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

[raptor3x]

wow. Canon desperately wants to avoid creating new sensors at all costs.

Have you intelligence that the C300II uses old an sensor? I would bet you sight unseen that it's new.

Canon likely wants to avoid retooling if the can, not "at all costs," but to avoid costs.

If they can improve their camera output without a new fab, why shouldn't they?

If it really is a hardware implementation of the ML increase in DR, then it's added DR at the expense of resolution. So instead of coming up with a more sensitive sensor, they are sacrificing one aspect of the image for the sake of the other.

Not if they use a layout similar to what's used in the DPAF sensors where you have twice the resolution in one direction.

 

[9VIII]

But, but, all these friendly folks on the forums insist that more than 12-stops isn't necessary.

That hasn't changed, and yes, 12 Megapixels is excessive... for facebook.

The problem with people saying "such and such a spec isn't good enough", is they make their arguments blind to any application.

 

[3kramd5]

wow. Canon desperately wants to avoid creating new sensors at all costs.

Have you intelligence that the C300II uses old an sensor? I would bet you sight unseen that it's new.

Canon likely wants to avoid retooling if the can, not "at all costs," but to avoid costs.

If they can improve their camera output without a new fab, why shouldn't they?

If it really is a hardware implementation of the ML increase in DR, then it's added DR at the expense of resolution. So instead of coming up with a more sensitive sensor, they are sacrificing one aspect of the image for the sake of the other.

It's all speculation at this point, but I doubt they're just adjusting firmware. I bet they changed their signal chain to be able to amplify the entire sensor readout twice simultaneously.

 

[Jokke_r]

The mention of magiclantern doesnt meam theyre doing the every other line different iso, which was a hardware limitation.
Im fairly sure they are just reading the sensor once, duplicating the signal and pumping it through different adc's and then combining the two resulting exposures. It would be stupid and slow to try and do ML's implementation in camera since the interpolation algorithms are extremely CPU intensive and slow, unless they developed hardware process of doing that work for them in realtime, which is idiotic since that would be more work than simply developing the hardware to duplicate the sensor output which should be fairly trivial in comparison and result in a full resolution composite and be much much faster since no interpolation is required.

 

[IglooEater]

Um, as to 12 stops of DR being enough, what is exciting us, or me at any rate, isn't so much the 3 more stops, but the fact that the c300II maintains that dynamic range at higher iso's. 15 stops at ISO 100 is one thing, but 15 stops at ISO 102,400 is another thing altogether. I for one would love to have 3 more stops at low ISO, but I'd really lkill for that much higher up the ISO ladder

 

[LetTheRightLensIn]

They did have a dual-ISO per photosite read out patent put out a few years ago. It's just like what Dr. Martinec suggested a few years prior in the DPR forums. It's much better than the ML hack, since you don't lose any resolution. It's real.

AFAIK, it would require a better fab than what they have made all CMOS DSLR sensors through the 5Ds on though. Did they decide to bump P&S off the 180nm copper pipe fab (can this one even handle DSLR sensors)? Secretly build a new fab without anyone noticing the big $$$? Make a deal with some other fab to make it for them? Manage to somehow find a way to squeeze the patent out of 500nm?

Absolutely awesome new if this is true though. It would even beat the best the D810 can do by a little.

Of course it just makes the 5Ds look all the more Canon. There they have a camera that screams out for this tech more than anything but it's like marketing was nope, gotta only deliver one thing at a time, gotta save stuff for the 5Ds2. So we bump the MP way up but make sure to keep it on old sensor tech even though this camera, of all cameras, would be the one most begging for the new sensor tech and then we of course totally cripple the Crop mode so it can't dare become an all-around camera like the D810, let's make sure the crop mode won't apply to RAW shooting to nobody gets the benefit of larger buffer, more fps, storage savings, gotta save that for the Mark II!!

And then with the 5D4 they are like OK we will keep the MP count low like 22-24MP so no we can dare to use the new sensor tech. One hopes they don't also hold back 4k, so typically Canon.

But all the same, if this rumor is true, that really would be great news!! The 5Ds would look really like a silly marketing made wreck but at least we know that 5D4 and all the sensors will be awesome again and the 5D4 might end up potentially better than the D810 in all ways other than for MP count (and maybe color) and the next round in another three years or so should really be great stuff.

But before too much excitement, the rumor might be pure BS, might have been misinterpreted along the way, some of it might be fake reading into slog2, etc. etc. but if it is the dual-ISO patent that would likely be pretty awesome.

 

[LetTheRightLensIn]

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

At 8MP/4k equivalent (which is what DxO normalizes too so you can directly compare to their Print DR numbers) it certainly could. The Canon DSLR base sensors are supposed to be 15+ stops for a long time, but their old fab meant they had to do some of the read out stage off chip and that just killed all the DR. A simultaneous dual-ISO read of each photosite would fix things a lot and I wouldn't be surprised if it could get 15 stops at 8MP normalization out of it. Maybe they are using or farming out to a new fab or found some way to cram the dual-ISO similtaneous readout circuitry onto the old fab process.

 

[LetTheRightLensIn]

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.

It either kills a ton of resolution, if done the ML way, which would be a joke. Or they are using the dual-ISO simulatneous read per photosite patent, which would be awesome. Many think that patent requires better than their old 500nm fab to implement though so either they are able to make DSLR sensors on their 180nm copper pipe fab and bumped the P&S stuff off that finally, somehow secretly spent a lot of money on a new fab that nobody noticed, made a deal to farm it out to some other fab for production, found some scheme to manage to make it on 500nm fab.

Assuming off course this is real and not someone just speculating and then sending in their made up dreams as a supposed legit rumor and that if a real input something wasn't misinterpreted.

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

Canon holds the patent for it, so probably not.
They don't need to though since they use their on chip parallel ADC system.

 

[LetTheRightLensIn]

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

nope, although it would have made all the sense in the world, 5Ds cried out for it more than any model
it's already been measured and it definitely does not do this (only if by some miracle they held back a better sensor and could swap it in if too much public outrcy, but I mean, that seems like fanciful thinking! and it's due in stores so soon, basically no way)

 

[LetTheRightLensIn]

… 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.

They didn't copy ML if this is real, simultaneous per photosite ISO read out and if all it is is the ML hack in firmware, that won't fly! That's totally fake and hits resolution.

First, Emil Martinec first talked about such a scheme years ago on DPR.
Second, I'm pretty sure Canon had a patent out on this prior to ML's dual-ISO hack too.

 

[LetTheRightLensIn]

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

I suppose it could be that it is not the dual-ISO read patent that some think and something that only works for the slower shutter speeds of video and thus video only on the 5D4 too?
Hopefully not.

 

[LetTheRightLensIn]

... or as I read on FredMiranda, Sony patents prevent Canon from doing on-chip ADC.

I do not buy the Sony patent explanation. Samsung did it on the NX1, didn't they? Obviously there are many ways to achieve the same goal.

yes

all those who do that have modern fabs though

 

[LetTheRightLensIn]

If it is, I wonder how they manage to have reduced rolling shutter by 2x on the C300 II though, they'd have to be reading the sensor mad fast if it was just this sort of method.

I wonder if it is just some upgraded version of the HDR movie mode from the new T6s, which seems to have been overlooked by just about everyone. I have no idea how well it works but here's a brief bit about it - from this page:

http://www.usa.canon.com/cusa/consumer/products/cameras/slr_cameras/eos_rebel_t6s_ef_s_18_135mm_is_stm_lens_kit#Features

 

[LetTheRightLensIn]

But, but, all these friendly folks on the forums insist that more than 12-stops isn't necessary.


Kind of like how friendly Nikon folks insisted that more than 12MP wasn't necessary...until the D800 came out.


In all seriousness though, this is exciting. If this camera delivers on DR and video, I may stay with Canon rather than switching to Sony. This is your last stand Canon, make it a good one!

It is funny how excited all those trashing the 'Droners' are the second their is some talk about Canon possibly getting even just a trace more DR than Exmor! ;D ;D ;D ;D :

 

[LetTheRightLensIn]

If it takes 20.000 USD to get 9.4 megapix of this technology - I cannot see the relevance for a 5D-whatever.

Just does not match up at all.

It takes $16k to get a C300 II with a a pixel-for-pixel perfect 4k density with 15 stops of DR, yes..... Plus the bucket loads of other high end cinema dedicated features, XLR and SDI inputs/outputs, recording capabilities, 10 variations of built in ND filters, waveform/zebra/focus peaking/etc displays and all with the capability of recording out in 14 Bit RAW or 4:4:4 ProRes.

I was hoping they would have priced it closer to 10-12k but oh well. Still can't compare the two.

Plus the margin on the C300 II has gotta be HUGE too. HUGE.

 

[LetTheRightLensIn]

wow. Canon desperately wants to avoid creating new sensors at all costs.

Have you intelligence that the C300II uses old an sensor? I would bet you sight unseen that it's new.

Canon likely wants to avoid retooling if the can, not "at all costs," but to avoid costs.

If they can improve their camera output without a new fab, why shouldn't they?

If it really is a hardware implementation of the ML increase in DR, then it's added DR at the expense of resolution. So instead of coming up with a more sensitive sensor, they are sacrificing one aspect of the image for the sake of the other.

no, if it's in HW then they can apply it to ALL photosites and it's something new.

The ML version already IS in the HW which is how ML can change the firmware to make it happen already.

 

[tyger11]

To me, the timing of this leak is suspicious. I wouldn't be surprised if Canon saw the comments online about the too-expensive-for-most C300 Mk II and the general disgust at the XC10 and figured if that's all they've got for NAB, people will start looking at non-Canon offerings at NAB. So leak this news of some awesome 5D Mk IV with the C300 Mk II sensor (which doesn't at all imply the same camera capabilities), so try to get people to wait that much longer.

 

[lintoni]

To me, the timing of this leak is suspicious. I wouldn't be surprised if Canon saw the comments online about the too-expensive-for-most C300 Mk II and the general disgust at the XC10 and figured if that's all they've got for NAB, people will start looking at non-Canon offerings at NAB. So leak this news of some awesome 5D Mk IV with the C300 Mk II sensor (which doesn't at all imply the same camera capabilities), so try to get people to wait that much longer.

And the moon landings were faked; the Malaysia Airlines tragedies were a false flag operation organised by the illuminati and the US govt has incontrovertible proof for the existence of ETs, but is hiding it.

 

[LonelyBoy]

But, but, all these friendly folks on the forums insist that more than 12-stops isn't necessary.


Kind of like how friendly Nikon folks insisted that more than 12MP wasn't necessary...until the D800 came out.


In all seriousness though, this is exciting. If this camera delivers on DR and video, I may stay with Canon rather than switching to Sony. This is your last stand Canon, make it a good one!

It is funny how excited all those trashing the 'Droners' are the second their is some talk about Canon possibly getting even just a trace more DR than Exmor! ;D ;D ;D ;D :

I think all the excitement is just being able to tell everyone who's been threadcrapping all over every forum to finally shut up.

 

[quod]

And the moon landings were faked; the Malaysia Airlines tragedies were a false flag operation organised by the illuminati and the US govt has incontrovertible proof for the existence of ETs, but is hiding it.

And Canon doesn't cripple its cameras or time its releases... Um, yeah it does.