Let's discuss Quality of Camera's abilities, and find out the truth of today's technology
May 15, 2019 03:39:55 #
So, in a previous Post, I had some very under-exposed RAW files, that could be helped greatly in Lightroom. I did it with curiosity about the quality of the whole system, including the sensor range and depth.
So is your camera better or worse than an Olympus OM-D M5 II $899 body?
And after I finish this, I'm am going to get some slightly highlight-blown pictures, on purpose. NO WHITE PICTURES!!!! DO NOT OVER-EXPOSE TOO MUCH! THE SENSOR PIXELS ARE SENSITIVE. ONLY TEST CAMERAS UNDER WARRANTEE!!! I TAKE ZERO RESPONSIBILITY FOR YOUR CAMERA!
As an electronic reliability engineer, I know that this shouldn't hurt any camera. There should be plenty of designed-in and built-in margin. I have done things like make a defibrillator preform 10,000+ shocks over a few weeks, incrementally increase (and go in the negative direction) operating temperature, voltage, vibration, under test, (on a sample until destruction), to find the weakest part of the unit under test...
WE ARE NOT TESTING ANYTHING UNTIL DESTRUCTION HERE!!! LET ME MAKE THAT CLEAR! No White photos! No pointing cameras at a light source! Just a little extra highlights is all we want!
"Overall in general - no, it not better to underexpose in digital photograph. However, cameras like yours are adjusting this general exposure thinking where it can be immaterial assuming the photographer is shooting in RAW and planning to post process their images. It's getting harder to generalize given the highlight recovery options of RAW files and the noise performance of ISO invariant models.
Rather than testing underexposure, you should be testing where the upper limit exists for over-exposure, as in: how high is too high to recover highlight details. Take the image of the fresh chicken. If this image had been exposed so the highlights were blinking on the packages, how would that image recover in your processing in LR, moving the overall exposure to the left and / or lowering the highlight slider? You may not see it on your monitor, but this example is still rather dark, the colors are unsaturated with a slight overall yellow tint and the whites of the tags not very white. The complexity of your edit work will benefit from images where the 'white' of the image is closer to white in the file. But as said above, your camera has more flexibility in post than other camera models and brands." - by CHG_CANON (Thanks).
We'll do under-exposure first, then a little over-exposure? Is anyone interested?
Please note, this is at your own risk. Please use cameras under warrantee, if you are going to even slightly over-expose the sensor. NO WHITE PHOTOS PLEASE!!!! Pixels are analog, sensitive, and might not have enough margin? I hope they are properly designed. Never point your camera directly at a light source, including all lights and never, ever at the sun. I take no responsibility for anyone's camera. The over-exposure doesn't have to be a lot in order to detect the quality of the file to recover details in the highlights. No white photos, please!!!!
So is your camera better or worse than an Olympus OM-D M5 II $899 body?
And after I finish this, I'm am going to get some slightly highlight-blown pictures, on purpose. NO WHITE PICTURES!!!! DO NOT OVER-EXPOSE TOO MUCH! THE SENSOR PIXELS ARE SENSITIVE. ONLY TEST CAMERAS UNDER WARRANTEE!!! I TAKE ZERO RESPONSIBILITY FOR YOUR CAMERA!
As an electronic reliability engineer, I know that this shouldn't hurt any camera. There should be plenty of designed-in and built-in margin. I have done things like make a defibrillator preform 10,000+ shocks over a few weeks, incrementally increase (and go in the negative direction) operating temperature, voltage, vibration, under test, (on a sample until destruction), to find the weakest part of the unit under test...
WE ARE NOT TESTING ANYTHING UNTIL DESTRUCTION HERE!!! LET ME MAKE THAT CLEAR! No White photos! No pointing cameras at a light source! Just a little extra highlights is all we want!
"Overall in general - no, it not better to underexpose in digital photograph. However, cameras like yours are adjusting this general exposure thinking where it can be immaterial assuming the photographer is shooting in RAW and planning to post process their images. It's getting harder to generalize given the highlight recovery options of RAW files and the noise performance of ISO invariant models.
Rather than testing underexposure, you should be testing where the upper limit exists for over-exposure, as in: how high is too high to recover highlight details. Take the image of the fresh chicken. If this image had been exposed so the highlights were blinking on the packages, how would that image recover in your processing in LR, moving the overall exposure to the left and / or lowering the highlight slider? You may not see it on your monitor, but this example is still rather dark, the colors are unsaturated with a slight overall yellow tint and the whites of the tags not very white. The complexity of your edit work will benefit from images where the 'white' of the image is closer to white in the file. But as said above, your camera has more flexibility in post than other camera models and brands." - by CHG_CANON (Thanks).
We'll do under-exposure first, then a little over-exposure? Is anyone interested?
Please note, this is at your own risk. Please use cameras under warrantee, if you are going to even slightly over-expose the sensor. NO WHITE PHOTOS PLEASE!!!! Pixels are analog, sensitive, and might not have enough margin? I hope they are properly designed. Never point your camera directly at a light source, including all lights and never, ever at the sun. I take no responsibility for anyone's camera. The over-exposure doesn't have to be a lot in order to detect the quality of the file to recover details in the highlights. No white photos, please!!!!
May 15, 2019 07:24:59 #
May 15, 2019 08:39:00 #
May 15, 2019 11:59:21 #
IDguy
Loc: Idaho
bleirer wrote:
A lot of that kind of testing has been done. Here for example: http://www.photonstophotos.net/
Great link!
Here is what interests me most for the three cameras I own and the D800 I last sent away. Trend is as expected: dynamic range follows sensor size. The Z6 is better than D800 because of larger pixels.
Now I know why I prefer D5600 images over GM5. But GM5 wins on size and weight.
May 15, 2019 21:19:38 #
Ah, forget this whole post. I'm going to concentrate on properly exposed photos, and enjoy my new camera. Have a nice day.
May 15, 2019 22:01:40 #
Sorry, I didn't intend to rain on the parade, I just thought it was interesting to see how other testers report their results.
May 15, 2019 22:32:18 #
bleirer wrote:
Sorry, I didn't intend to rain on the parade, I just thought it was interesting to see how other testers report their results.
You didn't rain on any parade. It's fine. I'm just going to enjoy taking the best photos I can in MANUAL. I'm an enthusiastic beginner in photography, and an electronic engineer. I just got an Olympus OM-D E-M5 II about 9 days ago. I'm really enjoying my new camera. Thanks for posting/commenting.
May 15, 2019 22:40:03 #
So as an engineer, what is your take on sites like DXOMark? Are they a useful way to decide on a camera?
May 15, 2019 23:04:38 #
bleirer wrote:
So as an engineer, what is your take on sites like DXOMark? Are they a useful way to decide on a camera?
Very interesting. I will have to research this. I have worked in defense, and hard drives, and integrated circuit companies, not optical or cameras. So I will check this out.
May 15, 2019 23:06:01 #
IDguy wrote:
Great link!
Here is what interests me most for the three cameras I own and the D800 I last sent away. Trend is as expected: dynamic range follows sensor size. The Z6 is better than D800 because of larger pixels.
Now I know why I prefer D5600 images over GM5. But GM5 wins on size and weight.
Here is what interests me most for the three cameras I own and the D800 I last sent away. Trend is as expected: dynamic range follows sensor size. The Z6 is better than D800 because of larger pixels.
Now I know why I prefer D5600 images over GM5. But GM5 wins on size and weight.
Thank you very much for this information. Have a great day.
May 16, 2019 07:44:58 #
May 16, 2019 09:00:47 #
catchlight..
Loc: Wisconsin USA- Halden Norway
Whats the point?
ETTR is exposing to the extreme right of the histogram, but not crossing the fine line of over exposure for the end result. This is considered a skill I think.
In reality, multi exposure is a much more practical approach, with more usable information captured.
Digital camera sensors tend to score their best, and make the most advancement in shadow detail and data information recovered from the dark side of the histogram.
As over exposure grows, it consumes information. The further you push it, the more you loose.
ETTR is exposing to the extreme right of the histogram, but not crossing the fine line of over exposure for the end result. This is considered a skill I think.
In reality, multi exposure is a much more practical approach, with more usable information captured.
Digital camera sensors tend to score their best, and make the most advancement in shadow detail and data information recovered from the dark side of the histogram.
As over exposure grows, it consumes information. The further you push it, the more you loose.
May 16, 2019 09:18:52 #
catchlight.. wrote:
Whats the point? br br ETTR is exposing to the ex... (show quote)
I thought the original reason for ettr was because the bit level of capture increased to the right, so one could capture a shadow at higher bit level and dial it back in post with more detail. But that was for 12 bit sensors. Now with 14 bit sensors, it is not as much an issue.
May 16, 2019 09:19:45 #
bleirer wrote:
I thought the original reason for ettr was because the bit level of capture increased to the right, so one could capture a shadow at higher bit level and dial it back in post with more detail. But that was for 12 bit sensors. Now with 14 bit sensors, it is not as much an issue.
More data is always better that less data ...
May 16, 2019 09:34:56 #
You can calibrate or test your camera’s sensor to determine how much you can overexpose a shot and not blow highlights. I believe my body was about 2 to 2 1/3 stops using the Sekonic calibration software.
If you want to reply, then register here. Registration is free and your account is created instantly, so you can post right away.