Canon 5D Mark IV -- Time to Discuss!

My memory and watching the history of the cameras has said the 1D cameras has always been dual processors and an independent processor for each meter and AF functions = 4. The 7D was dual processors and another that was unspecified for metere/AF functions. The 5D has always been single processor for image and another single for AF/Metering... except the 5DS which has dual processors.

Edit: I do believe the XXD and rebels use a single processor for imaging but they use a generic processor for the AF function.

This is the 5D3 https://static1.square​space.com …Main-PCB.jpg?format=2500w

The elpida chip is a memory module used in all the canon cameras but I belive elpida went bankrupt in 2012 or 13 something like that. Not sure where the source is. The other chip labeled "F74965A" is the AF/metering processor. So the 5D3 used a digic 5+ and a generic metering/af module that was used the same in the 7D, 60D... etc. The 1DX at the time used dual digic 5+ with a digic 4 for AF/Metering.

I did alot of research awhile back and this is what I uncovered.

For the 5D4 to use a 6+ for imaging and digic 6 for af/metering then this is why the 5D4 AF works better. Much more powerful processor handling the AF functions. This is a step closer to the 1D line. The 7DII I know for a fact uses the dual Digic processors for imaging but I thought uses the same F74965A chip for AF/metering

Colorblinded wrote in post #18135450
Doesn't really seem like there's enough information yet, it's a possibility, but I think I'll just keep an open mind and follow it until someone does a more in depth analysis.

If nobody has by the time I get a camera, I'll do it, although I don't keep my C1 up to date anymore and that would be interesting to compare as well.

All I need is a RAW to figure out how correctable the banding is. My question is, is the banding easily correctable by the camera but Canon has decided not to, or is the difference in blackpoint so small that 14 bits is not enough to mostly correct it in.

Any banding that is consistent all the way from the left to right will also be in the black border at the left edge of the RAW, where it can be measured, and heavily corrected, with extended precision.

Some converters may correct banding more than others, so this banding may be more visible with some converters than others, but that is not because they are creating it; it is because they are not correcting it, while other converters may correct it.

Talley wrote in post #18135608
My memory and watching the history of the cameras has said the 1D cameras has always been dual processors and an independent processor for each meter and AF functions = 4. The 7D was dual processors and another that was unspecified for metere/AF functions. The 5D has always been single processor for image and another single for AF/Metering... except the 5DS which has dual processors.

The xxD line and rebels I thought all use just a single processor for all functions.

what do you mean by single processors for image and another single for AF/metering? isn't that considered dual processors? or do you just mean "dual" = two identical processors?

mike_d wrote in post #18135601
Did the 5D3 have two processors also or is this new in the 5D line?

I'm wondering the same thing.

Technically the 7D Mk2 has dual processors, but the 5D Mk3 does not. It seems the 5D Mk4 does not either (technically).

John Sheehy wrote in post #18135643
All I need is a RAW to figure out how correctable the banding is. My question is, is the banding easily correctable by the camera but Canon has decided not to, or is the difference in blackpoint so small that 14 bits is not enough to mostly correct it in.

Any banding that is consistent all the way from the left to right will also be in the black border at the left edge of the RAW, where it can be measured, and heavily corrected, with extended precision.

Some converters may correct banding more than others, so this banding may be more visible with some converters than others, but that is not because they are creating it; it is because they are not correcting it, while other converters may correct it.

I haven't seen a raw file to look at, but even though it would indeed have to be inherent in the file, the amount of difference in output between the programs is interesting.

ilumo wrote in post #18135644
what do you mean by single processors for image and another single for AF/metering? isn't that considered dual processors? or do you just mean "dual" = two identical processors?

What I mean is teh rebel line and the xxd was always using a single processor to handle the imaging procesing and the AF/metering functions all at the same time. This could be wrong information though look at the 60D:

https://d3nevzfk7ii3be​.cloudfront.net/igi/rS​JGg3hHCTYBkass.huge

I haven't ever been able to figure out what the toshiba module is for. The other modules are memory chips.

The 5D3 uses a Digic 5+ for it's imaging and the FP module for it's AF and metering functions. When a camera company says "this has dual processors" it's referring JUST to it's imaging processors. The 1D series, 7D series and the 5DS all uses dual imaging processors to handle either the super high megapixel or the super fast framerate. The 1D has always had seperate processors in addition to the imaging processors to handle each metering and focusing. The 7D used dual digic 4 processors and an unknown processor to handle both AF/Metering = 3. The 7D2 appears to use dual digic 6 imaging processors and a spartan 6 processor to handle it's AF/metering. The 5D4 uses the digic 6+ for it's imaging and another digic 6 for it's metering/AF = 5D4 would have faster AF/metering computing.

Jared5 wrote in post #18135656
I'm wondering the same thing.

Technically the 7D Mk2 has dual processors, but the 5D Mk3 does not. It seems the 5D Mk4 does not either (technically).

See my post above. I'm not 100% accurate because the information is scarce but I did waste alot of my life researching it at one point. If I'm wrong shoot me but I'm pretty close to at least the processor count and what they are for.

gjl711 wrote in post #18135570
I take that back. both a 6 and a 6+. I was expecting them to look more or less the same.
http://www.lense.fr …k-IV-Circuit-Board-02.jpg

Ah, but in that drawing/schematic, where you see a Digic 6 written on the IC, that is actually a memory chip, it is a Micron Elpida IC. I don't trust that circuit board drawing, as we have a real bonafide teardown showing something else.

EDIT: I didn't see Talley's response, but we are in agreement. I see only one processor on the motherboard, as I run each chipset through google as best I can. Also Micron picked up the Elpida line not too long ago, it seems.

Why are people so intent on trying to create a banding problem that would never be present in real world shooting conditions?
If you expose properly you will never experience any banding problems with this camera..
Ok, so every now and again you will need to pull some shadows maybe a couple of stops but this sensor is more than capable of doing that....

don1163 wrote in post #18135691
Why are people so intent on trying to create a banding problem that would never be present in real world shooting conditions?
If you expose properly you will never experience any banding problems with this camera..
Ok, so every now and again you will need to pull some shadows maybe a couple of stops but this sensor is more than capable of doing that....

Experimentation, pushing the limits of the camera to see how it performs against other manufacturers.

gjl711 wrote in post #18135692
Experimentation, pushing the limits of the camera to see how it performs against other manufacturers.

overclocking is overclocking....

why have 1,000hp when the speed limit is 55mph.

...etc.

don1163 wrote in post #18135691
Why are people so intent on trying to create a banding problem that would never be present in real world shooting conditions?
If you expose properly you will never experience any banding problems with this camera..
Ok, so every now and again you will need to pull some shadows maybe a couple of stops but this sensor is more than capable of doing that....

Someone needs to be the first to discover what easter egg problem Canon intentionally cooked into the 5D4, so far all anyone's come up with is how it was "crippled" for not including an Ice Cold Gin dispenser, so now they are looking elsewhere.

Talley wrote in post #18135697
overclocking is overclocking....

why have 1,000hp when the speed limit is 55mph.

...etc.

You can get to 55mph faster than the other guy.

Talley wrote in post #18135697
overclocking is overclocking....

why have 1,000hp when the speed limit is 55mph.

...etc.

Bandimere speedway is open to the public on the right days here in CO:

http://www.bandimere.c​om …t-to-the-track-test-night

Talley wrote in post #18135697
overclocking is overclocking....

why have 1,000hp when the speed limit is 55mph.

...etc.

Whee is the speed limit only 55? Here in TX we have 85 in places.