HDV to SD deinterlace: surprise!

johnmeyer wrote on 12/8/2008, 11:13 AM

I was doing some tests to prepare a response over in this thread:

HDV to SD Workflow

In that thread -- and in many others in this forum -- people suggest that when rendering HDV footage to MPEG-2 in order to create a standard SD DVD, you MUST set the Vegas "Deinterlace Method" (found in the Project Properties) to something other than "none," or you will get horrible results.

This has never made sense to me, so I did some tests, and boy, did I ever get a surprise!!

I won't bore you with all the steps, and just cut right to the chase:

Setting deinterlace to "blend" or "interpolate" and then rendering to MPEG-2 using the standard DVD Architect Widescreen template (720x480) gives you interlaced footage, but if you set deinterlaced to "none" you get progressive footage, and this footage has all sorts of bizarre artifacts that look like interlace combing except that they are multiple scan lines thick!!

None of this makes sense to me, at any level. I used Vegas 7.0d for the test. Perhaps this behavior has changed in 8.x.

So, just to make sure you understand what I did, I captured some HDV footage directly from my FX1. I put it on the timeline, and set the Vegas project properties to the "HDV 1080-60i (1440x1080, 29.970 fps)" preset. I then set "Deinterlace Method" to none. I then selected Render As, and rendered using the "DVD Architect NTSC Widescreen video stream" MPEG-2 template, without any modifications.

I then did the same thing, except before rendering, I changed the Deinterlace Method (in the Project Settings) to "Blend."

I took the resulting two MPEG-2 files and put them into this AVISynth script, which I read into VirtualDub:

mpegsource("e:\test (HDV widescreen, BFF default SD, no deinterlace).mpg")

separatefields()

When I had deinterlace set to blend, each field was from a different instant in time than the previous field, i.e., it was interlaced. So, in fact, Vegas did NOT deinterlace anything!!!! At least not in the sense that most of us talk about when we use the word "deinterlace."

I can't show that in a still photo, but I can show in this photo the other thing, namely that each field looked smooth and free from artifacts:

In the photo above, each time I pressed the right arrow key in VirtualDub, which takes me to the next field (not frame, but field, because I'm reading the video via the AVISynth script), I got horizontal movement, because each successive field was from a different moment in time.

By contrast, when deinterlace was set to none, each pair of fields was from exactly the same moment in time (i.e., it had been deinterlaced!!!!), and it looked absolutely terrible:

This picture (above), like the previous picture, is one field of video, not a frame (that's why it appears vertically squished). Thus, we shouldn't be seeing interlaced "herring bones." But also note that the herring bones are more than one scan line, so in fact I don't think they are interlace artifacts at all, but rather some sort of strange Vegas bug. If so, who knows how deeply this affects other video rendering situations.

Also, as I went from one field to the next, I got no movement (other than slight up/down) between each pair of fields. This is what you get with progressive video.

So I have no idea what is going on here, and this is one place where it sure would be wonderful if there were an actual live body in Madison who still cared about this stuff and could provide input and possibly a fix, if this is indeed a bug.

I think some of this was covered about a year ago in another thread, but the focus there was in getting rid of the horrible jaggies. I am not sure whether anyone at that point in time figure out that setting deinterlaced to "blend" or "interpolate" actually results in creating interlaced results, whereas setting it to none results in progressive footage.

BTW, I redid these test four times just to make sure I didn't screw up the first time.

Back to post

Comments

johnmeyer wrote on 12/9/2008, 8:29 AM

I certainly hope that the scaling of (interlaced material with the same framerate) is done so that the source fields are scaled separately, and then just used as the source for the final output (with correct field order), instead of FIRST deinterlacing the source material to one full resolution frame, that is then scaled and re-interlaced at the output format. Probably it is done according to the first method, othervice you would lose some temporal information.Perfectly said. But of course, because Sony has gone home and has chosen to never again engage with its customers, we'll never know what is really going on, or exactly what to do about it.

Coursedesign wrote on 12/9/2008, 8:58 AM

Whatever your focussing is (or optical limitation), you have 1080 vertical lines in the image

Strange, I have 1080 horizontal lines. :O)

If there is anyone left who hasn't realized that Vegas is an interlaced-focused NLE, they haven't read this thread.

Note also that the MPEG-2 codec (the only choice for SD DVDs) was designed only for progressive footage.

But MPEG-4 as used in AVC for Blu-Ray truly supports interlaced.

I guess this must mean that the future is interlaced? :O|

(Seriously, I don't think we'll get rid of the frickin' interlaced video until the last TV engineer has been put to his final rest. I would recommend supergluing his casket lid and putting 100+ bolts around the edge to prevent any last minute surprises.)

Christian de Godzinsky wrote on 12/9/2008, 12:45 PM

Couldn't agree more on the detail about the TV-engineers... Should we put the bolts around the coffin horizontally or vertically ??? ;) Thx for correcting my error :)

Interlacing was at the time the only (and clever) way to stuff video signals though tubes and radio transmissions with limited bandwidth - at reasonable frame rates. That's a very loooooong time ago so that's its almost ancient technology. Interlacing - RIP as soon as possible. You will not be missed.

Christian

PS: What do you mean with "...the MPEG-2 codec (the only choice for SD DVDs) was designed only for progressive footage..." ?

johnmeyer wrote on 12/9/2008, 1:20 PM

Note also that the MPEG-2 codec (the only choice for SD DVDs) was designed only for progressive footage.I don't understand that at all. Obviously MPEG-2 supports interlaced, and was designed with that in mind. If you doubt this, remember that MPEG-1 (think VCD) only supported progressive. Interlacing support was one of several main additions when going from MPEG-1 to MPEG-2.

Interlacing was at the time the only (and clever) way to stuff video signals though tubes and radio transmissions with limited bandwidth - at reasonable frame rates. That's a very loooooong time ago so that's its almost ancient technology. Interlacing - RIP as soon as possible. You will not be missed.That is a very correct reading of the history of how modern TV was invented during the 1940s. However, it overlooks the extremely important fact that less than a decade ago, when the digital HD standards were developed and adopted, lots of engineers made the decision to provide 1080i, an interlaced standard, as one of the main HD delivery formats. Reason? Because despite what some people (incorrectly) believe, interlaced video looks damn good. I don't disagree for a moment with those that point out the problems some editing programs have with it. But, properly edited and shot, interlaced video looks absolutely great.

Now, if you can do 60 frames a second, progressive, at 1920x1080, would that look better? Yup. But that's another 2x bump up in data rate, storage, processor power, etc., and except for multiple cores and CPUs, we are nearing the end of processor performance improvements, so editng something that big is going to be a bitch for some time to come.

farss wrote on 12/9/2008, 1:29 PM

This is a very old topic. It first came up around V5. At least for me that's when I first got involved in it. It was found that during PAL to NTSC conversion better results were obtained by setting a de-interlace method. I don't recall for certain but I think a Sony person might have been involved.

Since then I've done some investigation of how broadcast equipment does this. All scalers and standards converters use the same technique, you have to involve both fields to derive each field. As I said previously such equipment may use adaptive motion compensation to align the two fields. Temporal resolution is not lost. If you watch any SD broadcast that was shot in interlaced HD it has been through much the same process.

If you want to study what is going on using footage from a HDV camera is confusing. The motion blur hides the process. I used Vegas generated interlaced HD test footage. If I can find the time I'll do the same tests again and post some images. What you will see happen using Blend is each field will have what looks like motion blur that was not in the original footage created with a 0 deg shutter angle. Temporal separation between the fields IS maintained.

Bob.

Christian de Godzinsky wrote on 12/9/2008, 2:36 PM

John, you are right on the point here, interlace lives on -but only due to bandwith limitations - or should I say bits per second limitations.

That is the ONLY reason we still have interlace. Ok, going to 25p would solve the bandwidth problem, but you would then loose temporal information. However, some people that are in favor of 25p would not have any problems with this....

I also fully agree that 50/60 fields per second (interlaced) gives you a very realistic feeling, and just looks stunning great. Some people want the "film" look, some people want the "realistic" look... Go figure.

It seem that the apetite is always greater than the available food. We always need more BW that currently available. It is the delivery medium or media that still dictates the limits.

I have in my living room a TV that can show 1080p ! Such cams exists. Just the delivery media is missing...

Still I think that interlacing is not a blessing, taking in consideration all the mess it causes in this digital era...

Christian

PS: SCS forum admin, do you monitor us at all ? Would be nice to really hear your comments about the original subject....

johnmeyer wrote on 12/9/2008, 5:56 PM

If I can find the time I'll do the same tests again and post some imagesWell, I created some footage, admittedly a pathological case. I used AVISynth to create 1440x1080 video with the odd scan lines red and the even scan line blue. I then ran this through another script which separated the video into fields, which gives me twice the number of frames, each half the height of the original video. As expected, the video alternated between red and blue frames.

I then rendered the original, non-separated video (which looks purple from a a distance) using various templates, and then put the results back into the script which separates fields. The results were surprising.

The Cineform codec broke down completely and I ended up with green. I have no idea what is going on there. I haven't used this codec for awhile so maybe some update has screwed it up. I know that many users have posted problems with Cineform. Since that's not part of this exercise, I'll look into that later.

I rendered to m2t, and to my surprise, when I put that through the separator, I ended up with perfect red/blue separation.

I then got to the main event, which was to render to the NTSC Widescreen template. I first tried this with deinterlace set to "none." Sure enough, when I separated fields, I ended up, within each field, with a pattern of horizontal red and blue lines, as if the individual field was made up of alternating fields, except that the horizontal lines were more than one scan line thick. This is exactly what I saw in the stills I posted at the beginning of this thread. So, this does not work at all.

I then changed the deinterlace to "blend" and rendered again to the NTSC Widescreen template. This time when I separated the fields, I got perfect red then blue then red, etc. fields, with no distortion or any artifacts. Thus, Vegas does a perfect job of spatially separating the fields.

What I did not do in this test was make any attempt to introduce motion. I guess what I should do is repeat the test except this time have a blue ball move diagonally across a red field, with the blue and red, as before, being on alternate scan lines.

If I have time ...

johnmeyer wrote on 12/9/2008, 6:11 PM

I created a test case with a blue ball, created with only even scan lines, moving over a background of red, which was only on the odd scan lines. When rendered to NTSC DV, with deinterlace set to blend, I didn't see any obvious problems.

Setting to interpolate, however, created an absolute mess, with things turning green and much more.

I don't have any more time to look into this. Since it is a pathological case, it may not reveal much, but certainly I will be using Blend and not Interpolate when going from HDV to SD.

Laurence wrote on 12/9/2008, 8:20 PM

Just to clarify, you are talking about going from 1080i to interlaced SD right? If you go from HD interlaced to SD progressive, there is a difference between using the "blend fields" and interpolate options, but when I go from 1080i to SD interlaced, I can see no difference whatsoever. Both options look the same, but if I uncheck the "select deinterlace method" tab I get all sorts of resized interlace artifacts.

Tim L wrote on 12/9/2008, 8:47 PM

I'm so far out of my league in this discussion that I hesitated to post, but maybe I'll learn something:

Is there any advantage to setting the project properties to "59.940 (Double NTSC)"? Does that make Vegas treat the individual fields like individual frames -- maybe avoiding mixing the two?

I'm just wondering if the "double ntsc" or "double pal" settings result in scaling that uses just one field. (Or, if I don't understand correctly can anybody explain what the "double" frame rates are used for?)

Tim L

Laurence wrote on 12/9/2008, 9:03 PM

The doubled rates are used for high frame rate progressive video which is where we are all headed eventually.

farss wrote on 12/10/2008, 5:05 AM

Finally found something about this process:

http://www.sensoray.com/support/video_scaling.htm

I haven't studied it in depth however it does sound very much like what we're seeing Vegas do. At least this article has pictures that helps to understand the processing.

Bob.