Originally posted by MOTT the HOOPLE
Jpg compression errors my arse! I have worked with renders and graphic software for years and I have seen hundreds of errors, But nothing like that!
With modern computers and software compression errors should be rare event.So what is NASA using to process there images a "commodore 64 and delux
[edit on 28-1-2010 by MOTT the HOOPLE]
It's easy to spot the pretenders. They start off with feigned incredulity and swear words, and then claim to have experience.
Let's just rub this in. Mott the Hoople has "worked with renders and graphic software for years". Yet he isn't aware that image processing, such
as sharpening and compression, creates false detail?? He isn't aware that compression, by DEFINITION, alters the original, especially around edges??
He isn't aware that the level of compression is entirely up to the person/software doing the compression, and that high levels of compression (and
the resulting poor quality) may be used where other constraints (eg slow data transmission, file size) are important?? Obviously he hasn't done any
video on a mobile phone.. He isn't aware that JPEG compression produce false detail around edges with high contrast, and doesn't understand the
characteristic blocky appearance that highly compressed jpeg files have??
Yep, he's got lots of useful experience. May I quote him:
Here's a quick example of what happens with quite routine processing - took me all of 5 minutes. NOTE - as above, I don't know exactly what
routines NASA uses, so this was just a WAG. It is by no means intended to match the quality of those initial images. It just gives an example
of the effects, using the same sort of processes used on the low-res images.
Click on it to get to the full-screen version, then you may need to click again to see it at 100%. Feel free to save it to your machine and view it
at huge, meaningless magnifications.
Note the blocky appearance. Note the haloing around the bright spots. Note the weird patterns that appear around any high contrast areas. Mott the
Hoople and other 'experts' have never seen this. They need to get out more...
Would you like to VERIFY that these effects are inevitable using similar processing to NASA?
Well, here's how to do it, step by step. FTR, I used XnView (freeware) to do all the processing.
1. Load the original file, namely: 20100118_053530_n4euB_195.jpg
2. Increase brightness (I used 22)
3. Increase contrast (I used 32)
4. Reduce image size to 512 pixels using the bilinear algorithm (this roughly corresponds to the 'pixel binning' reduction that nasa uses for the
initial streamed files)
5. Use 'Focus Restoration' to (over)sharpen.
6. Enlarge image to 1024 pixels (nasa resizes the streamed files (don't ask me why))
So there's an example. No, it doesn't match the NASA one exactly as I don't have their software or specifications. Someone with more patience
than me, might wish to try other settings to better match the initial files. But it does show all the effects that these so-called experts say can't
happen, and you can verify this for yourself.
Now after all this, you are of course perfectly entitled to still think there are huge spheres out there. And I'm sure there will be new images that
can be similarly misinterpreted by those without a shred of imaging knowledge and with the motivation to believe that it is something unexplainable.
It must feel wonderful to believe that you have made a huge discovery that all of us fools didn't spot...