Started Apr 14, 2013 | Discussions thread
TrojMacReady Veteran Member • Posts: 8,729
Re: Comparison at high ISO (6400)

123Mike wrote:

The viewing software scaled down the image you silly.

No it didn't, you "silly". I obviously did, using a Lanczos algorithm. There is no NR or "enhancing" in that process. And again, it doesn't matter which equal size you scale them to (upscaling the A57 works too, printing both at the same size as well), the relative differences is as displayed.

Yes it enhances! There is an inherit sharpening effect with the Lanczos algorithm!

First, you can't sharpen detail back that wasn't there. Second, Lanczos doesn't enhance anything since there is still information lost. It minimizes the loss. Don't confuse preserving detail from the original file, with sharpening.

It's called normalization and there is no better way to do it than the above mentioned methods. So how did you do it? Since you had some claims yourself too.

My point is, that you are comparing a 1:1 A57 output with a scaled down ENHANCED A58 output. That is not a fair comparison.

Nonsense, see what I wrote above. Educate yourself on scaling algorithms before making claims such as these.

That means it went through whatever algorithms it has to do a good job. I have an idea that might address this, but it may seem strange... Take the image, load it in a photo editing program like Photoshop (something descent, not Microsoft paint or something stupid like that). Make the image 2.5 times larger. 2.5 as wide I mean. Save that image, high quality possible. Heck, save as PNG. Then load that into the viewing software you were using. Show the A58 as you did before, where it got to scale down the image to make it visually the same scale as an A57 output. Then do the same to the upscaled A57 shot. Have it show it scaled down as if it were the original size. Now it underwent the same algorithms. I bet it comes out sharper. But I could be wrong.

Great if you're trying to kill time. See above.


Yes, you are in denial of your absolute ignorance on the subject, obviously. And more importantly, you're falling back to your old pattern. Demanding pointless "proof" for claims, when the proof is in front of you, yet making claims yourself that you can't back up. Hence why you carefully edited out my question asking how you compared the files.

When you pan around, one can easily cherry pick bits where the A57 looks better as well.

Only the far left of the A58 shows a focus or lens (curvature ?) issue with a sudden drop off in sharpness. The rest is consistent with the crops posted.

So, when the A57 is sharper it is because of the lens/focus, and when the A58 is sharper it is because the A58 is better? That's cherry picking on your part!

I didn't say the A57 was sharper, read again. I said the A58 sharpness has a sudden drop off towards the left. But this was about jpeg engines and how they compared. The RAW files show similar sharpness in the center, the correspondingjpegs do not. Thus the extra loss can be ascribed to the jpeg engine of the A57. Let's not let cognitive dissonance get in the way of that one.

It was not the "jpeg engine" that caused the out of focus effect. It was the test not being done right.

Face it, you can't explain the difference between the RAW files and corresponding jpegs and now you're trying to dodge it with loose claims of things not being done "right". Cognitive dissonance, it took you over.

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