-= 90D owners unite! Discuss and Post Photos

Eric K. wrote in post #18935366
John,

Your points are taken, but a lot of what you're talking about is too far into the weeds for me: the only purpose of my personal testing (and sharing with the forum) was a simple test of the camera's ISO range that I felt would be useful for those reviewing the thread who are wondering whether or not they should change or upgrade their camera.

The primary point of my post was to say that the ISO performance of the 90D was quite good from what I was able to find during these simple unimaginative tests, even up to the highest ISO settings.

I was not attempting to directly test the 90D with the 5D Mark IV. I only mentioned that I compared it with the 5D Mark IV because it's my other camera right now and I found it useful as a reference as I reviewed the output of the 90D. I wasn't trying to compare pixels to pixels but rather how do the overall photos look next to each other. I didn't want to get too lost in the details like upsampling or ensuring my perspectives were absolutely the same for each camera and shot...

IMO you did a wonderful job. Thanks for posting it.

Rod

RodS57 wrote in post #18935403
IMO you did a wonderful job. Thanks for posting it.

Rod

Thanks! As a “Just Lurking” member, it’s tough to know what to post.

Eric K. wrote in post #18935437
Thanks! As a “Just Lurking” member, it’s tough to know what to post.

Now you are "Mostly Lurking"!

Inspeqtor wrote in post #18935440
Now you are "Mostly Lurking"!

LOL!

I wonder what’s next: Peeping Tom?

markesc wrote in post #18935364
I'm waiting for Brian @ thedigital picture to give us the side by side results of every iso with those drop down menus so you can compare / contrast every conceivable iso and camera body.

That's useful for comparing the same sensor sizes, but it won't do what I was talking about, though, and I consider that a big problem. There's no question that unless your FF camera is 15 years old and your APS-C is current, that the FF camera's full images will have less noise at the same ISO. That has no guaranteed value, though. ISOs are not a direct response to ambient lighting; they are the result of photographic parameters which change between different sensor sizes.

We really need tools to compare cameras in equivalence and focal-length-limited, where a lot of photography happens, but is not tested by these sites that test a lot of cameras.

Pending still: "...I have no reservations about the image quality I'm seeing from this camera but will wait to critically evaluate image quality until we can lab test this model. Canon said I should expect no loss of image quality or increased noise levels from this imaging sensor relative to the previous 24 MP models."

https://www.the-digital-picture.com/Reviews/Ca​non-EOS-90D.aspx

Really curious how it looks compared to the 80d and Nikon D500 (that 500pf is looking tasty).

In the DPR studio comparison tool, the 90D is cleaner at high ISOs than any other Canon APS-C (and has less visible noise with the same total exposure times area than the Canon FF cameras) and just slightly noisier than the full D500 frame, but that frame is a little bigger, so the focal-length-limited noise for cropped subjects is probably close enough with the 90D that it isn't worth getting the much lower pixel density of the D500, IQ-wise. Obviously, the 90D can't compete with the D500 for AF, but a 7D3 with the same sensor as the 90D might, if it is ever released.

John Sheehy wrote in post #18935501
That's useful for comparing the same sensor sizes, but it won't do what I was talking about, though, and I consider that a big problem. There's no question that unless your FF camera is 15 years old and your APS-C is current, that the FF camera's full images will have less noise at the same ISO. That has no guaranteed value, though. ISOs are not a direct response to ambient lighting; they are the result of photographic parameters which change between different sensor sizes.

We really need tools to compare cameras in equivalence and focal-length-limited, where a lot of photography happens, but is not tested by these sites that test a lot of cameras.

In the DPR studio comparison tool, the 90D is cleaner at high ISOs than any other Canon APS-C (and has less visible noise with the same total exposure times area than the Canon FF cameras) and just slightly noisier than the full D500 frame, but that frame is a little bigger, so the focal-length-limited noise for cropped subjects is probably close enough with the 90D that it isn't worth getting the much lower pixel density of the D500, IQ-wise. Obviously, the 90D can't compete with the D500 for AF, but a 7D3 with the same sensor as the 90D might, if it is ever released.

You may wish to look at the definition of ISO- no mention of sensor size is mentioned nor other parameters.
https://www.iso.org …:std:iso:12232:​ed-3:v1:en

That's like saying 110 instamatic film at ISO 100 and 35 mm ISO 100 film are different ISO.

Capn Jack wrote in post #18935519
You may wish to look at the definition of ISO- no mention of sensor size is mentioned nor other parameters.
https://www.iso.org …:std:iso:12232:​ed-3:v1:en

That's like saying 110 instamatic film at ISO 100 and 35 mm ISO 100 film are different ISO.

I don't think you understand what I wrote. What I am saying is that comparing full images of different sensor sizes at the same ISO is pretty pointless, except as an academic exercise to investigate the obvious-but-irrelevant. A Nikon D500 has very similar noise (as measured by DxO) to a Canon 5D (original), but that's almost 15 years difference in QE and read noise technology. For cameras of recent vintage, saying that a FF camera is cleaner at ISO 6400 than APS-C is a rhetorical fact with no practical implications to rational photography.

I don't look at the lighting in a place, look at my desired composition, and think, "I would use ISO 6400 here, whether I use APS-C or FF". There is no reason to assume that equal ISOs means equal photographs with different sensor size, or that the same ISOs would ever be chosen for a good reason with different sensor sizes. Someone just pulled this "same ISO means equal" stuff out of a hat, and it gets apparent support just because it accidentally makes sense when all sensor sizes are the same. ISO is not a lighting challenge. ISO is just an inverse to exposure for a middle grey, and exposure varies with optics which are different to get the same photo with different sensor sizes.

The whole idea of comparing APS-C to FF at the same high ISO is of imaginary universal significance. The only time it applies practically is when there is shallower DOF in a composition on a larger sensor, and then it is just as much or even more a case of the shallower DOF being the real cause of increased SNR. If there isn't shallower DOF, it means absolutely doodly-squat, and the lower noise does not occur. No shallower DOF, no increased total light, with a given shutter speed.

"The same ISO" may have no practical photographic "sameness" at all, as is the case with required DOF photography, and focal-length-limited photography, where less of the sensor is used, the larger the sensor.

That's why I keep questioning why we keep talking about FF sensors being better at high ISOs as if it had some kind of wide-reaching practical meaning. They are better for an academic, arbitrary image at the same ISO. Not necessarily so for a real, needed photo with real, available glass.

There is "the same high ISO", and then there is also "the same need for a higher ISO than base due to needed shutter speed and limited aperture"; the latter results in the same PHOTO, but not necessarily the same ISO.

I have no idea what you said; but you said it very well!

John Sheehy wrote in post #18935546
I don't think you understand what I wrote.

Trying to follow along but I'm getting lost as well.

Are you saying that

- The same subject
- Captured with the same lens
- At the same distance
- In the same lighting conditions
- With the same iso/aperture/shutter speed settings
- Framed to fill the lens + APS-C field of view

using FF and APS-C bodies in a side-by-side comparison have roughly equivalent ISO performance because the FF image needs to be cropped to match the composition of the APS-C image?

John Sheehy wrote in post #18935546
That's why I keep questioning why we keep talking about FF sensors being better at high ISOs as if it had some kind of wide-reaching practical meaning.

Hi John,

While much of your post was quite technical and some of it was indecipherable to me as I don’t know all the acronyms, I think I get the gist of your point.

I agree that it might seem a rather pointless academic exercise comparing the same ISO performance of different cameras side-by-side and that this type of pseudo “apples to apples” type of comparison would run into the technical challenges and inconsistencies you outlined.

However, I wonder if most of us who are using ISO charts do so to get a feel for how their camera performs.

And for this second reason, I do think the comparisons can have relevance as a point of reference for the photographer using their equipment and gauging their (likely) results in different scenarios in the real world.

For example, when I’m out in the field and after I review my camera settings and the situation requires me to bump up my ISO to get my preferred shutter speed or aperture opening for the scene I’m trying to capture on my sensor - it’s useful to know what the result of increasing ISO is likely going to be (and the review of charts now let’s me think of this performance ahead of time in comparison to another camera I’m familiar with). I’ll understand where I start to get reluctant to increase the camera ISO on a FF camera and compare that to what I’ve seen comparing results from my new AP-S sensor camera.

I’ve never used a Nikon or Fuji camera, so looking at charts of their ISO performance is something that would serve no purpose to me - however, comparing a new camera with Canon cameras I’ve used before (regardless of frame size) provides a useful reference for my use of the new camera and how my ISO selections are going to impact my final results.

Not trying to flagellate a deceased equine...

John Sheehy wrote in post #18935546
I don't think you understand what I wrote.

I believe I did understand, and the statement is still incorrect:

John Sheehy wrote in post #18935501
ISOs are not a direct response to ambient lighting; they are the result of photographic parameters which change between different sensor sizes.

That isn't a correct statement as ISO is defined.

John Sheehy wrote in post #18935546
What I am saying is that comparing full images of different sensor sizes at the same ISO is pretty pointless, except as an academic exercise to investigate the obvious-but-irrelevant. A Nikon D500 has very similar noise (as measured by DxO) to a Canon 5D (original), but that's almost 15 years difference in QE and read noise technology. For cameras of recent vintage, saying that a FF camera is cleaner at ISO 6400 than APS-C is a rhetorical fact with no practical implications to rational photography.

"Read noise" technology? I think you mean also analog-digital conversion as well, which is separate from "read noise". Back then, a photodiode would be used rather than a CMOS today.

John Sheehy wrote in post #18935546
I don't look at the lighting in a place, look at my desired composition, and think, "I would use ISO 6400 here, whether I use APS-C or FF". There is no reason to assume that equal ISOs means equal photographs with different sensor size, or that the same ISOs would ever be chosen for a good reason with different sensor sizes. Someone just pulled this "same ISO means equal" stuff out of a hat, and it gets apparent support just because it accidentally makes sense when all sensor sizes are the same. ISO is not a lighting challenge. ISO is just an inverse to exposure for a middle grey, and exposure varies with optics which are different to get the same photo with different sensor sizes.

One needs to start somewhere. ISO gets us into the ball-park under a given set of lighting conditions, as it did with film of differing sizes. You seem to be harping on sensor size, and that really isn't important. Pixel size (light gathering area, pitch) is more important, as is pixel signal-to-noise, and pixel-dynamic range.

John Sheehy wrote in post #18935546
The whole idea of comparing APS-C to FF at the same high ISO is of imaginary universal significance. The only time it applies practically is when there is shallower DOF in a composition on a larger sensor, and then it is just as much or even more a case of the shallower DOF being the real cause of increased SNR. If there isn't shallower DOF, it means absolutely doodly-squat, and the lower noise does not occur. No shallower DOF, no increased total light, with a given shutter speed.

I'm assuming "SNR" = signal-to-noise ratio. As we are discussing light and sensitivity, I'm not understanding how depth-of-field (DOF) plays into this.

John Sheehy wrote in post #18935546
"The same ISO" may have no practical photographic "sameness" at all, as is the case with required DOF photography, and focal-length-limited photography, where less of the sensor is used, the larger the sensor.

True of film, too. For example, compare 35 mm and medium-format cameras.

John Sheehy wrote in post #18935546
That's why I keep questioning why we keep talking about FF sensors being better at high ISOs as if it had some kind of wide-reaching practical meaning. They are better for an academic, arbitrary image at the same ISO. Not necessarily so for a real, needed photo with real, available glass.

ISO is a starting point of sensitivity that allows one to determine if they can take a photo under a certain set of conditions. Again, full-frame doesn't matter. Pixel-pitch, pixel-signal-to-noise, and pixel-dynamic range are more important. Take a 7D2 sensor, build it so that it is a full-frame (more pixels of the same size as the 7D2- I don't fell like doing the math to determine the # pixels for such a sensor) and you'll find the ISO behavior will be the same between them.

John Sheehy wrote in post #18935546
There is "the same high ISO", and then there is also "the same need for a higher ISO than base due to needed shutter speed and limited aperture"; the latter results in the same PHOTO, but not necessarily the same ISO.

Look again at the definition of ISO. The ISO is the same, but the quality of the image may not be.

I have a 7DII, and I have it because I like to photograph birds. I use it for everything else (landscapes, Milky Way, the Moon, macro, grandkids, etc.) and it does all of those to my satisfaction. I am sure that the 90D would do them as well, or better. But after reading this thread, I am not so sure about BIF. I am especially concerned about the observation that the AF while tracking a bird is better (excellent, even) in live view, but not with the OVF. This would seem to be a deal breaker, as I am not going to hang a 150-600 on a 90D and try to track birds at arm's length (or any length) in live view. I handhold my bird shots, as I am tromping around fields and woods, and am not going to be toting around a tripod and gimbal.

In the last few pages of this thread a couple of people have asked about the tracking AF performance in live view vs. OVF, but there were no responses. Is there consensus on the relatively poor performance with the OVF? Is it worse than the 7DII?

This thread has taken a turn toward discussions of noise and other interesting topics, but that is all irrelevant if the images are not in focus, and if BIF are not going to be in focus using the OVF, this may not be a camera for me.

pknight wrote in post #18935798
I have a 7DII, and I have it because I like to photograph birds. I use it for everything else (landscapes, Milky Way, the Moon, macro, grandkids, etc.) and it does all of those to my satisfaction. I am sure that the 90D would do them as well, or better. But after reading this thread, I am not so sure about BIF. I am especially concerned about the observation that the AF while tracking a bird is better (excellent, even) in live view, but not with the OVF. This would seem to be a deal breaker, as I am not going to hang a 150-600 on a 90D and try to track birds at arm's length (or any length) in live view. I handhold my bird shots, as I am tromping around fields and woods, and am not going to be toting around a tripod and gimbal.

In the last few pages of this thread a couple of people have asked about the tracking AF performance in live view vs. OVF, but there were no responses. Is there consensus on the relatively poor performance with the OVF? Is it worse than the 7DII?

This thread has taken a turn toward discussions of noise and other interesting topics, but that is all irrelevant if the images are not in focus, and if BIF are not going to be in focus using the OVF, this may not be a camera for me.

That's a good summary of what is known about the AF capabilities of the 90D for now. Comparing tracking ability is so hard to do because every BIF is different. So one goes by what folks experience and their feelings.

I too like bird photography and do BIF, but decided to buy the 90D despite the reports about tracking. This is because the 90D seems to be stellar for macro (big interest for me too) and great for static birds... and because BIF shots, while of interest, are a small part of my photography. With 10 frames per second, I might get some good shots anyway.

I did buy the 90D last Thursday and it has been snowing here since then so no opportunity to do any testing of the new body outdoors. I will do some checking when the weather improves, which should be soon.

pknight wrote in post #18935798
I have a 7DII, and I have it because I like to photograph birds. I use it for everything else (landscapes, Milky Way, the Moon, macro, grandkids, etc.) and it does all of those to my satisfaction. I am sure that the 90D would do them as well, or better. But after reading this thread, I am not so sure about BIF. I am especially concerned about the observation that the AF while tracking a bird is better (excellent, even) in live view, but not with the OVF. This would seem to be a deal breaker, as I am not going to hang a 150-600 on a 90D and try to track birds at arm's length (or any length) in live view. I handhold my bird shots, as I am tromping around fields and woods, and am not going to be toting around a tripod and gimbal.

In the last few pages of this thread a couple of people have asked about the tracking AF performance in live view vs. OVF, but there were no responses. Is there consensus on the relatively poor performance with the OVF? Is it worse than the 7DII?

This thread has taken a turn toward discussions of noise and other interesting topics, but that is all irrelevant if the images are not in focus, and if BIF are not going to be in focus using the OVF, this may not be a camera for me.

I went from a 7DII to the 90D, and while I don't do much BIF shots, I do spend a lot of time at airshows so the OVF tracking is important to me. However, as it's not airshow season yet here in AZ I haven't had much of a chance to test it out in that regard. Like you, I used the 7DII for everything else as well but mainly wildlife and macro shooting with the occasional landscape or portrait making its way in. For both wildlife and macro I've been extremely impressed and very happy with the move from the 7DII. There were a few compromises going to the 90D, mainly in the customization options as the 7DII had more abilities to assign custom functions to buttons and of course had a third C option on the mode dial. The AF system on the 7DII also had more options to work with compared to the 90D, but I honestly rarely ever touched those on the 7DII. I also liked the extra size of the 7DII as well, but that's an easy thing look past as the 90D does fit in my hand very well.

One thing I have not had a chance to try on the 90D is the built-in focus stacking. I'm very interested in that, just haven't had the right opportunity to try it out as of yet.

If BIF are your main focus, I can understand the hesitation and hopefully some more answers about the OVF tracking will come out possibly with some more in-depth third-part reviews as well. I know I've been keeping an eye out for detailed reviews, but as of yet there aren't very many other than the ones that have been published outside of the ones from the Canon intro events.

John Sheehy wrote in post #18935342
It doesn't even make sense to compare FF and APS-C, IMO, without being very specific about optics and usage.

How often would one actually shoot both cameras at the same ISO, if they were not cropping the FF? The subject distance and/or the lens would likely be different. If they did crop the FF, it would lose its alleged IQ advantage. The 5D4 has more visible noise at ISO 25600 when it is cropped to 1.6x than the D90. The D90 provides a superior "film" or "emulsion" in a smaller format. If you want to compare the 5D4 to the 90D, you really should upsample the 5D4 to 167% to compare to the 90D at 100% pixel view.

You have to get closer, and change perspective, if possible, with the same lens, or get a bigger lens 1.6x as long and with 1.6x the entrance pupil diameter, to get the alleged FF benefit above base ISO.

.
I make such comparisons, and it makes sense to me.

I am not "reach challenged". . When I shoot wildlife or birds with a full frame, I do not crop into the image to get the composition I want. . I frame the images the same way whether I am using a full frame, a 1.3 crop, or a 1.6 crop. . If I am shooting full frame then I just zoom in further with my lens or I just get closer.

For someone who shoots wildlife the way I do, it makes sense to compare full frame results with crop sensor results. . Comparing downsampled FF files or cropped FF files would not make sense, because that is not what I am doing with my FF photos.

If, due to more modern sensor tech, Canon makes a 1.6 crop body that is cleaner than their older FF bodies, then that is something I would want to take advantage of. . This is why I am interested in seeing how uncropped full resolution 90D files compare with uncropped full resolution FF files.

If a 2019 1.6 crop with 34 megapixels gives better, cleaner results than a 2012 full frame with 20 megapixels, then that is something that would be of great interest to me. . If such were the case, then I would consider getting a 90D as a backup to replace my 6D.

If a 5D Mark 4 gives cleaner images than a 90D (both at the same ISO and at full resolution and uncropped), then when I have money to upgrade I would buy a 5D4 instead of a 90D. . If a 90D gives better, cleaner files than the 5D Mark 4, then I would buy the 90D.

.