Sony a73 vs Sony a7R3 - Part 3: Dynamic Range | 4K

0:00
Hi I'm Dustin Abbott and I'm here today to give you a look at the how the the sensor of the new
0:12
Sony a7 III performs when pushed towards the extremes in the dynamic range. Now in previous
0:21
comparison series I've compared cameras like the a7r mark III with the Canon 5d mark IV the Nikon
0:29
D850, and out of all of those cameras, I found that to this point, the winner when it comes to
0:37
dynamic range has been somewhat of a toss-up between the D850 and the A7R Mark III with a
0:45
fairly similar performance. Today, I'm going to be comparing the A7R Mark III with the new A7 III
0:52
to see if perhaps we have a new champ as we did when it came to high ISO performance
1:00
And if you missed that episode, check it out here. In that episode, we found that the A7 III dethroned the A7R Mark III at the very ends of the limit
1:12
And so today we're going to look at the same thing when it comes to the dynamic range
1:16
And so let's jump in and let's see how these cameras shake out and see if we have a new winner
1:22
Okay, we're going to look at the dynamic range starting at the base exposure. This just gives
1:26
you a baseline for looking at how everything should look, and so obviously equal exposure
1:32
value as we climb up through the dynamic range here. And so this just gives you a look at
1:39
you know, everything. Obviously there's a huge difference in the resolution of the two cameras
1:43
and so just note that difference. Also don't read too much into any kind of differences in
1:49
and sharpness in that they may not be focused identically, and so as a byproduct of that
1:57
this is about the dynamic range. So let's see what happens when we move one stop
2:02
So here on the A7 III, we have got an overexposure by one stop
2:08
and then in post, I have reduced that back down one stop
2:13
And so let's see how it plays out here. You can see, obviously, it looks too bright
2:18
looks overexposed and so on the the left side but we can see on the right side that we were able to
2:24
pull back that you know fairly easy one stop of exposure without any issues at all and everything
2:30
looks nice and clean there's no hot spots and so byproduct there no problem now in this example
2:38
we have overexposed by two stops and then in post have pulled it back down and so looking at the
2:43
image globally, you can see obviously still no problem. I mean, the image looks just fine
2:49
If we look at it at a pixel level, we can see that while everything was quite blown out here
2:55
as we look back here, it doesn't appear to me that we have any hotspots, any issues in anything
3:01
here. Everything here looks pretty sound to me still at this point. Now at three stops, obviously
3:08
our original image is looking pretty rough at this point. It's very clearly overexposed
3:14
And on the right side, we can see that after pulling it back down in post, it once again
3:19
on a global level, it looks quite clean and no problem there. If we take a look here a little
3:26
bit closer, I think we're going to find that, for example, in this skull here, there is some detail
3:31
that has been lost and also potentially on this front fascia of the camera that has collected
3:39
a lot of light. There's just some texture information that is starting to be lost here with a three-stop overexposure
3:47
Over on this lens, you'll also find probably a few hotspots that have not fully recovered
3:52
as you can see here. But for the most part, unless you're examining it on a very, very detailed looking at it
4:01
at a pixel level, you probably could get by with this and people wouldn't necessarily notice
4:07
Here at a four-stop overexposure, however, is where we see we have lost it
4:12
And you can see image looks very blown out in the original, but you can also see looking at the recovery here
4:18
where I've reduced it by four stops that we have lost a lot of color information
4:23
The hotspots have very clearly, texture has been lost. and there's a practical limit to how much can be recovered from the camera here
4:36
And so that shows you your upper limit. We'll compare that to the a7R Mark III
4:42
So on the left, you have the four-stop recovered image from the a7R
4:46
On the right, you have from the a7 III. And so looking at them at a pixel level, I would say that it would be hard for me to be able to distinguish a difference between the two of them
5:03
They look roughly equal, and I'm not necessarily seeing anything that's been recovered on one that has been lost on the other
5:12
And so somewhere between that three and four stops, you'll see the practical limits of recovery on both of these cameras
5:21
And so around three stops, you can get away with. Somewhere between three and four, you lose it
5:27
But all told this is a strong performance It is about a stop better than what we have seen on the high end from previous cameras and models and things like the 5D Mark IV Now if we start moving the opposite
5:39
direction, here we have an underexposure by one stop and then the recovery in post. And so in this
5:47
we're going to be looking to make sure that we don't start to introduce a lot of noise into the
5:53
image. And so as we see as we recover here, we can look at text and color. Everything looks still
5:59
nice and clean. No issues that I can report here. And that's to be expected. We see a strong
6:07
performance here in our out of focus area. There's no weird color banding that's been introduced from
6:13
the recovery. Everything looks as it should. Now with a two-stop underexposure here, we see that
6:20
the original image is starting to look quite dark and kind of a little bit murky in terms of the
6:25
colors. Here we see a much more pleasing image in the recovery. Everything looks nice and crisp
6:32
And if we look here at the details, I'm not really seeing any additional noise that's introduced
6:39
here. Everything still continues to look nice and clean in our out of focus area. No odd color
6:46
banding that's introduced everything looks quite nice there now at a three-stop under exposure you
6:51
can see that the original image is you know it looks in some ways like it could not be redeemed
6:58
but as we look on the right it surprisingly can be redeemed quite well and on a global level you
7:05
would have no issues with that recovery there and that's certainly a very valuable thing in terms of
7:11
the introduction of noise or color banding. I'm looking, of course, on a 4K monitor. I am
7:18
you know, looking at the original images and I still see a very, very clean result. Looking at
7:25
it as a whole, there's no place where I'm seeing any kind of color banding that's been introduced
7:30
and the noise levels as well are certainly in a very, very acceptable zone. Now at a four-stop
7:37
under exposure here you can see that the original is you know almost disappearing
7:42
here and once again the recovery looks much much more convincing than it did at
7:47
the four stops of over exposure which you know goes to illustrate a fairly
7:52
consistent point with cameras in that if you're going to miss exposure it's
7:57
typically best to under expose than it is to over expose and that you have more
8:02
latitude with an underexposed image than you do with an overexposed and so
8:07
So in this case, we can see even at four stops, there is just the faintest bit of banding
8:13
that's been introduced here in the shadow area. Probably a little additional noise that's showing up
8:20
but the 24 megapixels of the a7 III are fairly forgiving of that
8:26
in that it doesn't show up in a highly noticeable way. I can see that it's there now
8:32
but once again, looking at the image on a global level, could probably get away with this under a lot of different applications. Now, if we compare to the
8:41
four-stop recovered image from the a7R3, I think that what we're going to find is that there is
8:48
definitely a little bit more banding noise that has been introduced. And of course, once again
8:53
we are looking at this at a much higher resolution level. And so even so on our actual subject
9:01
everything looks quite crisp and clean here still. I do think that there maybe is a little
9:07
bit more kind of brightness that has been recovered on the a7 III and so once again it is showing its
9:13
merit here and definitely looking at the out of focus area it does seem like the a7R III has a
9:19
little bit more tendency to go green. There's green noise banding that's been introduced into
9:25
the image, whereas the a7 III is delivering a more neutral result, I would say. And so as a
9:32
byproduct of that, I do think that when it comes to shadow recovery that I'm giving an edge to
9:38
the a7 III in this comparison over the a7R III. Now, if you take this out into the real world
9:45
what it means is that it gives you some latitude when it comes to shooting a variety of subjects
9:50
In this case, I've chosen a landscape subject. And so this is set at what the camera deemed to be the proper metering here
10:00
And so 1 400th of a second, I'm using the Loudwa 15mm F2.0D lens for the shot, a nice landscape option
10:10
And so here the camera has metered and says this is the proper metering
10:14
On the right side, and what I'll do is I'm going to jump in here to the develop module
10:21
and I'll just show you what I've done. I've basically just shown you how much you have to play with just using the sliders here
10:28
We've gone, pulled the highlights completely down. We have pulled the shadows completely up, and I've done nothing else
10:34
Of course, you could play with the image more in terms of localized areas
10:38
You could play with the exposure a little bit, but I'm mostly wanting to just demonstrate to you
10:43
how much latitude you have to play with when it comes to recovering an image here
10:48
So coming back to the images to compare them, we have the original on the left
10:53
we have the one that I've played with a bit on the right. And so we can see a few things here
10:57
Number one, we can see that we have a more consistent sky here on the right side and in that we had kind of a blown out area and we had also lost maybe some of the proper color tone to the sky We can see that that has been recovered here and so
11:12
we can also see more detail around the clouds and what information was there and that to me is a huge
11:17
thing. If there are if there's features in the sky I want to be able to pull those back in landscape
11:22
images. We can also see if we look at the right compared to the left that we have recovered some
11:27
of the basic color of the fresh leaves that are there, spring leaves, and also a little more
11:33
definition around their edges. And so that's a positive there. Now, if we look towards the area
11:39
that was shadowed, and I intentionally chose a scene that had some kind of blacked out lost areas
11:46
we can see that we have also recovered quite a lot of information when it comes to like the
11:50
textures on the structure here. If we look inside this here, you can see on the original information
11:56
was basically lost and it's recovered there. And I'm not able to see any additional noise
12:02
that was introduced. Here at the edge of the frame, we can see that there's crushed blacks here
12:07
on this evergreen tree here. And we can see that we can recover the detail there
12:13
without any kind of issue. And even things so simple as in branches
12:18
just bringing back some information. Now, looking at the image as a whole
12:22
I mean, in this case, you may prefer the recovered image. just note that when you have cameras that have as good of dynamic range as what these cameras do
12:31
that running everything to the extremes, while it may recover the most information
12:35
it won't necessarily produce your best image. Remember that things like shadows, and in some cases even blown out areas
12:43
are part of the storytelling of photography. But at the same time, I would prefer to have the latitude to play with than to not have it
12:51
Now what I've done here is I've just used the actual EV dial on the camera itself
13:00
And so, you know, twisted it all the way in this case to the plus three. And so this is a three-stop overexposure on this scene
13:07
And then on the right, I have brought that back down, brought that back down three stops
13:12
And then what I've also done in the process of bringing it back down is try to keep those shadows open as well
13:18
So in terms of the shadow information, of course, obviously there was lots of shadow information there before
13:24
and I could clearly recover that even further if I played with it a little bit more
13:28
But here I've kind of gone to try to recover a more, you know, kind of balance scene there
13:33
Now, looking at the sky itself, we see that while we have recovered sky, you know, the color isn't quite right there
13:43
At the same time, we also see that there are some of these kind of leaves that have caught a lot of light that we haven't fully recovered
13:52
At the same time, we have also recovered a whole lot more information, a lot of the color that's back here
13:59
And if you look at the image at large, I think that while it's technically imperfect, it is credible
14:06
It's usable for some applications, whereas our original completely blown out is not usable
14:13
Now in this image, I did the opposite. I twisted the dial to the three stops of underexposure. And
14:19
so as you can see, the original here is essentially unusable. It's very dark. But as we saw in our
14:26
kind of laboratory test, there is a lot of latitude to play with the A7III's image when
14:31
it comes to shadow recovery. And so in this case, I was able to pull the exposures back and even dial
14:37
in a little bit of shadow recovery. And as we can see, areas that were completely lost before have
14:43
been recovered without really a lot of introduction of noise. Now, I think that if you could go in
14:49
further than this, you would be able to see that there is a little bit of noise that's there on
14:53
those board textures. But the reality is, of course, that we have recovered these very successfully
14:59
without a whole lot of penalty. And over on this side, whereas this is completely crushed, we can
15:07
see that those details have re-emerged. So the plus for this is if you're dealing with an image
15:13
or for example, a lens that has heavy vignette, you know, hopefully you're not making a habit of
15:19
underexposing your whole images like this, but let's just say you've got an image or a lens that
15:24
you know, vignettes heavily, as we have an example of here in the corner. If you have a camera that
15:28
doesn't have great dynamic range, even just using a profile to recover that means you're going to
15:34
introduce noise to the image. What we have seen here is that we have been able to recover this
15:38
at a very, very low noise penalty. The byproduct of this is it just adds to the flexibility of what
15:45
you're able to accomplish. And so certainly on the low end of the range, a lot of pluses there
15:51
And so once again, if you're going to miss a shot or purposely go one way or the other
15:58
it's typically better to purposely underexpose a bit to make sure that you have good sky
16:04
information and then you have a lot of latitude to play with shadows with the a7iii. One final
16:10
real world example to show you is in this portrait shot here and so this was not an intentional
16:16
under exposure so much as the conditions were very very bright when I shot this image and so as a
16:23
result it was somewhat tough to properly determine exposure value and so here is where the advantage
16:29
of having dynamic range comes in. Obviously on the left side the original image there just some weird kind of crushed areas And of course the skin itself is pretty lost here in the image But the plus is that here where we have brought
16:46
back, we have got very, very clean results where we've brought a lot of highlights back to the hair
16:52
and we have created an image globally that looks good. In some of these kind of crushed areas of
16:58
shadow we've been able to recover quite clean and and so byproduct is is that for portrait shooters
17:04
it will allow you to do a lot more available light type shooting with the dynamic range of
17:10
this camera simply because you're paying very very little penalty for restoring portions of
17:16
the image or even targeting particular portions of the image to either add exposure or in some
17:21
cases even to take some away. And so certainly a lot of plus there. And so as you can see, whereas
17:29
at the top end of the range, when it comes to overexposure, we got a roughly similar performance
17:36
and I didn't see an advantage for the a7 III. They were basically about equal, but on the shadow
17:44
recovery, the underexposure end of the frame, we found that the a7 III produced even cleaner
17:50
results in the a7r mark iii another thing i think that's important here to note when it comes to the
17:57
color science of the two cameras whereas they're producing pretty similar colors i think for the
18:03
most part what i am finding is that at the extremes both at high iso but then also here in a shadow
18:11
recovery that there is less of a tendency to skew towards green and we saw that at the ends of the
18:19
the dynamic range, how that, particularly in recovering shadows, that it introduced somewhat
18:24
of a green cast to areas on the A7R Mark III. The A7 III is doing a much better job of staying
18:31
neutral. And that tells me that maybe when playing with things like skin tones, you might also get a
18:37
little bit more neutral, a little bit more flexible color science out of that. But certainly, I'm very
18:44
impressed by what I see out of the dynamic range of the a7 III. Let me just say on that note
18:50
you know, whenever you talk about dynamic range, you get people that, you know, basically say
18:55
who needs dynamic range? If you're a good photographer, then you're not going to be
18:59
you know, having to recover these situations. The point I want to make here is not about
19:04
missing something by three or four stops. I mean, sure, as a photographer, you need to learn to
19:09
to do a better job than ending up in those situations. But the reality is that there are sets of circumstances
19:16
to where you have such an extreme of a single scene of shadows and highlights
19:22
that the ability to effectively recover either of those things can make a huge difference to the quality of your end result
19:31
So I would certainly rather have more dynamic range and it doesn't make you a bad photographer to have a camera with good dynamic range
19:39
A good photographer will make the best of their equipment, whatever it is
19:44
On top of that, however, we have reached a place with dynamic range in cameras
19:48
that I think that you do have to show some caution in the way that you approach things
19:54
and that it is possible to create images that are less pleasing to the eye
19:59
because of dynamic range and that you can recover shadow so much
20:03
that you can end up with images that are somewhat flat, almost like a log type profile for video, where there is some of the spark that comes from having higher contrast
20:17
And so I guess the word to the wise is that you're going to need a little bit more of a delicate touch when it comes to processing images
20:25
At the same time, you have more creative options than you ever have before
20:30
So kudos to Sony for giving us a new camera that really has a pretty amazing sensor on it
20:37
and that combined with a very effective focus system and a great price ends up with a camera
20:44
that is very, very compelling. I'm gonna be back to give you my final review on that
20:49
and I will cover autofocus as a part of that final review
20:53
so stay tuned for that. In the meantime, you can look in the description down below
20:59
to find linkage to the ongoing image gallery that I'm adding to. You can also find buying links
21:05
You might wanna get one of these for yourself, and I am strongly considering, right now I'm filming on the a7 III, I'm strongly considering
21:12
selling my a6500 and just using the crop mode on the a7R III for when I'm, you know, reviewing APS-C
21:21
Sony glass and, you know, maybe buying an a7 III for, you know, my primary video rig. And then also
21:29
I like it for shooting events because of the resolution is kind of a sweet spot for when I
21:34
don't need the higher resolution. And of course, you know, Sony could help that along with the A7R
21:40
Mark III and other type high resolution bodies by adding in a medium raw setting, which is kind of
21:46
a gross oversight on their part. I digress. Thanks for watching today. And of course, you can follow
21:51
me on social media. You can sign up for my newsletter. And if you haven't already clicked
21:56
that subscribe button, please do so. Thanks for watching. Have a great day

Sony a73 vs Sony a7R3 - Part 3: Dynamic Range | 4K

Dustin Abbott

Anycubic Is BACK! – Photon P1 LOOKS INSANE! + Kobra X & S1 Max First Look - FormNext 2025

Mr. Cast Iron Interviews In The Kitchen With Karen

Parking Brake Stuck in the Cold? Do NOT Force It!

Automatic Startup Repair not working in Windows 11

[EP 36] Interview with Emily V

Lodge Sportsman Grill Pork Chops

I Tested Viral Kitchen Gadgets — WORTH IT or TOTAL JUNK?

GOAT: Exclusive Deleted Scene

Orlando Miner: quiet quitting, job insecurity, and the housing market

Negotiation Mastery: Busting the 5 Most Dangerous Myths

Blake Weber (The Weber Woodshop): My First Wood Lathe Project!

Money Mike’s Syndicated Podcast: Caitlin Clark's Impact on WNBA, Fever's New Offense, CBA Progress

8 Best Robotic Lawn Mowers That Will TAKE OVER in 2026

Golden Rules of Kayaking for Beginners

10 Best Home & Outdoor Saunas For 2026

Up next in 10

Sony a73 vs Sony a7R3 - Part 3: Dynamic Range | 4K

Dustin Abbott