Longer zoom, or larger sensor & crop in post?

[Kawika Nui]

I am mathematically challenged, so I hope someone will help me out here.

The goal: shoot pictures of subjects far away with as much detail as possible.

The choice:

12mp bridge camera with up to 600mm equiv. zoom, 1/2.3" sensor (FZ150) or

20mp camera with up to 250mm equiv. zoom, 1" sensor (ZS100).

The question: will I get longer "reach" by using maximum zoom on the FZ150 or by using the maximum zoom on the ZS100 and cropping? In other words, which camera will allow me to get the most detail in a photo of the same subject at the same distance?

 

[Kawika Nui]

Thanks to everyone for helping me "get" this. I still am having a bit of trouble understanding the answers, so please bear with me.

I came across digicamdb.com which has a nice sensor size comparison feature. Here's what I got:

1in sensor@20MP vs. 1/ 2.3in sensor@12MP

1 inch vs. 1 /2.3 inch

Sensor size 13.2 x 8.8 mm vs. 6.16 x 4.62 mm

Sensor resolution 5492 x 3661 vs. 4011 x 3016

Diagonal 15.86 mm vs. 7.70 mm

Surface area 116.16 mm² vs. 28.46 mm²

Pixel pitch 2.4 µm vs. 1.54 µm

Pixel size 5.76 µm² vs. 2.37 µm²

Pixel density 17.31 MP/cm² vs. 42.4 MP/cm²

Crop factor 2.73 vs. 5.62

1 inch sensor has 308% more surface area

1 inch sensor pixel pitch is 56% higher

1 inch sensor pixel size is 143% larger

1 /2.3inch sensor pixel density is 145% higher

Working from the crop factor, 5.62/2.73 = ~2 (2.0586). Shouldn't this mean that by cropping about half of the ZS100 image one could attain a ~500mm EFL at about 10mp (half of the original 20mp? And of course with larger pixels?

Working from the surface area, it would seem that one could crop to about 1/3 of the original ZS100 image and have about 750mm EFL at about 7mp (about 1/3 of the original 20mp). Again, with larger pixels.

These would be a lot higher resolution than those presented on this thread; what am I doing wrong?

Thanks!

BTW, please excuse any delay in responding to answers, since for some reason dpreview is ignoring the fact that I have checked the "Subscribe to email updates" box.

 

[Kawika Nui]

Here's a good tool I came across:

http://www.digicamdb.com/compare/panasonic_lumix-dmc-fz1000-vs-panasonic_lumix-dmc-fz200/

You can substitute any cameras you want.

 

[kkardster]

1 inch sensor has 308% more surface area

I think this is 408% (~4X)

--
Bruce
You learn something new every time you press the shutter

 

[chezchuck]

I assume what cainn24 did was to say that to achieve the same angle of view of 600 mm with a 250 mm lens you need to crop by a factor 600/250 ie 2.4 But that is a linear dimension and when we speak of Megapixels they are spread over an area. So assuming the same shape ie aspect ratio we need to divide the 20MP by (2.4) squared. Hence 20MP / 2.4 squared = 3.5 MP.

Rated #1 answer - easy to understand and makes sense. Assuming we are talking about 35mm equivalent focal lengths. Gets a little more complicated when actual focal lengths are used.

 

[kkardster]

These would be a lot higher resolution than those presented on this thread; what am I doing wrong?

I think you may be improperly imposing linear focal length math onto 2-dimensional sensor area. Squaring [or square rooting] is required, depending on what's being compared.

--
Bruce
You learn something new every time you press the shutter

 

[kkardster]

1 inch sensor has 308% more surface area

I think this is 408% (~4X)

Sorry - I tried withdrawing my response but it was too late. I see now that you meant 308% more as a difference, making the 1" sensor 408% (4X) the size of the 1/2.3" sensor.

 

[Kawika Nui]

I am mathematically challenged, so I hope someone will help me out here.

The goal: shoot pictures of subjects far away with as much detail as possible.

The choice:

12mp bridge camera with up to 600mm equiv. zoom, 1/2.3" sensor (FZ150) or

20mp camera with up to 250mm equiv. zoom, 1" sensor (ZS100).

The question: will I get longer "reach" by using maximum zoom on the FZ150 or by using the maximum zoom on the ZS100 and cropping? In other words, which camera will allow me to get the most detail in a photo of the same subject at the same distance?

There are a lot of variables involved, but the FZ1000 can (1", 400mm EFL) can essentially match the FZ150/200/300 (1/2.3", 600mm EFL)

Hello Bruce,

Theoretically, this sounds plausible. In practice, it may not work, depending on one's threshold of acceptability when cropping.

If I filled the frame with the FZ200, and photographed from the same distance with the FZ1000, the crop to match the FZ200 framing usually turned out OK.

I said "usually" because in the case of certain wildlife, noticeable detail is sometimes lost when cropping the FZ1000 image to match that of the FZ200.

But if my FZ200 image did not fill the frame and I cropped for the final image, then the FZ1000 crop to match was not acceptable to me.

More and more, I pass up opportunities if a large crop is involved -- I don't click the shutter. With any lens, the closer you are, the more pixels you have to work with.

Each photographer will have her/his threshold of acceptability regarding cropping.

Agreed - this particular FZ1000 vz FZ150/200/300 comparison is right at the edge and falls on one side of the fence or the other depending on conditions and needs. I only offered it here as an example of how if the FZ1000 can somewhat/barely match the smaller FZ's, there's no way the ZS100 can possibly come close.

I don't have the FZ1000 but from reading threads here I understand that for many FZ1000 owners, shooting the FZ1000 using picture size M [possibly with i.Zoom] closely matches the 600mm EFL FZ models for their purposes.

yes, this is the "ETC" (Extra Tele Conversion) mode which uses sensor crop to increase the FL without any resizing or resampling. I use it all the time on my GX7--especially for video, since the video is only 2mp frames anyway. For stills, "M" yields 8mp and "S" yields 4mp (from a 16mp sensor); max effect is ~2x (max effect for video is 2.4x).

Of course - as with direct cropping in post - YMMV.

--
Bruce
You learn something new every time you press the shutter

 

[Kawika Nui]

1 inch sensor has 308% more surface area

I think this is 408% (~4X)

Sorry - I tried withdrawing my response but it was too late. I see now that you meant 308%

I just copied and pasted from the digicam database. Not my numbers.

more as a difference, making the 1" sensor 408% (4X) the size of the 1/2.3" sensor.

--
Bruce
You learn something new every time you press the shutter

 

[kohinoor]

Thanks to everyone for helping me "get" this. I still am having a bit of trouble understanding the answers, so please bear with me.

I came across digicamdb.com which has a nice sensor size comparison feature. Here's what I got:

1in sensor@20MP vs. 1/ 2.3in sensor@12MP

1 inch vs. 1 /2.3 inch

Sensor size 13.2 x 8.8 mm vs. 6.16 x 4.62 mm

Sensor resolution 5492 x 3661 vs. 4011 x 3016

Diagonal 15.86 mm vs. 7.70 mm

Surface area 116.16 mm² vs. 28.46 mm²

Pixel pitch 2.4 µm vs. 1.54 µm

Pixel size 5.76 µm² vs. 2.37 µm²

Pixel density 17.31 MP/cm² vs. 42.4 MP/cm²

Crop factor 2.73 vs. 5.62

1 inch sensor has 308% more surface area

1 inch sensor pixel pitch is 56% higher

1 inch sensor pixel size is 143% larger

1 /2.3inch sensor pixel density is 145% higher

Working from the crop factor, 5.62/2.73 = ~2 (2.0586). Shouldn't this mean that by cropping about half of the ZS100 image one could attain a ~500mm EFL at about 10mp (half of the original 20mp? And of course with larger pixels?

Working from the surface area, it would seem that one could crop to about 1/3 of the original ZS100 image and have about 750mm EFL at about 7mp (about 1/3 of the original 20mp). Again, with larger pixels.

These would be a lot higher resolution than those presented on this thread; what am I doing wrong?

To double the EFL you have to divide by four, to triple the EFL divide by nine....!

Thanks!

BTW, please excuse any delay in responding to answers, since for some reason dpreview is ignoring the fact that I have checked the "Subscribe to email updates" box.

 

[davesurrey]

Thanks for your kind comments. I think that despite what some think negatively about "equivalence" it's a good metric and useful for such calculations. But there are also different ways of looking at these things.

 

[spike29]

im not a realy experianced shooter but there are more then just pixels numbers that give the diverence.

the focus in the distance view (Cropping) or focus on close view (optical zoom) i think the optical zoom is more accurate in light and object focus. (only the blur from movement will grow.)

Depence how to watch it:

on photopaper, cloth xl printer, FHD or 4K 50" tv

TV is easy, when the photo resoultion is higher then the pixel count of the tv it will be fine.

Other wise the auto zoom correction of the TV is adding pixels and that will gonabe visible.

 

[papa natas]

I am mathematically challenged, so I hope someone will help me out here.

The goal: shoot pictures of subjects far away with as much detail as possible.

The choice:

12mp bridge camera with up to 600mm equiv. zoom, 1/2.3" sensor (FZ150) or

20mp camera with up to 250mm equiv. zoom, 1" sensor (ZS100).

The question: will I get longer "reach" by using maximum zoom on the FZ150 or by using the maximum zoom on the ZS100 and cropping? In other words, which camera will allow me to get the most detail in a photo of the same subject at the same distance?

On the bench the new ZS100 looks better with a larger sensor and higher quality screen. It also has multi focus zone and 5 axis shake reduction. Then again I wish they had supplied a flash hotshoe. I'll wait for the tests once the ZS100 has hit the market. Mind you, I own the FZ50, an oldie, but a good workhorse.

 

[cainn24]

I am mathematically challenged, so I hope someone will help me out here.

The goal: shoot pictures of subjects far away with as much detail as possible.

The choice:

12mp bridge camera with up to 600mm equiv. zoom, 1/2.3" sensor (FZ150) or

20mp camera with up to 250mm equiv. zoom, 1" sensor (ZS100).

The question: will I get longer "reach" by using maximum zoom on the FZ150 or by using the maximum zoom on the ZS100 and cropping? In other words, which camera will allow me to get the most detail in a photo of the same subject at the same distance?

On the bench the new ZS100 looks better with a larger sensor and higher quality screen. It also has multi focus zone and 5 axis shake reduction.

5-axis IS is only available for standard video recording. You don't get it for stills, and you don't get it for 4K. In such situations the camera falls back on the same Power OIS system that the FZ150 uses.

 

[papa natas]

Thanks for clearing that up.

 

[cainn24]

Thanks for clearing that up.

No problem. Panasonic have been annoyingly vague about it since its introduction, and therefore so have many camera review sites. I wonder if it would be unreasonable to suggest that being vague is sometimes a good marketing strategy.

 

[ANAYV]

I am mathematically challenged, so I hope someone will help me out here.

The goal: shoot pictures of subjects far away with as much detail as possible.

The choice:

12mp bridge camera with up to 600mm equiv. zoom, 1/2.3" sensor (FZ150) or

20mp camera with up to 250mm equiv. zoom, 1" sensor (ZS100).

The question: will I get longer "reach" by using maximum zoom on the FZ150 or by using the maximum zoom on the ZS100 and cropping? In other words, which camera will allow me to get the most detail in a photo of the same subject at the same distance?

On the bench the new ZS100 looks better with a larger sensor and higher quality screen. It also has multi focus zone and 5 axis shake reduction.

5-axis IS is only available for standard video recording. You don't get it for stills, and you don't get it for 4K. In such situations the camera falls back on the same Power OIS system that the FZ150 uses.

I stated this, back when the FZ300 was released, and got lots of flack for it.

Panasonic seems to beat around the bush with info on this, but I came up with same conclusion you did, cainn.

did you read any statement that CLEARY spells this out?

I had to put pieces together to come up with my conclusion.

I don't believe there is 5 axis mechanical IS, only 2 or 3 axix.

Some electronic/digital correction must be happening, IMHO.

ANAYV

 

[gardenersassistant]

5-axis IS is only available for standard video recording. You don't get it for stills, and you don't get it for 4K. In such situations the camera falls back on the same Power OIS system that the FZ150 uses.

I stated this, back when the FZ300 was released, and got lots of flack for it.

Panasonic seems to beat around the bush with info on this, but I came up with same conclusion you did, cainn.

did you read any statement that CLEARY spells this out?

I had to put pieces together to come up with my conclusion.

I don't believe there is 5 axis mechanical IS, only 2 or 3 axix.

Some electronic/digital correction must be happening, IMHO.

ANAYV

I just spotted http://eng-ca.faq.panasonic.com/app...s-hybrid-ois-+-and-how-does-it-work?-hc-x1000

I seem to recall (but don't have a source for it) that OIS deals with rotational movements, and that the extra axes the hybrid stabilisation provide are lateral (side to side and up and down). The diagram on the linked page associated with "4 & 5 Arm Movements - For zooming. All area of the sensor that are not used for zooming are electronically controlled to comensate for handshake." seems highly suggestive to me.

My guess has previously been that the video "frame" is smaller in pixel dimension than the sensor, and that the frame is moved around so as to mirror the detected lateral movement of the camera, and that statement and the associated diagram seem to be compatible with this. If this is the case, then of course it could only be used for video, not stills.

 

[cainn24]

I am mathematically challenged, so I hope someone will help me out here.

The goal: shoot pictures of subjects far away with as much detail as possible.

The choice:

12mp bridge camera with up to 600mm equiv. zoom, 1/2.3" sensor (FZ150) or

20mp camera with up to 250mm equiv. zoom, 1" sensor (ZS100).

The question: will I get longer "reach" by using maximum zoom on the FZ150 or by using the maximum zoom on the ZS100 and cropping? In other words, which camera will allow me to get the most detail in a photo of the same subject at the same distance?

On the bench the new ZS100 looks better with a larger sensor and higher quality screen. It also has multi focus zone and 5 axis shake reduction.

5-axis IS is only available for standard video recording. You don't get it for stills, and you don't get it for 4K. In such situations the camera falls back on the same Power OIS system that the FZ150 uses.

I stated this, back when the FZ300 was released, and got lots of flack for it.

Panasonic seems to beat around the bush with info on this, but I came up with same conclusion you did, cainn.

You can always expect some flack when you try to tell people something they don't want to hear, and people don't want to hear that a camera they just bought, or are planning to buy, doesn't have a feature that someone, somewhere, erroneously implied that it did.

Ultimately I blame Panasonic for this for being vague, and camera review sites for making assumptions and jumping to conclusions.

did you read any statement that CLEARY spells this out?

Here's a better idea: see if you can find a place where Panasonic themselves have ever explicitly stated that 5-axis IS is available in any sort of stills mode.

In their most recent spec sheets (FZ300, ZS100, TZ80) you will find the following note: "*5-Axis compensation works in video recording except for 4K video or high-speed video recording."

No mention of stills.

In older spec sheets (FZ1000) Panasonic have been a little more careless by virtue of not qualifying the feature listing in the same way. But in the advanced manual you will find the following remark: "The 5-axis Hybrid Image Stabilizer function automatically functions while recording motion pictures. This decreases the jitter of the image when recording motion pictures while zooming or walking, etc."

Again, no mention of stills.

It also bears mentioning that 5-axis stabilization has only ever been marketed as a video feature from the very beginning. Specifically, as an augmentation of the existing Power OIS system. See here: https://web.archive.org/web/20131231202830/http://panasonic.net/avc/lumix/ois/index.html

"The POWER O.I.S. is upgraded with an inclusion of HYBRID O.I.S. + for video recording with new 5-axis correction"

I had to put pieces together to come up with my conclusion.

I don't believe there is 5 axis mechanical IS, only 2 or 3 axix.

No, it's definitely not mechanical. Panasonic themselves even explicitly state as much here: http://eng-ca.faq.panasonic.com/app...s-hybrid-ois-+-and-how-does-it-work?-hc-x1000

"All area of the sensor that are not used for zooming are electronically controlled to compensate for handshake."

Some electronic/digital correction must be happening, IMHO.

In addition to all of the above information we know the following: with OIS switched off the cameras focal length range is the same for standard HD video (1080p, not 4K) as it is for stills. But with OIS enabled it is not. This is because, as Panasonic states, Hybrid OIS (which sacrifices peripheral image data so it can have a buffer for performing electronic corrections) is automatically enabled in that video mode in addition to Power OIS. So the effective focal length range is changed.

If 5-axis was active in stills mode why don't we see the same change in effective focal length with OIS enabled?

With all that in mind I feel confident enough to declare the matter settled. There is no 5-axis stabilization for stills mode. And any camera review site that suggests otherwise is in error.

 

[cainn24]

5-axis IS is only available for standard video recording. You don't get it for stills, and you don't get it for 4K. In such situations the camera falls back on the same Power OIS system that the FZ150 uses.

I stated this, back when the FZ300 was released, and got lots of flack for it.

Panasonic seems to beat around the bush with info on this, but I came up with same conclusion you did, cainn.

did you read any statement that CLEARY spells this out?

I had to put pieces together to come up with my conclusion.

I don't believe there is 5 axis mechanical IS, only 2 or 3 axix.

Some electronic/digital correction must be happening, IMHO.

ANAYV

I just spotted http://eng-ca.faq.panasonic.com/app...s-hybrid-ois-+-and-how-does-it-work?-hc-x1000

I seem to recall (but don't have a source for it) that OIS deals with rotational movements, and that the extra axes the hybrid stabilisation provide are lateral (side to side and up and down).

There's no way for an in-lens stabilization mechanism to correct for roll because there's literally nothing a floating lens element can do to change the orientation of the image relative to the sensor (just as you can't change the orientation of the image you see when looking through the rear element of a lens just but turning the lens itself).

What you need is a way to "decouple" things. That means either a sensor shift component, or equivalent digital corrections.

Presumably Panasonic haven't implemented sensor-shift stabilization in any of their fixed-lens cameras since there has been no song and dance about it as we would reasonably expect there to be. So I would suggest that correction for roll is being done electronically as well.

The diagram on the linked page associated with "4 & 5 Arm Movements - For zooming. All area of the sensor that are not used for zooming are electronically controlled to comensate for handshake." seems highly suggestive to me.

My guess has previously been that the video "frame" is smaller in pixel dimension than the sensor, and that the frame is moved around so as to mirror the detected lateral movement of the camera, and that statement and the associated diagram seem to be compatible with this. If this is the case, then of course it could only be used for video, not stills.

Sounds like a good explanation to me.

 

[papa natas]

But, of course!!

And nobody is asking questions any more.