-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
optical zoom vs honeycom zoom
- Thread starter JohnChu
- Start date
SFJP
Veteran Member
I have never seen that, but I would think it might be "Honeycomb zoom".
I would also think that this appellation is a way for Fuji to say that used as a 3 mp camera, the Fuji 6900 is not a 6 x optical zoom with in addition a digital zoom but something much more close to a 13.2 X real zoom in 1 mp picture mode, and 8.4 x real zoom in 3 mp mode.
Let me try to explain why I write that. The original picture extracted from the honeycomb pattern of the Fuji 6900 superCCD has an initial size of 6 mp and no less! When you ask for a 3 mp picture, this initial 6 mp picture is then resized to 3 mp (and not the contrary as many people would think, the 6 mp does not coms from an interpolated resizing of an original 3 mp picture!).
Notice that used in 6 mp mode, the Fuji does not offer any digital zoom.
In 3 mp mode it offers a 1.4 x digital zoom (2.2 x digital zoom in 1 mp mode). But it is not a simple usual digital zoom enlarging an initially smaller picture and cropping it like do other cameras. The "digital" zoom of the Fuji in fact is resulting from applying a smaller reduction to the initial 6 mp picture. At full digital zoom, in fact no reduction is applied and the original 6 mp picture is just cropped. Because there is no artificial enlargement of the initial picture, the 6 x optical zoom x 1.4 x digital zoom can really presented as a real 8.4 x zoom closde to the quality that would be given by an purely optical 8.4 x zoom (idem in 1 mp mode, but replacing 8.4 then by 13.2).
How to call such a new kind of mix of optiocal zooming with a digital zoom that in fact does not result from an enlargement but from a lesser reduction? A good idea would be to call it a Honeycomb zoom because all this is due to the special honeycomb pattern of the superCCD.
Well, interesting, with this appelation due to the creativity of some marketing people in Fuji and introduced to palliate the fact that Fuji has no longer the right to call its camera a 6 mp camera, Fuji recalls us that the full zoom range it provides is more than 6 x and without any artificial enlargement... In fact, people using it as a 3 mp camera be proud to own a real 8.4 x zoom camera (or 13.2 real zoom camera if you are using it as a 1 mp camera0. This beats the UZI 2100 and the Canon Pro 90 at similar resolution... and without picture enlargment!!!
By the way, if the "digital" zoom is only 1.4 in 3.3 mp mode to not alter the quality of the 3.3 mp pictures when digital zoom is used, this means that the real quality of the initial pictures produced by the 6900 is estimated in the average by Fuji to be 1.4 x 3.3 = 4.62 mp, which correspond well to the current observations: the Fuji 6900 is not really equivalent to a full 6 mp camera, not to a 3.3 mp camera but rather to a 4.62 mp camera... No?
Jean-Paul
So what does this means: that as a 3 mp camera
I would also think that this appellation is a way for Fuji to say that used as a 3 mp camera, the Fuji 6900 is not a 6 x optical zoom with in addition a digital zoom but something much more close to a 13.2 X real zoom in 1 mp picture mode, and 8.4 x real zoom in 3 mp mode.
Let me try to explain why I write that. The original picture extracted from the honeycomb pattern of the Fuji 6900 superCCD has an initial size of 6 mp and no less! When you ask for a 3 mp picture, this initial 6 mp picture is then resized to 3 mp (and not the contrary as many people would think, the 6 mp does not coms from an interpolated resizing of an original 3 mp picture!).
Notice that used in 6 mp mode, the Fuji does not offer any digital zoom.
In 3 mp mode it offers a 1.4 x digital zoom (2.2 x digital zoom in 1 mp mode). But it is not a simple usual digital zoom enlarging an initially smaller picture and cropping it like do other cameras. The "digital" zoom of the Fuji in fact is resulting from applying a smaller reduction to the initial 6 mp picture. At full digital zoom, in fact no reduction is applied and the original 6 mp picture is just cropped. Because there is no artificial enlargement of the initial picture, the 6 x optical zoom x 1.4 x digital zoom can really presented as a real 8.4 x zoom closde to the quality that would be given by an purely optical 8.4 x zoom (idem in 1 mp mode, but replacing 8.4 then by 13.2).
How to call such a new kind of mix of optiocal zooming with a digital zoom that in fact does not result from an enlargement but from a lesser reduction? A good idea would be to call it a Honeycomb zoom because all this is due to the special honeycomb pattern of the superCCD.
Well, interesting, with this appelation due to the creativity of some marketing people in Fuji and introduced to palliate the fact that Fuji has no longer the right to call its camera a 6 mp camera, Fuji recalls us that the full zoom range it provides is more than 6 x and without any artificial enlargement... In fact, people using it as a 3 mp camera be proud to own a real 8.4 x zoom camera (or 13.2 real zoom camera if you are using it as a 1 mp camera0. This beats the UZI 2100 and the Canon Pro 90 at similar resolution... and without picture enlargment!!!
By the way, if the "digital" zoom is only 1.4 in 3.3 mp mode to not alter the quality of the 3.3 mp pictures when digital zoom is used, this means that the real quality of the initial pictures produced by the 6900 is estimated in the average by Fuji to be 1.4 x 3.3 = 4.62 mp, which correspond well to the current observations: the Fuji 6900 is not really equivalent to a full 6 mp camera, not to a 3.3 mp camera but rather to a 4.62 mp camera... No?
Jean-Paul
So what does this means: that as a 3 mp camera
Jean-Paul, you are well versed in the Fuji Super CCD technology. Did you read up somewhere that explains how the HoneyComb technology works? I was only able to find a short explanation from the Ritz camera catalog about the Super CCD technology. I think your explanation is very good. To set things straight, the label on the side of the lens tube says "13.2x Honeycom Zoom" and not "Honeycomb". Somebody at Fuji messed up on the spelling. Thanks.
ianR
Veteran Member
Hi JP
Strange, I replied to this before but my browser tells me I didn't.
A really excellent piece of deduction once again JP.
That makes a lot of sense and explains a lot. The 4.62mp resolution in particular.
Thanks
Ian
Strange, I replied to this before but my browser tells me I didn't.
A really excellent piece of deduction once again JP.
That makes a lot of sense and explains a lot. The 4.62mp resolution in particular.
Thanks
Ian
SFJP
Veteran Member
A good explanation of the superCCD can be found in a review of the S1 at http://www.picturetrail.com/reviews/PRODS/S1/S1aPT-1a.shtml .
You can see also an animated presentation from Fuji at http://home.fujifilm.com/products/digital/sccd/index.html
But of course, a simple logical reasoning allows to understand much about superCCD. The honeycomb arrangement of photosites on the superCCD obviously cannot directly produce a square grid of pixels as needed by our software that represent pictures as a square grid of pixels.
In this, there is quite a difference with conventional CCD that are themselves a square grid of photosites.
With a conventional CCD, after color Bayer interpolation to guess the RGB value of each photosite, there is the generation of 1 software pixel per hardware photosite with same position on the grid.
To get a square grid of pixels, with the Fuji superCCD it is necessary to complete the honeycomb pattern rows and column with "deduced" intermediary pixels. So, with a 3.3 mp superCCD, you necessarily get a 6 mp (with pixels arranged in a square grid pattern) software image from the superCCD, you cannot get a 3.3 mp image directly (otherwise, translating each photosite into one pixel shifted in postion to be aligned with the square grid would result in a geometrically distorted image) . The closeness and size of the photosites on the superCCD grant that these added "deduced" pixels bear some significant information and really contribute to the increased resolution people observe on images produced by superCCD.
From this, to understand how works the special "digital" zoom of the Fuji 6900 as I explained it in my previous message is obvious, considering that there is no digital zooming at all at the true nominal pixel resolution that is 6mp only.
Note that this is an explanation I have purely built on logical consideration, I have never seen it written anywhere, but seems obvious to me.
But, alas, I have made a big error of computation in my previous message about why the number 1.4 for the digital zoom magnification at 3.3 mp. Note that my explanation of how works the Fuji "digital" zoom is not compromised by the correction of this error, but on the contrary it becomes even more obvious.
My error is that I negected to consider that a zoom magnification is measured by the enlargment of the linear dimensions of the picture and that the resulting enlargement in number of pixels is the square of this linear magnification, not the magnification itself (I have to go back to primary school one of these days). In fact, 1.4 is just the ratio of the lenght (and width) of the 6 mp picture with respect to the 3.3 mp picture. So, naturally Fuji uses a 1.4 magnification just to get a cropped version of the 6 mp picture without any reduction, as a 6 mp picture (2832 x 2128 ) is approximatively 1.4 times larger in its linear dimensions than a 3.3 mp picture (2048 x 1536), i.e approximatively 2832 = 1.4 x 2048 and 2128 = 1.4 x 1536. This explains completely and obviously why 1.4 magnificaiton (in fact 1.38...), Fuji did not limit itself in considering its initial image worth only a 4.6 mp resolution as I very wrongly stated in my previous message (I apologize) but uses well the full 1:1 6 mp image cropped to get its alledged 1.4 X digital zoom magnification in 3.3 mp.
Idem, the linear ratio between the dimensions of the 1 mp picture (1280 x 960) and the 6 mp picture (2832 x 2128 ) is approximatively 2.2 (in fact 2.2125), and this is why the "digital" zoom in this case is up to X 2.2, which gives for 1 mp pictures a real zoom of 2.2 X 6 = 13.2.
Now, with the correction of this error, my explanation of how works Fuji so-called "digital" zoom is perfect and then it becomes clearer why, when used as a 3 mp camera, the Fuji 6900 has a real zoom 8.4 X and not 6X, and a real zoom 13.2 X and not 6 X when used as a 1 mp camera. In other words, you use the extra resolution provided by the 6 mp native images to get the extra 1.4 magnification of the "digital" zoom, this by reducing less, or not at all, and cropping the initial 6 mp image to get the 3,3 mp image.
QDE (Quod Demonstratum Est)
So I repeat, persist and sign: no enlarging interpolation at all on the image is made when using the "digital" zoom with a Fuji 6900, which I am convinced, even more now after the correction of my error than one hour ago, is true.
I cannot believe that Fuji marketing never pushed such an information except by the cumbersome "13.2 X Honey Zoom" mention on some cameras, which would speak to noone except your devoted local Sherlock.
Fuji Marketing people are not as good as Fuji technical people... But that we knew!
Jean-Paul
You can see also an animated presentation from Fuji at http://home.fujifilm.com/products/digital/sccd/index.html
But of course, a simple logical reasoning allows to understand much about superCCD. The honeycomb arrangement of photosites on the superCCD obviously cannot directly produce a square grid of pixels as needed by our software that represent pictures as a square grid of pixels.
In this, there is quite a difference with conventional CCD that are themselves a square grid of photosites.
With a conventional CCD, after color Bayer interpolation to guess the RGB value of each photosite, there is the generation of 1 software pixel per hardware photosite with same position on the grid.
To get a square grid of pixels, with the Fuji superCCD it is necessary to complete the honeycomb pattern rows and column with "deduced" intermediary pixels. So, with a 3.3 mp superCCD, you necessarily get a 6 mp (with pixels arranged in a square grid pattern) software image from the superCCD, you cannot get a 3.3 mp image directly (otherwise, translating each photosite into one pixel shifted in postion to be aligned with the square grid would result in a geometrically distorted image) . The closeness and size of the photosites on the superCCD grant that these added "deduced" pixels bear some significant information and really contribute to the increased resolution people observe on images produced by superCCD.
From this, to understand how works the special "digital" zoom of the Fuji 6900 as I explained it in my previous message is obvious, considering that there is no digital zooming at all at the true nominal pixel resolution that is 6mp only.
Note that this is an explanation I have purely built on logical consideration, I have never seen it written anywhere, but seems obvious to me.
But, alas, I have made a big error of computation in my previous message about why the number 1.4 for the digital zoom magnification at 3.3 mp. Note that my explanation of how works the Fuji "digital" zoom is not compromised by the correction of this error, but on the contrary it becomes even more obvious.
My error is that I negected to consider that a zoom magnification is measured by the enlargment of the linear dimensions of the picture and that the resulting enlargement in number of pixels is the square of this linear magnification, not the magnification itself (I have to go back to primary school one of these days). In fact, 1.4 is just the ratio of the lenght (and width) of the 6 mp picture with respect to the 3.3 mp picture. So, naturally Fuji uses a 1.4 magnification just to get a cropped version of the 6 mp picture without any reduction, as a 6 mp picture (2832 x 2128 ) is approximatively 1.4 times larger in its linear dimensions than a 3.3 mp picture (2048 x 1536), i.e approximatively 2832 = 1.4 x 2048 and 2128 = 1.4 x 1536. This explains completely and obviously why 1.4 magnificaiton (in fact 1.38...), Fuji did not limit itself in considering its initial image worth only a 4.6 mp resolution as I very wrongly stated in my previous message (I apologize) but uses well the full 1:1 6 mp image cropped to get its alledged 1.4 X digital zoom magnification in 3.3 mp.
Idem, the linear ratio between the dimensions of the 1 mp picture (1280 x 960) and the 6 mp picture (2832 x 2128 ) is approximatively 2.2 (in fact 2.2125), and this is why the "digital" zoom in this case is up to X 2.2, which gives for 1 mp pictures a real zoom of 2.2 X 6 = 13.2.
Now, with the correction of this error, my explanation of how works Fuji so-called "digital" zoom is perfect and then it becomes clearer why, when used as a 3 mp camera, the Fuji 6900 has a real zoom 8.4 X and not 6X, and a real zoom 13.2 X and not 6 X when used as a 1 mp camera. In other words, you use the extra resolution provided by the 6 mp native images to get the extra 1.4 magnification of the "digital" zoom, this by reducing less, or not at all, and cropping the initial 6 mp image to get the 3,3 mp image.
QDE (Quod Demonstratum Est)
So I repeat, persist and sign: no enlarging interpolation at all on the image is made when using the "digital" zoom with a Fuji 6900, which I am convinced, even more now after the correction of my error than one hour ago, is true.
I cannot believe that Fuji marketing never pushed such an information except by the cumbersome "13.2 X Honey Zoom" mention on some cameras, which would speak to noone except your devoted local Sherlock.
Fuji Marketing people are not as good as Fuji technical people... But that we knew!
Jean-Paul
SFJP
Veteran Member
Ian,
Look at my new answer to John. The 4.62 mp deduction was pure error from me, I am ashamed on that. The magnification of a zoom is measured in linear magnification, the resulting augmentation of the number of pixels is the square of this linear magnification. So, when linearly magnifying 1.4 times (in fatc 1.38.. times) a 3.3 mp image, you just get an image with 1.4 x 1.4 =1.96 times more pixels, just a ... 6mp image, not a 4.63 mp image as I wrongly stated!!! Shame on me, and that noone can know that I have been a mathematics teacher in high school when I was very young and that I have even been teaching mathematical logic at Ecole Centrale de Paris a bit later, my former students would die laughing...
However, my explanation of how works the Fuji so-called "digital" zoom is not invalidated by the correction of this terrible error, on the contrary my explanation becomes even more obvious (see my new answer to John in this thread where I develop this: http://www.dpreview.com/forums/read.asp?forum=1012&message=1861431 ).
Jean-Paul
Look at my new answer to John. The 4.62 mp deduction was pure error from me, I am ashamed on that. The magnification of a zoom is measured in linear magnification, the resulting augmentation of the number of pixels is the square of this linear magnification. So, when linearly magnifying 1.4 times (in fatc 1.38.. times) a 3.3 mp image, you just get an image with 1.4 x 1.4 =1.96 times more pixels, just a ... 6mp image, not a 4.63 mp image as I wrongly stated!!! Shame on me, and that noone can know that I have been a mathematics teacher in high school when I was very young and that I have even been teaching mathematical logic at Ecole Centrale de Paris a bit later, my former students would die laughing...
However, my explanation of how works the Fuji so-called "digital" zoom is not invalidated by the correction of this terrible error, on the contrary my explanation becomes even more obvious (see my new answer to John in this thread where I develop this: http://www.dpreview.com/forums/read.asp?forum=1012&message=1861431 ).
Jean-Paul
ianR
Veteran Member
Well JP no one here would have noticed the maths error!
Great explanation and I'd like to refer everyone to this page:
Check out the bottom of the page first.
http://home.fujifilm.com/news/n010716.html
Firstly because it shows the high esteem the SuperCCD is held in by the industry, but secondly because it gives a very nice diagram showing exactly how the squares fit around the honeycomb sensor
Another point if understand it correctly is that on a conventional CCD a pixel size of light can only ever be interpreted by one sensor. However on the SuperCCD it is always going to be between 2 or more sensors and each has the ability to interpret part of that light. Hence the ability to interpret the light more accurately. The same way as aircraft can find their exact position by coordinating between 2 beacons.
Regards
Ian
Great explanation and I'd like to refer everyone to this page:
Check out the bottom of the page first.
http://home.fujifilm.com/news/n010716.html
Firstly because it shows the high esteem the SuperCCD is held in by the industry, but secondly because it gives a very nice diagram showing exactly how the squares fit around the honeycomb sensor
Another point if understand it correctly is that on a conventional CCD a pixel size of light can only ever be interpreted by one sensor. However on the SuperCCD it is always going to be between 2 or more sensors and each has the ability to interpret part of that light. Hence the ability to interpret the light more accurately. The same way as aircraft can find their exact position by coordinating between 2 beacons.
Regards
Ian
That's why I love this site - excellent technical discussions. I'm glad I bought the FP6900... with such a large user-base I can rest easy that somebody somewhere knows the answer to my problem. No longer do I have to solely rely on my local shop's salesman to help with my problems (I probably know more about this camera than he does by now). Thanks once again JP and Ian for your excellent contributions to this site.
John.
John.
Tim O
Well-known member
I've had my 6900Z for 2-1/2 months now. I initially used it at the 3mp mode since a lot of other owners were doing the same. Afraid of handling large files on their computers? Anyway, while fooling around taking pictures of the moon about a month ago I starting thinking like Jean-Paul and realized that using something "crazy" like VGA (640x480) resolution and setting the zoom to its maximum at the high end of the digital range simply crops the original 6mp image that must be derived from the honeycomb array. No quality is lost! However, if you use less than maximum digital zoom (as available on the 3mp, 1mp, and VGA resolutions) you are telling the camera to not only crop the original 6mp image but to downsize what's left. This downsizing throws away some quality. Therefore, if you're shooting for the highest quality, use the 6mp mode where there is no digital zoom. But, if the 6mp mode includes more in the scene than you want (like my moon pictures) then don't hesitate to got to 3mp, 1mp, or VGA resolutions to crop your scene. Just BE SURE to use the maximum digital zoom at these lower resolutions to keep all the original 6mp quality. There are two advantages I can see in using lower resolutions and maximizing the digital zoom: (1) your file sizes are much smaller since you're cropping away stuff you didn't want in the first place (2) the view in the EVF is much magnified, making it easier to check focus and know if you're holding the camera still as you shoot.
I am very happy with the way Fuji has implemented this digital zoom. It allows me to get 4.43x210=929mm focal length at VGA resolution and this is the full original quality of the 6mp image, simply cropped.
I am very happy with the way Fuji has implemented this digital zoom. It allows me to get 4.43x210=929mm focal length at VGA resolution and this is the full original quality of the 6mp image, simply cropped.
Peter Heberer
Forum Enthusiast
Just plan awesome information!!
I knew I had a really great camera. Now I know I have a truly remarkable camera. Wait till Phil finds this out. He may just have to redue his review.
Enjoy!
Pete H.
I knew I had a really great camera. Now I know I have a truly remarkable camera. Wait till Phil finds this out. He may just have to redue his review.
Enjoy!
Pete H.
--Pete H
RJNedimyer
Veteran Member
I've been following this conversation and am somewhat amazed. However, I've looked all over my camera not just the lens and can't find HoneCom or 13.2. My camera was purchased in early July could this be the difference???--RayRJNedimyer
Michael J. Hußmann
Veteran Member
You did a great job in explaing how the SuperCCD achieves its improved resolution, but the real reason for the 1.4 factor is that the honeycomb pattern improves the resolution in the horizontal and vertical direction by the square root of 2 = 1.4142135624. If one translates about 3 million pixels (3.3 MP doesn't mean that all 3.3 million pixels are actually used) arranged in a honeycomb pattern into pixels arranged into lines and columns, it has to be 1.4142135624 as many pixels per line and 1.4142135624 as many pixels per column in order to account for the increased resolution, so the total number of pixels in the resulting image will be square root of 2 squared as many -- i.e. twice as many.
- Michael--------------------------------------------------------------- http://www.michael-hussmann.de
Ray,
On newer cameras, the label "13.2x Honeycom Zoom" has replaced the "6x Optical Zoom". The reason I know this is because I bought a 6900 (a week ago) that had "6x Optical Zoom" label. Because I found there was a mechanical malfunction with the len, I exchanged it for another 6900. The new camera has the label "13.2x Honeycom Zoom". That's the reason I'm posting this question. So far I'm pretty satisfied with JP's explanation. That's the only difference I found between the two versions.
John.
On newer cameras, the label "13.2x Honeycom Zoom" has replaced the "6x Optical Zoom". The reason I know this is because I bought a 6900 (a week ago) that had "6x Optical Zoom" label. Because I found there was a mechanical malfunction with the len, I exchanged it for another 6900. The new camera has the label "13.2x Honeycom Zoom". That's the reason I'm posting this question. So far I'm pretty satisfied with JP's explanation. That's the only difference I found between the two versions.
John.
Anand67599
Well-known member
Sorry Jean Paul,
Despite your elaborate explanation I could not get this. Would you like to challange your explanation skills again (against an analytically challenaged fellow
1st I 'll tell you what i had in my mind about zoom stuff:
1. Optical zoom changes perspective, digital zoom just (crop)zoom in to central portion. And hence less file size with digital zoom if Same degree of interpolation used (as opposed to optical zoom).
2. Honeycomb patterns let fuji expand eqv. 6 MP resolution file. Now, if you want 3 mp, well fine. crop in 1/1.4 times each side. central part only appears hence looks bigger with same degree of interpolation 6mp would have used.
3. fuji get 6*1.4=8.4x (opticodigital) zoom at 3MP resolution (produces better pics than native 3mp camera, very good). However, it does not give perspective of more than 6X optical zoom.
4. Now, you can crop in further similarly (4.4x max, thanks to honeycomb) but as similar from here as other camera would do. I fail to understand what is different here once fuji already got interpolation advantage. May be i did not realize what you mena by 'crop in' as oppose to 'reducing less'. thanks, -anand.
Despite your elaborate explanation I could not get this. Would you like to challange your explanation skills again (against an analytically challenaged fellow
1st I 'll tell you what i had in my mind about zoom stuff:
1. Optical zoom changes perspective, digital zoom just (crop)zoom in to central portion. And hence less file size with digital zoom if Same degree of interpolation used (as opposed to optical zoom).
2. Honeycomb patterns let fuji expand eqv. 6 MP resolution file. Now, if you want 3 mp, well fine. crop in 1/1.4 times each side. central part only appears hence looks bigger with same degree of interpolation 6mp would have used.
3. fuji get 6*1.4=8.4x (opticodigital) zoom at 3MP resolution (produces better pics than native 3mp camera, very good). However, it does not give perspective of more than 6X optical zoom.
4. Now, you can crop in further similarly (4.4x max, thanks to honeycomb) but as similar from here as other camera would do. I fail to understand what is different here once fuji already got interpolation advantage. May be i did not realize what you mena by 'crop in' as oppose to 'reducing less'. thanks, -anand.
SFJP
Veteran Member
Anand,
Some remarks.
My messages just pointed to something interesting in Fuji's approach with respect to digital zoom. For most other manufacturers, "digital zoom" means enlarging via interpolation a part of the original picture. For Fuji, "digital zoom" means reducing less the original full resolution picture and then cropping it. As a consequence, you will not find any "digital zoom" in the Fuji if you are using it at full resolution, Fuji "digital zoom" is only available when using a lower resolution resulting from a reduction of the original full resolution images taken by your camera. This is true of the 6900, but also apparently even of other Fuji cameras that do not use the superCCD (e.g the 2600).
When I wrote that used in 3 mp mode the Fuji 6900 as a "real" zoom of 6 x 1.4 = 8.4, I meant that at this 3 mp resolution it gives you a maximal magnification of 8.4 without any loss of image quality, as there is no enlargement of the 3 mp image to get this extra magnification, exactly like if this was due to a 8.4 trully optical zoom.
However, you are right to raise the point that, nevertheless, the extra 1.4 magnification being due to cropping, the 8.4 X opticodigital zoom does not exactly provide exactly the same perspective at full tele than a purely optical 8.4 X zoom. But, there would be a lot to write about optical zoom and perspective, especially with consumer digicams as the small focal length of the lenses give them a perspective rendering closer to a wide-angle in 35 mm camera than a tele even at full tele. Moreover, the depth of field exact width being also morevisible at higher resolution, cropping the full size image may result in a visible reduction of the depth of field exactly like it is the case when using extra optical magnification.
Jean-Paul
Jean-Paul
Some remarks.
My messages just pointed to something interesting in Fuji's approach with respect to digital zoom. For most other manufacturers, "digital zoom" means enlarging via interpolation a part of the original picture. For Fuji, "digital zoom" means reducing less the original full resolution picture and then cropping it. As a consequence, you will not find any "digital zoom" in the Fuji if you are using it at full resolution, Fuji "digital zoom" is only available when using a lower resolution resulting from a reduction of the original full resolution images taken by your camera. This is true of the 6900, but also apparently even of other Fuji cameras that do not use the superCCD (e.g the 2600).
When I wrote that used in 3 mp mode the Fuji 6900 as a "real" zoom of 6 x 1.4 = 8.4, I meant that at this 3 mp resolution it gives you a maximal magnification of 8.4 without any loss of image quality, as there is no enlargement of the 3 mp image to get this extra magnification, exactly like if this was due to a 8.4 trully optical zoom.
However, you are right to raise the point that, nevertheless, the extra 1.4 magnification being due to cropping, the 8.4 X opticodigital zoom does not exactly provide exactly the same perspective at full tele than a purely optical 8.4 X zoom. But, there would be a lot to write about optical zoom and perspective, especially with consumer digicams as the small focal length of the lenses give them a perspective rendering closer to a wide-angle in 35 mm camera than a tele even at full tele. Moreover, the depth of field exact width being also morevisible at higher resolution, cropping the full size image may result in a visible reduction of the depth of field exactly like it is the case when using extra optical magnification.
Jean-Paul
Jean-Paul
Michael J. Hußmann
Veteran Member
Perspective depends on distance, and nothing else. This is something artists have known for centuries, long before photography was invented. You don't change the focal length to change perspective, but to change the field of view. Only if you change the focal length and then also change your distance to the object to again achieve the same field of view, the perspective will change -- but that's due to the change in distance. Whether you choose a longer focal length or crop and scale, the effect -- different resolutions of the resulting pictures aside -- will be the same. Disregarding distortions, a longer focal length will just magnify everything, just like scaling will.
- Michael--------------------------------------------------------------- http://www.michael-hussmann.de
ianR
Veteran Member
Dear Michael
What you say is very interesting, and extremely well reasoned, and although I would be the first to admit I don’t know enough on this subject to argue with you there is still something that doesn’t quite feel right about what you say, so perhaps you’d like to put me right.
Firstly, I think perspective is as much about perception as actual fact, and it’s quite easy to fool the eye in terms of perspective. We all know that if the horizon isn’t straight for example, anything against that horizon will not look straight either. It seems to me that at the wide angle setting we can use more distance markers in the field of view than we can at telephoto so it’s easier to judge perspective.
Secondly, the distance aspect of perspective is more true for the human with two eyes and stereoscopic vision than the one-eyed camera.
Thirdly is just observation. When we see images of sports events on our television screens taken with very powerful telephoto lenses with very good depth of field, two objects which are in reality far apart can seem as if they are right next to each other. Take for example F1, where the cameras are positioned facing down the straight so that the cars look as if they are together when in reality they can be several hundred meters apart.
regards
Ian
What you say is very interesting, and extremely well reasoned, and although I would be the first to admit I don’t know enough on this subject to argue with you there is still something that doesn’t quite feel right about what you say, so perhaps you’d like to put me right.
Firstly, I think perspective is as much about perception as actual fact, and it’s quite easy to fool the eye in terms of perspective. We all know that if the horizon isn’t straight for example, anything against that horizon will not look straight either. It seems to me that at the wide angle setting we can use more distance markers in the field of view than we can at telephoto so it’s easier to judge perspective.
Secondly, the distance aspect of perspective is more true for the human with two eyes and stereoscopic vision than the one-eyed camera.
Thirdly is just observation. When we see images of sports events on our television screens taken with very powerful telephoto lenses with very good depth of field, two objects which are in reality far apart can seem as if they are right next to each other. Take for example F1, where the cameras are positioned facing down the straight so that the cars look as if they are together when in reality they can be several hundred meters apart.
regards
Ian
SFJP
Veteran Member
I'd eventually agree with both of you Michael and Ian
I'd r agree with you Ian, as, apparently, the reduction of the relative size of the objects proportionally to their distance is not the same according to the focal length you use. I have always read that 90 mm was the focal length that gives the perspective the closest to human eye perception, that if you want to amplfy the relative distances of the object in your picture you should use a wide-angle and that to get dramatic effects due to closing the apprent distance between the background and foreground you should use a tele, like a 200 mm.
This being said, Micheal might be right also, because when switching from a tele to a wide angle, to keep a portrait for instance, we have to get closer to the subject, which makes the relative distance of the foreground greater, as well as the relative distance of the nose and the ears might become greater for a face portrait which will amplify the reduction in size of the ears (or amplify the impression of a big nose!). Another example: I take a picture of a personage on the beach in front of a 5 feet high wave that is 6 feet behind the personage. If I use a 400 mm tele, then I have to be far of the personage, say 40 feet, and the relative distance of the personage and the wave is almost the same (46/40) and the wave will appear as if it was with its full size on the personage, ready to engulf him. If I use a wide-angle, I have to be close to the personage, say 6 feet, which means the relative distance of the wave becomes twice the distance of the personage and then the wave will appear twice smaller than its real size with respect to the size of the personage.
However, there might be more than the remark from Michael suggest. The "standard" lems on 35 mm cameras is 50 mm. The "standard" lens on medium format (say 6X9 cms) camera is 90 mm. As 50 mm on 35 mm camera and 90 mm on medium format camera give the same angle of field of view, when taking a portrait at the same distance with both types of cameras we get the same ratio of the size of the personage with respect to the size of the picture. If the remarks from Michael were complete, we should get also the same perception of perspective. Most photographers would tell you it is not the case, that the wide-angle deformation of perspective due to the 50 mm focal length of the 35 mm camera operates there and is visible when compared to the picture taken with the medium format camera and its 90 mm lens.
Is there any professional in the attendance of this forum who could give his/her say on that?
Jean-Paul
I'd r agree with you Ian, as, apparently, the reduction of the relative size of the objects proportionally to their distance is not the same according to the focal length you use. I have always read that 90 mm was the focal length that gives the perspective the closest to human eye perception, that if you want to amplfy the relative distances of the object in your picture you should use a wide-angle and that to get dramatic effects due to closing the apprent distance between the background and foreground you should use a tele, like a 200 mm.
This being said, Micheal might be right also, because when switching from a tele to a wide angle, to keep a portrait for instance, we have to get closer to the subject, which makes the relative distance of the foreground greater, as well as the relative distance of the nose and the ears might become greater for a face portrait which will amplify the reduction in size of the ears (or amplify the impression of a big nose!). Another example: I take a picture of a personage on the beach in front of a 5 feet high wave that is 6 feet behind the personage. If I use a 400 mm tele, then I have to be far of the personage, say 40 feet, and the relative distance of the personage and the wave is almost the same (46/40) and the wave will appear as if it was with its full size on the personage, ready to engulf him. If I use a wide-angle, I have to be close to the personage, say 6 feet, which means the relative distance of the wave becomes twice the distance of the personage and then the wave will appear twice smaller than its real size with respect to the size of the personage.
However, there might be more than the remark from Michael suggest. The "standard" lems on 35 mm cameras is 50 mm. The "standard" lens on medium format (say 6X9 cms) camera is 90 mm. As 50 mm on 35 mm camera and 90 mm on medium format camera give the same angle of field of view, when taking a portrait at the same distance with both types of cameras we get the same ratio of the size of the personage with respect to the size of the picture. If the remarks from Michael were complete, we should get also the same perception of perspective. Most photographers would tell you it is not the case, that the wide-angle deformation of perspective due to the 50 mm focal length of the 35 mm camera operates there and is visible when compared to the picture taken with the medium format camera and its 90 mm lens.
Is there any professional in the attendance of this forum who could give his/her say on that?
Jean-Paul
Michael J. Hußmann
Veteran Member
Hi Ian,
Well, no, it's rather the other way round. Perspective is about depicting a three-dimensional reality in two dimensions. As long as we have stereoscopic vision, we don't care about perspective at all -- with stereoscopic vision, we can judge distances reasonably well, and our brain does a great job of correcting any distortions that might arise because one object is near and appears to be quite big, whereas another object is far away and seems to be small. We just "see" that the apparently small object near the horizon is actually much bigger than the small object in front of our eyes. That this is so easy -- at least it doesn't require any conscious effort -- is precisely the reason why it took painters several centuries before they finally mastered perspective. But perspective isn't a subjective matter (or it doesn't have to be), but just an application of the laws of geometry.
This is actually a good example for the effect of great distances. Suppose an image shot with a telephoto lens shows two soccer players that are 10 meters apart -- but as the distance to the camera is, let's say, 100 meters for the first player and 110 meters for the second, both will be about the same apparent size, and in the absence of any other depth clues, they could just as well be only 1 meter apart, for all we know. If, on the other hand, the distance between the first player and the camera was just 1 meters, the second player would be twice as far (instead 1.1 times as far as in the first case), so (a) the second player would seem to be much smaller than the first, and (b) the depth of field might not suffice to include both.
- Michael--------------------------------------------------------------- http://www.michael-hussmann.de
Keyboard shortcuts
- f
- Forum