ISO 3200 performance

But how can it be so ? There must be some magic involved.

I have been told that there are the airy discs and diffraction limits and pixel fairies that all increase noise on Sony cameras. You must have shot this on a Pentax and changed exifs to fool us. That can be the only possible explanation. :-).
 
Thank you. I am not sure if the A580 has the same sensor as the A57. The A55 has the same sensor as the A57. Actually, I bought the A77 first before the A57. But after testing the A77's high ISO performance, I returned it. I was not satisfied with the A77's high ISO performance even at ISO 800.
Pics look great! I hope my incoming A77 performs this well at ISO3200. Does the A57 have the same sensor as the A580 (my current camera).
--
Emmanuel
"Vanity, ALL is Vanity and chasing after the Wind" - King Solomon
 
: -) No
But how can it be so ? There must be some magic involved.

I have been told that there are the airy discs and diffraction limits and pixel fairies that all increase noise on Sony cameras. You must have shot this on a Pentax and changed exifs to fool us. That can be the only possible explanation. :-).
--
Emmanuel
"Vanity, ALL is Vanity and chasing after the Wind" - King Solomon
 
Thanks. I have the A57 for a month now and I am glad I made the right choice. Now, the price even went down to $618 at B&H. If the A57 has the same features as that of the A77, it would have been a formidable photographic tool.
That's a very impressive showing of the a57 at ISO3200. Thanks for sharing some real world information with us.
--
Emmanuel
"Vanity, ALL is Vanity and chasing after the Wind" - King Solomon
 
Those are excellent but do you realize that because you used + 0.3 EV comp that the true iso was 1/3 stop less than iso3200?
I thought that meant the camera would have picked 1/65 but the photographer asked for a +0.3Ev exposure over what the camera suggested resulting in the 1/80. It sounds like you are suggesting that the +0.3Ev compensation is done electronically after the image is captured?
On a modern camera, noise almost completely depends on the amount of light, not on the actual ISO setting.

If you set your exposure parameters so you get 1/3 stop more light than the camera suggested, the resulting noise will be equal to setting your ISO 1/3 stop lower and use the camera's suggested exposure parameters.
I see this a lot but I dont think it actually applies. The noise is more noticeble in some areas, but less in others. The following crops show my point.













As you can see, the -1ev actually shows less noise in the speaker area, and more noise in the paint on the wall. The grain size of the noise is the same for each though, and this will distort edges and lines equally. The reason noise is often more noticable in low light is obviously bc lighter color chroma and luminance noise is further away in color than a black or a dark. If you blow every pixel on the frame and its all white, there is still noise.

If you never want to shoot a shadow, or have to lose shutter speed by +EV then that sucks. I didnt buy in to the theory however, and this shows its not actually true.
 
Emmanuel--just looking for more info about your findings with a77 and high iso. you said you find 800 not satisactory. Was this Jpeg or raw? Just curious as i have a77 in the mail this tuesday and am a current owner of a57 (love the camera, just want more control)
 
Those are excellent but do you realize that because you used + 0.3 EV comp that the true iso was 1/3 stop less than iso3200?
--
So again, by this logic a picture that is 3 stops higher iso yet 4 stops of EV more should have less noise, by one stop. This is the exact comparison I have here, and as you can see, the lower iso and -2EV shot still has less noise.









If this is the basis for the many arguments of sony noise issues, then I want nothing to do with that line of thinking. Boosting available light causes more noise, EV changes shutters which has nothing to do with the sensor as long as it does not end up a very long exposure (several seconds or more).
 
Unravel,

It is jpeg. Let me know what are your findings when you get your A77 on jpeg. Sometimes, I am thinking to buy the A77 back for more control and features because I think I outgrown my A57. But on the second thought, I will invest on a lens like the 70-200G and 70-400G and wait for the full frame - A99
Emmanuel--just looking for more info about your findings with a77 and high iso. you said you find 800 not satisactory. Was this Jpeg or raw? Just curious as i have a77 in the mail this tuesday and am a current owner of a57 (love the camera, just want more control)
--
Emmanuel
"Vanity, ALL is Vanity and chasing after the Wind" - King Solomon
 
Actually it appears to me that the amount of noise is pretty much the same in all areas in both photos. How apparent it is varies with how bright it is. Considering that the crops are highly magnified and at ISO 6400 I don't think the noise is all that objectionable. Your obsession with high ISO noise is a bit over the top. I am much more concerned with color, sharpness and dynamic range all of which are compromised with every camera made at such high ISOs. Because of that I avoid using high ISO at any cost and rarely shoot above ISO 1600 and never above 3200. In poorer light I use fast lenses. I guess it's a habit that goes back to never using film above ISO 800.
--
Tom

Look at the picture, not the pixels

http://www.flickr.com/photos/63683676@N07/
http://www.flickr.com/photos/25301400@N00/
 
I am very impressed by A57's high ISO performance. I took my A57 with 16-50mm in our family vacation at Hershey's Chocolate Tour. Here are some images taken with no post processing involved:
Is this jpg files from the camera? Then what camera settings?

I think this is the results you could expect from any new aps-sized sensor with proper exposure at ISO 3200.
 
How about equalizing exposure? It's easy to eliminate noise if you just make everything black...
If you blow every pixel on the frame and its all white, there is still noise.
or white. If every pixel is white, then where's the noise? Nothing has been captured except that the exposure was sufficient to saturate the sensor and/or electronic path.
 
Those are excellent but do you realize that because you used + 0.3 EV comp that the true iso was 1/3 stop less than iso3200?
Nope, ISO is electronic gain (as the variable with sensor sensitivity as the constant). Exposure compensation is a modifier for the exposure calculation algorithm and has nothing to do with sensitivity (except in that it may cause the algorithm to pick a different gain).
This is the exact comparison I have here, and as you can see, the lower iso and -2EV shot still has less noise.
It's also invisible (while the other is blown out), not really a useful comparison. Again, equalize luminosity in PP for a useful comparison.
If this is the basis for the many arguments of sony noise issues, then I want nothing to do with that line of thinking.
How about this. Take the same picture with the same lighting and as many different exposure/ISO settings as you want. Then PP them to look the same! Then we can usefully compare how the changed settings affect the final result.

This is like pixel peeping a 36MP sensor and comparing it to an 8MP sensor. You have to do something to equalize them. What to do (upsample, downsample, print...) is a separate debate, but without equalizing the outputs it's very difficult to compare them.
Boosting available light causes more noise
Uh?
 
Yes, it is jpg. I always use an Aperture Priority mode, DRO Auto, AWB, Standard Creative Style (Contrast 1,Saturation 1 and Sharpness 3)
Is this jpg files from the camera? Then what camera settings?

I think this is the results you could expect from any new aps-sized sensor with proper exposure at ISO 3200.
--
Emmanuel
"Vanity, ALL is Vanity and chasing after the Wind" - King Solomon
 
Invest in better lens is far more sensible to invest in a digital camera body because digital tech moving really fast. After a yr or two sure have must better 24mp camera out there.
It is jpeg. Let me know what are your findings when you get your A77 on jpeg. Sometimes, I am thinking to buy the A77 back for more control and features because I think I outgrown my A57. But on the second thought, I will invest on a lens like the 70-200G and 70-400G and wait for the full frame - A99
Emmanuel--just looking for more info about your findings with a77 and high iso. you said you find 800 not satisactory. Was this Jpeg or raw? Just curious as i have a77 in the mail this tuesday and am a current owner of a57 (love the camera, just want more control)
--
Emmanuel
"Vanity, ALL is Vanity and chasing after the Wind" - King Solomon
 
You are the man. You are few of them willing to share your setting. :)
Yes, it is jpg. I always use an Aperture Priority mode, DRO Auto, AWB, Standard Creative Style (Contrast 1,Saturation 1 and Sharpness 3)
Is this jpg files from the camera? Then what camera settings?

I think this is the results you could expect from any new aps-sized sensor with proper exposure at ISO 3200.
--
Emmanuel
"Vanity, ALL is Vanity and chasing after the Wind" - King Solomon
 
Photoviewer,

No problem. I read my Bible everyday and one of the things I learned is to share what I have. There is a blessing and joy to help somebody.

Emmanuel
Yes, it is jpg. I always use an Aperture Priority mode, DRO Auto, AWB, Standard Creative Style (Contrast 1,Saturation 1 and Sharpness 3)
Is this jpg files from the camera? Then what camera settings?

I think this is the results you could expect from any new aps-sized sensor with proper exposure at ISO 3200.
--
Emmanuel
"Vanity, ALL is Vanity and chasing after the Wind" - King Solomon
--
Emmanuel
"Vanity, ALL is Vanity and chasing after the Wind" - King Solomon
 
So again, by this logic a picture that is 3 stops higher iso yet 4 stops of EV more should have less noise, by one stop.
At equalized viewing conditions this would be a likely result. But you will run into other issues, especially when working from jpeg.
This is the exact comparison I have here
You do not have equalized viewing conditions.
as you can see, the lower iso and -2EV shot still has less noise.
Impossible to draw any conclusions from your photos, especially regarding noise. Brighter exposure will make the existing noise less visible, while at the same time resulting in more noise if the exposure is increased by ISO boost.
If this is the basis for the many arguments of sony noise issues, then I want nothing to do with that line of thinking.
We know you're not honest, but thanks for spelling it out.
Boosting available light causes more noise
Literally, you are so incorrect in this statement it's not even funny. Boosting available light - which means to provide more light onto the subject - does not in any way increase noise. It always decreases noise.

And if that wasn't what you meant, why don't you stop and think before you hit post so that you're certain you've written what you actually mean?
EV changes shutters which has nothing to do with the sensor as long as it does not end up a very long exposure (several seconds or more).
Which is also blatantly incorrect - and goes against all your noise pedantic ranting over the months.

Jesper
 
On a modern camera, noise almost completely depends on the amount of light, not on the actual ISO setting.

If you set your exposure parameters so you get 1/3 stop more light than the camera suggested, the resulting noise will be equal to setting your ISO 1/3 stop lower and use the camera's suggested exposure parameters.
I see this a lot but I dont think it actually applies. The noise is more noticeble in some areas, but less in others. The following crops show my point.

As you can see, the -1ev actually shows less noise in the speaker area, and more noise in the paint on the wall.
I will repeat myself from another post:

"The ISO vs. over/under exposure discussion is only really relevant if you darken or brighten a photo in post processing to compensate for wrong exposure. "

In your examples, you have not done any post processing to compensate for wrong exposure. If you did that, you would see much more noise in the picture taken at -1EV.
 
Yes the a57 ISO 3200 performance is very impressive!!
These shots confirm only that the ETTR method work perfectly well with ALL cameras (including Sony)

Impressive, compared to what?
No comparison shots with native sensitivity, nor another camera/s.
WB is so so.

Contrast is to high (so, middle tones areas wherever the noise can appear are almost inexistants).

This test says nothing. A missed assessment imho.

Regards,

--
Michel J
 

Keyboard shortcuts

Back
Top