Moving to Mirrorless

Hello, John.

Canon R5! Wow. That's car buying territory for me. I'd suggest drag the thing out and get clicking. Glad you found the info useful.

Angus

The R7 has quite a slow electronic shutter. I've read about and seen photos of issues with shutter lag. Indeed, if you select electronic shutter on the R7 it has a warning message about possible distorted images.

However, shutter lag does not appear to be an issue with electronic 1st curtain. Don't ask me what this means, or I'll whimper.

Following the engineer's Zen philosophy of 'forget the manual, prod it and see what happens', I took my R7 out to the field for a test run on the first available opportunity - about two days after I got it. Still pretty much factory settings. Odd one (e.g. AI Servo, AF Case 3, Single point AF) set by me. I did not have AI tracking on at this stage.

The day was very overcast and manky.

Here are Tufted ducks, having just taken off from Colebrook lake (south) on Moor Green Lakes nature reserve. Lens is at 600mm.  No distortion on wings, despite Tufty ducks flap them at a rate of 20 to the dozen.

Still no distortion with electronic 1st curtain, though the birds are side on.

This is where shutter lag issues will become more apparent, but so far looks good.  I think Duade did say, in one of his videos, that slight distortion could occur with electronic 1st curtain.

Is this a tiny bit of distortion on the outer wing bits of the Tufty on the left? Can't tell, but looks alright to me. Shutter was 1/1000 with ISO ranging about the 800 to 1000 mark.

After first tests on Moor Green Lakes, I wandered over to Manor farm restoration, where I knew Lapwing were nesting. I could test out more BIF capabilities as they chased crows.

Again, still first outing, still lots of factory settings, but I had now turned on AI tracking - eye and animal Sill lousy light conditions.

First skill is acquire target. A bit tricky as I still wasn't used to the view finder. Once AI tracking has locked onto target, simply follow it and keep pressing shutter.

With my 80D, I would take a photo, then hunt around for the subject and get focus on it, before photographing - then repeat. With AI tracking, simply follow target as smoothly as possible and press shutter. AI tracking does all the hard work of keeping things in focus.

Here we are, I've acquired a target. AI tracking has locked onto the crow. All image taken from DPP 4, which will show focus points.

I did have AF Case 3 which will focus on anything that comes into focus areas, which tends to pick up closest object. However, AI tracking is doing all the hard work for me, as I could see the red rectangle pulsing around the crow.

Is the crow in focus? Clicking on the image in DPP 4, takes you down to pixel level, I believe. Not entirely in focus, but the day was really manky, the bird was over 100m away, and moving fast.

It looks as if AI tracking has lost its target in the next photo. Not entirely sure, as the birds were commencing a dive.

The Lapwing is still in reasonable focus.

This, I feel, is where AI tracking comes into its own. Both birds have now entered a cluttered area of trees. Edge detection is now quite hard due to cluttered background and low light making everything merge into each other, with nothing really standing out. Again, no effort on my part. Just keep following subjects and pressing shutter with gay abandon.

Not only was I beginning to lose the birds against the trees, so was AI tracking. However, damn good effort by AI tracking.

Cropping out the birds, still pretty good. No post processing going on here. I just cropped out the birds. I could go on the sharpen the image a touch, put contrast up a notch, etc.

As with any good engineering type, I didn't pay too much attention to Daude's video, especially when he got to the bit about customising camera buttons. The drift I got was that his approach is to start with AI tracking off, acquire a target and then thumb AI tracking on; or was it the other way around?

I tend to leave AI tracking on, with spot or single point AI Area for subjects either close up or in cluttered areas. I switch to 'Whole Area' when a subject is in clear sky. Very lazy shooting, as I let AI pick up the subject.

I will switch AI tracking off, occasionally.

A last word about Auto Focus. It is bloody fast. Perhaps too fast, in some circumstances with Single point or spot AF. More on this later, as I figure it out more.

For this next lot, still on first outing, I set my lens to about 400mm. My thinking was this.

1. I'm trying to photograph tiny, fast moving Sand Martins, which change direction quickly. Therefore seeing a much of the sky as possible to pick up the little buggers is probably a good strategy.

2. The Sigma lens can produce soft images at 600mm

3. The R7's sensor has 25% more pixels than the 80D, therefore I can crop photos more to achieve, at worst, a similar outcome as with my 80D.

AI tracking was on. I can't remember if AF Area was Single point, Expand AF Area: around or Whole Area. Probably the middle one. Once a target was acquired by AI tracking, I just tried to keep up with these birds, who were very frisky in the cool weather as they hunted insects.

Oh, I think I was still using electronic 1st curtain. Not sure, I have a strong feeling I was on full electronic.

AI tracking doing its business. Again, screen capture from DPP 4 to show focus points.

AI tracking doing its best to stay locked onto the bird.

I think AI tracking is possibly losing it. On the other hand, the bird is moving pretty fast, a diving banking manoeuvre.

Cropping out the bird in FastStone.

Errrm, who knows what the R7 is focused on. It was cloudy.

Obviously thinks a speck in the sky looks like an eye.

Though the Martin is still pretty much in focus. Not bad for such a manky day. The sensor is really good.

No post processing, not that you can do a huge amount. Perhaps a little sharpening, lightening up a tad.

Bearing in mind this was my first outing, with a camera I'd had for barely three days; during which time I'd photographed subjects in the confines of our small garden, I think the R7 excelled itself.

Auto focus is lightening fast, once you acquire a subject. I get the feeling that this is where the R7 appears a little slower than the 80D. I'd point the R7 in the approximate direction of a bird in the sky. All is out of focus, but then the R7 slowly brings a subject into focus, at which point AI tracking takes over. I feel that the initial getting a subject in focus is slower on the R7 than 80D - but it could be my imagination.

Unknown said:

Hello, John.

Canon R5! Wow. That's car buying territory for me. I'd suggest drag the thing out and get clicking. Glad you found the info useful.

Angus

I can't wait to play with it.

I've had the R5 a couple of weeks, though my father inlaw passed away soon after I joined the forum, so things have been a bit preoccupied. Thankfully, for a country man, he lead an organised life and that has made the finals more or less straightforward.

Did you see the info about how the sensor works, and the resultant reflection difference?

That could explain why action shots of people can look very distorted, or at least that was what I was warned of when making the purchase....

The image below is taken from that video, look at the wings of the swallow on the right and the reflection of the same swallow on the water. I've circled the two different representations.

What a palaver to upload images!

Hello John,

Edited to move "top to bottom (I think)" as it was a bit vague where it was.

I think that is rolling shutter - I've come across the term shutter lag as well. It takes time to scan across the sensor, top to bottom (I think). The swallow is beating its wings very quickly. By the time it takes to scan from the wing to the reflection, the wing has moved - thus the wing reflection does not align with the wing when the scanner passed it.

When switching to fully electronic shutter the R7 has a warning message about this.

My R7 has a particularly slow scan time - bit strange as it was designed for wildlife photography. Thus I should see a lot of rolling shutter.

I think the R5 has a much shorter scan time than the R7, therefore these distortions should be less apparent, if show up at all.

Switching to electronic 1st curtain or mechanical shutter should mitigate or remove this distortion altogether. Thus, in the Tufted duck photos, there was no evidence of rolling shutter, even though tufties flap their wings very rapidly. I was using electronic 1st curtain.

Unknown said:

Hello John,

Edited to move "top to bottom (I think)" as it was a bit vague where it was.

I think that is rolling shutter - I've come across the term shutter lag as well. It takes time to scan across the sensor, top to bottom (I think). The swallow is beating its wings very quickly. By the time it takes to scan from the wing to the reflection, the wing has moved - thus the wing reflection does not align with the wing when the scanner passed it.

When switching to fully electronic shutter the R7 has a warning message about this.

My R7 has a particularly slow scan time - bit strange as it was designed for wildlife photography. Thus I should see a lot of rolling shutter.

I think the R5 has a much shorter scan time than the R7, therefore these distortions should be less apparent, if show up at all.

Switching to electronic 1st curtain or mechanical shutter should mitigate or remove this distortion altogether. Thus, in the Tufted duck photos, there was no evidence of rolling shutter, even though tufties flap their wings very rapidly. I was using electronic 1st curtain.

I'll put my hand up here, camera tech is not my domain, whereas electronics is, or was now I'm retired.

The scan time makes perfect sense, and for those who've tried to take photos from the old TV cathode ray tv screens will know exactly what is being said. if the shutter speed was too fast, then there would be drop outs on the final picture because the CRT refresh rate was too slow.

Not a problem with today's faster refresh rates and processing for tv screens, using LCD, LED, TFT or OLED.

I would also ponder the thought that some delay may be down to processor speeds, which may seem negative, but isn't when you grasp the full process. That is also why sometimes when taking multiple shots on one shutter press, the TFT screen at the back shows that the camera is working its way through processing all those photos you've taken, and at frame rates of 30 per sec, that can be a lot of data to process.

OK, I've been playing a lot with the R5 settings, and getting, well, interesting results.

While at a nearby Wildlife Trust reserve, I did meet up with another chap, who also has an R5, and it was good to share experiences, and, they are experiences to build on, not hide away from.

Even better, his understanding of electronics and mirrorless is on a similar level to mine, so we're not quite two steps forward one step back, yet, but we are at the one step forward on step backward stage. So that has to be a bonus for now

Interestingly, we both have the same issue, when taking photos on bright sunny days, the image is dark. I ditched all mine, thinking they were beyond editing, but I've since found out, using the Canon camera download software, this can easily be manipulated, or even using a reputable third party editing program, likewise it can be edited.

It seems I may have innocently ditched a lot of photos, when they could have been saved.

The eye tracking has been a very interesting experience, but still one that needs more work from me as the operator, not the camera. Ideally I need to work more with moving subjects, panning was the old term in sports photography. That is yet to come....

Using the eye tracking

Uncropped, a black bird on a neighbours chimney pot

Cropped

One final ask for now, to those with retirement experience, how do you find the time to play!

I now wonder how I managed to find the time to work, be a family man, all among many other duties, and there was me thinking life would be at a slower pace.....

Oh well, back to the drawing board as they used to say.....

I'm not sure if it's good or bad news, but there's not a general problem with the R5 underexposing images in bright sunshine, so it must be something you're doing :-)
Worth making sure exposure simulation is switched on, at least you'll quickly see in the viewfinder if there's something wonky with your exposure settings that way. Without knowing how you're using the camera, it's difficult to suggest why you're getting dark pics, but I'd take a stab at guessing you're using spot metering & the camera is metering on a bright part of the image, resulting in most of the pic being dark. If so, try centre-weighted instead. That generally gives more consistent results