Leaked: Canon EOS M50 Image & Specifications

[Don Haines]

Whoa, 16-bit stills will be first rolled out to out to a mid-level crop camera? Isn't this a heretofore only available in medium format sort of spec?

If so, why on earth would Canon first launch this in the middle of the crop portfolio? If/when this ever happens at Canon, it would be an exclusive 1-series level feature (perhaps a rebirth of the 1Ds3 'pure IQ, screw the throughput' sort of product), would it not?

- A

Lots of stuff gets introduced in lower end bodies.... the update cycle is quicker and the market is more competitive. Besides, do you really think that they would wait for the 1DX3 to introduce the next new feature?

 

[neuroanatomist]

Where's the codec, Harry? Where is Canon's new medium format camera that looks like a big 1D X II, Harry?

Got my CODEC running on a 7D mark2 right now - Works great! Does 16-bit (kinda faked since the 7D ADC's don't actually go that high), 14-bit, 12-bit, 10-bit, 8-bit 4:4:4, 4:2:2, 4:1:1 and 4:2:0 colour sampling at various bit-rates at various fps.

I bricked the 7D a few times which means back to the test bench and a forced BIOS re-upload. Taking apart and putting back together a 7D takes about 4 hours each time! I had the same problem Magic Lantern did with their BIOS hook-in strategy.

I will get the 5D mk3/4 version up next and then publicly release the codec on GitHub for the mentioned cameras. The 1Dx series, C200, C300, C500, C700 will come later once I am sure I don't/can't brick our $10,000+ camera systems!

It's much cheaper to test/brick/destroy a 7D, 6D and 5D than a $40,000+ C700!

--- I should note that I actually DON'T MIND bricking the C700 since I can take apart that one for a forced BIOS re-upload A LOT EASIER than the 7D to 1D series!

Oh, ok.

 

[mistaspeedy]

What I am more interested in is the ability to do great low ISO dynamic range AND nice mid-high ISO dynamic range at the same time.
So far, we have had APS-C sensors tweaked to do one OR the other, but not both at the same time.

{dxomark source of data} If we compare the 200D (whose sensor is a bit better than the 80D), with the 7D mark II... we see a low ISO dynamic range advantage for the 200D, but they are equal by ISO 800, with the 7D mark II taking the lead at ISO 1600 and upwards.

7D mark II has a tiny bit less noise from ISO 800 onwards (the difference getting greater the higher you go).
Similar situation with tonal range and color sensitivity... high ISO favors the 7D mark II, whilst the low ISO 100 setting (less sensitive on the 200D) gives it an advantage in all areas over the 7D mark II.

Long story short:
200D wins low ISO
7D mark II wins high ISO

Can we get both in the same camera? The Sony a6500 obliterates both these cameras across the entire range, both low, mid and high ISO.

Maybe this new Canon sensor will close the gap at least a little bit.

 

[privatebydesign]

What I am more interested in is the ability to do great low ISO dynamic range AND nice mid-high ISO dynamic range at the same time.
So far, we have had APS-C sensors tweaked to do one OR the other, but not both at the same time.

{dxomark source of data} If we compare the 200D (whose sensor is a bit better than the 80D), with the 7D mark II... we see a low ISO dynamic range advantage for the 200D, but they are equal by ISO 800, with the 7D mark II taking the lead at ISO 1600 and upwards.

7D mark II has a tiny bit less noise from ISO 800 onwards (the difference getting greater the higher you go).
Similar situation with tonal range and color sensitivity... high ISO favors the 7D mark II, whilst the low ISO 100 setting (less sensitive on the 200D) gives it an advantage in all areas over the 7D mark II.

Long story short:
200D wins low ISO
7D mark II wins high ISO

Can we get both in the same camera? The Sony a6500 obliterates both these cameras across the entire range, both low, mid and high ISO.

Maybe this new Canon sensor will close the gap at least a little bit.

No the 200D is better at low iso but they are the same above 800iso.

http://photonstophotos.net/Charts/PDR.htm#Canon%20EOS%20200D,Canon%20EOS%207D%20Mark%20II

 

[mistaspeedy]

It depends which source you use and how you interpret the data.
The one thing missing from the source you linked (but is taken into account in dxomark's graphs) is the fact that the 7D mark II is closer in sensitivity to the marked ISO rating. (7D mark II exposure will be brighter if all other factors are equal).
But even in the link you supplied, you can see the small advantage for the 7D mark II.

This can even be seen in a real world example on dpreview's studio shot comparison scene. When set to low light and both cameras set to ISO 3200 or above. Just look anywhere over in the darker left side, and you can see the 7D mark II is cleaner.

 

[privatebydesign]

It depends which source you use and how you interpret the data.
The one thing missing from the source you linked (but is taken into account in dxomark's graphs) is the fact that the 7D mark II is closer in sensitivity to the marked ISO rating. (7D mark II exposure will be brighter if all other factors are equal).
But even in the link you supplied, you can see the small advantage for the 7D mark II.

This can even be seen in a real world example on dpreview's studio shot comparison scene. When set to low light and both cameras set to ISO 3200 or above. Just look anywhere over in the darker left side, and you can see the 7D mark II is cleaner.

I trust my linked source more because his methodology is open and honest and repeatable, DXO calculations are black box with no open methodology and no repeatability.

I don’t trust much at DPReview either as the comparisons rely on Adobe basic processing and that is often very easy to best. The only way to make fair comparisons is to take open source raw data or work on a best possible processed file. I linked to the only one of either I know of.

P.S. At 3200iso there is 0.14 of a stop difference in DR, yes the 7D MkII is the better but, seriously, 0.14 of a stop!

 

[Jack Douglas]

Oh, ok.

 

[ritholtz]

It depends which source you use and how you interpret the data.
The one thing missing from the source you linked (but is taken into account in dxomark's graphs) is the fact that the 7D mark II is closer in sensitivity to the marked ISO rating. (7D mark II exposure will be brighter if all other factors are equal).
But even in the link you supplied, you can see the small advantage for the 7D mark II.

This can even be seen in a real world example on dpreview's studio shot comparison scene. When set to low light and both cameras set to ISO 3200 or above. Just look anywhere over in the darker left side, and you can see the 7D mark II is cleaner.

I trust my linked source more because his methodology is open and honest and repeatable, DXO calculations are black box with no open methodology and no repeatability.

I don’t trust much at DPReview either as the comparisons rely on Adobe basic processing and that is often very easy to best. The only way to make fair comparisons is to take open source raw data or work on a best possible processed file. I linked to the only one of either I know of.

P.S. At 3200iso there is 0.14 of a stop difference in DR, yes the 7D MkII is the better but, seriously, 0.14 of a stop!

SL2 noise pattern is also little better, I think. If you look at 100%, shadows look bad even at lower ISO levels for crop cameras.

https://www.the-digital-picture.com/Reviews/Camera-Noise.aspx?Camera=1141&Test=0&ISO=3200&CameraComp=1044&TestComp=0&ISOComp=3200

 

[ahsanford]

Whoa, 16-bit stills will be first rolled out to out to a mid-level crop camera? Isn't this a heretofore only available in medium format sort of spec?

If so, why on earth would Canon first launch this in the middle of the crop portfolio? If/when this ever happens at Canon, it would be an exclusive 1-series level feature (perhaps a rebirth of the 1Ds3 'pure IQ, screw the throughput' sort of product), would it not?

- A

Lots of stuff gets introduced in lower end bodies.... the update cycle is quicker and the market is more competitive. Besides, do you really think that they would wait for the 1DX3 to introduce the next new feature?

No, just the reeeeeeeally juicy ones. That would be a pretty major milestone, wouldn't it?

- A

 

[mistaspeedy]

According to DXOmark's data, there is an absolutely huge difference between the 200D and Canon M3.
When you set the camera to ISO 3200, the actual ISO on the 200D is only 2125! Whilst the actual ISO on the M3 is close-to-ideal ISO 3133. This is nearly a quite massive 50% difference in sensitivity (3133 is 47.4% more than 2125).
7D mark II gives 2458 ISO at the same setting.

This has real-world implications for shooting. If you were using a certain ISO and shutter speed on the Canon M3, you can no longer use it on the 7D mark II or 200D... you would possibly need to bump the ISO up one stop to get a proper exposure.

We are not doing an apples to apples comparison in dynamic range tests if we set both cameras to 'ISO 3200' in the settings.

We could arbitrarily slap the 'ISO 3200' label onto a camera with a real ISO 100 setting, and it would beat all other currently manufactured cameras.

This is why comparisons like this give a clearer picture than the link you offered, which is clearly labelled "Photographic Dynamic Range versus ISO Setting'... unfortunately there is a discrepancy between "ISO setting" and "actual ISO sensitivity", a huge one.

https://ibb.co/n0zLQx

 

[rrcphoto]

According to DXOmark's data, there is an absolutely huge difference between the 200D and Canon M3.

that was a long post about something not that relevant anymore, considering the M3 has been replaced by the M5 and M6 a while back.

 

[mistaspeedy]

That's my whole point.... the old M3 is better than the new 200D, M5 and M6 at high ISO dynamic range.

https://ibb.co/gJCwyH

Canon M5, Canon 200D and all the newer APS-C cameras have been tweaked for maximum low ISO dynamic range, at the expense of high ISO dynamic range!

The M3 and 7D mark II are still the best APS-C cameras for high ISO dynamic range!
We can see concrete side-by-side comparisons here:
https://ibb.co/n5uLQx

We can have:

1) Good low ISO dynamic range
2) Good high ISO dynamic range

Not both!

So I'm wondering where this new Canon sensor will stand in this whole balance. Will we get both? Will it continue the trend of great low ISO dynamic range at the expense of high ISO dynamic range?
It will be interesting to see.

 

[Talys]

That's my whole point.... the old M3 is better than the new 200D, M5 and M6 at high ISO dynamic range.

https://ibb.co/gJCwyH

Canon M5, Canon 200D and all the newer APS-C cameras have been tweaked for maximum low ISO dynamic range, at the expense of high ISO dynamic range!

The M3 and 7D mark II are still the best APS-C cameras for high ISO dynamic range!
We can see concrete side-by-side comparisons here:
https://ibb.co/n5uLQx

We can have:

1) Good low ISO dynamic range
2) Good high ISO dynamic range

Not both!

So I'm wondering where this new Canon sensor will stand in this whole balance. Will we get both? Will it continue the trend of great low ISO dynamic range at the expense of high ISO dynamic range?
It will be interesting to see.

Is this not also a function of megapixels? ie the more megapixels, the harder it is to have great high ISO performance.

 

[eosuser1234]

Seems that the M5 only point of difference would only have the ability to touch and drag AF.
Other than that, seems this M50 may be a better spec camera all around. Will be interesting to see the buffer, but with the RAW CR3 files being able to compress better, may be able to hold more in buffer when shooting in raw.

 

[Talys]

Seems that the M5 only point of difference would only have the ability to touch and drag AF.
Other than that, seems this M50 may be a better spec camera all around. Will be interesting to see the buffer, but with the RAW CR3 files being able to compress better, may be able to hold more in buffer when shooting in raw.

If the CR3 format is smaller or has the option of compression (like Sony A7R3), that could really help with buffer clear times.

For me, I'm still waiting for the high(er) ISO performance unicorn. Unless mid-range ISO (800-3200) performance increases quite a lot, it could be a while before I buy another APSC camera.

White seems like an odd color for M50 (instead of silver)

 

[Sharlin]

Seems that the M5 only point of difference would only have the ability to touch and drag AF.

As has been said repeatedly, it's clear that there are major ergonomic differences. The M50 is missing at least two dials and probably C modes, just like x00D bodies compared to x0D.

 

[Sharlin]

Whoa, 16-bit stills will be first rolled out to out to a mid-level crop camera? Isn't this a heretofore only available in medium format sort of spec?

If so, why on earth would Canon first launch this in the middle of the crop portfolio? If/when this ever happens at Canon, it would be an exclusive 1-series level feature (perhaps a rebirth of the 1Ds3 'pure IQ, screw the throughput' sort of product), would it not?

Wasn't this 16-bit stuff pure speculation? Moving to 16 bit shouldn't necessitate a major file format update anyway AFAICS. CR2 is just a container format, like TIFF, capable of storing multiple datasets of all kinds of formats.

 

[IglooEater]

Whoa, 16-bit stills will be first rolled out to out to a mid-level crop camera? Isn't this a heretofore only available in medium format sort of spec?

If so, why on earth would Canon first launch this in the middle of the crop portfolio? If/when this ever happens at Canon, it would be an exclusive 1-series level feature (perhaps a rebirth of the 1Ds3 'pure IQ, screw the throughput' sort of product), would it not?

- A

I agree it does seem like a bit of a stretch. Let’s think of other reasons they might upgrade to CR-3. Because well, wild speculation is fun. ;D

 

[Diltiazem]

Dumb question.

Why call it CR3?
Canon could have made all the changes and still could have called it CR2. What am I missing?

 

[Diltiazem]

CR should have provided a link to original source, especially because last two bullet points are CR's own opinion and has nothing to do with original rumor.

http://www.nokishita-camera.com/2018/02/eos-m50.html