This article should be read by everybody

Started Dec 12, 2013 | Discussions thread
artlmntl Senior Member • Posts: 1,804
Re: We can't conclude anything from this.

Allan Olesen wrote:

artlmntl wrote:

I think something is not right with these A77 test pictures.

Something is never right with Dpreview's noise test photos.

Dpreview does not take these photos with the same amount of light. They take them with the same ISO setting, no matter which amount of light this results in. This is an incredibly flawed testing method for two reasons:

  1. It is not the ISO which creates the noise. It is the amount of light.
  2. For any photographic purpose it is the amount of light which is interesting, not the ISO. It doesn't have any relevance that Camera A has cleaner ISO 3200 than Camera B if Camera A wants ISO 1600 with the shutter speed and f-stop (or T-stop) where Camera B wants ISO 3200.

I downloaded and converted the RAW files myself using the same settings. The A77 pictures look underexposed and the noise looks crunchy, even amplified. Downsampling does not produce parity. These A77 RAWs look horrible. If you go by these test pictures, the A57 has a huge advantage in high iso performance, but I'm not sure you can go by these. If you told me the sensor in the A77 DPR used for these tests was defective, I'd probably believe you. I'd like to know why this seems so screwed up. I'd like to see some other comparisons shot with different sample cameras using same lens, settings and setup.

There were a lot of discussion of exactly the comparison between the a57 and the a77 when the a57 was new. I have a faint memory that these observations were made, but I haven't checked if I remember correctly:

  • The shutter speed and f-stop for the a57 test photos allowed more light into the camera.
  • The test photos for the a77 were made with a pre-production camera. Not that I think that this necessarily matters, but:
  • The test photos for the a65 were better than the test photos for the a77 even though both cameras use exactly the same sensor.

Thank you for the additional information. It does cause a problem for people who try to use these pictures to decide whether the camera will do what we want them to do.

-- hide signature --

Hunter

Post (hide subjects) Posted by
tex
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow