EOS 5D Mark IV To Use EOS C300 Mark II Technology? [CR2]

[Canon Rumors]

We’re told that the upcoming Canon EOS 5D Mark IV sensor will employ the same dual-channel readout as the EOS C300 Mark II.</p>
<p>The source says:</p>
<blockquote><p>… this essentially works as a hardware implementation of what the Magic Lantern guys did with software, and actually takes it a bit further.</p></blockquote>
<p>There has been no real confirmation that the EOS 5D Mark IV will shoot 4K video like the EOS C300 Mark II. Although, a lot of people almost expect it to be able to.</p>
<p>An announcement for the EOS 5D Mark IV is not expected until late summer at the earliest.</p>

 

[3kramd5]

If so I hope it's still raw.

 

[Maximilian]

Sounds promissing.

 

[bmwzimmer]

If True, then take my money Canon!!!

 

[IglooEater]

Any chance the 5d iv will have 15 stops of latitude as well?

 

[Orangutan]

the upcoming Canon EOS 5D Mark IV sensor will employ the same dual-channel readout

Simultaneously innovative and conservative! An interesting way to provide a significant DR boost without huge capital expenditure. If it works well, and doesn't kill too much resolution, this should serve as a decent bridge for another few years.

Question for the chip-heads: can Sony sensors do dual readout, or is that precluded by the on-chip A/D, or something else...?

 

[9VIII]

Yay Magic Lantern!

Any chance the 5d iv will have 15 stops of latitude as well?

That's basically what the article says.


Exciting as that thought may be, if more DR means less resolution, I would rather have 12 stops of DR in a 30MP+ 1DxMk2

 

[bmwzimmer]

Any chance of Canon implementing this on the 5DS/5DSR?

 

[3kramd5]

Any chance the 5d iv will have 15 stops of latitude as well?

I doubt the output will be >14 bits. Maybe if this option requires tiff it may be 16 bit.

 

[Random Orbits]

Any chance of Canon implementing this on the 5DS/5DSR?

Most likely not. The Westfall videos released when the 5DS/5DSR were released indicated that DR will be similar for 5DIII.

 

[siegsAR]

That being hardware based and then some I'd expect it would be much "cleaner".

Software interpolation might cause the resolution bit w/ ML, hardware should should fix that.

 

[dolina]

I wonder if they'll be using CFast... In all seriousness it should only be applied to Cinema EOS first before trickling it down to the 1-Series and 5-Series.

 

[Light_Pilgrim]

If True, then take my money Canon!!!

I actually do not understand the announcement. So what are you willing to pay for? Read it a couple of times and did not understand. Thanks for an explanation

 

[lintoni]

I haven't looked at the C300 specs... is this going to be an implementation of Canon's dual pixel chip design?

 

[-1]

No 5D4 at all then... Say hi to the 5Dc!

 

[andrewflo]

That would be a very nice addition to the 5D series. Good to see some innovation coming from Canon (well... really from the Magic Lantern guys).

 

[Marsu42]

… this essentially works as a hardware implementation of what the Magic Lantern guys did with software, and actually takes it a bit further.

Unbelieveable, I didn't think Canon would copy any ML innovation because af sheer principle. I hope, now they have to officially ack that ML isn't there to threaten their brand and maybe be of more assistance in development. Giving an "unofficial" sdk to the guys would save manyears of dumb reverse engineering.

And it gives a hint to the anti-drones 'round the CR forum that having more dynamic range esp. in video isn't something to be ignored ... now please do raw video too, will you, Canon?

That being hardware based and then some I'd expect it would be much "cleaner". Software interpolation might cause the resolution bit w/ ML, hardware should should fix that.

I wouldn't be so sure there - an in-camera "software" implementation actually relies on the PC hardware, i.e. much more computing power and it can take all the time it takes to get optimal iq. An in-camera "hardware" design is limited by time and computation constraints.

If it's the ML dual_iso idea, there's no 'round the fact that some deep shadows and highlights are missing in each other scanline and have to interpolated. It just depends on the algorithm to do it, not where it's done.

 

[Stu_bert]

… this essentially works as a hardware implementation of what the Magic Lantern guys did with software, and actually takes it a bit further.

Unbelieveable, I didn't think Canon would copy any ML innovation because af sheer principle. I hope, now they have to officially ack that ML isn't there to threaten their brand and maybe be of more assistance in development. Giving an "unofficial" sdk to the guys would save manyears of dumb reverse engineering.

And it gives a hint to the anti-drones 'round the CR forum that having more dynamic range esp. in video isn't something to be ignored ... now please do raw video too, will you, Canon?

It also indicates that this is either more cost effective than going the same route as Sony's solution, or as I read on FredMiranda, Sony patents prevent Canon from doing on-chip ADC.

But in true Canon fashion, the perception will be to puzzle & frustrate anyone thinking of the 5Ds cameras if the IV does get it. Given that a lot of this feature set was agreed internally probably about 18 months ago, and given the gap between the C300 II and the 5Ds, it is bizarre if this tech appears in the mk IV and not in the 5Ds. One can perhaps only conclude that the target market for the 5Ds was not asking for this... hmmm....

Even more interesting if the IV has it and the 1DX II does not....

 

[horshack]

Is the 5DM4 rumor of dual-ISO readout for video or stills or both?

 

[PureClassA]

Any chance of Canon implementing this on the 5DS/5DSR?

Most likely not. The Westfall videos released when the 5DS/5DSR were released indicated that DR will be similar for 5DIII.

If it really is a simple matter of firmware implementation like ML, then I don't see why they wouldn't. There is also the possibility that at least in part, this is why the delay between announcement and release is so long. It seems the capability is there UNLESS of course this is something geared towards DPAF, which the 5DSR will not have so far as we know. I would expect the 5D4 however to be DPAF. Since they say "its a hardware issue" I suspect it IS directly related to DPAF.