G9 digital zoom vs cropping

[Sangoma]

Topic came up a while back. To my mind they should both have the same result.


These were taken and then one brick (475' away) cropped from the result.



This one is using the digital zoom



And this one is without the digital zoom


The cropping isn't exact, but the length of the brick is pretty much the same.

I can't see any difference that makes one preferable to the other.


While I was outside, I also compared the G9 +400 to the 5D + 300, and again cropped the same area.
The 5D + 300 was surprisingly good.

I have the contrast and saturation set right down on the G9



5D + 300







G9 + 400

 

[lindsay]

interesting experiments

 

[JohnMcL7]

In general the results should be the same since the digital zoom is simply cropping the image. The one area I've seen the digital zoom is claimed to be different is that the camera calculates exposure for the zoomed in area so if you have exposure set to measure for whole frame, potentially it could be more accurate when digital zooming vs a crop afterwards. It's not something I've tried myself though and prefer to capture the full frame and then I have the flexibility to crop as I want after.

 

[Nod]

As lindsay says, interesting experiment and as John says, the results should be the same but PP cropping adds flexibility in the final framing.

 

[ajophotog]

Doesn’t digital zoom use jpg rather than raw, l think on my sony to use it you need to shoot jpg, just a thought that might well be incorrect.
l would test a raw crop against digital zoom if that was the case to see how that performs.

 

[the black fox]

I’ve started using this on the OM1 with a 300mm f4 + 1.& tc if you then use FF terms of equivalence it gives a reach of 1680 mm at F5.6 and thanks to i.b.i.s is fully hand holdable .. yes the result is a .jpg but with a bit of menu tweaking in camera the results are really close to a edited raw file ..
it’s all subjective and the usual rules apply I.e get as close as you can to your subject ,it’s a camera not a Jodrell Bank telescope

 

[Sangoma]

I’ve started using this on the OM1 with a 300mm f4 + 1.& tc if you then use FF terms of equivalence it gives a reach of 1680 mm at F5.6 and thanks to i.b.i.s is fully hand holdable .. yes the result is a .jpg but with a bit of menu tweaking in camera the results are really close to a edited raw file ..
it’s all subjective and the usual rules apply I.e get as close as you can to your subject ,it’s a camera not a Jodrell Bank telescope

Bit puzzled, 300X1.7X2=1020, what am I calculating wrong?

 

[the black fox]

Bit puzzled, 300X1.7X2=1020, what am I calculating wrong?

300+1.4 = 420mm x 2 crop =840mm x2 digital crop = 1680mm

 

[Sangoma]

300+1.4 = 420mm x 2 crop =840mm x2 digital crop = 1680mm

Ah, OK I didn't see the digital crop, sorry.
And I translated 1.& as 1.7 as the 7 is on the unshifted & key.

I don't have any lenses for the G9 that can take the TC, so I'm stuck with a measly 3200mm with a 4X digital zoom, or 1600mm with the 2X digital zoom

 

[Micro-four-thirds]

I think I’m going to put my neck on the block here but I’m sure I read somewhere recently, and I can’t be certain where, that with Olympus and using the digital zoom on the om1 mk1 the jpeg still retains the 20mp instead of reducing it by half.
Can anyone confirm if this is correct or have I misread or imagined something?

 

[sk66]

I think I’m going to put my neck on the block here but I’m sure I read somewhere recently, and I can’t be certain where, that with Olympus and using the digital zoom on the om1 mk1 the jpeg still retains the 20mp instead of reducing it by half.
Can anyone confirm if this is correct or have I misread or imagined something?

Not really... you simply cannot crop and retain the same resolution. What some cameras do is upscale the cropped image so you have the same MP; but that doesn't increase the actual resolution/detail recorded.

 

[the black fox]

I think I’m going to put my neck on the block here but I’m sure I read somewhere recently, and I can’t be certain where, that with Olympus and using the digital zoom on the om1 mk1 the jpeg still retains the 20mp instead of reducing it by half.
Can anyone confirm if this is correct or have I misread or imagined something?

There will always be people that try to look knowledgeable with these things . It’s totally irrelevant as the only thing that matters in the end is the Finished photo . The kingfisher I did a week or so ago decided to hide in deep cover in the bushes trying to get it in raw was a nightmare as the AF was picking up the surrounding twigs ,by using the 2x digital it excluded them from the frame and got the shot ..
however I will look at the file sizes later as I have both and confirm one way or the other

 

[Micro-four-thirds]

There will always be people that try to look knowledgeable with these things . It’s totally irrelevant as the only thing that matters in the end is the Finished photo . The kingfisher I did a week or so ago decided to hide in deep cover in the bushes trying to get it in raw was a nightmare as the AF was picking up the surrounding twigs ,by using the 2x digital it excluded them from the frame and got the shot ..
however I will look at the file sizes later as I have both and confirm one way or the other

That will be great, thank you. I understand fully where your coming from with regards the end result being what matters, I am just inquisitive as to whether the final image is boosted so to speak to 20mp or halved.

 

[Micro-four-thirds]

Not really... you simply cannot crop and retain the same resolution. What some cameras do is upscale the cropped image so you have the same MP; but that doesn't increase the actual resolution/detail recorded.

That’s interesting, thank you.
That’s the bit I was struggling to understand when I found the article mentioning it last week. I couldn’t see how the image could be cropped in camera and still retain 20mp in the same way a standard image would.

 

[the black fox]

That’s interesting, thank you.
That’s the bit I was struggling to understand when I found the article mentioning it last week. I couldn’t see how the image could be cropped in camera and still retain 20mp in the same way a standard image would.

and the totally wrong answer .. just done a factual check of file sizes from my OM1 and the actual results are

RAW FILE SIZE 5184x 3888 uncropped image

.JPG FILE SIZE 5184x3888 uncropped image

you can't argue with the figures the digital TC 2x image on olympus cameras at least retains exactly the same file size in both raw and jpeg files . as I said earlier it matters not one iota about supposed resolution etc the file size is what you work from .end of argument. I think where Steven sk66 is going wrong is referring to a CROPPED image which is factual

 

[Sangoma]

There will always be people that try to look knowledgeable with these things . It’s totally irrelevant as the only thing that matters in the end is the Finished photo . The kingfisher I did a week or so ago decided to hide in deep cover in the bushes trying to get it in raw was a nightmare as the AF was picking up the surrounding twigs ,by using the 2x digital it excluded them from the frame and got the shot ..
however I will look at the file sizes later as I have both and confirm one way or the other

Sorry, I think that is a little unfair, the thread was about the resolution and image quality compared between using digital zoom and cropping in post.

The finished photo didn't come into it, though I agree in your case where you would not have got two shots to compare, it was a useful feature that enables you to get the shot you wanted.

Someone else also mentioned focus and exposure, and depending on camera features there might be other advantages to use it or not use it.

 

[sk66]

RAW FILE SIZE 5184x 3888 uncropped image

.JPG FILE SIZE 5184x3888 uncropped image

Are you saying the images are the same composition (uncropped)? In that case they would be the same regardless.

If you are saying the jpeg used the 2x digital TC and the compositions are different, then there are only two possible answers as to how that is possible. The camera either used the central 1/4 of the sensor and upsampled it, or the camera used the central 1/4 and pixel shift (it can, but it doesn't for the TC function). The OM1 and *G9 both resample; I'm sure others do as well.

It is quite possible that the resample is better than you might get with a simple resize... more like what you would get by using Topaz Gigapixel AI. There is certainly an advantage to resampling a low resolution image to prevent it from pixelating. And there may be no advantage to actually having 20MP of sensor resolution on such a small area (diffraction, etc); but they are certainly not the same thing.

The G9 has extra tele conversion mode which is only a crop (resulting in medium or small picture sizes) and works with raw+jpeg; and digital zoom mode, which is the same thing but resampled and works only with jpeg... and as shown in the OP; it doesn't really improve the results over just a crop.

 

[nandbytes]

Olympus 2x digital zoom is simply upscaling the cropped file to 20mp. There are other cameras and even phones that do the same thing.

I get the feeling you'll get better results with something like topaz gigapixel AI for upscaling/enlarging images than in camera upscaling.

Pretty easy to do comparisons tbh if anyone cares and actually has the camera

 

[the black fox]

What a load of over officious waffle . Simply no need for it . You take a photo of a distant bird in raw you then have to crop it till it fills the frame and do all sorts of adjustments to get it to fill the frame . Or you take the same shot ,fill the frame no cropping needed .and very little if any adjustments . The results are to close to call only difference is one takes five minutes the other takes 30 seconds . .
I have no idea or desire to find out how the digital crop works in Panasonic , james asked a question it was answered incorrectly I checked the FACTS in case I was wrong . But the FILE sizes are identical however you wish to put it it’s still a 20mp file

 

[Micro-four-thirds]

Thanks for checking the file sizes, that’s certainly interesting. I’m going to have a go this weekend and see how I get on with the em5.3. I’ve tried the dtc before but never gone further than actually taking the photo using it.

and the totally wrong answer .. just done a factual check of file sizes from my OM1 and the actual results are

RAW FILE SIZE 5184x 3888 uncropped image

.JPG FILE SIZE 5184x3888 uncropped image

you can't argue with the figures the digital TC 2x image on olympus cameras at least retains exactly the same file size in both raw and jpeg files . as I said earlier it matters not one iota about supposed resolution etc the file size is what you work from .end of argument. I think where Steven sk66 is going wrong is referring to a CROPPED image which is factual

 

[nandbytes]

its not waffle and may seem so to anyone not interested in finding out how the process of digital crops works (and its fine to be not interested, but if you are read on.....).

I don't think anyone is disagreeing with the fact that olympus (and may be panasonic) given 20mp JPG with their respective digital crop feature.
What in fact happening is the camera is taking the readout of the full sensor, cropping to the middle 5mp (for a 2x crop) then upscaling that centre 5mp to give 20mp JPGs.

It is important to understand this as the camera is only really saving the middle 5mp of the sensor (for a 2x crop).
The final file size on disk or even pixels is mostly meaningless as you can upsample any number of ways and some methods are better than others.

Unless olympus/panasonic have some crazy amazing AI in camera I'd suggest just shooting in RAW, cropping in post and upscaling with something like gigapixel AI. I'd be really surprised if the incamera upscaling can come close let along match tools like gigapixel.

 

[rd6743]

This got me curious as I've played with digital TC on the OM1 and not sure if I set it up wrong etc as was trying to setup custom settings at the same time....

OM1 in the front room just now....digital TC for one then RAW opened in lightroom, exported as JPG, opened in photoshop, resized to 200% canvas then resized to 50%. Both were then canvas resized for the below.

 

[the black fox]

How it works is irrelevant ,the fact that it does and produces a full sized file is what matters . Raw is still a better quality image but in certain circumstances if applied correctly 2x digital produces a useable image .. the waffle that surrounds things like this when they are mentioned often by people that use other brands then stops olympus owners from actually trying ..
I have to admit I was on the side of it’s a gimmick myself till I took the time to try it out and found it works and is a usefull tool at times allowing shots that if shot in raw would need over cropping

 

[nandbytes]

The camera is just over cropping for you. It's not a magic bullet. Understanding how your tools work is not waffle, it helps people take better pictures.

I'm not stopping anyone one from trying anything in fact it's exactly the opposite.
Just like @rd6743 I suggest doing comparisons with Olympus/Panasonic upscaling versus something like topaz AI tools (trial versions are available).
Then use whichever people feel gives them the best result. All I'm suggesting is making an informed decision instead of simply using digital crop in camera without understanding how it works or it's limitations.

 

[Micro-four-thirds]

Thank you all for the help with my question mid post, and apologies to @Sangoma for interrupting the thread with my question.
I didn’t realise there was software that could boost the pixels of an image, this is something I’m going to look into as I have to admit to being one for pixel peeping an awful lot and I often get disgruntled when zooming in and seeing blurred details etc in certain photos mainly away from the main subject.

 

[Sangoma]

Thank you all for the help with my question mid post, and apologies to @Sangoma for interrupting the thread with my question.
I didn’t realise there was software that could boost the pixels of an image, this is something I’m going to look into as I have to admit to being one for pixel peeping an awful lot and I often get disgruntled when zooming in and seeing blurred details etc in certain photos mainly away from the main subject.

No apologies needed, relevant point.

As far as the software goes, I found on some images it was good, and on some it was not good, so worth getting a trial first.
I decided against buying it

 

[Micro-four-thirds]

No apologies needed, relevant point.

As far as the software goes, I found on some images it was good, and on some it was not good, so worth getting a trial first.
I decided against buying it

yes, I’ve been reading reviews this morning and some say it leaves a plasticky look. I’m definitely going to go for the trial first.

 

[nandbytes]

yes, I’ve been reading reviews this morning and some say it leaves a plasticky look. I’m definitely going to go for the trial first.

As a general rule of thumb, if your image has a lot of details the software (in camera or post processing) can work fairly well to upscale it.
if your subject is very small in the frame and/or your picture is lacking details in general it cannot save your image.

The way like to see it is, it cannot make a bad image into a good image but it can make a good image better.

 

[sk66]

james asked a question it was answered incorrectly

How it works is irrelevant ,the fact that it does and produces a full sized file is what matters . Raw is still a better quality image but in certain circumstances if applied correctly 2x digital produces a useable image .. the waffle that surrounds things like this when they are mentioned often by people that use other brands then stops olympus owners from actually trying ..
I have to admit I was on the side of it’s a gimmick myself till I took the time to try it out and found it works and is a usefull tool at times allowing shots that if shot in raw would need over cropping

Sometimes you can be quite the tool...

Actually James asked the question as to how it could crop and simultaneously RETAIN the full 20 MP; it was not answered incorrectly. It's what he didn't understand and wanted to know.

That’s the bit I was struggling to understand when I found the article mentioning it last week. I couldn’t see how the image could be cropped in camera and still retain 20mp in the same way a standard image would.

I never said the resulting file wasn't 20MP. And I did not say it was entirely pointless... I specifically mentioned a couple possible use benefits (post 17).

And I don't beat up on Olympus or any other brand/type. I currently own several small sensor cameras from Fuji, Panasonic, and Nikon; I have owned/used Olympus as well. I simply point out facts when asked for... Similarly I noted that you should stay with your 4/3 kit over going to APS when you asked.

 

[wiltw]

If you think about the matter, the LENS only can deliver just a certain amount of detail content onto a frame...let is assume that we have a truly superb camera lens which managed to record onto the 135 format frame 120 line-pairs of resolution PER MILLIMETER onto the finest grained film. For digital to capture that same detail requires (according to the Nyquist Limit) about 241 pixels per millimeter, or a 50 MegaPixel sensor. So let us, for discussion purposes,

1. assume we have such a high performance sensored camera that happens to offer a 'selectable area digital zoom' ...if we specify a 1mm x 1.5mm frame area, it will have 241 x 361 pixels in that digital zoom area, and 120 line-pairs per millimeter of detail within that area.
2. assume instead that we shoot the full frame with the same 50 MegaPixel camera, and in Lightroom we crop to the same 1mm x 1.5mm area. It still captured the same120 line-pairs of lens-delivered resolution -- ALL that the lens could deliver within that small area...no different than in #1 !

It does not matter if Lightroom might be able to make its JPG have 1000 x 1500 pixels, it still only has 120 line-pairs per millimeter of detail!!!

Folks always forget about the lens, and think ONLY about the sensor. A photo is really the PRODUCT of [lens * sensor] or [lens * film], which is why the 120 l-p/mm lens would not be able to achieve that same level of performance if we shot the test on Tri-X rather than on Pantomic-X !