DIGIC V [CR1]

[Stuart]

So two new chips - the Digic V and the sensor to go with it.
As suggested above a better ADC design to get dynamic range, the DXO marks always knocked Canon down on these so this will be a welcome improvement.
More throughput is great but this may not equate to more fps as the sensor mPixels will increase too.
Industry revolutionary noice reduction - a totally different way of reducing noise, that's exciting, i hope its a pixel level sensitive electronics design and not just software. This will also give better dynamic range.

Nothing about global shutter for video or bent helicopter blades at certain shutter speeds?

“creative” tools for new DSLR and compact features - will we not need photoshop?
I'm hoping for eye detection and focus selection. Interesting that this core is also aimed at compacts, maybe less memory area in the compact version.

Re: comments about not buying a 60D but one of these instead - This solution for a 5dmk3 will initially be 3 times the cost of a 60D, so whilst i agree its tempting to wait, this next product will be in a different market are to the 60D.
I hope in mid 2012 it filters down to a 7Dmk2 then maybe i can afford it.

 

[WarStreet]

Any word if this digic V processor will be in the 5D mkIII ?

It will have to be. In development for 4 years? There's no other way for them to increase the MP and FPS unless they keep adding more DIGIC IVs, which is probably more expensive than a more modern, efficient chip.

Digic V will be a Dual Core 1ghz proc. making it possible for 8FPS on single chip. and 12FPS on two chips (1D mkV)

At what 5DIII & 1DV MP count did you estimate these ? Any info we don't know ?

 

[BennyJ]

It is unfortunate that Canon, Nikon etc refuses to make their RAW formats public though. In practice they are thanks to reverse engineering in dcraw, but it would be a nice gesture from the manufacturers to actually publish the specs.

I would even go further and would really like to see Canon and other majors switch to DNG as their default RAW format, or at least add support for it. I think the time of proprietary RAW formats is coming to an end. They're unflexible, not future-proof and offer no significant advantages over open standards like DNG.

 

[DuLt]

So two new chips - the Digic V and the sensor to go with it.
As suggested above a better ADC design to get dynamic range, the DXO marks always knocked Canon down on these so this will be a welcome improvement.
More throughput is great but this may not equate to more fps as the sensor mPixels will increase too.
Industry revolutionary noice reduction - a totally different way of reducing noise, that's exciting, i hope its a pixel level sensitive electronics design and not just software. This will also give better dynamic range.

Nothing about global shutter for video or bent helicopter blades at certain shutter speeds?

“creative” tools for new DSLR and compact features - will we not need photoshop?
I'm hoping for eye detection and focus selection. Interesting that this core is also aimed at compacts, maybe less memory area in the compact version.

Re: comments about not buying a 60D but one of these instead - This solution for a 5dmk3 will initially be 3 times the cost of a 60D, so whilst i agree its tempting to wait, this next product will be in a different market are to the 60D.
I hope in mid 2012 it filters down to a 7Dmk2 then maybe i can afford it.

Looking forward for 650d.

 

[motorhead]

For the time being I'd much prefer to keep manufacturer specific processing methods. It keeps the door open for major advances. Locking everyone into a particular RAW "flavour" this early in the development of digital cameras seems like tying designers hands behind their backs.

I don't find it a problem. Canon provide me with DPP, which is designed to work with the RAW formats they use. I don't have to buy it, it's offered completely free and gets updated frequently (v3.9.4 is the latest release).

By the way, "Amateur Photographer" has a short article this week on Bayer pattern sensors which mentions a host of adjustments the processor has to do to the basic data it receives. So it's easy to see how the Digic5 might have taken 4 years of work, and why its essential it works well with whatever sensor it is harnessed to.

 

[BennyJ]

For the time being I'd much prefer to keep manufacturer specific processing methods. It keeps the door open for major advances. Locking everyone into a particular RAW "flavour" this early in the development of digital cameras seems like tying designers hands behind their backs.

I don't find it a problem. Canon provide me with DPP, which is designed to work with the RAW formats they use. I don't have to buy it, it's offered completely free and gets updated frequently (v3.9.4 is the latest release).

That's true but it forces you to actually use DPP which doesn't really fit into the Photoshop-oriented workflow of many photographers anyway. Sure you could also use Camera Raw which currently supports nearly all formats thanks to the enourmous effort Adobe put into it, but the point is: You just don't know how things will be in some years. Will the particular RAW format of some old EOS body still be supported? Will DPP still be around? Will Canon even still exist? OK, you could just try to install the old software, but will it still run and will you be able to get it when you need it? Many questions you can't answer today.

On the other hand: I don't believe there are so many innovations to be expected in terms of RAW formats. Basically, it's all about image data and meta data. The image data stays the same, there are different ways of encoding it, yes, but that should be easy to handle for an open format. Meta data is more tricky because the vendors put some undocumented stuff into it. But do you really need it if you don't know what it is anyway? Probably not. And if Canon switched to an open standard, they could easily integrate their specific meta data into it as well. So yea, I think it's about time for them to support DNG, at least as an option. Some other camera makers already do.

 

[shorthand]

You can bet that if Canon throws its weight behind a new image format, support for that format will appear VERY quickly.

In terms of image format, the gap isn't in compression, its in dynamic range. JPEG 2000 is only 20% more efficient than JPEG-1.

Therefore, I think we may instead see a new lossy compression format that captures more than 8 bits of dynamic range. (JPEG-HDR or Clarity HDR or ERI or something else similar) The pro community could really use this instead of "shunken" RAW files.

 

[motorhead]

BennyJ,

It's not simply the RAW format, its more complicated than that. I cannot see the digital camera sensor lasting in it's present form, it's simply far too expensive to produce. So assuming a better and cheaper method will arrive at some time, probably out of some eccentric geniuses shed, I believe we will be using very different processing methods.

Like you I don't trust computer software producers as far as I can throw them, but I'd trust Canon/Nikon etc more than I'd trust third parties like Adobe, who have shown a scant regard for existing customers over the years.

 

[BennyJ]

So assuming a better and cheaper method will arrive at some time, probably out of some eccentric geniuses shed, I believe we will be using very different processing methods.

Maybe, maybe not. From today's perspective that's rather unlikely. At the end of the day, it's all about pixels, so processing probably won't change that much. But we don't know. It's also a topic of its own that doesn't really have anything to do with today's RAW/DNG debate.

but I'd trust Canon/Nikon etc more than I'd trust third parties like Adobe

It's not about trust, its about having an open standard. While Adobe may have invented DNG and may have been pushing it for a while, it's an open standard and easy to understand. Everyone could write programs to handle these files even in 100 years from now. On the other hand, Canons format is a secret only they really know. They even create a new, slightly different format for every Body they release. Although there is software like CR that does a great job interpreting it, I wouldn't rely on it for long-time archiving even though I love Canon

 

[Olly]

I do not believe a single word of it. This is common marketing strategy from Canon. They are good at hyping things up. I have never ever seen a camera from Canon that was revolutionary compared to its predecessor. Never. They always make small incremental steps in IQ. ( If you believe everything Canon is saying you will probably disagree.. )

I hope for some improvements in DR and noise but I do not expect significant jumps, only minor improvements like we have seen before.

Canon is making a lot of money these days. They will continue to focus on video, that is what the majority of the customers want. Still-imaging is important of course, it will be improved but not to the extent that some of us wish. Video quality, improved focusing, more in-camera processing, 20+ megapixels, that is what we will get. For this Canon need the speed of the coming DIGIC V.

 

[LetTheRightLensIn]

It will have to be. In development for 4 years? There's no other way for them to increase the MP and FPS unless they keep adding more DIGIC IVs, which is probably more expensive than a more modern, efficient chip.

True. However, they should be mass producing the chip by now if they plan to use it in the 5D mkIII I guess. Strange that there hasn't been any rumors about the digic V earlier.

There were rumors about it on the long term rumors posting on Northlight a few years ago. Talk of properly sampling over the entire frame for video and so on.

 

[LetTheRightLensIn]

revolutionairy noise controll sounds good it almost sounds TO good making me wonder if this is just a load of software noise controll or some actualy improvement.
other then that im not really interested in any of the features except maby touch focus for life view noise and shutterspeeds higher iso gives me is my main concern.


ps. the post confirmation question doesnt know about yellow snow i suggest you educate it before you know what happens.

In terms of stills, yes, it's just marketing talk (unless you rely on in camera jpgs only). For video though, it could allow for much less noise since the current chips have not been fast enough to sample and bin properly at video speeds.

 

[DuLt]

I do not believe a single word of it. This is common marketing strategy from Canon. They are good at hyping things up. I have never ever seen a camera from Canon that was revolutionary compared to its predecessor. Never. They always make small incremental steps in IQ. ( If you believe everything Canon is saying you will probably disagree.. )

I hope for some improvements in DR and noise but I do not expect significant jumps, only minor improvements like we have seen before.

Canon is making a lot of money these days. They will continue to focus on video, that is what the majority of the customers want. Still-imaging is important of course, it will be improved but not to the extent that some of us wish. Video quality, improved focusing, more in-camera processing, 20+ megapixels, that is what we will get. For this Canon need the speed of the coming DIGIC V.

Doesn't digital "stills" reached some sort of peak in IQ?

I mean it's going to get better but it's sortoff the same in consumer cameras. Their flaws aren't limited by technology, they are mostly limited by marketing and design (I'm talking about the compact photographer that must have the 14mp camera but never prints a thing).

Only recently I've printed something that required more than 10mp, but still, the image has enough quality, more would be nice but not needed.