Canon EOS R5 Unite and Discuss!

Leigh wrote in post #19435070
The camera through Deep Learning tech. "knows" how to distinguish a subject from an obstacle.

If you're tracking a single Subject, TS at minus will prevent switching to an Obstacle, and if there are no other nearby Subjects, STS setting is irrelevant.

If you were tracking a single Subject among other nearby Subjects setting TS to minus will keep from switching to Obstacles, & STS to 1, or 0 would prevent switching to another Subject.

That is true. I guess I'm too old school and like to have total control of my system, but I'm fully open to new tech. Since we are having these excellent conversations l'll through this one in. If the camera's deep learning can distinguish between obstacles and subjects then why do you need Case numbers in STS? There is my little what if guy on why shoulder again.

digital paradise wrote in post #19435112
If the camera's deep learning can distinguish between obstacles and subjects then why do you need Case numbers in STS? There is my little what if guy on why shoulder again.

Even if the camera learns to distinguish one from the other, it still needs to be told how to respond to either one. Deep learning or not, the AF still needs directions.

Levina de Ruijter wrote in post #19435120
Even if the camera learns to distinguish one from the other, it still needs to be told how to respond to either one. Deep learning or not, the AF still needs directions.

I knew someone would say that and I agree.

Either way I'm still not going to change my set up for what I shoot. Machine learning or not I still don't get how Auto would know my intent. Do I want it to Case 2 or Case 3 for me. Seems like important instructiions to input.

I still think Canon should have completely separated the two and added Case numbers or a variant to STS. It is completely different because to the ability to recognise faces and eyes.

apersson850 wrote in post #19435040
TS set to -2 tells the camera to avoid focusing at irrelevant things (i.e. things not being faces, if you are tracking faces) that may obscure the view. STS instead controls the camera's willingness to switch to another relevant thing (i.e. yet another face when already tracking one face).

apersson850 wrote in post #19435044
Yes, if you add that the tracking sensitivity is really a time-out. In older 1D-series cameras Canon actually specified the time. From 1 second before switching to 0.125 seconds before switching across the scale.

These are the most concise and helpful explanations I've seen. Excellent!

digital paradise wrote in post #19435107

Reflecting on that over the years pretty much every bit I've read or watched about birding recommends to use Case 2 and typically TS @ -2. I don't look for sports and other people suggestions but I came across a pro sports shooter that recommended that for the 1DX.

I don't remember anyone recommending Case 3. In a made for Canon video about the 7D2 Art Morris had no use for Case 3 and used it to tune TS and other perimeters to his taste. TS was -2.

digital paradise wrote in post #19435133
Reflecting on that over the years pretty much every bit I've read or watched about birding recommends to use Case 2 and typically TS @ -2. I don't look for sports and other people suggestions but I came across a pro sports shooter that recommended that for the 1DX.

I don't remember anyone recommending Case 3. In a made for Canon video about the 7D2 Art Morris had no use for Case 3 and used it to tune TS and other perimeters to his taste. TS was -2.

You have got to stop listening to all these people. Canon isn’t talking so everything is opinion and speculation. Instead be your own guru, experiment with different settings and see what works best for you, for the way you shoot. Then just forget about this stuff.

digital paradise wrote in post #19435133
I don't remember anyone recommending Case 3.

Well, then I'll immediately do that for you. Imagine you are trying to take photos of motocross. The more spectacular things are the jumps, and then especially if you can get below the bike rider when he jumps. That implies being behind and below the hill they are jumping over, which in turn implies you can't start tracking the biker already before he jumps. He's not visible then. Not until he suddenly appears over the crest.

That's where Case 3 is the best, since it can snap into focus in the blink of an eye and start tracking the biker whilst airborne.

apersson850 wrote in post #19435144
Well, then I'll immediately do that for you. Imagine you are trying to take photos of motocross. The more spectacular things are the jumps, and then especially if you can get below the bike rider when he jumps. That implies being behind and below the hill they are jumping over, which in turn implies you can't start tracking the biker already before he jumps. He's not visible then. Not until he suddenly appears over the crest.

That's where Case 3 is the best, since it can snap into focus in the blink of an eye and start tracking the biker whilst airborne.

I used it exactly for those purposes. I’ve never seen it suggested for birding as I stated.

Levina de Ruijter wrote in post #19435140
You have got to stop listening to all these people. Canon isn’t talking so everything is opinion and speculation. Instead be your own guru, experiment with different settings and see what works best for you, for the way you shoot. Then just forget about this stuff.

I have used all the cases for their designed purposes including Case 3. Not very often but I have. Never for birding. I have no trouble understanding what they do. That was never my original question.

digital paradise wrote in post #19435153
I used it exactly for those purposes. I’ve never seen it suggested for birding as I stated.

Ah, I missed the connection with birds. I find them pretty uninteresting to take photos of, so I didn't read that part carefully enough.
Sorry! This means we share the same opinion here.

apersson850 wrote in post #19435172
Ah, I missed the connection with birds. I find them pretty uninteresting to take photos of, so I didn't read that part carefully enough.
Sorry! This means we share the same opinion here.

No problem. My mistake for making a blank statement with birds in mind. Does not represent all the members. Case 3 has been beneficial when I needed it.

apersson850 wrote in post #19435172
Ah, I missed the connection with birds. I find them pretty uninteresting to take photos of, so I didn't read that part carefully enough.
Sorry! This means we share the same opinion here.

there's other kinds of photography?

Deep Learning was originally developed for the 1DX series, and for sports.

Here's an excerpt from an article, on how it's done, and what it can accomplish, for anyone interested.

"By giving the Deep Learning algorithm access to a vast library of images, of everything from upside-down gymnasts to hockey players wearing pads and helmets, it is able to learn and differentiate the human form in an endless variety of situations – and is ultimately able to perform this 'head detection', so that even if the person's face is not visible, the head is always the primary point of focus."

https://www.digitalcam​eraworld.com …ai-powered-autofocus-work

And another:

"It's even more of a challenge when obstacles get in the way. Félicia used to struggle to maintain autofocus at weddings when tracking people on the move while flowers, confetti and rice were being thrown, often obscuring their faces. Now, she says, "it's incredible. With the EOS R5 and EOS R6 I was using, autofocus stays locked on the eyes and faces of people even when the camera can't see them."

https://www.canon-europe.com …gent-autofocus-explained/