Quote:
Originally Posted by mwardy I don't think anyone has posted any A-B comparisons of the encoders here, so try these. Which do you think is which?
A zSHARE - A1.png
B zSHARE - B1.png
Before anyone asks the next question, they are both B frames. Captured on a satellite stb then imported into the PC. |
OK, small drum roll...
A is the old encoder at 16 mbps, B the new one at 9. Thanks to the people who gave an opinion, all of whom were correct! In fact, Viewer 66 expresses it very similarly to how I did with a previous example:
BBC - BBC Internet Blog: Picture Quality on BBC HD: a response
@awesometeeth: it turns out that today isn't the day for trying to put video clips up, sorry. I'll have a go soon. But in the meantime a couple of things. First, I don't have access to HD atm, but when the previous Doctor Who was on lots of people pointed it out as a shining example of HD done right, and yes, this was on the new encoder. Why should that be when so much else is poor/indifferent? Who knows? Even more mysteriously, one episode of Paradox looked very good but the rest looked like clean SD.

Of course, it does show that not all problems can be simply reduced to with bitrate.
But on bitrate: Surely the blocking and posterisation in B are classic features of bit starvation? You say that you need the video itself because the encoders behave differently. Well, fair enough, and as I say, I'll try, but I've taken a lot of grabs now, and in no case has the reverse been true and the old encoder behaves like this while the new one is artefact-free, or shows fewer artefacts. Not even close. It *never* artefacts like this. I have noticed that I frames can look slightly sharper on the new encoder, but that's not to say the old one's I frames were artefacted. So, do you think there is still no justification for saying there is a bitrate problem? (That's a genuine question, not a rhetorical one, in case it's not clear.)