Sony A7R3 Noise

[Malchio]

Hi all,

one reason to upgrade from my trusty a6000 to a A7R3 was the noise. When photographing at night time,.I used the A6000 for northern lights photography and pictures quite good, but noise level was far to high to print any of the pictures.

So last weekend was the first time I tried the A7R3 at night.

Now Im curious if its only me or the photos are more noisy then they should be...

If they are im pretty sure its my bad, but i cant quite get it. The moon is not perfectly sharp because of the SS I do know that, but SS of 1/125 or higher is quite hard to achieve with such a dark moon.

All photos are:
manual focus
shot with A7R3+SEL70300g
no noise reduction/sharpening from LR




+1 exp






+1 exp






+2 exp






+1 exp




+0 exp



greetz

Jonas

 

[Beaverhelmet]

A couple of things got me puzzled. Only a couple of the shots have something in focus (pic 3 and 5 have the moon in focus) on the other pics, pretty much nothing is in focus. Maybe because all shots were done wide open, which leaves very little margin for focus error.

And the choice of ISO and exposure is not what I would pick. Did you shoot this hand held? For me, this should definitely be done from a tripod, which would allow far slower shutter speeds, plus closing down aperture a tad and pick a lower ISO to boot.

 

[BBQue]

Can't tell much from these images. Also depends on your camera settings regarding noise. Or how you process you RAWS and what your noise reduction settings are in software.

I had both A6000 and now A7riii. I can tell you the difference in noise is huge. Shoot some street scene with both cameras side by side and you will see.

Add to that IBIS and ability to reduce ISO by several stops due to slower shutter speeds.

Add to that the possibility to donwsample your 42MP files to 24 to make the comparison fair. And the difference will be eye-popping huge!

 

[Malchio]

all pictures are with tripod.

longer exposure instead of higher ISO would have been nice, but the moon is too fast moving. Normally the SS had to even faster.

 

[Malchio]

which noise reduction settings in camer do you mean? Long exposure NR is disabled as I found on the A6000 that it wasnt beneficial.

Software wise the RAWs are imported to LR and no NR was applied.

 

[Beaverhelmet]

all pictures are with tripod.

longer exposure instead of higher ISO would have been nice, but the moon is too fast moving. Normally the SS had to even faster.

Ok.

Regarding the noise, I suspect you do the same fault as most people when comparing noise between cameras/sensors. You absolutely need to normalize the photos to the same output size, be it 8MP jpegs or physical prints, it doesn't matter as long as the candidates have the same measurements. Comparing straight out of camera shots at pixel level will not tell you the truth about the actual low light capabilities of the two cameras.

 

[biza43]

which noise reduction settings in camer do you mean? Long exposure NR is disabled as I found on the A6000 that it wasnt beneficial.

Software wise the RAWs are imported to LR and no NR was applied.

LENR would not matter, since it only kicks in at lower shutter speeds than yours.

If you find noise too much, learn how to do noise reduction in LR.

 

[BlueCosmo5050]

I don't know if anyone commented this but there is something you have to keep in mind here.

When I got my first high megapixel camera, the original Sony A7R, it was 36 Megapixels. I'd zoom in 1:1 in Lightroom and I'd see noise at ISO's I wasn't used to seeing with the 5D Mark 3 I had before it.

The problem was, I was zooming into a larger photo. If you zoom into a 42 megapixel photo with the A7R3 and compare it to a 20 Megapixel photo of the A6300, *I forgot exactly how many MP it had but somewhere around there.* You will notice more noise sometimes even if technically the image is over a stop cleaner.

To show yourself that this is the case, export a Sony A7R3 photo at 20 Megapixels, then take a native A6300 photo at the same ISO, same lighting conditions. You'll see the A7R3 is cleaner, even when you zoom in 1:1.

However, now when you zoom in 1:1 you're seeing what you are used to.

You have to get used to looking at high resolution photos and keeping in mind the resolution is double what you're used to.

So, if you put both photos online, you wouldn't post a 42 Megapixel photo online more than likely, most websites are even less than 8 megapixels, you will then see how much cleaner the full frame is.

I didn't realize that at first either until one day I realized what was going on.

Some people hate when I bring this up, but the same applies to sharpness in shutter speed. On a 20 MP camera, the thing that you can get tact sharp at 1/160th a second may take 1/250th of a second on a 42 MP camera, but ONLY if you are talking 1:1 native resolution. If you are talking viewing them at the same size, then it does not take a faster shutter speed.

Since I never know if I'm gonna crop a lot or not, I always go for 1:1 tact sharpness. Which shouldn't be too hard for Sony users, considering you have IBIS. Being a 5D Mark IV user, I do not have IBIS. I have IS on the lenses but that's it.

 

[Malchio]

I don't know if anyone commented this but there is something you have to keep in mind here.

When I got my first high megapixel camera, the original Sony A7R, it was 36 Megapixels. I'd zoom in 1:1 in Lightroom and I'd see noise at ISO's I wasn't used to seeing with the 5D Mark 3 I had before it.

The problem was, I was zooming into a larger photo. If you zoom into a 42 megapixel photo with the A7R3 and compare it to a 20 Megapixel photo of the A6300, *I forgot exactly how many MP it had but somewhere around there.* You will notice more noise sometimes even if technically the image is over a stop cleaner.

To show yourself that this is the case, export a Sony A7R3 photo at 20 Megapixels, then take a native A6300 photo at the same ISO, same lighting conditions. You'll see the A7R3 is cleaner, even when you zoom in 1:1.

However, now when you zoom in 1:1 you're seeing what you are used to.

You have to get used to looking at high resolution photos and keeping in mind the resolution is double what you're used to.

So, if you put both photos online, you wouldn't post a 42 Megapixel photo online more than likely, most websites are even less than 8 megapixels, you will then see how much cleaner the full frame is.

I didn't realize that at first either until one day I realized what was going on.

Some people hate when I bring this up, but the same applies to sharpness in shutter speed. On a 20 MP camera, the thing that you can get tact sharp at 1/160th a second may take 1/250th of a second on a 42 MP camera, but ONLY if you are talking 1:1 native resolution. If you are talking viewing them at the same size, then it does not take a faster shutter speed.

Since I never know if I'm gonna crop a lot or not, I always go for 1:1 tact sharpness. Which shouldn't be too hard for Sony users, considering you have IBIS. Being a 5D Mark IV user, I do not have IBIS. I have IS on the lenses but that's it.

I think that exactly could be my problem. Thanks for pointing that out.

I will give it a try.

 

[BBQue]

So, if you put both photos online, you wouldn't post a 42 Megapixel photo online more than likely, most websites are even less than 8 megapixels, you will then see how much cleaner the full frame is.

This is true of course. You can downsample the A7rii or A7riii file to 12MP and its high ISO performance is just as good as the legendary A7s.

However, files are meant to be viewed at 100%. That's why we bought a 42MP camera and not a 10MP one, right? However, even at 100% mag the noise of the A7riii looks better (more even, less "wormy" compared to A6000).

Some people hate when I bring this up, but the same applies to sharpness in shutter speed. On a 20 MP camera, the thing that you can get tact sharp at 1/160th a second may take 1/250th of a second on a 42 MP camera, but ONLY if you are talking 1:1 native resolution.

Exactly my experience. In the old days, 1/50 for 50mm lens or 1/100 for 100mm tele lens was safe. With high res sensor, you need to double that, i.e. 1/100 as minimum shutter speed for 50mm lens, 1/200 for the tele lens. (That's without IBIS. How much extra the IBIS gives depends very much on your technique, and you MUST test that yourself cause no two photographers are alike).

Not sure what you mean by "only if you are talking 1:1 native resolution". We are ALWAYS talking about the resolution the camera can bring to the table, i.e. native 42MP for A7riii. It makes no sense at all to buy such a camera and then throw away that resolution by sloppy technique, i.e. using too slow shutter speeds or not knowing how to use IBIS) - and then saying, "ah, oh, well, my friends watch my pics on their cell phone screens anyway. It is just fine as is".

 

[HP1999]

Is flicker control on?

 

[Malchio]

nope, should it ?

 

[HP1999]

nope, should it ?

I’d keep it off

for the moon try using the spot meter and a monopod or tripod

 

[BlueCosmo5050]

So, if you put both photos online, you wouldn't post a 42 Megapixel photo online more than likely, most websites are even less than 8 megapixels, you will then see how much cleaner the full frame is.

This is true of course. You can downsample the A7rii or A7riii file to 12MP and its high ISO performance is just as good as the legendary A7s.

However, files are meant to be viewed at 100%. That's why we bought a 42MP camera and not a 10MP one, right? However, even at 100% mag the noise of the A7riii looks better (more even, less "wormy" compared to A6000).

Some people hate when I bring this up, but the same applies to sharpness in shutter speed. On a 20 MP camera, the thing that you can get tact sharp at 1/160th a second may take 1/250th of a second on a 42 MP camera, but ONLY if you are talking 1:1 native resolution.

Exactly my experience. In the old days, 1/50 for 50mm lens or 1/100 for 100mm tele lens was safe. With high res sensor, you need to double that, i.e. 1/100 as minimum shutter speed for 50mm lens, 1/200 for the tele lens. (That's without IBIS. How much extra the IBIS gives depends very much on your technique, and you MUST test that yourself cause no two photographers are alike).

Not sure what you mean by "only if you are talking 1:1 native resolution". We are ALWAYS talking about the resolution the camera can bring to the table, i.e. native 42MP for A7riii. It makes no sense at all to buy such a camera and then throw away that resolution by sloppy technique, i.e. using too slow shutter speeds or not knowing how to use IBIS) - and then saying, "ah, oh, well, my friends watch my pics on their cell phone screens anyway. It is just fine as is".

Because not everyone is talking about 1:1, you may say always, but some people will get in an argument over just bringing that topic up. Because they will claim it's not like you can see the difference compared to the other photo from the lesser resolution camera if at the same size. That might be true, but not at 1:1.