Advice about Canon vs *the others*

Hardware
49

No SLR lenses have that as well. Even my decades old Sony a65 had that .

Mirrorless lenses CAN be made smaller compared to what we’ve been used to from the SLR era.

The question is if that really happens. Or if lens makers use the weight to improve the quality of the glass (since modern sensors are more demanding for pixel peepers) .

A Tamron 70 to 180 2.8 e-mount is a massive difference in size and weight compared to the Sony a-mount 70 200 2.8 which i also still own . I call that lens 'the beast ’ (while i know 150-600 lenses or 400mm f4 lenses are even bigger).

Does the size/weight benefit come from more modern lens building? The 20mm missing at the long end ? Or the fact it’s for a mirrorless mount?
I don’t know. Probably all of the above. I do know i actually like taking the 70 180 with me, and i can’t remember the last time i used to 70 200 a-mount beast. It handles like a SLR style 70-300 f4-f5.6 ‘starter lens’, but gives you f2.8 . I think it’s awesome when you compare them.

But… To counter my own findings. As a smaller camera i went with an Oly omd em10 mark1. I loved how small it is, specially with a cheap entry level pancake lens. And I picked it because I wanted something that still gives me the 'photography experience’s , like back button focus , viewfinder , custom settings , etc…

But for a small camera ,the Oly has some metal elements and is heavier then ii think for its size.

Now, i replaced my Sony a65 SLR for a Sony a7mark2 some time ago . And that Sony a7 with a cheap entry level pancake lens… Did make me stop using my Oly for a while, because the Sony combo was actually not really different in size and weight !

I basically only use entry level lenses for my Oly (since i see it as the cheaper lighter 2nd camera). And i love that i can pack my whole kit in a little 9" camera insert. Body, effective 28mm pancake , effective 35mm pancake, effective 50mm, effective 90mm, pancake collapsible default kitlens, and a very small and light effective 85-300 entry level zoom . That whole kit fits in the same size as my Sony a7 with a 85mm 1.8.

So, i don’t think the mirrorless aspect is the real space saver, it’s more the thing if you want the best of the best or allow yourself to use cheaper and slower glass.

Still … my Tamron 28-75 2.8 e-mount lens is a bit smaller and quite light compared to my old 17-50 2.8 APS-C only a-mount lens . So steps have been made ! (Higher quality , full frame , smaller/lighter vs the older. But a no-compromise Sony 24-70 2.8 will be huge no matter what ).

1 Like
50

Sony a7 mark2 has the uncompressed format. It makes the files huge, though.

My older Sony a65 has that compressed format, and it is way overplayed what the impact is. Never had an issue with it, to be honest.

It loses out with edges with very high dynamic range.
But it’s not like it is a JPEG style compression where the pixels have artefacts.

You really have to try to make that old format break. Like in brightly lit neon signs at night.

It basically picks a base ‘exposure’ for every cluster of pixels (9 I believe), and then the pixels inside that cluster get only 7 bits of dynamic range. But those 7 bits are stored lossless.

So you get issues if there is a stark exposure difference inside a cluster. Because it can’t describe that the whole cluster is very dark, but an extremely bright pixel is still in it.

Anyway, that format is a bit old by now. The Sony a7mark2 got a firmware update way back to write files without any kind of compression.

Lately they have a 3rd format, best of both worlds. Lossless but compressed. But open-source support is still sketchy, I believe.

3 Likes
51

Hello Glenn,

might not be that interesting for you, but I just wanted to let you know, that your comment somehow stuck with me and lead to the decision to change systems.
Meanwhile I sold most of my canon stuff. I’m now experimenting with a Sony A7 iv, which got used for a fair price.

Currently I’m very happy with my decision and I just wanted to give back thanks for encouraging me, by sharing your experience.

Kind Regards
Robert

edit: somehow i got the wrong button, this should have been private… anyways I let it stand here.

3 Likes
52

That IBIS was such a game changer for me was totally unanticipated. I used to be sensitive to < 1/30sec, dating back from my film days in the '70s, but now that number is 1/10sec. I just take a breath, gently squeeze the shutter, and the camera gives me a sharp image, with all my lenses. Really, that too is about dynamic range, gathering more light rather than using ISO gain to make less light work with a hand-holdable shutter speed.

About evaluation: My very second job ever was selling cameras in a department store camera section. We had all the major brands and and all of their flagship models in the display case, available for handling. I don’t remember a customer asking to run a roll through any, but a lot of discussion was had on what one could see/feel just with it in their hands. The one camera we did keep loaded with film was the then-new Polaroid SX-70. I would pull that camera from the case, hand it to the customer, and tell them to take a few pictures. Proof’s-in-the-pudding, most walked away with the pictures they took, as well as a purchase…

These days, it’s hard to get a sense of confidence buying a device that costs north of $1KUS; so many factors to consider, just staring at a picture and product description on Amazon. The reviews help somewhat, if they’re of sufficient quantity and not too peppered with shills. My experience is Nikon-centric, but there are general things about it I feel are worth considering, no matter what brand one is looking at…

7 Likes
53

@qmpel regarding this image from the 80D where you tried the +4EV trick, if you still have the RAW, can you try enabling denoise (profiled) with the wavelets: chroma only preset in darktable and see if that easily cleans it up? Or try another technique for reducing chroma noise and see if one of those makes it a lot better. I know you switched to Sony, but I’d still be interested to see if this makes a difference for the above photo

1 Like
54

@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.

here is a 100% crop

image
image1278×872 498 KB

and a smaller view of the complete picture

image
image865×582 122 KB

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 … :thinking:

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).

I hope this helps somehow?!

3 Likes
55

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

56

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 :man_shrugging: 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.

57

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).

58

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.

1 Like
59

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.

60

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).

image
image2162×1249 884 KB

One starts to think if the difference justifies the expenditure :man_shrugging: 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 :wink:

1 Like
61

So, of course I had to do this comparison. :slight_smile: 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… :man_shrugging:

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.

image
image2560×1440 2.6 MB

3 Likes
62

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).

63

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.

64

That’s the conclusion I came to also.

While doing the tests, I was impressed with the performance of the 80D and you can probably get it for 1/4 of the price (second hand).

65

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.

66

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.

67

The 80D is from 2016, which was 9 years ago

There must be a time warp in the interwebs, since I’m reading this from 2023…

5 Likes
68

I forgot the “never do math before your morning coffee” rule. :blush:

6 Likes