Jump to content
Andrew Reid

EOS R official video specs discussion

Recommended Posts

EOSHD Pro Color for Sony cameras EOSHD Pro LOG for Sony CamerasEOSHD C-LOG and Film Profiles for All Canon DSLRs
35 minutes ago, Shield3 said:

Not confused at all.  We're talking about total light gathering and equivalent noise and ISO.  The A7r2 is a bad example as the S35 mode is cleaner since the "binning" of the FF means at ISO 3200 and above it's loses detail and gets noisy quickly.  But you'll notice the A7sII is the opposite - no crop and using the entire sensor for the 4k.

The Canon 50 1.2 while shooting 4k video is really a 85 F/2 equivalent.

The 'equivalence' is to do with the FOV and DOF and nothing to do with exposure or 'light gathering' 

Share this post


Link to post
Share on other sites

Equivalent aperture is there to illustrate what the depth of field is going to look like.

The low light performance remains as bright as the real aperture.

Separate issue:

The low light performance will dip if the sensor is doing a 1:1 readout, because the noise is sampled 1:1 and therefore is as large and intrusive as you can get. With oversampling or downscaling the noise is reduced and you can involve further noise reduction processing in the pipeline there, which you can't do with a 1:1 crop.

This has been true and plain for all to see as far back as the 1:1 ExTele crop mode on the GH1.

It shouldn't really be controversial.

The 5D Mark IV was already not so hot at high ISOs either for video or stills, so that is another area of disappointment to chalk up on the board for the EOS R.

Share this post


Link to post
Share on other sites

But yes, it's more about DOF and effective focal length than light gathering.  I would submit the larger the sensor area, all things being equal, would gather more "total" light, even with the same light intensity.  Exposure will not change; F/1.8 is F/1.8.  People intertwine "light intensity" and "light gathering" and they're not the same.

But again - shoot a 50 2.8 on a full frame and compare that to a 25mm F/1.4 on a micro four thirds.  Same DOF / focal length.

Share this post


Link to post
Share on other sites
6 hours ago, ntblowz said:

Well you can get lens adapter with variable ND, so that just leaves the XLR

Sound Devices MixPre3 or something like the Panasonic DMW-XLR1, or if being super ultra frugal.... the Tascam DR60Dmk2

6 hours ago, jonpais said:

I’m looking, but just see this:

 On the plus side, the EOS R does have 4:2:2 10-bit video output externally via the HDMI port (internally, it's 8-bit).

Not seeing any mention of HDMI 4K out...


That would be a most hilarious troll (and classic Canon!) if they gave us 10bit 422 over HDMI but limited it to 1080!! :scream:

5 hours ago, Django said:

Release schedule:

DmUJku4VsAAP5mH.jpg


Key point there is the variable ND adapter won't arrive until February 2019! Ah well
 

 

5 hours ago, no_connection said:

With that bad rolling shutter it's pretty much dead on arrival, what point is 10bit if you can't use it?

"– Still a huge ass crop in 4K recording. Looks the same as 5D IV. Heavy rolling shutter as well. (info via Jordan Drake)"

And no sign of any speed booster to help with crop factor ether (at least not electronic), at best maybe metabones could hack something unofficial. No 3rd party RF lenses ether.

"– Canon will not be opening up the RF mount specs to third parties (info via Jordan Drake)"


Same as Nikon :-/ 
Ah well, at least no one has an advantage here. 

I wonder which will be unlocked first by a third party figuring it out on their own?
 

 

Share this post


Link to post
Share on other sites

Andrew, I agree the crop on the Eos R is a big disappointment, but the high data rate is a positive not a negative, not that it IMHO balances the crop issue, in 2018.

Your statements about ALL-I, IPB appear contradictory ...

When you reviewed the GH5 you had this to say about ALL-I ...

Andrew ... GH5 review extract ...

“These high bitrates especially in ALL-I are needed for 10bit to truly get the benefit of the finer gradation, especially useful for V-LOG – otherwise macro blocking occurs on plain surfaces and skies, and you may as well drop back to 8bit. I also prefer the motion cadence of an ALL-I codec to the more common space saving IPB.”

Andrew ... And today, Canon EOS R extract  ...

“The ALL-I H.264 codec is an all-hype codec. Those high bitrate numbers might look attractive on paper. The reality is it does not offer higher bitrates PER FRAME vs IPB mode at 120Mbit. It’s just very inefficient and hard to edit compared to superior codecs (like ProRes 422). File sizes are as unmanageably big as MJPEG but the image is worse than MJPEG – it’s 4:2:0, not 4:2:2.”

 

The New Canon has log 420 8 bit.

ALL-I is actually easier on a machine for editing purposes than IPB, it does take up more storage space, but you can double that again for ProRes 422.

Share this post


Link to post
Share on other sites
43 minutes ago, Mattias Burling said:

Cropfactor applied to aperture is wrong unless you add a long ass sentence about shutter speed and/or iso, sensor generation etc, etc, etc. It makes zero sense.

F2 is f2 is f2.

No. f/2 is f/2 only in exposure triangle. f/2 is not f/2 when considering DoF or (total) amount of light per photo (not per pixel).

Think about this - aperture is the size of the lens opening. If you do not use the whole lens opening, the whole aperture size, that means that you are using smaller than maximum aperture. What does it matter to you if lens has a certain aperture when you are not using the light that that aperture passes through - you are using only a part of it? So, on crop sensor that f/2 aperture is effectively smaller than it would have been on full frame senzor. What is really important - physical size of an opening, or size of that opening that is actually used for taking a photo?

You have to realize that F-stop of f/2 (or F2) is a relative aperture - relative to the focal length! Without focal length you have nothing! On the other and, real aperture is what defines DoF and total amount of light per photo. In other words, when using 50 mm f/2 lens on a FF body, real aperture size is 25 mm. If you use f/2 on 100 mm lens, real aperture size is 50 mm. Wow! Mind blown! The same relative aperture (the same F-stop) equates to different real apertures on different focal length lenses!

But, if you use equivalent F-stops - 100 mm F4 on FF gives aperture size of 25 mm, which is the same as 50 mm F2 lens - real aperture size in both of those instances is 25 mm. F-stop (relative aperture size) is different, but real aperture size is the same. And we can expect the same FoV with that combination, the same DoF, and the same amount of light per photo (which means we should drop ISO on MFT camera by 2 stops - and obviously, FF cameras have about 2 stops better ISO performance - everything checks out).

Share this post


Link to post
Share on other sites
1 hour ago, Shield3 said:

No one is saying the EXPOSURE is different.  The same intensity of light is hitting the sensor.  I am saying all things being equal, a larger sensor gathers more total light and has less read noise than a smaller one.  Have you forgotten that the micro 4/3rds cameras struggle past ISO 1600?

Micro 4/3 cameras struggle with low light not because they have a smaller sensor that gathers less light, but because in that smaller sensor they have the same number of photosites (= megapixels) other camera have in bigger sensor, so each photosite is smaller. Its the size of the photosite that matters, not the total number of them on the sensor. Each photosite send its electrical signal to the processor regardless of how many other photosites are around it.

Share this post


Link to post
Share on other sites
21 minutes ago, 1Ale82 said:

Micro 4/3 cameras struggle with low light not because they have a smaller sensor that gathers less light, but because in that smaller sensor they have the same number of photosites (= megapixels) other camera have in bigger sensor, so each photosite is smaller. Its the size of the photosite that matters, not the total number of them on the sensor. Each photosite send its electrical signal to the processor regardless of how many other photosites are around it.

Not necessarily. Yes, there is size of photosites you don't want to go below, but the main reason is that MFT sensors in general receive less light per photo. Consider this - if you want to take a photo in low light, you might use 35mm f/1.4 on FF. To get the equivalent photo on MFT, you need 17,5mm f/0.7 lens. Does that kind of lens actually exist? If you use nifty fifty on FF (50mm f/1.8), to give the same amount of light to MFT sensor you would have to use 25mm f/0.9 lens. Does that lens even exist? Do MFT users use lenses that are that fast? No? No. They usually just use lenses with the equivalent focal lenght, but the same F-stop (which means 2-stop less light, and therefor get about 2-stop more noise at the same ISO).

So, that is the answer why MFT struggles with low light - because of the sensor size, and not having adequately fast lenses which would compensate for that small sensor size. For MFT cameras not to struggle with low light, used lens must have about 2-stops better F-stop than FF camera.

Share this post


Link to post
Share on other sites
56 minutes ago, IronFilm said:

That would be a most hilarious troll (and classic Canon!) if they gave us 10bit 422 over HDMI but limited it to 1080!! :scream:

It would be even funnier if they gave us a 4K crop smaller than APS-C, wasted valuable real estate on a dedicated on/off wheel; and a silent shutter that isn't silent for continuous shooting.

Share this post


Link to post
Share on other sites

Thinking about this on my workout today.

Aperture (hole) size can be 1.4 on Micro four thirds and full frame.  But which hole is physically larger and thus allows more (total) light in?  The FF lens, of course!  If an eye dr. told me you have your pick - see through a F/1.2 25mm FF lens or a quarter crop sensor F/1.2 25mm, the FF one would be twice as wide and total light coming in would be greater.  Canon has essentially made the "eye" smaller in UHD mode - the rest of the sensor is doing nothing to gather light.  So not all F/1.2's are equal - shutter speed, ISO, aperture all make up the exposure for the given sensor size.  In the EOS R's case, since they're using only the center 1/3rd of the sensor, it's just like having 1.2 @ 1.7x - for both focal length and DOF.   Yes the camera will still read F/1.2, but it's *not* the same as a FF sensor gathering that F/1.2.

23 minutes ago, 1Ale82 said:

Micro 4/3 cameras struggle with low light not because they have a smaller sensor that gathers less light, but because in that smaller sensor they have the same number of photosites (= megapixels) other camera have in bigger sensor, so each photosite is smaller. Its the size of the photosite that matters, not the total number of them on the sensor. Each photosite send its electrical signal to the processor regardless of how many other photosites are around it.

Which is why they released a 42.5mm F/1.2 - it's like a 85 F/2.4 on FF.

Share this post


Link to post
Share on other sites
27 minutes ago, Marekich said:

No. f/2 is f/2 only in exposure triangle. f/2 is not f/2 when considering DoF or (total) amount of light per photo (not per pixel).

Think about this - aperture is the size of the lens opening. If you do not use the whole lens opening, the whole aperture size, that means that you are using smaller than maximum aperture. What does it matter to you if lens has a certain aperture when you are not using the light that that aperture passes through - you are using only a part of it? So, on crop sensor that f/2 aperture is effectively smaller than it would have been on full frame senzor. What is really important - physical size of an opening, or size of that opening that is actually used for taking a photo?

You have to realize that F-stop of f/2 (or F2) is a relative aperture - relative to the focal length! Without focal length you have nothing! On the other and, real aperture is what defines DoF and total amount of light per photo. In other words, when using 50 mm f/2 lens on a FF body, real aperture size is 25 mm. If you use f/2 on 100 mm lens, real aperture size is 50 mm. Wow! Mind blown! The same relative aperture (the same F-stop) equates to different real apertures on different focal length lenses!

But, if you use equivalent F-stops - 100 mm F4 on FF gives aperture size of 25 mm, which is the same as 50 mm F2 lens - real aperture size in both of those instances is 25 mm. F-stop (relative aperture size) is different, but real aperture size is the same. And we can expect the same FoV with that combination, the same DoF, and the same amount of light per photo (which means we should drop ISO on MFT camera by 2 stops - and obviously, FF cameras have about 2 stops better ISO performance - everything checks out).

Fake news

Share this post


Link to post
Share on other sites
18 minutes ago, Marekich said:

Not necessarily. Yes, there is size of photosites you don't want to go below, but the main reason is that MFT sensors in general receive less light per photo. Consider this - if you want to take a photo in low light, you might use 35mm f/1.4 on FF. To get the equivalent photo on MFT, you need 17,5mm f/0.7 lens. Does that kind of lens actually exist? If you use nifty fifty on FF (50mm f/1.8), to give the same amount of light to MFT sensor you would have to use 25mm f/0.9 lens. Does that lens even exist? Do MFT users use lenses that are that fast? No? No. They usually just use lenses with the equivalent focal lenght, but the same F-stop (which means 2-stop less light, and therefor get about 2-stop more noise at the same ISO).

So, that is the answer why MFT struggles with low light - because of the sensor size, and not having adequately fast lenses which would compensate for that small sensor size. For MFT cameras not to struggle with low light, used lens must have about 2-stops better F-stop than FF camera.

Voigtlander makes lenses that fast with those focal lenghts for m4/3.

Considering the "light gathering" topic, just consider the sensor size as simply the total number of photosites included on that silicon matrix, nothing more. The size of the silicon matrix is what we call medium format, full frame, apsc, m4/3, 1 inch sensors, etc. Then each photosite, regardless of how many of them are on the sensor, sends its own signal to the camera processor. The bigger the photosite, the lower the light needed to send that signal, keeping other things like time and aperture equals. You don't need more photosites to gather more light.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×