Best 16:9 without anamorphic lens (YAT)

Joshua Provost said:
I prefer straight SQUEEZE to Photo Zoom. So, I guess you'd have to try it yourself and find out.
Click to expand...

Sorry for re-upping this thread, but I'm still confused about a few things.

Wouldn't uprezzing letterboxed footage yield better footage than uprezzing Squeeze to HD resolution? Barry Green explains it's better to do this in his book, since you stretch unmodified footage, rather than modified footage. But it doesn't seem to make sense to me, since when you uprez the Stretched footage, it's already 16:9 -- so you stretch/uprez LESS than if you did with Letterbox footage; so wouldn't it offbalance the whole "stretching untouched v.s. touched footage" issue? Plus, you get the added factor that Squeezed footage is footage that is stretched in camera, before compression, etc. So with uprezzing letterbox -- you uprez 100% (using these numbers as an example), whereas with Stretch footage, you uprez 70%, with the other 30% already stretched for you by the camera (probably better stretched also).

In the book, when Barry Green says "letterbox"; I assume he means either 4:3 or Letterbox footage (with black bars); I don't think he cares either way. But based off of this thread, it seems it actually would be BETTER to uprez to HD from Letterbox footage rather than 4:3 footage, since Letterbox probably contains just a little less compression.

Another issue is that when shooting with Letterbox mode, I've heard that the footage isn't actually true 16:9; it's a different aspect ratio -- is this just an issue with certain cameras, or with all DVX's?

My last concern is why someone needs to uprez stretched/letterboxed footage to HD res., especially when you plan on viewing the footage on DVD on your big screen TV -- the end result will still be 480 right? So you uprez the footage to HD resolution, and you burn it onto DVD which basically shrinks it back down to 480. Or is downrezzing HD back to 480 always better? This might make sense with footage originally in HD, but the DVX footage was uprezzed in the first place -- so to me there seems to be some unneeded messing with the footage.
 
Hey guys,

Been really busy with the hobby shop so I've pretty much just used my DVX100a to shoot promos for SpeedWings and not spent much time in any forums except RC. But, I'm looking at perhaps getting an HVX200 as I would love to have HD in 24P (some said it would never happen ;>) so I'm back reading up on the pros cons.

Anyway, as how to get the best 16x9 without a anamorphic lens, I started a thread on this some time back shortly after getting my own camera. Turns out I was pretty much on, according to an email Jan Critten sent me regarding when compression is actually done in the DVX100.

To summarize - Using squeeze mode, we can discard the datastream from 25% vertical we loose prior to compression. Compression then occurs and since DCT is always 3.5 MB per second no matter what is being compressed, in the end, the less information that needs to be compressed, the higher the quality of that information is when decompressed back to view or edit. So, theoretically at least, squeeze mode is superior to either 4:3 or letterbox when shooting for 16x9 work.

For any interested, here's a link:

http://www.dvxuser.com/V6/showthread.php?t=8880

All I shoot is squeeze and I see no reason to (or even possibility of) using an anamorphic adapter as I am zooming all over the place trying to keep up with our 100+ MPH flying wings. Now if I had a HVX200 with true 16x9 CCD's, that would be great as I have always kinda felt short-changed loosing so much resolution shooting 16x9 in the DVX100a.

Hello to Jarrad, Berry, and the rest of the guys.

Still waiting for my 100 bucks BTW (think it was a 100 bucks anyways ;>)

DVXUser.com is still the best on the web for sure.

Take care all.

-Rodger
 
IMHO, the letterboxing option has no use in this day and age. Shoot either 4:3 or 16:9 (squeeze or with the Anamorphic adapter). Delivery options almost always support this and it will be future-proofed for when even more people have 16:9 displays.

I'm not terribly sold on the HDV world, but 16:9 is a definite must in my opinion, whether SD or HD.
 
marjamar said:
To summarize - Using squeeze mode, we can discard the datastream from 25% vertical we loose prior to compression. Compression then occurs and since DCT is always 3.5 MB per second no matter what is being compressed, in the end, the less information that needs to be compressed, the higher the quality of that information is when decompressed back to view or edit. So, theoretically at least, squeeze mode is superior to either 4:3 or letterbox when shooting for 16x9 work.

For any interested, here's a link:

http://www.dvxuser.com/V6/showthread.php?t=8880
Click to expand...

Hey Rodger, thanks for confirming the whole bit about Squeeze mode stretching things BEFORE compression -- this confirms that I should probably use Squeeze if I want the optimal quality.

But, I personally plan to uprez my footage to HD resolution, prob. 720p; so it's better for ME to shoot in Letterbox/4:3 mode than Squeeze (according to Barry Green's book). So I need confirmation on one thing; hopefully you can tell me if you remember reading anything about this; is shooting Letterbox mode "better" than regular 4:3 mode? The first post by Greenlighted in this thread makes the assumption that Letterbox mode would yield slightly better image quality than 4:3 because less image gets recorded on tape; the rest is cropped out. Did Jan Crittenden mention anything about this theory? To me, I'm not sure if it makes sense, because the image isn't necessarily cropped; it just could be filled in with black bars -- so the same ammount of image still gets recorded to tape.
 
Gordon,

SQUEEZE produces a better 16:9 image, so it'll produce a better HD upscale. Not to say that any SD to HD upscale is pretty, but start with the better source footage and you'll get a better upscale.
 
Why not uprez the footage to 16:9 1080p and colorgrading and doing all post-effects in 1080p instead of SD. Then when you are to burn to a DVD, you use the 1080 source... gives a better compression result.
Also, just transfer the 1080p material into SD... or if you use Avid, change to SD and then do a Master tape.

Wouldn't that be even better then just uprezing to SD 16:9?
 
Going LetterBox means 25% Data Loss !!!

Going LetterBox means 25% Data Loss !!!

Achieving 16:9 from 4:3 CCD sensors is like covering the ocean with river volume. DVX100B is an SD Camera. There is absolutely no way of technically getting true or even closer to true 16:9 footage. Ofcourse there are certain wayouts to mimic that popular so-called 16:9 widescreen look.

1- Letterbox means 25% original data loss

2- Squeeze means implicit letterboxing + visually unaesthetic on 4:3 TV/monitors sets

3- (expensive) Anamorphic filter which utilizes the full CCD surface to achieve 12:9 i.e. full vertical resolution and about 75% horizontal to 16:9. Also anamorphic needs some skills to get the most out of it.

4- Convert SD to HD (via software or hardware) means introducing artifacts (may not look apparent) or noise.

Giving the above four options I would like to go with the third and fourth. i.e. Using anamorphic adaptor skillfully to get 12:9 original footage and then use HD software to make it 16:9 will introduce much less noise and artifacts to the uprised HD.

Any suggestions ? :-)
 
Joshua Provost said:
Gordon,

SQUEEZE produces a better 16:9 image, so it'll produce a better HD upscale. Not to say that any SD to HD upscale is pretty, but start with the better source footage and you'll get a better upscale.
Click to expand...

Joshua, to indirectly quote Barry Green (I don't have the book with me at the time), I believe he stated that uprezzing to HD from Squeeze footage is worse since you are basically uprezzing footage taht's already been uprezzed, whereas with normal letter box OR 4:3 footage, you get the regular unmodified footage to uprez into HD -- the results should actually be better this way.

But considering that Squeeze footage is stretched footage before DV compression, you may be right in this case. Looks like the only way to know is through testing.

Glazarus said:
Why not uprez the footage to 16:9 1080p and colorgrading and doing all post-effects in 1080p instead of SD. Then when you are to burn to a DVD, you use the 1080 source... gives a better compression result.
Also, just transfer the 1080p material into SD... or if you use Avid, change to SD and then do a Master tape.

Wouldn't that be even better then just uprezing to SD 16:9?
Click to expand...

This is what I've been asking, but I haven't yet gotten an answer. I'm not sure what you mean by changing to SD, and using a Master tape (on Avids) -- are you referring to an online / offline editing process, where you edit the offline footage first, than use the online HD footage (uprezzed)? Can't other NLEs do this as well?

I think one reasoning for not burning HD onto DVD is that it won't necessarilly be better; it may even be better to burn the non-uprezzed (non HD) footage, since you won't be uprezzing to such a high resolution, and basically compressing it back down into SD again -- when burning footage such as Squeeze or PhotoZoomPro, you burn what you get. This is hypothetical -- I think the only way to know is to test it. One camp says it's better to work with HD footage and downrez it to DVD, the other camp says it won't make a diff. -- it might even be worse.

And a third camp might say that it would be better if you actually had REAL HD footage, rather than uprezzed footage (which is not as good of course). Can anyone comment on this or have tested these options before? I've always wanted to finish my footage by uprezzing into 720p.

Glazarus; just a reminder, I think it's better to uprez to 720p rather than 1080p since you're not stretching the footage too much to a point where it starts looking noisy and artificial. It might even be better to uprez to an artificial resolution that is between 720 and 1080 (a sweet spot) since it doesn't matter if it's exactly 1080 or 720 -- when you burn it onto DVD it'll be SD anyway (it MIGHT probably matter if you display it on an HDTV though and you have some way you could view this footage, perhaps burning onto BluRay).
 
Last edited:
HaroonGhauree said:
Giving the above four options I would like to go with the third and fourth. i.e. Using anamorphic adaptor skillfully to get 12:9 original footage and then use HD software to make it 16:9 will introduce much less noise and artifacts to the uprised HD.
Click to expand...

Do you mean using the anamorphic adapter, and then uprezzing the 16:9 footage into 720/1080p? This is probably the best method, although I'm not sure if uprezzing is necessary and/or necessarilly considered "better". Read my post above this one.

But for those who can't afford the anamorphic lens, it's best to shoot Squeeze for 16:9. For HD uprezzing, if you need it (or if it's better to downrez HD to SD for DVD), then shoot Letterbox (slightly less compressed than 4:3, not really noticeable), and uprez that footage to 720/1080p using a PhotoZoomPro -- although Joshua Provist believes it is better to uprez from Squeeze (this issue is still undecided).

As for those who want 2.39:1 footage (the best results); I still have no clue. Probably the same as for 16:9, but cropping everything when you're done. Either shoot squeeze, crop to 2.39; or for HD, shoot Letterbox, crop to 2.39 (it might be better to crop to 2.39 before you uprez to HD since then, the HD has to do less work with less footage), then use PZP to uprez to 720/1080.

For anyone who knows, is it true that Letterboxed footage isn't really 16:9? I've heard this complaint come up a few times already. It's not just the EVF or LCD; they say it's not 16:9 aspect even in post.
 
I really don't understand then need of down-rising an HD footage when the question is about up-rising ...

Let me clear my POV here:
I've already told that "SD to HD is just like covering the whole ocean with the volume of a river ..." i.e. Data is literally low-resolution 720x (576 or 480) and getting 1920x1080 out of it is like performing a miracle ...

SD to HD must or should introduce more noise (artifacts) technically. But here I'm more interested in naked-eye viewing rather than being microscopic ...

There are two possible ways (shortest path for best results as longer path introduces more digital artifacts) to get HD out of SD ...

General Walk-through:

1-Select frame-mode of shooting:
a- You may shoot a letter-boxed (recommended way)
b- You may shoot a 4:3 full SD resolution
c- You may shoot with Squeeze mode (debatable, I wouldn't recommend it)

2-Get an Anamorphic Adapter and shoot with it (I need a technical advice from you guys on how get most out of Anamorphic Adapter) ...

3- Anamorphic adapter gives 12:9 footage i.e. much closer to 16:9 hence up-rising it to HD will introduce less artifacts/noise ...

4- For up-rising you may use:
a-A picture/frame level software like PhotoZoom Pro etc. In this case one needs to export the footage to a series of TIFF imgz. From there batch process them into the software.

b- A video-level software like InstantHD (RedGiant) to up-rise the footage.

(Note:It is still needed to find out which one produces better HD results. Until then I would go with option-b ... As more simpler n easy)

For no-Anamorphic use steps 1 & 4 ...

Any suggestions ...
 
Gordon JL said:
For anyone who knows, is it true that Letterboxed footage isn't really 16:9? I've heard this complaint come up a few times already. It's not just the EVF or LCD; they say it's not 16:9 aspect even in post.
Click to expand...

They are correct.

This really is not all that complicated. There are two independent variables: number of pixels, and aspect ratio:

SD (standard definition) has an aspect ratio of 4:3. That means that if the image is 4 feet wide, it will be 3 feet high. ED (extended definition) and HD (high definition) have an aspect ratio of 16:9.

SD has 720x480 = 345,600 pixels. The pixels are not square. ED also has 345,600 pixels. Those pixels are not square either, but their shape is not the same as the SD pixels. HD has more pixels, the exact number depending on which version of HD. I believe that HD pixels are not square either. I don't know whether pixels have the same shape in each variety of HD, nor whether their shape is the same as SD or ED. It doesn't really matter for now.

The point to keep in mind so far as your question is concerned is that the shape of an SD pixel is not the same as the shape of an ED pixel. A letterboxed image uses SD pixels, while a squeezed image uses ED pixel shapes.

In order to get a picture with an aspect ratio of 16:9 out of the DVX, you can:

1. Use squeeze mode, which produces an ED picture in the camera. That ED picture has 345,600 pixels. However, the camera generated it based on 259,200* pixels on the sensor.

2. Use letterbox mode (or, equivalently, mask the image in post), which produces an SD picture in the camera. That SD picture has 345,600 pixels. However, only 259,200 of them are not black. To get a 16:9 aspect ratio from the masked image, one has to discard the black pixels and uprez the remaining pixels. This is the same thing that the camera does in squeeze mode.

Either of those approaches will get an ED picture. One can then further uprez to get an HD picture.

Alternatively, if using method 2, one can do a single uprez step. This is certainly going to get you a superior image compared to using method 2 and doing two uprez steps. It is not clear to me how the image would compare to using method 1 and doing one uprez step.

The HD picture will have more pixels, and the pixel shapes may be different. The content of that picture will still be based on those 259,200 pixels on the image sensor, though. Converting to HD is only appropriate if you want an HD product (e.g., HD-DVD, or an HD quicktime movie). There is no point adding pixels only to take them away again later. In fact, it will most likely degrade the picture quality.

----
* If I did the math correctly! The exact number isn't important, though.
 
HaroonGhauree said:
3- Anamorphic adapter gives 12:9 footage i.e. much closer to 16:9 hence up-rising it to HD will introduce less artifacts/noise ...
Click to expand...

12:9 = 4:3.

What it does is squeeze a 16:9 image onto a 4:3 imager, and it's recorded as 4:3. You need to unsqueeze it manually in post.
 
this is a long thread for sure but there is a question unanswered which i wanna know

if my client wants the option of both end result being able to watch it as 16:9 format in normal 4:3 tv and widescreen tv, what are the viable/best options?
 
When you say the client wants it to be viewed on 4:3 and 16:9 TV's, do you mean full-frame on both... or letterboxed on 4:3 screens and full-frame on 16:9 screens? Shooting SQUEEZE and authoring a 16:9 DVD will achieve this. It will play back as 16:9 letterboxed widescreen on 4:3 TV's and full-frame on 16:9 TV's. That's what most people are looking for.

If they want full-frame on both 4:3 and 16:9 TV's (implying that you'll actually have more live picture area on 4:3 TV's, then you're only choice is to shoot normal 4:3. You'll have to create two versions, a full-frame 4:3 version, and another crop-and-zoom 16:9 version. Not a great workflow, and not the best quality.
 
Joshua Provost said:
When you say the client wants it to be viewed on 4:3 and 16:9 TV's, do you mean full-frame on both... or letterboxed on 4:3 screens and full-frame on 16:9 screens? Shooting SQUEEZE and authoring a 16:9 DVD will achieve this. It will play back as 16:9 letterboxed widescreen on 4:3 TV's and full-frame on 16:9 TV's. That's what most people are looking for.

If they want full-frame on both 4:3 and 16:9 TV's (implying that you'll actually have more live picture area on 4:3 TV's, then you're only choice is to shoot normal 4:3. You'll have to create two versions, a full-frame 4:3 version, and another crop-and-zoom 16:9 version. Not a great workflow, and not the best quality.
Click to expand...

I think most people will go with the first option of running letterboxed on 4:3 tv. And that's what I intend to do too... And what you mean here, the DVD player or the 4:3 tv will automatically make the picture fit into the tv with the letterbox? Means I only will have to author 1 DVD in 16:9 and play on both TVs.
 
wseng said:
And what you mean here, the DVD player or the 4:3 tv will automatically make the picture fit into the tv with the letterbox? Means I only will have to author 1 DVD in 16:9 and play on both TVs.
Click to expand...

Exactly, like most professional widescreen film in your DVD collection. It's one DVD with one version of the film in 16:9. It's the DVD player that converts it to letterbox as it outputs it in SD to your TV.
 
ok so when shooting 24p squeeze mode what is the best setting in fcp to import the footage into the software for 16x9? also even thou your shooting in sd do u want to use the hd timeline preset for importing or do you just use the regular sd? I just wanna make sure im getting the best quality importing to final cut then will run it thru instant hd.....
 
You must log in or register to reply here.
Back
Top