joeleejj
帖子: 4
注册时间: 2018-02-03 12:54

关于各种效果混合问题

2018-02-08 18:30

首先必须说清楚,本人小白一名,完全不懂avs语言但对于压制很感兴趣,在反复爬了各位大大的文章,很多事情始终没办法自己弄明白。。于是发这个贴问问大家。

首先我想对视频(anime)进行aa,dehalo,胡化降噪daband再回噪,非線條的亮場 denoise,非動態部分temporal stabilize,動態部分被motion blur之後的線條稍微做一點銳化,8bit 掺16bit filter, 最后 8bit 压制:

Dither_convert_8_to_16()

Dither_resize16nr (1280, 960, kernel="spline36")
nr16 = Dither_removegrain16(11, 11).Dither_removegrain16(20, 0)
noise = Dither_sub16(nr16, y=3, u=3, v=3, dif=True)
nr16 = Dither_sub16(noise, y=3, u=3, v=3, dif=True)
last = nr16
DitherPost ()

zero_l = 64
m_edge = mt_edge("sobel", 7, 20).mt_inflate()
m_luma = mt_lut("x "+String(zero_l)+" - Abs 255 * "+String(zero_l)+" 128 - Abs 128 + /")
m_motn = mt_motion()
m_aa = m_edge.mt_expand()
m_halo = m_aa.mt_expand().mt_lutxy(m_edge, "x y -")
m_fin = mt_logic(m_edge, m_motn, "max")

aa_clp = maa2(aa = 48)
dh_clp = aa_clp.Dehalo_alpha(rx = 2.0, ry = 2.0, darkstr = 0.8, brightstr = 1.3, lowsens = 50, highsens = 50, ss = 1.5)

Dither_convert_8_to_16()
db_clp = nr16.f3kdb(16, 40, 40, 40, 0, 0, input_mode=1, output_mode=1)
last = Dither_add16(noise, y=3, u=3, v=3, dif=True)
nr_clp = dfttest(lsb_in=true, lsb=false, tbsize=1)
DitherPost ()

flat = mt_merge(last, nr_clp, m_luma, luma=true)
non_hl = mt_merge(flat, aa_clp, m_aa)
spat = mt_merge(non_hl, dh_clp, m_halo)

stab = spat.TTempSmoothF()
sharp = spat.MSharpen(threshold=10, strength=26, highq=true, mask=false)
temp = mt_merge(stab, sharp, m_fin)
last

这样写能行吗?

回到 “理论讨论 / Theoratical discussion”