VideoHelp Forum




Poll: Which deinterlacer in 2023 produces the best quality in single-rate mode?

Be advised that this is a public poll: other users can see the choice(s) you selected.

+ Reply to Thread
Page 2 of 2
FirstFirst 1 2
Results 31 to 33 of 33
  1. Banned
    Join Date
    Nov 2022
    Search PM
    If you saying there is semantic difference between

    Originally Posted by Skiller View Post
    It basically works like this: first it does a dumb bob-deinterlace (with or without more intelligent spatial interpolation such as NNEDI), then temporally blurs the differences in bob-shimmer away and uses motion compensation to detect where to do that and where not to avoid a motion blur effect.
    and

    Originally Posted by Bwaak View Post
    as you said, the whole point of QTGMC was to take the result from a bobber, then to smooth it out
    I am not going to argue

    In any case, all this time I have been in agreement with you that QTGMC adds more functionality to the modules that it uses, and as such it cannot be compared to these modules taken separately. This is the topic of the thread.
    Quote Quote  
  2. Originally Posted by Bwaak View Post
    QTGMC adds more functionality to the modules that it uses, and as such it cannot be compared to these modules taken separately.
    Why not compare ? What is the rationale ?

    Yadifmod/ yadifmod2 with edeint interpolation using nnedi3 essentially uses yadif with nnedi3, why can you not compare the outputs of yadif, nnedi3 separately ? Same with bwdif plus edeint=whatever. or TDeint plus edieint=whatever. They all use the base model with alternate interpolation algorithm

    Sometimes doing too much - adding other processes - "filteritis"- or overprocessing makes the quality worse. QTGMC is guilty of that in some cases - there can be problems such as smoothing out details, introducing temporal artifact errors and ghosting artifacts, wobbly mvtools2 temporal inconsistencies. In some cases it might be better by some criteria, but in others it's clearly worse . It's situation and source dependent. There are well know textbook QTGMC "fail" situations, and others where BWDif excels

    Whatever you want to define "quality" as - it's important to evaluate and compare the output between whatever filter(s) /processes that you apply . In some cases , using an individual "module" might be better (there are clear cases of this, I can post examples) - but you'd never know that if you closed your eyes and never took the time to investigate. You would never know when to use one over another. Or what situations one is ideal for, or terrible for. You would never know the cost/benefit tradeoffs , the pros/cons etc...
    Quote Quote  
  3. Banned
    Join Date
    Nov 2022
    Search PM
    I should have stopped at "QTGMC adds more functionality to the modules that it uses"

    @poisondeathray, point taken, and I stand corrected. It is not a black box with other boxes inside, but a cocktail. What is better: whiskey, Red Bull, or whiskey with Red Bull? It depends. I am going to pour myself some Irish coffee.

    To add some meat to the message, I can say that as I am script-averse, I use VDub Yadif plugin for HD (fast and looks good enough) and VDub MSU Deinterlacer plugin for SD (slightly better than Yadif, but an order of magnitude slower), and I use Temporal Smoother separately. Maybe I should try using lollo's settings for QTGMC (they seem to be simple enough) and compare.

    Is there a comprehensive How-To about QTGMC settings for different content types?
    Quote Quote  



Similar Threads

Visit our sponsor! Try DVDFab and backup Blu-rays!