To Joe: software v optics (re - weak AA)

rovingtim · Feb 3, 2012

First of all, I am not trying to make you (or anyone) not like their E5.

This is a technical discussion not a qualitative discussion about the E5 v E3.

What I am asserting is that with proper processing a sensor with a strong AA is likely to produce just as many 'hairs on the bum' as one with a lighter AA. Additionally, I assert that at 100% there will be pixel level colour issues on a weak AA camera.

However, I also note that at normal viewing distances, you won't see many of them as the eye will automatically mix the colours (as it does with any ink jet print) to produce the right shade.

But this is about technicals. Is MTF a suitable measurement for resolution? Is the higher contrast of the lighter AA giving the illusion of more detail than other sensors when that isn't really the case? (Note: I'm not saying you or anyone shouldn't like the higher contrast of the weak AA cameras)

I think I have a strong case especially as people with the equipment to test my theory simply tell me I'm wrong (say and say and say) and refuse to post useful images.

I'm not here to say you shouldn't like your E5. I'm not even saying you shouldn't like your E5 over the E3 because of the AA filter.

What I seek are the facts.

EDIT: That is why I addressed this to Joe not bogus or riley. I wasn't looking for a fight.

ROC124 · Feb 3, 2012

rovingtim said:
First of all, I am not trying to make you (or anyone) not like their E5.

This is a technical discussion not a qualitative discussion about the E5 v E3.

What I am asserting is that with proper processing a sensor with a strong AA is likely to produce just as many 'hairs on the bum' as one with a lighter AA. Additionally, I assert that at 100% there will be pixel level colour issues on a weak AA camera.

Tim, may I barge into your thread with Joe? I'm not sure you really are making a technical assertion. You are questioning how camera/lens manufacturers and reviewers measure resolution, how sharpening comes into play, the role of other picture artifacts, and how it all relates to visibility and perception in real-world photo subjects.

The technical fact seems to have been settled long ago - sharpening can't increase resolution, but it makes the limits of resolution easier to see. Strong AA filters actually do reduce resolution, and blur what can be resolved versus light or no AA filters. Good sharpening can improve the appearance of resolution, but not the actual measurable resolution. Established testing methods of industry confirm that.

You seem to be challenging the significance (perception) of those facts in real-world photography, along with the addition of other artifacts introduced from light or no AA filters. Those are aesthetic/perception issues not technical arguements, as there is no one right answer. It involves judgements and preferences of the individual.

You have said several times you are making a technical assertion, and have challenged others to show photo examples proving you wrong. But the industry has already proven you wrong on the technical grounds - i.e. sharpening doesn't iimprove resolution, but can improve the perception of resolution. Lab tests have well-documented that the lighter AA filter of the E-5 gives a measurably significant resolution advantage, irrespective of sharpening.

Again, your assertion is really not a technical one, it is based on perception and aesthetics, not measurement. Therefore, there is no way to prove your assertions for real world photography. You are right from your perspective, AND Joe and others are right from the technical perspective. In the final analysis, the only disagreement is your statement that your really are making a technical assertion. You aren't, even if you think you are.
--
Bob Cole
http://www.robertcolephotography.com

CollBaxter · Feb 3, 2012

Tim I was not looking for a fight as well.

And yes I understand the technicalities involved with a AA filter. Most of us who used the E-520 and E-3 did notice a definite lack of detail when coming for in my case a E-510 to E-520 . And you are correct it did sharpen well but there was a definite lack of fine detail ( The hairs on the bum of a fly). The decision by Olympus to lighten the AA filter gave back some of the detail on a senor that was not current is the correct move.

DPR tested most of the Olympus 4/3 lenses on a L10 (lighter AA) due to the lack of detail on the E-3.

--
Collin

(Aficionado Olympus DSLR )

http://collinbaxter.zenfolio.com/
http://www.pbase.com/collinbaxter

Life is not measured by the number of breaths we take, but by the moments that take our breath away. (George Carlin)

New Seventh Wonder of the World.

rovingtim · Feb 3, 2012

CollBaxter said:
DPR tested most of the Olympus 4/3 lenses on a L10 (lighter AA) due to the lack of detail on the E-3.

Resolution is measured by contrast and contrast can be introduced in PP-ing or via a weak AA.

Some time ago I wondered if the softness we have been seeing are the result of shots not quite in focus. When I my E3 hits focus, the image is very detailed. The reason I started posting is I didn't see anything significantly different in E5 shots that I haven't seen in my in-focus E3 shots. I also saw alias artifacts.

As such, originally I wanted people with the E5 and a normal AA filtered Oly camera to post an image they think definitively demonstrated greater E5 detail and then let me have the RAWs so I can play.

But no would. But everyone is willing to SAY I'm wrong.

As to you liking your E5 over all previous Oly cameras: fine, I have nothing to say to that.

Big Ga · Feb 3, 2012

Oly E10 said:
You do have it wrong, but we all make mistakes.
Oly E10 has no connection with Marin.
In fact it was myself Oly E10 that originated the original thread.
I was just suggesting to "rovintim" that if my image was used the sensible thing
to do was ask for my approval. Nothing more than that.

Someone's making a mistake, but I don't think its me (unless I've missed something!)

I'm fairly sure the image Rovingtim posted originated from a thread/post by Marin where he was comparing E5 v E3 images.

Now you re saying that your image was used and it would have been good to ask for approval.

So either I've got my facts wrong, or you are Marin, or you are somehow making a mistake
???

Big Ga · Feb 3, 2012

You are correct. But a Question for you .... (and others ...)

CollBaxter said:
DPR tested most of the Olympus 4/3 lenses on a L10 (lighter AA) due to the lack of detail on the E-3.

So if the L10 was to the E3, what the E5 is to the E30 ..... why wasn't everyone here going apeshit over the L10 in the same way they are now for the E5
???

rovingtim · Feb 3, 2012

ROC124 said:
The technical fact seems to have been settled long ago - sharpening can't increase resolution, but it makes the limits of resolution easier to see.

True

ROC124 said:
Strong AA filters actually do reduce resolution,

Show me the proof that a well tuned AA significantly reduces detail.

ROC124 said:
Good sharpening can improve the appearance of resolution, but not the actual measurable resolution.

True.

ROC124 said:
You seem to be challenging the significance (perception) of those facts in real-world photography, along with the addition of other artifacts introduced from light or no AA filters. Those are aesthetic/perception issues not technical arguements, as there is no one right answer. It involves judgements and preferences of the individual.

Your missing the technical point: so the industry tests actually measure resolution effectively? Does a well tuned AA filter blur detail or does is simply flatten the contrast?

ROC124 said:
You have said several times you are making a technical

Indeed.

ROC124 said:
and have challenged others to show photo examples proving you wrong.

Absolutely. If it is so obvious that I'm wrong, why doesn't someone put up to shut me up? Why all the saying and saying and saying ....

ROC124 said:
But the industry has already proven you wrong on the technical grounds - i.e. sharpening doesn't iimprove resolution

I never said it did, so this just proves me right.

ROC124 said:
Lab tests have well-documented that the lighter AA filter of the E-5 gives a measurably significant resolution advantage, irrespective of sharpening.

Show me the proof. And it had better not be MTF because MTF measure contrast and contrast adjustments will alter the MTF as DPReview has already noted.

And here's the technical article about that:

http://www.lenstip.com/...le-Why_the_Lenstip_reviews_differ_from_others_.html

Read this and get back to me.

ROC124 said:
Again, your assertion is really not a technical one,

So you say.

CollBaxter · Feb 3, 2012

L10 = Panasonic , no IS , Poor JPG engine , Small Raw Buffer , Pretty boring.

At the time.

--
Collin

(Aficionado Olympus DSLR )

http://collinbaxter.zenfolio.com/
http://www.pbase.com/collinbaxter

Life is not measured by the number of breaths we take, but by the moments that take our breath away. (George Carlin)

New Seventh Wonder of the World.

windsprite · Feb 3, 2012

I remember it exactly that way myself. Those crops were definitely posted by msusic / Marin a year or two ago.

Strange!

Julie

Big Ga said:
Someone's making a mistake, but I don't think its me (unless I've missed something!)

I'm fairly sure the image Rovingtim posted originated from a thread/post by Marin where he was comparing E5 v E3 images.

Now you re saying that your image was used and it would have been good to ask for approval.

So either I've got my facts wrong, or you are Marin, or you are somehow making a mistake
???

Timskis6 · Feb 3, 2012

Big Ga said:
So if the L10 was to the E3, what the E5 is to the E30 ..... why wasn't everyone here going apeshit over the L10 in the same way they are now for the E5
???

Your suggestion of the L10 is to the E-3 as the E-5 to the E-30 is restricted to the narrow view of sharpness only. As you and everyone else know, there's much more to a camera than sharpness.

If you broaden the analogy to also include weather sealing, reliability, ergonomics, viewfinder, etc. etc., then the L10 is no longer to the E-3 as the E-5 is to the E-30.

If you restrict it to sharpness only, then the analogy makes sense. And the reason why people (well, myself) aren't crazy over the L10 are for all those other reasons above.

Cheers,
--
Tim
http://www.developemotion.com
'I haven't been everywhere, but it's on my list.'
E3/7-14/12-60/35-100/150/EC20
http://www.flickr.com/photos/timskis6/

Big Ga · Feb 3, 2012

Fascinating.

CollBaxter said:
L10 = Panasonic ,

= Huge company. Invest in lots of R&D instead of ... er ... instead of ... er ... ok, we'd better not go there ;-)

CollBaxter said:
no IS

No IS is a pain. But hey .... the E10, E20, E1, E300, E500, E330, E410, E420 ... none of those had IS, didn't really put people off those at the time. I certainly can't see it having put people off if they had some magical 30% or so increase in resolution over the alternatives ?!?

CollBaxter said:
Poor JPG engine

Hey ... we all know really serious shooters don't care about JPG ;-)

CollBaxter said:
Small Raw Buffer

Indeed. Could have been better. I can't remember what the write speed is like though with a modern fast card

CollBaxter said:
Pretty boring.

Hey ... come on.. the E5 is pretty boring in most ways. Low pixel count. Probably the slowest frame rate of all cameras with any sort of pro aspirations. Possibly worst in class focus .... about the most exciting thing about it is that you can throw up a bit of vomit over the thing when reviewing your ISO6400 shots and not worry too much about whether any moisture from your pea and ham soup is going to get inside.... At least the L10 could to +- 2 stops at a time bracketing, and could live view focus CDAF lenses ...

But jesting aside. See my point? all this hullabaloo about how fantastic the E5 sensor is supposed to be, however I've just got this huge sneaking suspicion that the L10 did exactly the same thing for IQ back then ... but because it didn't have Olympus on the front, people didn't want to know.
Fascinating.

Great Bustard · Feb 3, 2012

rovingtim said:
Here is an E5 image posted with moire.

http://forums.dpreview.com/forums/read.asp?forum=1022&message=40386348

Where? I don't see it.

rovingtim said:
Now, notice how everyone in this thread was talking about the obvious rainbow moire on the right.

No one talked about the rainbow moire below that. More importantly to my point, no one talked about the green and/or pink shades appearing where all the thin branches are intersecting, (note the thick branches have no colour deviations) or the colour deviations on just about all the verticals, what appears to be a significant green deviation on the lower left (though this might be the real colour ... hard to tell when a weak AA is throwing up colour deviations everywhere).

In a nutshell, you can see colour deviations all over this image.

Are you sure it's not CA/PF?

rovingtim said:
Now, as the respondents say, you can correct much of this in post. Also, the green and pink shades being thrown up by the micro detail will be cancelled out by the eye when the image is seen as a whole.

But then the entire debate centers on micro detail does it not? If that is the case, the colour of the micro detail of E5 files look really off to me. It is not a file I would like to have for display of micro detail.

The thing is, how can you know that these attributes are all a function of the AA filter?

rovingtim said:
I should also note that someone else contacted me about this image. He was gobsmacked about the colour deviations and he is someone I have been debating about weak AA's and sharpness.

Show me 100% file with fine detail from an E5 and I'll show you colour deviations.

Well, that's what we need -- side-by-side examples of the same scene at the same settings using the same lens.

rovingtim said:
As to your point about sharpening creating artifacts: I agree. It does and it is hard to get right. In addition, everyone's eye seems different with regards to sharpness. You criticise my work for not being sharp enough, but you can see by my thread I can do it if I want.

I don't remember criticizing your work for not being sharp enough. I remember Rriley making such a comment about a photo you posted in the SST, which you reposted after sharpening, but I felt it was oversharpened. Somewhere between the two would have been more to my tastes.

rovingtim said:
So I guess this entire debate comes down to whether your eye is more insulted by poor colour or soft detail. The strength of the AA might be one of those things that can be only subjectively optimal.

For sure, but, first, let's make sure that the effects are due to the AA filter.

rovingtim said:
Obviously, I am disappointed by Oly's change in philosophy (best optical file possible) whereas others are very pleased (if we can get close with software and its cheaper, why not?).

What say you?

I say that if Olympus lenses are all what people say they are, then Olympus should have been doing everything they could to get sensors with as many pixels as they could on them to make the most of them.

Yes, more pixels will result in better performance regardless of the lens, but you will get more out of more pixels behind a good lens than a lesser lens. In addition, the more the pixels, you use a weaker AA filter for the same relative strength. In short, no, it's not "all about the lens".

ROC124 · Feb 3, 2012

rovingtim said:
Show me the proof that a well tuned AA significantly reduces detail.

I didn't say a "well tuned" AA filter "significantly reduces detail", and don't know what would constitute proof to you since you reject industry standard MTF measurements. How do you define "well tuned" and "significantly"?

I said a strong AA filter does reduce detail, and that should be evident. Another poster already took you through the thought experiment on that: The AA filter reduces moire' and other artifacts through blurring of detail. Blurring includes a reduction of contrast. When contrast is at maximum (100%) there is no blurring, so no loss of detail. When contrast is at minimum (0%) there is maximum blurring and no detail.

The AA filter designer tries to strike a balance between those extremes, maintaining as much detail as possible consistent with the desired degree of elimination of moire'.

How much loss of detail is significant is in the eye of the beholder. I am not aware of a commonly accepted threshold of significance. So you can say the E-5 AA filter doesn't allow significantly more visible detail than a well-sharpened E-3 image, and I can say the opposite, and we can both be right.

rovingtim said:
Your missing the technical point: so the industry tests actually measure resolution effectively? Does a well tuned AA filter blur detail or does is simply flatten the contrast?

No, I understand that technical point. It just seems moot. I am confident science and industry have developed rational means of defining and measuring resolution within current technical capabilities. I have no rational basis to belive they haven't.

How do YOU define a well-tuned AA filter? The filters both flatten contrast and blur detail by scattering light. That softens edges and that softness can be partially ameliorated with downstream sharpening, but this is happening where the detail hasn't already been measurably lost.

If there was no scattering of light, only a reduction in contrast, then sharpening could make measurable detail more obvious to the human eye without introducing other artifacts. But AA filters cause their own difraction, i.e. a scattering of light as well as reducing contrast in the process. The scattering of light creates slight tonal variations in locations that don't correspond correctly to the unfiltered image. Those slight tonal variations, i.e. differences in contrast, can be sharpened but create false detail. Test it yourself: shoot a sharply focussed image out the window, then the same image out of focus through a window screen. Sharpen the latter. Does it look like an accurate recreation of the former?

rovingtim said:
Absolutely. If it is so obvious that I'm wrong, why doesn't someone put up to shut me up? Why all the saying and saying and saying ....

You are not willing to accept industry standards as proof. The test charts and resolution measurements are available on this site, the reviews state that they played with sharpening on the E-3 files to makes sure they couldn't get more resolution (and couldn't) and there was measurably less resolution than a Panasonic with the same sensor and lens, and other 10 mp cameras, hence the conclusion that a stronger than average AA filter was at work.

The E-5 had disproportionately higher resolution and moire' than other 12mp cameras, hence a reasonable conclusion of a much lighter filter. Since they said they tried sharpening to get more resolution out of the E-3 and couldn't, what other explanations are possible? While these differences were measurable and numerically substantial, that doesn't mean those differences will be obvious to any particular person in any particular image. It does demonstrate the cameras potential, even if it can't always be realized in a given image.

rovingtim said:
I never said it did, so this just proves me right.

You were referring to my statement that sharpening can't increase measurable resolution. I don't know what that proves you right about.

rovingtim said:
Show me the proof. And it had better not be MTF because MTF measure contrast and contrast adjustments will alter the MTF as DPReview has already noted.

Not accepting MTF as proof of resolution is like not accepting a mathematical proof that 2 times 6 equals 12. How can you NOT accept industry standard resolution tests as proof of resolution found in those tests? Again, the results are documented on this site. You can certainly disagree with the conclusion that the AA filter had anything to do with the results, but you are in no position to demand proof of a rational conclusion when you can't offer any proof that the conclusion was wrong.

You have offered images taken under uncontrolled or documented circumstances in which you don't see significant differences. I do see differences, and whether they are significant are up to the individual. Since you have offered those unmeasurable images as part of your assertions, that is partly why I say your assertions are not technical.

What measurable proof not subject to individual perception would satisfy you? The burden is on you to identify it since you reject commonly accepted industry standards.

rovingtim said:
And here's the technical article about that:

http://www.lenstip.com/...le-Why_the_Lenstip_reviews_differ_from_others_.html

Read this and get back to me.

I've read it several times prior to today. So what?

Back to you.
--
Bob Cole
http://www.robertcolephotography.com

Great Bustard · Feb 3, 2012

ROC124 said:
ROC124 said:

Show me the proof. And it had better not be MTF because MTF measure contrast and contrast adjustments will alter the MTF as DPReview has already noted.

Click to expand...

Not accepting MTF as proof of resolution is like not accepting a mathematical proof that 2 times 6 equals 12. How can you NOT accept industry standard resolution tests as proof of resolution found in those tests? Again, the results are documented on this site. You can certainly disagree with the conclusion that the AA filter had anything to do with the results, but you are in no position to demand proof of a rational conclusion when you can't offer any proof that the conclusion was wrong.

This is central to the whole discussion. A manufacturer MTF chart for a lens measures contrast at a particular resolution (lp/mm on the sensor). A lens (system) test measures resolution (lw/ph) at 50% contrast.

So, the point Tim is making is that since the processing can affect the contrast, it will also affect the MTF-50 score. In addition, he is making the assertion that applying the same processing to both systems is not optimal, especially if they have different AA filters (and pixel counts).

This is the exact same reason the DxOMark has different measures for DR than IMATEST. DxOMark measures the DR of the unprocessed RAW file, whereas IMATEST measures the DR of a processed photo. And, yeah, the differences can be huge .

So, Tim is rejecting the MTF-50 results on this basis. I have to say that this is a very valid objection, and the reason I am so interested in this discussion. I've always been under the impression that MTF-50 scores were comparable across systems, but now I'm thinking that differences in processing make that a problem, and a problem worth sorting out.

Great Bustard · Feb 3, 2012

Here's the thread:

http://forums.dpreview.com/forums/read.asp?forum=1022&message=37080141

Something weird is afoot.

windsprite said:
I remember it exactly that way myself. Those crops were definitely posted by msusic / Marin a year or two ago.

Strange!

Julie

Big Ga said:

Someone's making a mistake, but I don't think its me (unless I've missed something!)

I'm fairly sure the image Rovingtim posted originated from a thread/post by Marin where he was comparing E5 v E3 images.

Now you re saying that your image was used and it would have been good to ask for approval.

So either I've got my facts wrong, or you are Marin, or you are somehow making a mistake
???

Click to expand...

Geo Wharton · Feb 3, 2012

rovingtim said:
Show me 100% file with fine detail from an E5 and I'll show you colour deviations.

Well:

And Cropped:

So, where is the color deviation? I'll accept that if you point it out. But I don't yet see it.

(12-60mm SWD)

Rriley · Feb 3, 2012

Great Bustard said:
ROC124 said:

Great Bustard said:

Show me the proof. And it had better not be MTF because MTF measure contrast and contrast adjustments will alter the MTF as DPReview has already noted.

Click to expand...

Not accepting MTF as proof of resolution is like not accepting a mathematical proof that 2 times 6 equals 12. How can you NOT accept industry standard resolution tests as proof of resolution found in those tests? Again, the results are documented on this site. You can certainly disagree with the conclusion that the AA filter had anything to do with the results, but you are in no position to demand proof of a rational conclusion when you can't offer any proof that the conclusion was wrong.

Click to expand...

This is central to the whole discussion. A manufacturer MTF chart for a lens measures contrast at a particular resolution (lp/mm on the sensor). A lens (system) test measures resolution (lw/ph) at 50% contrast.

what you describe is the system of measurement

Great Bustard said:
So, the point Tim is making is that since the processing can affect the contrast, it will also affect the MTF-50 score. In addition, he is making the assertion that applying the same processing to both systems is not optimal, especially if they have different AA filters (and pixel counts).

this is a really infantile way of expressing what happens. You people seem to imagine they take a photo of a chart, open it up in PS, process it with secret herbs and spices, and pull away with a result. The underlying suggestion is that a camera could cheat the system with processing, quite frankly Im amazed that people apparently do not know better

Great Bustard said:
This is the exact same reason the DxOMark has different measures for DR than IMATEST. DxOMark measures the DR of the unprocessed RAW file, whereas IMATEST measures the DR of a processed photo. And, yeah, the differences can be huge .

IMATEST uses both RAW and jpeg shot off a standardised ISO chart, IMATEST has available to it a system of evidence, DxO OTOH is a black box with no system of evidence.

Great Bustard said:
So, Tim is rejecting the MTF-50 results on this basis. I have to say that this is a very valid objection, and the reason I am so interested in this discussion. I've always been under the impression that MTF-50 scores were comparable across systems, but now I'm thinking that differences in processing make that a problem, and a problem worth sorting out.

what by swiping images

--
Riley

any similarity to persons living or dead is coincidental and unintended
support 1022 Sunday Scapes'

Great Bustard · Feb 3, 2012

Rriley said:
Rriley said:

Rriley said:

Rriley said:

Show me the proof. And it had better not be MTF because MTF measure contrast and contrast adjustments will alter the MTF as DPReview has already noted.

Click to expand...

Not accepting MTF as proof of resolution is like not accepting a mathematical proof that 2 times 6 equals 12. How can you NOT accept industry standard resolution tests as proof of resolution found in those tests? Again, the results are documented on this site. You can certainly disagree with the conclusion that the AA filter had anything to do with the results, but you are in no position to demand proof of a rational conclusion when you can't offer any proof that the conclusion was wrong.

Click to expand...

This is central to the whole discussion. A manufacturer MTF chart for a lens measures contrast at a particular resolution (lp/mm on the sensor). A lens (system) test measures resolution (lw/ph) at 50% contrast.

Click to expand...

what you describe is the system of measurement

Rriley said:

So, the point Tim is making is that since the processing can affect the contrast, it will also affect the MTF-50 score. In addition, he is making the assertion that applying the same processing to both systems is not optimal, especially if they have different AA filters (and pixel counts).

Click to expand...

this is a really infantile way of expressing what happens. You people seem to imagine they take a photo of a chart, open it up in PS, process it with secret herbs and spices, and pull away with a result. The underlying suggestion is that a camera could cheat the system with processing, quite frankly Im amazed that people apparently do not know better

As usual, you are wrong. You see, this was discussed rather thoroughly in the last thread:

http://forums.dpreview.com/forums/read.asp?forum=1022&message=40470963

Let me summarize, since you appear to have missed it the first time around: DPR uses the same processing for all systems which is likely not the best approach for anyone seeking optimal results.

Rriley said:
Rriley said:

This is the exact same reason the DxOMark has different measures for DR than IMATEST. DxOMark measures the DR of the unprocessed RAW file, whereas IMATEST measures the DR of a processed photo. And, yeah, the differences can be huge .

Click to expand...

IMATEST uses both RAW and jpeg shot off a standardised ISO chart, IMATEST has available to it a system of evidence, DxO OTOH is a black box with no system of evidence.

The RAW photos are still processed with IMATEST. You see, the point was that the processing affects the results. DxOMark does not base its results on processed photos, IMATEST does.

Here ya go:

http://forums.dpreview.com/forums/read.asp?forum=1018&message=39686135

along with many other posts in that thread.

Rriley said:
Rriley said:

So, Tim is rejecting the MTF-50 results on this basis. I have to say that this is a very valid objection, and the reason I am so interested in this discussion. I've always been under the impression that MTF-50 scores were comparable across systems, but now I'm thinking that differences in processing make that a problem, and a problem worth sorting out.

Click to expand...

what by swiping images

Tim used a photo that was posted some time back by Marin:

http://forums.dpreview.com/forums/read.asp?forum=1022&message=40498785

and then Oly E10 called him on it, saying it was his, and saying Tim should have asked. Well, sure, Tim should have asked, but his using the photo was well within the realm of "fair use".

What makes you such a hypocritical ***** is how many times you post photos and charts not your own that you don't credit the source for.

Please, seek "entertainment" elsewhere. As Tim said:

http://forums.dpreview.com/forums/read.asp?forum=1022&message=40495927

That is why I addressed this to Joe not bogus or riley. I wasn't looking for a fight.

odl · Feb 3, 2012

People dont seem to take the time to read what others have written.

in this thread you have a conversation progressing, but round and round it goes, here is an example why:

Tim says here Emphasis mine):

rovingtim said:
What I am asserting is that with proper processing a sensor with a strong AA is likely to produce just as many 'hairs on the bum' as one with a lighter AA. Additionally, I assert that at 100% there will be pixel level colour issues on a weak AA camera.

To which Bob replies (emphasis mine):

rovingtim said:
rovingtim said:

The technical fact seems to have been settled long ago - sharpening can't increase resolution, but it makes the limits of resolution easier to see. Strong AA filters actually do reduce resolution, and blur what can be resolved versus light or no AA filters.

Click to expand...

Where Tim replies (emphasis mine):

rovingtim said:
Show me the proof that a well tuned AA significantly reduces detail.

To which Bob replies (emphasis mine):

rovingtim said:
rovingtim said:

I didn't say a "well tuned" AA filter "significantly reduces detail" , and don't know what would constitute proof to you since you reject industry standard MTF measurements. How do you define "well tuned" and "significantly"?
I said a strong AA filter does reduce detail, and that should be evident.

Click to expand...

What the hell is a strong AA, a weak AA, a light AA etc etc etc.

Th E-5 resolves more detail than the E-3, you dont believe it, you feel the E-3 matches it... That is great, you have nothing to get the E-5 for, just saved $1700

As others have said, while you can sharpen a file, if the detail isnt there you will not get there with sharpening... Nevermind the E-3 E-5, look at the a850, it has a stronger AA filter...

What did the different pixel counts:
a850 - Absolute Resolution: 2700

E-5 - all nine lines of our chart are accurately described by the E-5 up to approximately 2600Lph

That is impressive, as my a850 was described as:

...the Alpha 900 sets a new standard for resolution, edging past the EOS-1Ds Mark III by a whisker, and leaving its 12-ish megapixel competitors in a cloud of dust. Next to the Canon models the output looks soft...

It doesnt have to be E-3 vs E-5, and as i have said before the E-3 was a very good camera (still is)... As to if this matters compare it to the a850 for hair detail, the E-3 does alright (the A850 should destroy it, but does it destroy it in terms of detail?)... still doesnt change the fact the E-5 outresolves the E-3 (ignore the SD1):

all 100% crops.

--
alatchinphotography.com

“Imagination is more important than knowledge. For
knowledge is limited to all we now know and
understand, while imagination embraces the entire
world, and all there ever will be to know and
understand.” - Albert Einstein

ROC124 · Feb 3, 2012

Great Bustard said:
ROC124 said:

Great Bustard said:

Show me the proof. And it had better not be MTF because MTF measure contrast and contrast adjustments will alter the MTF as DPReview has already noted.

Click to expand...

Not accepting MTF as proof of resolution is like not accepting a mathematical proof that 2 times 6 equals 12. How can you NOT accept industry standard resolution tests as proof of resolution found in those tests? Again, the results are documented on this site. You can certainly disagree with the conclusion that the AA filter had anything to do with the results, but you are in no position to demand proof of a rational conclusion when you can't offer any proof that the conclusion was wrong.

Click to expand...

This is central to the whole discussion. A manufacturer MTF chart for a lens measures contrast at a particular resolution (lp/mm on the sensor). A lens (system) test measures resolution (lw/ph) at 50% contrast.

So, the point Tim is making is that since the processing can affect the contrast, it will also affect the MTF-50 score. In addition, he is making the assertion that applying the same processing to both systems is not optimal, especially if they have different AA filters (and pixel counts).

This is the exact same reason the DxOMark has different measures for DR than IMATEST. DxOMark measures the DR of the unprocessed RAW file, whereas IMATEST measures the DR of a processed photo. And, yeah, the differences can be huge .

So, Tim is rejecting the MTF-50 results on this basis. I have to say that this is a very valid objection, and the reason I am so interested in this discussion. I've always been under the impression that MTF-50 scores were comparable across systems, but now I'm thinking that differences in processing make that a problem, and a problem worth sorting out.

I'm aware of all those issues and the complications of comparing tests with different methodologies and input assumptions. Nature of the beast, but it doesn't invalidate any of them, just makes direct comparisons more difficult. So what? Seems like there has already been general agreement that the differences often don't matter much in practical application.

One of the original justifications for this protracted discussion is that some here claim substantial, visible resolution differences between the E-3 and E-5, with the speculation that the AA filter is largely but not enitirely responsible. Others say they don't see substantial differences. Big deal, humans are prone to hyperbole in both directions. There isn't any standard for "substantial" and it doesn't matter what test numbers show - if you don't perceive the differences you aren't inclined to believe the tests. That really isn't a technical issue, it is a human issue, and no technical proof is likely to be satisfactory.

--
Bob Cole
http://www.robertcolephotography.com

To Joe: software v optics (re - weak AA)

Veteran Member

Leading Member

Forum Pro

Veteran Member

Forum Pro

Forum Pro

Veteran Member

Forum Pro

Veteran Member

Veteran Member

Forum Pro

Forum Pro

Leading Member

Forum Pro

Forum Pro

Senior Member

Forum Pro

Forum Pro

Senior Member

Leading Member

Similar threads

Keyboard shortcuts