16M or 10M? or...?

With my Fuji I drop it from 12M to 6M--recommended for the 2/3" sensor to give better IQ. As I understand Fuji EXR system, the pixels kind of "fill in"/replicate in "M" size v. "L" and the detail actually looks better unless you -really- enlarge.

What about Casio? Is there any advantage to dropping to 10M? I tried it and the photos do seem to look better, but maybe it's just wishful thinking. :-|

I was wondering the same ... I know sizing images in my work flow improves the look for slide shows on an HDTV.

The Casio's images when reduced are only in 4X3 format. If I read the manual correct. I guess they could then be cropped to any other format.

I recently reduce mp's on one of my Panny's to see if overall quality suffered. Still undecided. I think back to an old Oly UZ750 (4mp's) and the images were great and that was a 2/5 sensor.

Will be experimenting to say the least. Thanks for keeping the thought in my head.

vK

I would have assumed that dropping the resolution from 16MP to 4MP would have conveniently allowed four sensor pixels to be amalgamated into one screen or printer pixel, with resulting greater low light clarity. However, 4 MP is not an option with the ZR800, so my idea of how these things work clearly does not match the arcane necessities of electronics engineering.

--

Cyril

Cyril Catt said:

I would have assumed that dropping the resolution from 16MP to 4MP would have conveniently allowed four sensor pixels to be amalgamated into one screen or printer pixel, with resulting greater low light clarity. However, 4 MP is not an option with the ZR800, so my idea of how these things work clearly does not match the arcane necessities of electronics engineering.

--

Cyril

Click to expand...

It's all done by interpolation, so is not a matter of simply combining pixels to make a smaller pixel count.

The camera itself does that interpolation when you choose a lower resolution. It always takes the 16MP shot then crunches it down to something smaller as requested.

Better by far, in my opinion, is to always take 16MP (or whatever the full resolution happens to be) and then reduce the pixel count as needed for whatever purpose by whatever method suits best.

Programs like FastStone Viewer (donateware but so good we should always donate) offer a few methods of interpolation, such as this....



Ideally the method used should match the end purpose and end resize, up or down. After resizing the images always should be examined again for sharpness, downsizing usually needs more sharpness added.

Regards..... Guy

Guy Parsons wrote: Cyril Catt said:

Guy Parsons said:

I would have assumed that dropping the resolution from 16MP to 4MP would have conveniently allowed four sensor pixels to be amalgamated into one screen or printer pixel, with resulting greater low light clarity. However, 4 MP is not an option with the ZR800, so my idea of how these things work clearly does not match the arcane necessities of electronics engineering.

-- Cyril

Click to expand...

It's all done by interpolation, so is not a matter of simply combining pixels to make a smaller pixel count.

The camera itself does that interpolation when you choose a lower resolution. It always takes the 16MP shot then crunches it down to something smaller as requested.

Better by far, in my opinion, is to always take 16MP (or whatever the full resolution happens to be) and then reduce the pixel count as needed for whatever purpose by whatever method suits best.

Programs like FastStone Viewer (donateware but so good we should always donate) offer a few methods of interpolation, such as this....



Ideally the method used should match the end purpose and end resize, up or down. After resizing the images always should be examined again for sharpness, downsizing usually needs more sharpness added. Regards..... Guy

Click to expand...

Thanks for the update, Guy. I had based my conclusion on a fading memory of a Fuji camera, which I think halved resolution and pixel binned to increase low light capability. But I can see that using more computing power on a RAW data set can provide a range of alternatives.

--
Cyril

Cyril Catt said:

Thanks for the update, Guy. I had based my conclusion on a fading memory of a Fuji camera, which I think halved resolution and pixel binned to increase low light capability. But I can see that using more computing power on a RAW data set can provide a range of alternatives.

Click to expand...

I'm not talking RAW, it works on jpegs of course. In fact I've never bothered to use the RAW on my ZR1000, but do use only the RAWs on my Olympus Pen.

Regards........... Guy

I agree with you Guy.

Joao

Guy Parsons wrote: Cyril Catt said:

Guy Parsons said:

Thanks for the update, Guy. I had based my conclusion on a fading memory of a Fuji camera, which I think halved resolution and pixel binned to increase low light capability. But I can see that using more computing power on a RAW data set can provide a range of alternatives.

Click to expand...

I'm not talking RAW, it works on jpegs of course. In fact I've never bothered to use the RAW on my ZR1000, but do use only the RAWs on my Olympus Pen. Regards........... Guy

Click to expand...

I've seldom used RAW, because - for MY needs - it seems a lot more work, more time to capture, and more media needed to store, for little difference from jpegs. But I assumed that, even from small sensors, the extra data that RAW files provide, combined with the presumably greater power of PCs and their specialised apps, do allow a wider scope for the amelioration of unsatisfactory features in the final image than is provided by in-camera processing.

I started out shooting raw on the ZR (but am very satisfied with the Oly jpegs, Guy), but switched to 16M because I hate having to get both a raw and a jpeg shot when choosing raw. Now I am down to shooting 10M. I do not see any difference on my MacBook Pro Retina screen between 16M and 10M, so I went for 10M. I won't be printing any of these so. . .

BTW, at 10M I think the super tele goes all the way up to 1200x, the 16M being limited to 900X. Hard to hold anything steady at either length, but it still is kinda cool and I do use it.

rube39 said:

I started out shooting raw on the ZR (but am very satisfied with the Oly jpegs, Guy), but switched to 16M because I hate having to get both a raw and a jpeg shot when choosing raw. Now I am down to shooting 10M. I do not see any difference on my MacBook Pro Retina screen between 16M and 10M, so I went for 10M. I won't be printing any of these so. . .

BTW, at 10M I think the super tele goes all the way up to 1200x, the 16M being limited to 900X. Hard to hold anything steady at either length, but it still is kinda cool and I do use it.

--
Rube

Click to expand...

Interesting! Is that with the digital zoom for ZR700? I can't hold mine steady, but instead of 18x - 36x - 72xs at 16M in Premium Auto Pro, at 10M it's going from 25 mm to 18x (420mm equiv) to 45x (900?) to 90x (?). (Multi SR Zoom seems to have the same range, unlike the other modes).

(see small cat statue on pond by back fence at 25mm)


#1 Cat statue on bird bath


#2 18x


#3 45x


#4 90x

rube39 said:

I started out shooting raw on the ZR (but am very satisfied with the Oly jpegs, Guy), but switched to 16M because I hate having to get both a raw and a jpeg shot when choosing raw. Now I am down to shooting 10M. I do not see any difference on my MacBook Pro Retina screen between 16M and 10M, so I went for 10M. I won't be printing any of these so. . .

Click to expand...

Thanks for the feedback, Rube.

The 10MP does make sense when you think that the 1/2.33 inch sensors are all living in the zone of diffraction limiting, so it is basically impossible to get 16MP worth of resolution anyway. I hadn't thought of that, but was stuck in M4/3 size sensor mindset where the turning point is about f/8. With the smaller sensor it is probably f/2.8.

rube39 said:

BTW, at 10M I think the super tele goes all the way up to 1200x, the 16M being limited to 900X. Hard to hold anything steady at either length, but it still is kinda cool and I do use it.

Click to expand...

Hmmm, must play further with the ZR1000 at 10MP and see what happens with the super tele, it of course has the 24-300mm equivalent lens so only goes to a lesser super tele limit.

Reading the manual it seems that some functions/best shots automatically take the output to 10MP anyway despite initially setting to 16MP. Also I see this table for the ZR1000 digital zoom limits.....


Page 60 of ZR1000 manual, optical zoom is 24-300mm equivalent.

So 199.3X seems to be my absolute limit for a 640x480 result.

Also looked at the ZR800 manual and found this......


Page 52 of ZR800 manual, has 25-450mm equivalent lens.

So I only see 286.9X as the highest possible zoom factor. Where did you see the 1200x/900x numbers, Rube? Maybe thinking of mm equivalent where the 16MP is limited to 72X which means 72 x 25mm = 1800mm equivalent, and at 10MP limit is 90.9 x 25mm = 2272.5mm equivalent.

Anyway, time for some experiments and some critical 16MP vs 10MP evaluation. Not that I doubt you, Rube, it's just that the ZR1000 (or my wife's ZR200) plus our PCs & monitors may treat the images differently to your combination.

[Later.... Trying the advanced diffraction calculator at http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm I see that at 16MP on the 1/2.3 inch sensor that f/2.8 seems to be the start of diffraction limiting while at 10MP it is f/4. So in all truth 16MP is nonsense with that sensor size and available apertures of that lens, so it's down to 10MP for me to make it more sensible].

Regards.... Guy

Guy,

Excellent technical analysis, as usual!

Lisetta1 said:

With my Fuji I drop it from 12M to 6M--recommended for the 2/3" sensor to give better IQ. As I understand Fuji EXR system, the pixels kind of "fill in"/replicate in "M" size v. "L" and the detail actually looks better unless you -really- enlarge.

What about Casio? Is there any advantage to dropping to 10M? I tried it and the photos do seem to look better, but maybe it's just wishful thinking. :-|

Click to expand...

Using my ZR1000 I did try a few 16MP vs 10MP shots in a very casual fashion and while the result is close, I find that on some fine detail I do see a whisker better result with the 16MP so will stick with that for now.

More careful testing needed though, but pixel peeping the small sensor cameras is a bit futile, lots of noise and artifacts, I'm so used the Micro 4/3 cameras that the Casio is rather disappointing at 100% pixel peep. So best to avoid doing that.

In the past I have printed crops from the 16MP Casio images to simulate what 16x20 inch print detail might look like, and shots taken at lowest ISO out in the sun do print up very well indeed despite looking grim at 100% pixel peep.

Summary... storage is cheap so 16MP is what I'll do.

Regards...... Guy

Guy Parsons said:

Lisetta1 said:

With my Fuji I drop it from 12M to 6M--recommended for the 2/3" sensor to give better IQ. As I understand Fuji EXR system, the pixels kind of "fill in"/replicate in "M" size v. "L" and the detail actually looks better unless you -really- enlarge.

What about Casio? Is there any advantage to dropping to 10M? I tried it and the photos do seem to look better, but maybe it's just wishful thinking. :-|

Click to expand...

Using my ZR1000 I did try a few 16MP vs 10MP shots in a very casual fashion and while the result is close, I find that on some fine detail I do see a whisker better result with the 16MP so will stick with that for now.

More careful testing needed though, but pixel peeping the small sensor cameras is a bit futile, lots of noise and artifacts, I'm so used the Micro 4/3 cameras that the Casio is rather disappointing at 100% pixel peep. So best to avoid doing that.

Click to expand...

Yes. It's good people here seem able to do that, while still offering constructive criticism and advice.

Guy Parsons said:

In the past I have printed crops from the 16MP Casio images to simulate what 16x20 inch print detail might look like, and shots taken at lowest ISO out in the sun do print up very well indeed despite looking grim at 100% pixel peep.

Click to expand...

That's the thing, how they look as you want to display them. I'm looking forward to making some prints--curious about that. 100% pp isn't everything (or even, in some cases, important at all).

I know these Casios have small sensors with IQ drawbacks compared to many, but I'm looking forward to learning more about the relative strengths, for what it is. I'm hoping there'll be more testing/examples from various settings/modes in this forum to help make using the camera more rewarding, even within the limitations. (For example, while it may be of little use most of the time, it's still valuable imo to know that using 10MP extends the telephoto/digital cropping to over 2000mm equivalent in SR Zoom and Premium Pro Auto modes. Little tricks of the camera are interesting to know and hopefully to see in use as well. Likewise, understanding settings and situations that are most likely to yield best results with this little camera, ultimately, will be fun.)

Lisetta1 said:

With my Fuji I drop it from 12M to 6M--recommended for the 2/3" sensor to give better IQ. As I understand Fuji EXR system, the pixels kind of "fill in"/replicate in "M" size v. "L" and the detail actually looks better unless you -really- enlarge.

What about Casio? Is there any advantage to dropping to 10M? I tried it and the photos do seem to look better, but maybe it's just wishful thinking. :-|

Click to expand...

The Fuji is unique, the sensor is designed to work at 6mp in order to improve DR, no other camera can do this trick. Think of the Fuji as a 6mp camera and not a 12 mp camera being downsized to 6mp