@garibaldi, sorry for the late answer. Yes, I’m able to clean this image up. And for family photos I’m actually not too concerned about some noise - most family members view the pictures on their phone anyways and when exporting to 1080p you can obliterate lots of noise.
In this picture the “denoise chroma only” does a good job.
Still, the picture shows the limitations of my (old) cam. For shots of my children I prefer 1/250 or less exposure times. Even with F2.8 that brings me in shadowy indoor situations into (even more) high iso areas. Which leads to more noise, than I’m willing to accept. What I didn’t show here, are 80% of my christmas pictures with really high noise level. Had to work a lot to get something presentable …
The other main point is the auto focus. I really wanted a fast eye focus to get good focused pictures of playing children without spamming burst shots (I find these very distracting).
Thanks for the detailed response - yes this helps! I’ve also run into this challenge with low light indoors and I’ve started to wonder if the solution (for me) is to get a faster lens (e.g. 50mm f/1.4) to avoid the high ISO. In certain situations I’ve found my speedlite (bouncing off the wall behind me of course for diffusion) really helps too.
Regarding the eye focus on fast moving children, this is something I am interested in as well. You found that the 80D was not good at this? Reviews online really praise the 80D’s AF system and even it’s eye tracking, so I’m surprised to hear this
Yes, a faster lens certainly helps. I had the 50mm F1.4 canon lens (now sold). It gets more light, but it also produces (of course) a tiny depth of field. So, for my family photos I didn’t use it too much. And regarding flash light … I’m not a fan (I like to take pictures from the back, not getting noticed - so that I get a few authentic moments and this makes it of course harder for the camera).
Actually you should take this with a grain of salt. I have to admit, I’ve apparently lived under a rock for some time (at least regarding photo technology) and the unsatisfactory focus is probably a combination of my inability and my old lenses (stuff had mostly slow focus motors, i realize now). A modern lens might have helped there.
It was mainly selling the old gear I wasn’t even using, that provided me the opportunity to start from scratch and make new decisions for my current needs.
AFAIK with the Canon cameras when you use a fast lens (opens at f/2.8 or more) the camera can use a more accurate AF sensor at the center (in the XXXD series) and possibly elsewhere (XXD, XD).
AF systems in consumer-grade MILCs have improved a lot in the past few years. Assisted with machine learning (which, for some reason, manufacturers call “AI”), you can just tell the camera what to track and it will be in focus most of the time, finding features like eyes if possible. Sony is still the best at this, probably, but other manufacturers are not far behind. Yes, the lens matters, and sensors with PDAF give an advantage, but Panasonic’s DFD and Canon’s Dual Pixel tech are also competitive.
This feature is mostly about computational capacity of the dedicated CPU in the camera, and algorithms making use of it. It is probably the largest difference between current and previous models for most manufacturers, if you photograph things that move, especially people. It is, unfortunately, not something you can boil down to a single number, like megapixels. It is best to read reviews (DPreview now tests this, shooting stills with a biker that goes left and right), or take an overactive toddler to a camera store and try out various cameras.
If you want to stay in DSLR waters 90D is a great camera with many features of the mirrorless world. In live mode the focusing is super fast, has focus peaking (great for manual lenses), can detect the eyes, touch sensitive etc.
I’m not entirely sold on Canon mirrorless, mostly because I’d need an adapter like 100% of time and R native lenses are 2x expensive than EF ones. And batteries run down so fast, while on 90D I used to make over 1400 RAWs out of a single charge.
FYI … just another comparison regarding low light / noise performance
I took a low light picture with both cams:
Sony A7 iv + Sigma 28-80/F2.8
Canon EOS 80D + Sigma 17-50/F2.8
I tried a fair comparison using the following settings:
1/25 sek. shutter speed on both
F4 on the A7 vs F2.8 on the Canon (same DoF)
ISO400 on the A7 vs IS160 on Canon (comparable ISO)
50mm on the A7 vs 35mm on Canon (same perspective)
I took the formulas from here
Nearly no post processing, just a crop to the same area (the A7 has more Megapixel, so on exporting to the same size, it gets an advantage here) and a little WB on the A7 picture to match the white a bit better (so, no sharpening or denoising in any form).
Ah yes, and I bumped the exposure +3EV.
The resulting picture, note the noise level, looks as follows (might need to zoom, to see differences).
One starts to think if the difference justifies the expenditure But then, the A7 has probably still more headroom going to F2.8 and also better performance in higher iso areas… would be subject for another test
So, of course I had to do this comparison. I got me a little bit more darkness and made a similar test for the Sony A7 iv (with the Sigma 28-70 / F2.8), the Canon EOS 80D and 50D (both with Sigma 17-50 / F2.8).
At this point its probably tiring for most people, but I’m dumping my result here anyways - perhaps someone out there is interested…
As above I tried to do a fair comparison with regards to field of view, depth of focus and ISO Settings (== equal shot), using the formulas in the above referenced article.
So the columns should be comparable regarding their settings (I think) and the rows contain the A7, the 80D and the 50D.
Thanks, this is actually very helpful to me. I can see the usual Canon “banding” is clearly present on the 50D but that seems to be gone on the 80D which is great. Other than that the 50D and 80D don’t look much different at high ISO.
It does seem like at ISO below 1600 there isn’t a ton of difference between the Sony and 80D (so if you plan to shoot at high ISO a lot, the Sony is a worthwhile upgrade).
This was in fact exactly what I did yesterday, and the results on the 80D are pretty underwhelming - maybe 1/3 in focus using the viewfinder and even less than that using subject tracking and the screen. Maybe this is to be expected in burst mode, e.g. if you were to instead shoot in Continuous AF and instead just snap a single shot, maybe it could keep up? I’d be shooting mostly with an EF-S 17-55 f/2.8 (e.g. the almost-L-quality EF-S lens) so I’d hope that one at least would be able to keep up.
This is the exact same dilemma I have; Canon seems to be focusing mostly on full frame for mirrorless and the price to jump all in to RF is high. Plus the need to buy the EF to RF adapter, battery life, etc. My understanding is that the 90D is a relatively modest upgrade from the 80D; it also doesn’t seem to be readily available used, whereas the 80D is widely available so I think I’d lean that route unless there’s a specific reason to really search for a used 90D.
You can pretty much predict tracking AF performance based on when the camera was designed. The 80D is from 2016, which was 9 years ago. Tracking needs computational power, and the older the chip, the less of that you have available.
This is the exact same dilemma I have; Canon seems to be focusing mostly on full frame for mirrorless and the price to jump all in to RF is high. Plus the need to buy the EF to RF adapter, battery life, etc. My understanding is that the 90D is a relatively modest upgrade from the 80D; it also doesn’t seem to be readily available used, whereas the 80D is widely available so I think I’d lean that route unless there’s a specific reason to really search for a used 90D.
To be fair, Canon started issuing cropped sensor ML cameras lately, R7, R10. There are other models too I think.
In my work I use both cropped and full frame, having 5DmkIV and 90D. I love them both, each one has its own beauty and use.
I’m not sure about used cameras, 90D can be purchased new as it is still in production. You can do online comparison of 80D and 90D and see if it’s viable.
This is true, but RF glass seems to be pretty expensive and there are only a couple of RF-S lenses available so far.
I suppose another option would be to consider the EOS M6 mk ii, which is newer and has good focus tracking, but then you’re buying into the end-of-the-line EF-M ecosystem. Maybe that’s not really a problem though?
Indeed. So far I have been more than underwhelmed by current RF glass offerings. It’s either budget-friendly or super high-quality, pro grade with insane prices. Nothing in-between or semi-pro.
On the other hand, RF 50mm f1.8 is inferior to the same EF 1.8 STM lens optically, yet almost 2x the price. A ripoff.
There is nothing like EF 50mm f1.4 on the horizon. And when it eventually arrives, it will probably cost 2 to 3x the price of EF 50mm f1.4.
If I were you I wouldn’t buy any M6, as you said it’s end of the line. Depending on your budget you can check R7 or R10. I remember there were promotions with a free adapter. But then again 90D was on sale too, costing less than 1k euros with 4y warranty.
@maboleth you articulated it perfectly; there seems to be no RF solution for the “prosumer”.
That’s true, but the R50 really seems to be geared towards people upgrading from using a smartphone for photos, not a “prosumer”. From what I read, its options for doing more manual adjustments are limited
What do people think about Canon’s policy on 3rd party lenses for the RF mount? I have two EF mount Canons and have been very happy with them, but if I could justify buying another camera, which I can’t, I’d think twice about the RF ones. Also Canon weren’t nice over CR3 files of course. Sony’s open approach to 3rd party lenses seems a lot better.
and this makes it of course harder for the camera).
I got me a little bit more darkness and made a similar test for the Sony A7 iv (with the Sigma 28-70 / F2.8), the Canon EOS 80D and 50D (both with Sigma 17-50 / F2.8).
