5Ds vs 5Ds R

[HDNitehawk]

Ok, posting this I know its two days before the rumored notification.
Maybe it will turn out to be biggest hoax of the year.

But just in case it is real;

So the question, which to go with.

I have read about the difference with and without the AA filter.
Nikon didn't include the filter when they released the D810.

B&H had this article comparing the D800 to the D800e (without filter)
http://www.bhphotovideo.com/explora/...d800-and-d800e
But then the D810 was released and I wonder if B&H advice turned out to be valid.

That said I know what the AA filter does and moire.
I understand the effect in fabrics, made made items etc.

I have some questions;

1. If you are a bird photographer wouldn't you see moire in the fine detail of the birds feathers?
2. Perhaps hair of animals as well?
3. I have read it is not noticeable in nature, but surely with grasses and trees it could happen?
4. Why did Nikon drop there filter, was it because of the high pixel density it wasn't needed?
5. If so, wouldn't it be probable that Canon will do the same thing as Nikon, once it is accepted drop the one without the AA filter?

I could wait for the reviews, but prefer to pre-order as soon as it is released. I suppose I could pre-order both and later cancel the one I decide against.

[Kayaker72]

Interesting question. Thinking about it, I would be tempted by the "R".

Best quote I saw on moire: "The patterns are generated when an image with too high a spatial frequency is captured by a sensor. To avoid moire, the sampling frequency of the sensor has to be double the spatial frequency that reaches it."

So it stands to reason that the higher resolution your sensor, the less likely you are to have moiré. So if patterns in nature are only causing moiré infrequently with ~20mb sensors, wouldn't they even be less frequent with the 50mb sensor?

[HDNitehawk]

So if patterns in nature are only causing moiré infrequently with ~20mb sensors, wouldn't they even be less frequent with the 50mb sensor?

Maybe true, but how do we know how much moire were getting with the 20mb sensor if it has an AA filter.

[Kayaker72]

It shows up sometimes.

Btw,
http://www.luminous-landscape.com/re...mparison.shtml

http://photo.net/nikon-camera-forum/00amZu

More resolution, no moire. I am still looking, but if that is representative, I understand why the 810 doesn't have an AA filter.

[conropl]

When the D800E came out there was a lot written about not needing an AA filter with higher MP's (although I do believe the 800E still had a filter). That was the theory, but I have not read that as much anymore. From what I have seen, most of the landscape folks bought the 800E. The lack of (or weaker) AA filter should provide sharper pictures (or at least the appearance of being sharper). If it was me, I think I would be ordering the 5Ds R, but I was also wondering about feather and hair detail on birds and animals. Sounds like you have found that to not be a problem. However, it does make me wonder... if the AA filter is not needed as much when pushing MP's in the 36 range, then why would Canon be adding one at all with the 5Ds at 50+ MP's ?

Pat

[Kayaker72]

When the D800E came out there was a lot written about not needing an AA filter with higher MP's (although I do believe the 800E still had a filter). That was the theory, but I have not read that as much anymore. From what I have seen, most of the landscape folks bought the 800E. The lack of (or weaker) AA filter should provide sharper pictures (or at least the appearance of being sharper). If it was me, I think I would be ordering the 5Ds R, but I was also wondering about feather and hair detail on birds and animals. Sounds like you have found that to not be a problem. However, it does make me wonder... if the AA filter is not needed as much when pushing MP's in the 36 range, then why would Canon be adding one at all with the 5Ds at 50+ MP's ?

Pat

From what I read last night, a typical AA filter has a filter that distorts in the horizontal direction, then the IR filter, then a filter that distorts in the vertical direction. The Nikon D800E had a similar set up in that it had a filter that distorts in the horizontal direction, then the IR filter, but, then the change, the filter that corrects for the distortion in the horizontal direction.

An image of that is here. Actually, that is a pretty good write up comparing the D800 to the D800e. The D800e is sharper but does generate more significant moire than the D800 (i.e. the D800 has moire, but the D800e is worse).

One potentially relevant point, as moire is a function of the demosaicing process, so it is a function the programing. So moire with Canon may be a bit different from Nikon.

After my reading last night, I think I at least understand moire much better. Take a grouping of four pixels (G, G, R, B), then assume a two pixel wide line with that grouping. If a blue colored line wider than that two pixel line hits that grouping, the line of blue pixels fills up, the other pixels get nothing and the demosaicing registers that color as blue, and all is well. If a wider than 2 pixel line of purple hits our two pixel wide line a portion of R, B, etc will fill up, and the demosaicing will register the color as purple. But now, if we have two lines of different colors that are thinner than the 2 pixel line the light filling up the pixels will be inconsistent with either line. At different points in the line/pattern the demosaicing program may try to blend the two colors, at others it may register one color but not the other, etc. It is easy to see how this could baffle the program and lead to artifacts as the program tries to make sense of what it is seeing.

If this was my money, I would likely wait for reviews. But if I was to pre-order, I would read up on this a bit more, but right now, for landscapes and wildlife, I would likely go with the 5Ds R. The extra sharpness seems to be there all the time, but the it will likely take some very fine patterned detail to cause moire. And the AA filter doesn't remove the moire when it happens, at least not all of it, it just lessens the moire. You will likely see it sometimes, but I've read several instances of people trying to cause it with the D810 and failing, and those that did find it, it doesn't look so bad.

[Kayaker72]

BTW...I think this also gets into the whole "does the lens out resolve the sensor or vice versa" question. Basically, if lenses are your limiting factor, almost by definition, they would be doing the same thing as an AA filter.

[Anthony M]

Am I to assume that the 5DS R then does have an AA filter, but the "Low-Pass Cancellation Filter" allows you to bypass it (toggle)?
That's sort of what I'm reading into the description, anyway. I thought this was how the D800E works - I need to ask my friend with one now! Anyway, if that is case, then $200 extra seems like a no brainer once you're in this price range.

[HDNitehawk]

From what I have read the D800e only had a minimal increase in resolution.
When the D810 came out the AA was completely removed and better than a 10% increase was realized.
Hopefully Canon completely removed the filter rather than mask it's effect like I have read Nikon tried initially.

[Anthony M]

From what I have read the D800e only had a minimal increase in resolution.
When the D810 came out the AA was completely removed and better than a 10% increase was realized.
Hopefully Canon completely removed the filter rather than mask it's effect like I have read Nikon tried initially.

What's got me (and this could all be marketing gibberish) is their description as "cancellation filter". To me, you don't need to cancel something that's not there.
Oh well, no reason to get my panties all in a bunch until we see some real specs tomorrow and previews. Thanks...