube.

I will try to simplify with a picture of the Canon patent design that perhaps you can understand. Only one set of 1.4x TC optics flips out of place, though it does so in two pieces to minimize the size of the overall optic. When the rear split 1.4x group ('B') moves out of the optical path, the front 1.4x group ('A') slides further back into position but remains in the optical path. At no time is there an empty tube.

View attachment 228800

This is still not the thing you think is happening, i.e. there is no 1x form for this design or any other. As already stated, an extension tube precludes infinity focus, so a '1x' option in a switchable TC would require reducing optics to flip in when the TC optics flip out.

Lenses with built-in TCs are designed to be optically correct at 1x in the first place, with open internal air gaps as part of the prescription. An air gap is just another design variable, and large air gaps are not unusual in supertelephoto optics.

The 1.4x mode is then created by inserting a matched negative afocal converter group into that already-corrected 1x system. That group modifies the ray angles and pupil imaging so the effective focal length and f-number increase, while the final image plane remains at the sensor.

So there is no physical requirement for some separate "1x compensating group" when the TC is out. The 1x state is the native solved state. The 1.4x group is the added perturbation that was specifically designed to work within it.

You are asserting a physics constraint that does not exist.

 

[neuroanatomist]

Lenses with built-in TCs are designed to be optically correct at 1x in the first place, with open internal air gaps as part of the prescription. An air gap is just another design variable, and large air gaps are not unusual in supertelephoto optics.

The 1.4x mode is then created by inserting a matched negative afocal converter group into that already-corrected 1x system. That group modifies the ray angles and pupil imaging so the effective focal length and f-number increase, while the final image plane remains at the sensor.

So there is no physical requirement for some separate "1x compensating group" when the TC is out. The 1x state is the native solved state. The 1.4x group is the added perturbation that was specifically designed to work within it.

You are asserting a physics constraint that does not exist.

Did you rephrase what I already stated and illustrated all on your own, or did one of your 'friends' like Claude or Grok help you?

Let me remind you of what you asserted:

A 1.4x TC built into a lens does not have to be any different to a separate 1.4x TC added to a lens. The lens is simply designed to focus onto the sensor without the 1.4x TC in place, and the 1.4x TC optics are designed so that they don't screw with that.

The discussion was about a switchable separate 1x-1.4x TC (echoing previous discussions about a switchable 1x-1.4x-2x TC that resulted from online misinterpretation of a Canon patent). The desire is for a standalone optic that is added to and mounts behind the lens like the current 1.4x and 2x TCs. The explanation was about why that is not technically feasible without additional optics for use at '1x'.

No one but you is arguing that a built-in switchable TC, which is something that has been done for decades by multiple manufacturers, is technically feasible and requires nothing flipped in when the TC optics are flipped out…that would be obvious even to an idiot.

Either you failed to understand the point of the discussion, or you failed to understand the response. Either way, fail.