Quote:
Originally Posted by TheFluff
BZZZT! Worng.
Some fun with REAL PICTURES from REAL FANSUBS:
1
2
Compare the above two shots. 1 is pretty obviously blurrier than 2 (compare the lines), and 1 also seems rather more blueish. 2 is obviously the better encode! Or...?
Spoiler: They are actually both from the EXACT SAME ENCODE. The shots are just two frames apart, and still they're so different. If you don't have the exact same frames, any and all encoding comparisons become completely meaningless. Actually, all encoding comparisons based on single screenshots are pretty meaningless anyway but eh, it gives people something to stir up drama about so why not.
Moral of this story: don't overuse temporal softening, it fucks stuff up. |
There's no subs in the picture though...
I think some people are missing the point, here. Hell, you'll notice on the example pages I posted, the dated VKLL subs are obviously not as clean looking. The difference to note is the style of the subs. The coloring, the size, etc...
I think a lot of people have misunderstood the intent of this as comparing encodes, translation, and other aspects. This was not the aim. The aim is to compare typesets, mostly. Perhaps after I put in some more work on the infrastructure and pop up some more examples and some templates, the purpose will be more easily recognized.
However, I do thank you all for this input. A lot of ideas I had not thought of are being brought to light.
