Canon EOS 5D Mark IV Talk [CR1]
- Thread starter Canon Rumors
- Start date
Mr Majestyk said:
As many have already covered, the 5D line (not 5DS) has historically only gotten one chip, and 36 MP x 8 fps would require more than two at present throughput -- that would be moving more data than the 1DX II!
Tempering your best possible case just a bit, 30 X 8 is about in line with the 5DS's 50 x 5, but again, that would bring a second chip to a $3500 professional camera line that heretofore only gets one for no reason whatsoever. Why would Canon do something practical and useful like that?
- A
Upvote
0
The real question is: presuming your throughput is capped at one DIGIC chip, and let's presume that throughput is half of the 1DX II, what would prospective 5D4 owners want more?
24 MP x 6 fps?
28 MP x 5 fps?
32 MP x 4.5 fps?
(Hint: none of those should get you super excited.)
Again, I think Canon is batsh-- crazy to put two chips in a 7D2 and 5DS -- two completely different rigs -- and then not do same for the 5D4. Put two chips in the 5D4 and make a compelling product -- 24 x 10 fps, 28 x 8 fps, etc.
- A
24 MP x 6 fps?
28 MP x 5 fps?
32 MP x 4.5 fps?
(Hint: none of those should get you super excited.)
Again, I think Canon is batsh-- crazy to put two chips in a 7D2 and 5DS -- two completely different rigs -- and then not do same for the 5D4. Put two chips in the 5D4 and make a compelling product -- 24 x 10 fps, 28 x 8 fps, etc.
- A
Upvote
0
IMO 32 MP is too much!
I could live with it, if there were an easily recognizable improvement in IQ over the whole ISO range and ISO 3200 looking like 1600 on the 5D3, hopefully even better. But if not, Canon, please stay below 28 MP.
If IQ is not improving enough, I'd 100% skip this generation.
You know that your high res camera without AA filter already exists? It's called 5 DS R!
If you're not pleased with it's IQ performance, it seems that you'll have to skip one generation, too.
Or you'll have to jump boat.
And If you're not pleased with the FPS then be sure that with 32 MP it will be at max at the level of the 5D3.
Everything always comes with a tradeoff as well.
I could live with it, if there were an easily recognizable improvement in IQ over the whole ISO range and ISO 3200 looking like 1600 on the 5D3, hopefully even better. But if not, Canon, please stay below 28 MP.
If IQ is not improving enough, I'd 100% skip this generation.
Hi Guys!RGF said:
You know that your high res camera without AA filter already exists? It's called 5 DS R!
If you're not pleased with it's IQ performance, it seems that you'll have to skip one generation, too.
Or you'll have to jump boat.
And If you're not pleased with the FPS then be sure that with 32 MP it will be at max at the level of the 5D3.
Everything always comes with a tradeoff as well.
Upvote
0
So stop whining and buy yourself a 5DS.Mr Majestyk said:
Most photographers will not need huge number of megapickles. Something in the mid-20s is sufficient for most photographers. It is all about the image quality, not the image quantity. The other side to ever growing megapickles is the need for faster computing power. Do you want to drop $3500 on a new camera body just to find out you need to drop as much on a new computer, or more?
I'd be happy with the MP count going down if the image quality goes up.
Upvote
0
Simon Young said:
Actually, the 5Ds and 5DsR have the best dynamic range of all Canon sensors ever produced besides the 2 brand new 80D and 1DX mark II with onboard ADCs.
If you filter the results to show only APS-C sensors, then again, the newest Canon 750D and 760D have the best dynamic range of all Canon APS-C sensors... pixel density seems to have zero influence on dynamic range... if anything, averaging down a large number of pixels to a smaller size increases dynamic range. The same goes for Nikon (Sony sensors)... their highest resolution body also has the highest dynamic range (Nikon D810) with 36 megapixels.
http://www.dxomark.com/best-canon-cameras-for-landscape
http://www.dxomark.com/best-nikon-cameras-for-landscape
The Canon G7X uses a Sony sensor.
Upvote
0
It could be a slight crop to allow more fps and not waste quite as much space when shooting reach limited wildlife/sports?
24MP out of 32MP is a little odd, certainly not APS-H nor APS-C sized crop.
24MP out of 32MP is a little odd, certainly not APS-H nor APS-C sized crop.
Upvote
0
bdunbar79 said:
Years ago, before CR, Northlight actually used to have a great track record. Often getting things well hinted at even years before release with their roadmap leaks.
Granted the times have changed.
Upvote
0
ahsanford said:
the years old 5D3 already does 24MP at 6fps more or less, that would be kinda crazy specs after all these years, especially since the DR seems it won't be exmor level (even if better than 5D3 for sure), the sams 24MP and 6fps?
Upvote
0
The higher megapixel count means smaller individual pixels on the physical sensor, meaning less light gathered by that pixel, meaning less sensitivity to light.
If they can make a 32MP camera with exactly (if not better) light sensitivity at ISO100, then I'm on board. Otherwise, I will gladly take it as a 24MP camera.
I don't get the option of choosing different megapixel modes. That part sounds sketchy.
Looking forward to seeing what the official announcement is. I know Canon must feel like they're under the microscope with this camera.
If they can make a 32MP camera with exactly (if not better) light sensitivity at ISO100, then I'm on board. Otherwise, I will gladly take it as a 24MP camera.
I don't get the option of choosing different megapixel modes. That part sounds sketchy.
Looking forward to seeing what the official announcement is. I know Canon must feel like they're under the microscope with this camera.
Upvote
0
Correct, those numbers are based on 140-145 Mpixels/sec. From the 1Dx2, we know that a single DIGIC6+ should be able to push around 175Mpixels/sec. So the MP/fps numbers should be:LetTheRightLensIn said:
24 MP x 7.3 fps
28 MP x 6.25 fps
32 MP x 5.5 fps
Upvote
0
kaihp said:
Coincidentally, the best rumour we've got for the 5D IV so far is that it's going to be 24 MP and do 7 FPS.
I'd be happy with that, and happier with 8 FPS (software improvement?), but that's mostly because I think eight is a nicer number, not because 8 FPS would be dramatically better than 7 FPS.
I sometimes think that my ideal camera would be something like 80 MP with a 20 MP mode, where the 20 MP images were exactly what they'd be if it was a native 20 MP sensor. I'd use 20 MP most of the time, but every so often it's nice to have lots of pixels.
Upvote
0
If the chip is the limiting factor, then I think the numbers should be slightly higher.
Those numbers are based on 14fps.... but don't forget that the 1DX mark II is capable of 16fps in mirror lockup mode (the mirror being the limiting factor for 14fps mode).
So if we do the math, we can have the following with a single chip (obviously they will round down to a full or half FPS):
24 MP x 6.733 FPS
28 MP x 5.77 FPS
32 MP x 5.05 FPS
It is possible that the 1DX mark II isn't hitting the limits of dual digic chips, and that there is slightly more bandwidth that could be squeezed out of it. So... I think at least 7fps at 24 MP and 6 FPS at 28 MP should be possible.
However, even with the above numbers, were are forgetting something very important. (if we wish to speculate about chip power).
The Canon 80D can do 7 FPS @ 24.2 MP = 169.4 MP per second.
The Canon 80D does this with an ordinary 'Digic 6' chip, whilst the chip in the 1DX mark II is a 'Digic 6+', which should be even better.
Don't forget about video processing too...
4096 x 2160 at 60fps = 530 MP/s for the 1DX mark II so the bandwidth available is even higher.
My guess of what is possible with a single Digic 6+ chip is as follows (I'm not saying Canon will decide to implement this though):
24 MP at 11 FPS
28 MP at 9 FPS
32 MP at 8 FPS
50 MP at 5 FPS
If these numbers seem unrealistic to you, I would like to point out that the Sony a6300 does 11 FPS at 24.2 megapixels.
It seems logical that Canon's top chip should be able to perform similarly.
Those numbers are based on 14fps.... but don't forget that the 1DX mark II is capable of 16fps in mirror lockup mode (the mirror being the limiting factor for 14fps mode).
So if we do the math, we can have the following with a single chip (obviously they will round down to a full or half FPS):
24 MP x 6.733 FPS
28 MP x 5.77 FPS
32 MP x 5.05 FPS
It is possible that the 1DX mark II isn't hitting the limits of dual digic chips, and that there is slightly more bandwidth that could be squeezed out of it. So... I think at least 7fps at 24 MP and 6 FPS at 28 MP should be possible.
However, even with the above numbers, were are forgetting something very important. (if we wish to speculate about chip power).
The Canon 80D can do 7 FPS @ 24.2 MP = 169.4 MP per second.
The Canon 80D does this with an ordinary 'Digic 6' chip, whilst the chip in the 1DX mark II is a 'Digic 6+', which should be even better.
Don't forget about video processing too...
4096 x 2160 at 60fps = 530 MP/s for the 1DX mark II so the bandwidth available is even higher.
My guess of what is possible with a single Digic 6+ chip is as follows (I'm not saying Canon will decide to implement this though):
24 MP at 11 FPS
28 MP at 9 FPS
32 MP at 8 FPS
50 MP at 5 FPS
If these numbers seem unrealistic to you, I would like to point out that the Sony a6300 does 11 FPS at 24.2 megapixels.
It seems logical that Canon's top chip should be able to perform similarly.
Upvote
0
mistaspeedy said:
I already accounted for the 16fps
The 1Dx2 pushes 20.2 MP x 16 fps in mirror lockup mode with two DIGIC6+. That's 323.2 MP/sec, or just under 162Mpix/sec per processor. So the ~175Mpix/sec is above that number, but is close to what the ratio of Mpix/sec between the 1Dx and 5D3 (253.4Mpix/sec : 133Mpix/sec = 1.90:1) indicates it should be (170Mpix/sec).
The argument for upping the single-processor number slightly is the speculation that it is the 4K capability that is really the pace-setter, not the stills performance.
Upvote
0
Which side of the processor is the buffer?
If it's between the sensor and processor, then the speed of the processor doesn't directly limit FPS, it just determines how quickly the buffer is read.
If it's between the processor and storage, then it's only needed when cards aren't fast enough, which should never be the case when using today's fastest cards, so I'm guessing this isn't where the buffer is.
If it's between the sensor and processor, then the speed of the processor doesn't directly limit FPS, it just determines how quickly the buffer is read.
If it's between the processor and storage, then it's only needed when cards aren't fast enough, which should never be the case when using today's fastest cards, so I'm guessing this isn't where the buffer is.
Upvote
0
I totally agree that the video processing is what takes most of the power, and that stills performance is not really a factor anymore... it is how much the chip can send and how fast the mirror can flip (and can the autofocus keep up with this)... it's up to Canon to set the framerate, but today, the 'bottleneck' isnt going to be the processor for stills performance - in my opinion.
Upvote
0
keithcooper
Canon Rumors Premium
If you'd read the quoted link you'd see I also put it in a similar context. As ever, I take most suggestions from people unwilling to use their name, with a general large pinch of salt ;-)tron said:
I don't have a problem with 32MP, it helps differentiate from a 6D2, doesn't quite reach the 5Ds and makes a 100+MP 5Ds2 look more of a jump...
Upvote
0
Similar threads
