You can also encode to DNxHD in Adobe Media Encoder, If you choose the mxf type rather than mov it's much quicker because it bypasses the quicktime engine.
With ffmpeg, for 1920x1080 use a command line containing this: -vcodec dnxhd -b:v 145M -acodec pcm_s16le
Add any other parameters you like. If you don't specify a bitrate it defaults to 220. If you don't specify a raster size it defaults to source. Interlace/progressive also defaults to source.
Try StreamFab Downloader and download from Netflix, Amazon, Youtube! Or Try DVDFab and copy Blu-rays! or rip iTunes movies!
+ Reply to Thread
Results 31 to 37 of 37
Thread
-
-
This is the first time I have ever heard anyone make such a claim. Are you telling me back when DVD was the best we could get in the consumer market and the highest end cameras shot in 1080i that the professional studios would de-interlace the content, resize it, then re-interlace it for dvd?
Does this make any sense to you? Have you ever experienced any such problems because I sure haven't. -
Yes, that is 1 way to get around interlacing issues:
Temporarily convert the interlaced to progressive, resize, and then re-interlace.
There are a number of tools that don't anticipate the complications of working with interlaced footage, so this is a workaround to accommodate them.
In common use, even now.
If you haven't heard of it, I'm surprised, but it could be that you just didn't recognize it as such.
Scott -
This subject discussed a lot, in any forum I basically came across, folks notice that using NLE and making DVD out of 1080i is flickering, shimmering etc. It was discussed even here. I found this to be a problem also.
There is a term used like "low pass filter" for video to handle this problem. Deinterlace 1080i into 1080p double frame rate, resize it to SD. Using some techniques like blurring vertical resolution, or other scripts to make it ready for re-interlacing, so it is not flickering that much after reinterlacing. Sure video is slightly blurred then, but overall feel is much better.Last edited by _Al_; 2nd Sep 2014 at 14:33.
-
Perhaps I haven't been 100% clear. The "need" for upsizing arises from YouTube requiring square pixels. I don't understand fully what YouTube does with uploaded video, but Youtube does recommend basic settings for best results. IOW, YouTube doesn't say, "Don't worry if you give us non-square pixels, we will make them square." Rather they seem to want square pixels up front.
The "need" for deinterlacing arises from the fact that I don't like mixing native progressive with interlaced video on my NLE. Plus all of my video will end up on YouTube with little of any being burned to DVD or BluRay.
You obviously have a much better command of sharpening and de-noising than me. Right now I am very dependent on QTGMC doing the thinking for me. But could you be so kind as to post your deinterlacing workflow? It sounds quite bespoke and I am sure educational for this novice!Last edited by SameSelf; 2nd Sep 2014 at 16:14.
-
There is so much being discussed in this thread now it is hard for me to keep up. So forgive my late response. As you and poisondeathray point out QTGMC is a bob deinterlacer. However, I assumed that QTGMC was using motion and edge detection in combination with bob. IOW it is so much more than just bob. And as sophisticles has pointed out, it seems to do a fair bit of sharpening and de-noising as well. Puts your footage through the ringer to quote sophisticles.
I don't know the answer to the question about upsizing myself vs letting Youtube do it. Perhaps I will get a chance to test soon.