Canon EOS R5 Specifications

[AlanF]

I wasn't doubting you, just surprised that they were going for that much cheaper than what I see listed in the used dept. at places like B&H and Adorama as well as reputable resellers like KEH and lensauthority.

I guess there is a difference, though, between what they're paying for them and what they are reselling them for. I don't do eBay for anything over about $100, buy or sell.Too many scammers here. An "excellent plus" listing on ebay might be "excellent" at KEH or a"9" at B&H, or it might be a "fair" or "7". You never know, because too many sellers tend to overrate them on eBay.

Perhaps you have had bad experiences eBay, as I am sure some have, but I have bought and sold a lot of high value items on it and so far have been unscathed. WEX has a "manager's special" today of a grade 9 used 300/2.8 II for £3,099 ($3,900). I am afraid that us owners of big whites have been royally done over by Canon. Long gone are the days when buying glass was an "investment". Maybe it's a good time to buy used and a sensible one to keep on with our old EF glass to use with adapters.

 

[Michael Clark]

Perhaps you have had bad experiences eBay, as I am sure some have, but I have bought and sold a lot of high value items on it and so far have been unscathed. WEX has a "manager's special" today of a grade 9 used 300/2.8 II for £3,099 ($3,900). I am afraid that us owners of big whites have been royally done over by Canon. Long gone are the days when buying glass was an "investment". Maybe it's a good time to buy used and a sensible one to keep on with our old EF glass to use with adapters.

I've never considered cameras or lenses "investments." They've always been expenses to me. Same thing with vehicles and everything else the marketers try to sell to us by fooling us into thinking it's an "investment" rather than an "expense." If you depreciate it on your taxes, it's not an investment.

 

[Quarkcharmed]

Choosing a view size is not an arbitrary definition of DR. It's an arbitrary but necessary input into the formula for DR.

The classic DR is defined as "ratio between the maximum output signal level and the noise floor at minimum signal amplification". There's no viewing distance or print size involved.
What they have on DxOMark and PTP is a special photographic DR that's good for comparison between sensors, but again and again its absolute value is unusable.

 

[Joules]

What they have on DxOMark and PTP is a special photographic DR that's good for comparison between sensors, but again and again its absolute value is unusable.

Since you mention unusable so often, I'm curious what you even want to use the number for. I don't think I've read that here, but with how many posts this thread has, I skipped a good few of them.

 

[Kit.]

That's exactly why absolute DxO or PTP values are meaningless - they're based on arbitrary choices for normalisation. If the measurement depends on one's arbitrary interest and definition of the DR, you can't do it scientifically.

There is no other definition of DR. DR always describes a signal, and the choice which signal is of interest to you is always "arbitrary". When you know which signal is of interest to you, you can do it scientifically.

You get drastically different results on DxO and PTP (up to two stops I believe). The DR of your sensor reduces as you come closer to the print. There's something intrinsically wrong about it.

That's because "the DR of your sensor" is not even a thing. There are no DRs of "sensors". There are DRs of signals transmitted through channels. And your sensor, being a part of your transmission channel, affects the DRs of different signals differently.

"As you come closer to print", you just refine the idea what signal you are actually interested in.

Per-pixel DR however is invariant.

It's not invariant (it's still different for different signals). It's just irrelevant, at least when we are talking about photography.

Imagine two different sensors with the same resolution and the same "per-pixel DR" (for the number of incoming photons(*) as the signal you are interested in). Both sensors have their DR mostly limited by read noise. But in one sensor the read noise is uncorrelated between different pixels, in the other it results in random banding. Do you really have no preference which one to choose?

*) let's abstract from photons' own quantization noise, as it is the same for both sensors.

 

[SecureGSM]

Not really. At venues where there are armys of photographers who use 300/2.8 or 400/2.8 lenses, there's more than enough light to use a 1.4X with a 300. It's a LOT lighter, a LOT cheaper, more flexible, and works almost as well for the few longest distance shots to carry a 300/2.8 + 1.4X as it does to carry a 400/2.8. In the U.S., outdoors we're talking primarily baseball and american football. The pace of both of those and the way one shoots them give plenty of time to decide whether one wants 300mm or 420mm on the "long" body. The other bodies have 70-200s and/or 16-35s hanging on them. A 400 does make more sense for baseball, where 400 + 1.4X can be useful at times. But college baseball gets very little coverage, and there are only 30 MLB teams, compared to 32 NFL and 254 Division 1 (129 FBS + 125 FCS) football teams.

Not many I know or have seen shoot in indoor gyms with 300s for most sports. They're too long too, unless one is shooting from the rafters in a very large arena (or from the upper seats in the end of a mid-size gym like with volleyball). Maybe gymnastics, but there are a LOT more shooters on the baselines for basketball and on the sidelines and end lines for football, even with mid-level colleges, than the number shooting even major college gymnastics and volleyball. I don't know what hockey shooters use. But again, there aren't near as many hockey shooters in the U.S. as there are shooters covering football, basketball, and baseball.

"... Not many I know or have seen shoot in indoor gyms with 300s for most sports...."

Yes, that is why I shot with 120-300 / 2.8 indoors

 

[Michael Clark]

"... Not many I know or have seen shoot in indoor gyms with 300s for most sports...."

Yes, that is why I shot with 120-300 / 2.8 indoors

120-300mm is a lot different than 300mm prime. We were talking about EF 300mm f/2.8 or EF 400mm f/2.8.

 

[Quarkcharmed]

Since you mention unusable so often, I'm curious what you even want to use the number for. I don't think I've read that here, but with how many posts this thread has, I skipped a good few of them.

In DxO or PTP, the DR numbers are good for comparison between the cameras measured with the same method.
But DxO and PTP give different DR values for the same cameras, so the numbers are useless outside the respective database (from DxO or PTP).

How else would you use say a DR value of 13 stops from DxO? Your target print size changes or viewing distance changes and the number becomes irrelevant.

So for example, Sony's alleged 15-stop DR of A7RIV is also a useless gimmick, as it's meant to be the photographic DR similar to DxO, when 61Mp is downsampled to 8Mp. Perhaps they could downsample to 1Mp and get 16 stops, or not downsample and get some 13 stops.

A usable value is the per-pixel DR that tells me the max contrast in a real scene and how it's reflected in the resulting RAW image when viewed 1:1.

 

[Quarkcharmed]

There is no other definition of DR. DR always describes a signal, and the choice which signal is of interest to you is always "arbitrary". When you know which signal is of interest to you, you can do it scientifically.

Per-pixel DR of CMOS sensors is well defined. Only it's hard to use for comparison between sensors with different resolution and crop factor. But it basically shows shadow recoverability when viewed 1:1.

It's not invariant

It doesn't depend on the print size and viewing distance.

Imagine two different sensors with the same resolution and the same "per-pixel DR" (for the number of incoming photons(*) as the signal you are interested in). Both sensors have their DR mostly limited by read noise. But in one sensor the read noise is uncorrelated between different pixels, in the other it results in random banding. Do you really have no preference which one to choose?

But I think DxO and PTP don't take banding into account either.

I try to make my images look good at 1:1 so I do care about the per-pixel DR. Some of them I only put on Instagram, some I print small, some print large, some use large sized (say one of them made it to a page of a tourist book - they cropped it but needed 3000 pixels on the longest side).

DPReview has good charts where you can view samples at 1:1, btw.

 

[AlanF]

I've never considered cameras or lenses "investments." They've always been expenses to me. Same thing with vehicles and everything else the marketers try to sell to us by fooling us into thinking it's an "investment" rather than an "expense." If you depreciate it on your taxes, it's not an investment.

Very true. However, there was a time a few years back when inflation was higher and lenses were not updated frequently that price rises in new lenses meant that used lenses could be sold for their original purchase price or more. Nowadays, Canon reduces some prices with time. There were loads of posts here about “investing” in glass.

 

[Kit.]

A usable value is the per-pixel DR that tells me the max contrast in a real scene and how it's reflected in the resulting RAW image when viewed 1:1.

Except that it doesn't. What it reflects is the characteristics of image viewed pixel by pixel, with all spatial information about particular pixels lost.

Per-pixel DR of CMOS sensors is well defined.

Based on an arbitrary choice of a "signal" that has neither intrinsic nor practical meaning for a sensor with an AA filter (let alone Bayer mask).

Which reminds me of the streetlight effect.

But it basically shows shadow recoverability when viewed 1:1.

"Shadow recoverability" is an arbitrary subjective value. What is "recovered shadows" for one viewer is not for another.

It doesn't depend on the print size and viewing distance.

It doesn't depend on the actual end result? Cool... it still doesn't mean that it's invariant, though.

But I think DxO and PTP don't take banding into account either.

If they put different weights on different spatial frequencies of the noise, they do. Something that's impossible to do with "per-pixel DR".

I try to make my images look good at 1:1 so I do care about the per-pixel DR.

"Per-pixel DR" doesn't tell you how good your images look at 1:1, unless by "1:1" you mean "one pixel at a time".

 

[Quarkcharmed]

That's just not true between these two cameras. The gap is not that wide.

It's 0.8 stops or so. That means I can push shadows on D850 0.8 stops farther. Sometimes it means the difference between usable and unusable image. It doesn't happen too often, but it happens.

 

[BillB]

Canon users aren't going anywhere. Remember, the Body pros use is frankly the most replaceable element of their craft. Glass has far more longevity as do other accessories. They will buy RF cameras as soon as they feel they can use them to replace their existing cameras, and they will maintain their existing glass libraries until they feel the need or desire to upgrade

Fair enough. Photographers can and will use AF glass after they buy RF cameras, but what Canon is interested in is how many more EF lenses they will buy once they have RF cameras.

 

[BillB]

Not sure if you've seen this trend as well but I think 3rd party EF lens mfg.s are starting to discount more so than in the past. Gotta get on the RF bandwagon.

Not only that. As the market keeps shrinking, manufacturers are getting squeezed, likely some more than others.

 

[BillB]

The classic DR is defined as "ratio between the maximum output signal level and the noise floor at minimum signal amplification". There's no viewing distance or print size involved.
What they have on DxOMark and PTP is a special photographic DR that's good for comparison between sensors, but again and again its absolute value is unusable.

True, but for some of us, the question is whether there is a discernible difference in print quality at relevant viewing distances.

 

[AEWest]

That other 25% is still worth millions of dollars in sales. EF won't go away until those millions are not there anymore. I doubt it will be ten years (though I think many EF lenses will still receive service/support in 2030), but it will be far more than a year or two. It will be at least five years before Canon no longer sells EF lenses, probably longer, and they'll service whatever they sell for around seven additional years.

Canon will provide service and support for EF lenses, I have no doubt. But there will come a point when enough R cameraa have been sold that they will stop production of EF version of R lenses. It will make no sense to produce both.

 

[slclick]

indoors 400 is a bit too long, outdoors 300 is a bit too short.

Whoa! When I think indoors my mind goes to Basketball and typically you're at 35 under the rim, 50-85 baseline and MAYBE an occasional 135 across court for tight. Of course, indoors could be Kart racing at 400!

 

[AEWest]

Perhaps you have had bad experiences eBay, as I am sure some have, but I have bought and sold a lot of high value items on it and so far have been unscathed. WEX has a "manager's special" today of a grade 9 used 300/2.8 II for £3,099 ($3,900). I am afraid that us owners of big whites have been royally done over by Canon. Long gone are the days when buying glass was an "investment". Maybe it's a good time to buy used and a sensible one to keep on with our old EF glass to use with adapters.

Unfortunately that is the market talking, not Canon. Over the past 5 years, ILC camera sales have been cut in half. That would tend to leave to an over supply in used lenses and thus falling prices.

 

[brad-man]

Canon will provide service and support for EF lenses, I have no doubt. But there will come a point when enough R cameraa have been sold that they will stop production of EF version of R lenses. It will make no sense to produce both.

In economics, there is the concept of Supply and Demand. As long as customers are willing to pay for EF lenses, Canon will produce and sell them. Canon makes a profit from the sales of EF lenses. Canon's plan is not to force people into switching to RF. They are producing RF lenses that are superior to their EF counterparts so folks will want to switch...

 

[AlanF]

Unfortunately that is the market talking, not Canon. Over the past 5 years, ILC camera sales have been cut in half. That would tend to leave to an over supply in used lenses and thus falling prices.

That is true but it is has been exacerbated by Canon. For example, the 400mm and 600mm III coming out after a relatively short time after version IIs.