I have yet to see a group erase parameters to try to hide good ones; it's always preventing others from seeing their idiocy.
This. The SEI takes so few bytes of space, it's not justified to take it out for whatever reason. Also, straight-out copying of settings from someone's SEI isn't usually a really sane thing to do, since you never know if certain settings were selected to better contain that specific source, or if they were even selected sanely. Not to mention that the x264's defaults, together with the presets and tuning settings should already deal with most of your needs. Some do the SEI data tampering for the lulz (and usually "properly"), while the ones who used zeroes for it actually ended up making non-standard streams.
*sigh*I've been doing encodes now for about 4 years (with two of them being avc related). The entire time I have been using simple avs scripts such as:
DirectShowSource("C:\Source", fps=23.976, audio=false, convertfps=true)
Lanczos4Resize("", "") # Lanczos4 (Sharp)
Now, there have been talks about various release groups and their "special" things that they do to their encodes like adding grain or some groups' releases being darker than others or what have you. I've been doing my avc encodes, like I said, for about 2 years with my simple scripts and have had great success. Am I missing something that's never spoken of in any of the Blu Ray guides or something like adding more plugins that will add to my encode quality or something?
*Sources are from Blu Rays, not some screener or whatever else kind of releases there are.
As far encoding goes, KISS is the main principle. If something isn't broken, don't touch it. But if you see something bad (anything that strikes your eye as something that shouldn't be there -- and couldn't be an artistic choice on that scene, such as halos from oversharpening/upscaling
[Well hello there DTB], way too excessive noise
[Well hello there Tayutama], striking aliasing
[Well hello there Bakemonogatari; I've got to love those low-res CG items on high(er)-res art], banding in gradients etc. etc.), you might try to fix it with something that does the least amount of evil to the other parts of video (which might be very much :effort: at times). Also, asking for help when trying to fix something is generally never a bad idea, just as I found out myself some time ago with my dehalo'ing needs. And then, lastly, you have to care about actually getting the final encode to look close enough to your input, which is mostly easy by just using a good-enough CRF value, but which might need some special tweaking at times. Consulting other people on this is usually a good idea as well.
As for that script, I personally dislike DSS for its shortcomings and non-frame-exactness, so I'd switch to Haali's DSS2 or Myrsloik's and friends' ffms2 for H.264/VC-1 (although the latter might need a remux to MKV with eac3to, not sure how problematic transport streams still are for ffms2), and to dgdecode for MPEG-2 sources.
Here are x264 scene rules and settings from 2007:
http://www.sbytes.info/NFOwb.php?id=high.def.x264.movie.standards.rev2
Their recommended encoding settings are at the bottom.
As far as the "x264 scene rules" go, they're just meant to keep the worst of the pack from doing
very stupid things, and they shouldn't really be kept for any reference nowadays -- since the x264 defaults are pretty good nowadays, and you'd mostly end up using command lines the likes of "--preset mumbo --tune jumbo --crf value" (with the --crf switched to 2pass bitrate VBR mode (with VBV usually) for when you have to limit yourself to some size (physical media etc.)).