sebboh wrote:
just to clarify what i wrote above: the staggered zipper pattern is caused by the lens being so far out of focus that it can no longer resolve the frequency of the pattern being shot (which is of course a much lower frequency than what any AA filter blocks out). the effective "sampling frequency" of the lens drops below the spatial frequency of the lines on the paper leading to an aliasing effect. the only way an AA filter would prevent this is if it was set so low that the camera could not resolve the lines on the paper when focused properly (meaning the paper would look like a solid grey sheet of paper instead of lined paper)....Show more →
Thanks, that's good to know and may explain why he got this effect only with fast lenses.
Tariq Gibran wrote:
"The reasons why Leica don't incorporate AA filters into their designs are quite simple:
1) they cannot afford it
2) since they're already using the cheapest, cheesiest materials available for the filter package, an AA filter would add to much thickness, and make the edge colour vignette even worse.
The reasons why MF manufacturers don't include AA filters:
1) they cannot afford it
2) they expect their users to be aware of the effect and do what they can when shooting to avoid it."
I couldn't quite make out if this was meant as sarcasm, or if it's your opinion, or a quote from somebody. Why can't these manufacturers of premium gear afford an AA filter that even the cheapest DSLR has?
Wonger wrote:
Judging by this conversation and the amount of passionate users who are so strongly anti No-AA I'd disagree. I agree more people would just be like "well we'll deal with it" but it'd still have people frothing with anger.
Thats besides the point of my question though, if it was such a common and massive issue on either of those cameras wouldnt people not have purchased them or have brought it up as a problem more often?
If its such an issue for fashion photography why do they still often opt for AA filterless MFDB's?
I can tell you that with my experience with the X100 I had for 6 months I didn't run into moire issues once. I wish more M9 users could join in the conversation and offer their input, one or two don't seem to be enough to sway any opinions. Maybe some pixel peepers can scan through these photos and find us a lot of examples of ruined shots.
I have pre-ordered a D800E but I'm still entirely open to the D800. When it comes to this camera we're all still working off speculation on the frequency of its occurrence but I've based my decision on how other non AA sensor camera users are reacting and I'm just not seeing it as the great negative some others are. Hopefully production cameras get in the right testers hands soon so we can all find out if we made the right choices for our style or not. Hell I'd be happy to find out the D800 is better suited for me so I could use that $300 and get a 85 1.8
The Fuji X100 actually does have an AA filter. I think it's a very weak one but does provide some protection against moire as I never ran into it myself with the X100.
"4Slim optical low-pass filter exploits the full potential of the lens
In order for the sensor to accurately reproduce the image from X100’s high-performance lens without loss of the high resolution performance, it is desirable to use an optical low-pass filter that is as thin as possible with a small separation width. However, when the separation width is reduced, false colours can be generated by sensor sampling. By optimising signal processing to minimise false colour generation while reducing the separation to the smallest possible width, the design lets the user take full advantage of the X100’s lens."
Original size, look at the collar. There's quite evident color moire there.
I dunno. It looks like chroma noise to me. Look at the dark area on the extreme right of the image. Color blotches and patterns roughly similar in intensity and length scale to those found on the collar.
Thorsten wrote:
I couldn't quite make out if this was meant as sarcasm, or if it's your opinion, or a quote from somebody. Why can't these manufacturers of premium gear afford an AA filter that even the cheapest DSLR has?
This is a direct quote from Joakim/ theSuede, not me.
My guess though - and Joakim would know better - is this:
AA filters are very expensive to begin with. Large ones for a MF digital back may cost many thousands of dollars just for the filter alone - I do believe a MF manufacturer had offered AA filters as an option in the past though. Since these backs sell at best in the hundreds per year - versus many hundreds of thousands per year for a cheap DSLR - I would guess the MF digital manufactures are not able to really obtain major discounts for large/ bulk quantities like the DSLR makers can (plus the DSLR sensors are much smaller to begin with so the filter is dramatically cheaper). Leica is in the same boat regarding the quantity problem - they don't sell a lot of M9's in the grand scheme of things - but, as Joakim/ the Suede has also said, AA filters present additional optical problems for wide angle rangefinder lenses so there is a real reason not to use one in this case.
RustyBug wrote:
BTW ... it should be noted that I shoot with M645 glass for my longer stuff on the SLR/c. I should run a test shot or two with SLR glass vs. MF glass to see if I can induce moire differently between them.
Just anecdotal (and thinking outloud), but many of us around here are likely using pretty darn good glass. The point has been made that a sensor without an AA filter will offer better micro-contrast resulting in an increase in perceived detail. So what happens when we use a lens which has really high micro-contrast to begin with on both the AA and AA-less sensor? Well, funny thing happened when I owned the original 5D and used it with certain Zeiss lenses. All of a sudden, I ran into moire issues I had previously never seen with other lenses. The 5D AA filter is known to be weak to begin with and my conclusion is that the increased contrast of the Zeiss lens basically cut right through the effect of the weak AA filter. Thus, it seems it would follow that if your using a very good lens with high micro-contrast, your more likely to run into moire issues with the sensor without the AA filter, particularly at F-stops near the lenses ideal performance - F4, 5.6 and even F8. I think the point has also been previously made that very few Nikkor lenses will run the risk of out-resolving the D800 sensor - and they don't have the extreme micro-contrast that Zeiss lenses have to begin with -so part of the choice of D800E/ D800 may also come down to which lenses one intends to use.
Ruahrc wrote:
I dunno. It looks like chroma noise to me. Look at the dark area on the extreme right of the image. Color blotches and patterns roughly similar in intensity and length scale to those found on the collar.
Chroma noise is pixel level, it cannot produce blotches of color. Well, unless you want to call "chroma noise" any kind of color artifacts in the image :-)
Thats good to know, I didn't intend to mislead anyone by my post. I was going off what luminous landscapes was saying. I'll still stand by my preorder of the D800E based on the M9 community, does anyone know for certain any other consumer priced digital cameras that dont have an AA filter from the factory to get more of an idea what we might see?
Check Pg 15, it confirms that AA will smashed the high frequency detail.
'D800E suit virtually complex subject' good for Landscape shooter.
For fabric, production shot, we are into that controversy loop again, you want max sharpness for detail but you may run into trouble to have moire
However, Nikon do have a good business model here for D800E owner, It hint you need Capture NX and Nikon Gold ring glass to take advantage of it The newly implemented Auto ISO for zoom and Face detect focus with adjusted backlite exposure are the features you have to use Nikkor to have.
Go through the document, I have found this D800 is an almost perfect camera for anything other than speed. Though many features I don't use it at all, but for professional, this is THE camera.
So what happens when we use a lens which has really high micro-contrast to begin with on both the AA and AA-less sensor?
5D is 1/3 of pixels density. What you see may not apply D800E.
It depend on your lens and AA filter setting. Both act like low pass filter, with poor lens, this AA or not AA is meaningless. With good Lens have higher row off freq, AA will be your bottle neck. With good lens have even higher row off freq that out solve your sensor, alias will happen but 'detectable' alias may or may not happen depend on scenario.
Wonger wrote:
Thats good to know, I didn't intend to mislead anyone by my post. I was going off what luminous landscapes was saying. I'll still stand by my preorder of the D800E based on the M9 community, does anyone know for certain any other consumer priced digital cameras that dont have an AA filter from the factory to get more of an idea what we might see?
The Ricoh GXR M mount is the only one I'm aware of and there just so happens to be a moire example posted in the current thread:
Fuji really got it right with the X100 with regard to matching the lens to the sensor perfectly imo. I'm sure specific lens-filter-sensor interactions play a huge role so that sort of customization that Fuji achieved with the X100 may not be possible with an interchangeable system where the manufacturer is not sure which lens will be used.
I have a preorder for both the D800 and D800E in at the moment because this whole AA issue is still an open question until we see more comparisons between the two cameras - and I'm sure we will before either ships.
zhangyue wrote:
See, you didn't get the point, your lens may not out solve the sensor, but it out solve the frequency at AA filter kick in. of course, you gained higher contrast here.
zhangyue wrote:
... with poor lens, this AA or not AA is meaningless.
KaaX wrote:
Ah. I see you've come to my point of view :-D
any $50 50/1.8 or similar lens will blow away the 36mp resolution at f/5.6. you do NOT need fantastic glass to get a benefit from that pixel density (which is the same as the pixel density of the d7000).
Tariq Gibran wrote:
Just anecdotal (and thinking outloud), but many of us around here are likely using pretty darn good glass. The point has been made that a sensor without an AA filter will offer better micro-contrast resulting in an increase in perceived detail. So what happens when we use a lens which has really high micro-contrast to begin with on both the AA and AA-less sensor? Well, funny thing happened when I owned the original 5D and used it with certain Zeiss lenses. All of a sudden, I ran into moire issues I had previously never seen with other lenses. The 5D AA filter is known to be weak to begin with and my conclusion is that the increased contrast of the Zeiss lens basically cut right through the effect of the weak AA filter. Thus, it seems it would follow that if your using a very good lens with high micro-contrast, your more likely to run into moire issues with the sensor without the AA filter, particularly at F-stops near the lenses ideal performance - F4, 5.6 and even F8. I think the point has also been previously made that very few Nikkor lenses will run the risk of out-resolving the D800 sensor - and they don't have the extreme micro-contrast that Zeiss lenses have to begin with -so part of the choice of D800E/ D800 may also come down to which lenses one intends to use. ...Show more →
+1 @ 5D and occasional moire' (my experience also), but the 5D is probably inidicatve of the kind of "occasional" that people are suggesting will come into play at a true non-AA, but it really is going to be quite a bit more frequent. I would assume by default that anyone considering using a non-AA filter system is already using great glass (or at least desiring great glass). It would really SUCK to be going along with a non-AA system and good glass and occasional moire, then to upgrade $$$$ to great glass and voila, the moire party begins.
Good/great normal/long glass seems to typically yield higher MTF's than wide glass does, so I wonder if that is why WA doesn't seem to have as much problem ... but when I picked up my 24L TS-E II (with its higher MTF than my Oly's), it showed up again. I haven't figured out all the nuances of combinations or can fully explain them ... I just know they are real for me ... and as such, it stands that they may be real concerns for others. What does that mean for the D800E ... theory/hype/kool-aid ... or totally awesome and smooth sailing. Time will tell ...
Maybe that's a good indicator? I've used the 5D original quite a lot, and having good technique and good lenses (or just a shitload of light available) that camera produces quite a lot of artefacting. Large pixels, weak filter.
This was one of the very few drawbacks with that camera, that otherwise is one of my all time favorites.
About aliasing at this (D800) pixel size....
Grabbed a quick backlit shot of the GF's dog yesterday, as it used one of the cats' rear end as a headrest. I missed "the moment" while finding a good hand-rest for me, she turned her head the other way, facing into the window in stead.
This is a D7000, with a 50/1.8AFS. At 1/30s, handheld with a good support. A very dull overview:
What this results in, in a camera WITH an AA filter, with a 200$ lens, handheld at 1/30s. 1:1 pixel crop.
As has been shown, this is easy to remedy at least in a superficial way, but it WILL destroy any color information taken from reality. It's now acceptable, but a totally flat surface.
The image above is a worst case sample with regards to the target (blue contrast on white/black or red contrast on white/black are the worst-case scenarios for moire) - but still. This is handheld, and not very favorable conditions for a sharp shot.
theSuede wrote:
Maybe that's a good indicator? I've used the 5D original quite a lot, and having good technique and good lenses (or just a shitload of light available) that camera produces quite a lot of artefacting. Large pixels, weak filter.
This was one of the very few drawbacks with that camera, that otherwise is one of my all time favorites.
About aliasing at this (D800) pixel size....
Grabbed a quick backlit shot of the GF's dog yesterday, as it used one of the cats' rear end as a headrest. I missed "the moment" while finding a good hand-rest for me, she turned her head the other way, facing into the window in stead.
The image above is a worst case sample with regards to the target (blue contrast on white/black or red contrast on white/black are the worst-case scenarios for moire) - but still. This is handheld, and not very favorable conditions for a sharp shot....Show more →
Good illustration! Looks like vibration won't help! I had doubt about it. Your example prove me wrong. What aperture setting?
Tariq Gibran wrote:
This is a direct quote from Joakim/ theSuede, not me.
My guess though - and Joakim would know better - is this:
AA filters are very expensive to begin with. Large ones for a MF digital back may cost many thousands of dollars just for the filter alone - I do believe a MF manufacturer had offered AA filters as an option in the past though. Since these backs sell at best in the hundreds per year - versus many hundreds of thousands per year for a cheap DSLR - I would guess the MF digital manufactures are not able to really obtain major discounts for large/ bulk quantities like the DSLR makers can (plus the DSLR sensors are much smaller to begin with so the filter is dramatically cheaper). Leica is in the same boat regarding the quantity problem - they don't sell a lot of M9's in the grand scheme of things - but, as Joakim/ the Suede has also said, AA filters present additional optical problems for wide angle rangefinder lenses so there is a real reason not to use one in this case. ...Show more →
Thanks, that makes sense. I had long suspected that Leica and MF manufacturers may not omit the AA filter because they think it has no merit, but because they just didn't bother to design one and can get away without it.
This shot was at F5.6. But unless I had a chair seat to rest my hands on, I highly doubt the shot would have been this sharp. The effect would no doubt have been a lot weaker (or at least smaller in area) at F2.8, since less of the curtains would have been in perfect focus. I probably would have aimed a few centimeters further back with focus too.
When I saw it I thought that it would be good to try an aperture-bracket and see what happens, but there's heavy skies and almost no light at all today...
And mind that this truly IS a worst case scenario. The blue channel in raw is seriously mottled by the warm light shining trough a dark blue mesh cloth.
theSuede wrote:
This shot was at F5.6. But unless I had a chair seat to rest my hands on, I highly doubt the shot would have been this sharp. The effect would no doubt have been a lot weaker (or at least smaller in area) at F2.8, since less of the curtains would have been in perfect focus. I probably would have aimed a few centimeters further back with focus too.
When I saw it I thought that it would be good to try an aperture-bracket and see what happens, but there's heavy skies and almost no light at all today...
And mind that this truly IS a worst case scenario. The blue channel in raw is seriously mottled by the warm light shining trough a dark blue mesh cloth....Show more →
I guess one question about this curtain example is: would the moire effect have been worse without the AA filter? I'm also curious if the AA filter in the Nikon 7000 is known to be on the weaker side. The one in the D3x appears to be on the weaker side but I believe previous Nikons generally used fairly heavy AA filters.
An example of moire from the D3x can be seen on Lloyd's free blog here:
By the way, which tool did you use to remove both the color moire and it's pattern? I know C1 Pro has the pattern slider in it's moire tool which would work with minor patterns like this. Unfortunately, I have not seen any "canned" tool which will work on worse patterns such as a strong Maze pattern (visible in the Nikon Kimono example).
I'm a wedding photographer using 5Dc's, I get moire every single wedding from the bridal veil. Not terrible but certainly there. As a wedding photographer I wouldn't touch the 800E with a bargepole, the advantages are outweighed for this profession by the downsides. We aren't printing as large as fashion shoots are neither are we lighting as hard, etc.
