Modifying scene files; do it in post?

[evolvnyc]

I'm shooting in 24p (s5) and am going for the most "film like" feeling. After reading Barry's book, I decided to alter the settings a bit, I switched the detail freq to thin as i intend to output to film, and set the detail level to -1 (from -3 I think). As a professional photographer, the rule of thumb when shooting digital is to do any softening or sharpening, color exaggeration, and contrast in post. I like to look of the default S5 but I worried about it being printed to film and naturally going even softer, thus making it twice as soft (following me?) So my thinking is if I have a bit more detail when shooting, when printed or shown on a large screen it will naturally fall a bit softer (to look about like th default -3 setting) Am I way off base here? Does anyone have experience in modifying the S5, what did you find and did you go out to film? I am also shooting in squeeze as I am mostly working alone and can't deal with the added headache of calculating anamorphic equations...

So if my theory is correct, are there software programs I can push the footage through which will bring a more "film look" ie. saturated and soft later? Thoughts?

 

[Scottdvx100]

As with still photography you want to minimize the amount of processing the camera does so you have the option of determining what you want to do and how much in post.
If there's too much sharpening (problem with lower cost video cameras) then you get a ringing effects around the image, especially on film.
Best bet is if you're truly planning to go to film contact some of the companies hat do this and get their advice.
It would also be worth shooting some short tests and having them go through the pipeline to film so yo ucan be sure of your settings.

 

[Matt Grunau]

I'm shooting in 24p (s5) and am going for the most "film like" feeling. After reading Barry's book, I decided to alter the settings a bit, I switched the detail freq to thin as i intend to output to film, and set the detail level to -1 (from -3 I think). As a professional photographer, the rule of thumb when shooting digital is to do any softening or sharpening, color exaggeration, and contrast in post. I like to look of the default S5 but I worried about it being printed to film and naturally going even softer, thus making it twice as soft (following me?) So my thinking is if I have a bit more detail when shooting, when printed or shown on a large screen it will naturally fall a bit softer (to look about like th default -3 setting) Am I way off base here? Does anyone have experience in modifying the S5, what did you find and did you go out to film? I am also shooting in squeeze as I am mostly working alone and can't deal with the added headache of calculating anamorphic equations...

So if my theory is correct, are there software programs I can push the footage through which will bring a more "film look" ie. saturated and soft later? Thoughts?

I have been taking pictures, videos, all of it for years. The number ONE fundamental rule I have with ANY medium that I have found is the simplest. If you have an idea what the final product will look like, and you know you can do it in post, do it in post! I can shoot in 60i and with a few tips I have learned through after effects make it look any way I want. Want to look like film? Well there are ways to do that. Want to enhance edges? Contrast? Saturation? Do it in post. Want to make your make look like it was shot on film? Then get the best possible base footage you can, white balanced, exposure settings, and simply get the video. If you let the camera make choices and the deciding factors on what the video will look like once capture, you are stuck with that look, and you can do nothing about it. There are people I know who can take film shot at 60i and make it look damn near identical to the DVX100's 24p or 24pa mode. It's all in understanding the capabilities of your post software and how to use it to your best advantage.

As a professional photographer, you statement "rule of thumb when shooting digital is to do any softening or sharpening, color exaggeration, and contrast in post." Is exactly on point.

I know I will have some die hard DVX users jumping on me for much of what I have said, but I stand by it, as that is the way it is done in every movie studio around. Get the best footage you can, and then tweak the hell out of it (barring things like shutter speed and slow mo which are a more difficult to fake) in post.

I bought the DVX 100a, not so much for its cinema settings or even 24p. I bought it because it is the best camera for sound, and has great color representation and video quality. I personally don’t give a hoot what the Cinema-D setting looks like, because I can replicate it in a post program.

Here is a good example of a nice cheat using 60I to look more like film. I have used it and in conjunction with some Curves and Levels in After Effects, have gotten surprisingly impressive results,

http://www.dvinfo.net/articles/filmlook/broadway1.php

Read or don't read the other info, but look at the use of displaced interlaced fields. It may seem trivial, but there are lots of ways to get a great film look, without jumping through hoops. The point of the link is, not that it is perfect, but it works.

Also, if you are shooting in Squeeze, why? Why not shoot normally, and make any letterboxing, 16x9 effects in post? You are not getting any more pixels on tape with squeeze or any other setting (in fact with squeeze you are only getting 75%), so you could simulate that look afterwards, if by only having your software alter the pixel aspect ratios, or by changing small degrees of scaling. Although squeeze does cut a few corners, they can be addressed elsewhere without losing anymore info than needed. The important thing being you still have the option of having untouched original footage (which you can do with what you like), as opposed to having something on tape that is unchangeable because that is the way it came out of the camera.

Again, I love this camera. I love it for its audio capabilities and for its video quality. I am NOT knocking that. I am simply trying to say that using in camera processes which can be done in post leaves you with fewer end options.

 

[MovieSwede]

One issue you must think of is that miniDV gives you compression artifacts and so on. Wich will make postproduction a bit harder than if you would shoot raw-video (4:4:4)

So its seems that get the right image on the spot will create the best picture.

On the other hand. I like myself to have full controll in post production (im a better editor than cameraman, and that doesn´t say am I good editor )So im looking myself for the ultimate DVX setting for postproduction.

As I understand DVX is actually shooting details -7 and everything above that is in-camera sharpening. So you actually isnt softening the image at -7. I could be wrong on that issue.

As for shooting squeeze mode. You will actually benefit from that becuase you can spend 100% of the bandwith on the image as for letterboxing you can only spend about 66% bandwith on the actuall image.

 

[Jack Daniel Stanley]

One issue you must think of is that miniDV gives you compression artifacts and so on. Wich will make postproduction a bit harder than if you would shoot raw-video (4:4:4)

So its seems that get the right image on the spot will create the best picture.

On the other hand. I like myself to have full controll in post production (im a better editor than cameraman, and that doesn´t say am I good editor )So im looking myself for the ultimate DVX setting for postproduction.

As I understand DVX is actually shooting details -7 and everything above that is in-camera sharpening. So you actually isnt softening the image at -7. I could be wrong on that issue.

As for shooting squeeze mode. You will actually benefit from that becuase you can spend 100% of the bandwith on the image as for letterboxing you can only spend about 66% bandwith on the actuall image.

I think this is page 102 in Barry's book:
anamorphic adaptor = 500 lines of resolution
letterboxing in camera or in post and/or uprezzing to 16X9 in post = 350 lines of resolution
squeeze = 330

Squeeze does NOT use the full processing power of the CCD's - only about 75% just like letterboxing, because squeeze doesn't actually work by squeezing the image horizontally, but by cropping the top and bottom then pulling -strecthing rather than squeezing- the image vertically so that the black bars are pushed off the screen. It's only working with the same portion of the screen as letterboxed footage.

The reason that upfuzzed or letterboxed footage is higher resoltion than squeeze is because the computer has a cleaner starting point to to crop/uprezz.

Use squeeze if you have limited time in post (like a 72 hour film contest maybe) or limited processing power. Otherwise see the sticky posted in the techincal section on uprezzing 4:3.

It may seem counter intuitive, but thinking that squeeze uses all of thr CCD's for the image is a big misconception that I see posted all the time. So much so that I've unintentionally memorized the page number that I cited above.

Everyone should really get Barry's book -- its a very valuable and accurate tool.

 

[Jay Rodriguez]

Everyone should really get Barry's book -- its a very valuable and accurate tool.

I totally agreee with that. It's helped me learn so much quicker.

 

[MovieSwede]

jdanstan im aware that you cant use the CCDs more at squeeze mode because of the 4:3 framing of them. But one thing U cant take advantage of att squeeze mode is the miniDV compression.

If you shoot letterbox about 33% of the footage will be black. That means 33% of 25mbs (more or less) will just record black frame. So 16,65 mbs will be used for the actual frame (sound isnt in my calculation so it will error some)

If you shoot squeeze you can use the entire MiniDV bandwith on the image. The full 25 mbs.

 

[Jack Daniel Stanley]

Ohhhhhhh

Ohhhhhhh

jdanstan im aware that you cant use the CCDs more at squeeze mode because of the 4:3 framing of them. But one thing U cant take advantage of att squeeze mode is the miniDV compression.

If you shoot letterbox about 33% of the footage will be black. That means 33% of 25mbs (more or less) will just record black frame. So 16,65 mbs will be used for the actual frame (sound isnt in my calculation so it will error some)

If you shoot squeeze you can use the entire MiniDV bandwith on the image. The full 25 mbs.

Ohhhhhhhh right! I had no idea what you meant by bandwidth. I was trying to apply bandwith in the shootingprocess. :cheesy:

(EDIT) Hey wait a minute -- don't you still get that if you use photo zoom pro to crop the letterboxed footage -- thereby making squeeze not the way to go because once you crop letterboxed footage in post (so the frame is only teh size of what's between the black bars) then you get all the bandwidth agility of Squeeze, without the rez loss, plus the ability to re-frame some in post??? :undecided

 

[MovieSwede]

Not really because you still would have the artifact problem even if you get a better resolution.

The best way would be to make a testfootage. Were U shoot letterbox and squeeze. Then uprez letterbox to anamorph with photozoom.

That would give best way to see what footage gives the most pleasing image.

 

[evolvnyc]

wow guys, i admit my eyes have glazed over and i have completely lost track of what we're saying here... let me see if i get it now... squeeze, while cropping and thus lessening quality, uses the entireccd to record whereas letterbox records just black on much of it, so when i print to film, is there going to be much of a noticable difference? I know ideally I would do a test, but the thought of that is so overwhelming and complicated that i doubt my ability to do it. So does it really make a difference? and if it does, as 1/3 of my film has been shot in squeeze, would it be noticibly different if i shot the rest of it with the anamorphic or in ketterbox? would my editor hate me for having to intersperse and match different footages?

 

[Barry_Green]

Shooting Squeeze, vs. blowing up from letterbox, would make a perhaps small difference in a film transfer. I'd say you'd be better off blowing up from letterbox footage than from squeeze footage.

 

[Jack Daniel Stanley]

... squeeze, while cropping and thus lessening quality, uses the entireccd to record whereas letterbox records just black on much of it, so when i print to film, is there going to be much of a noticable difference? ... as 1/3 of my film has been shot in squeeze, ... would my editor hate me for having to intersperse and match different footages?

Squeeze and letterbox both use the same percenatge of the ccd's.

Squeeze is worse than letterbox - period. Although I understand people use it and get great reults.

For 16X9 the Anamorphic adaptor is the only thing that uses all of the ccd's and gives 500 lines of resolution in 16X9, because it truly SQUEEZES the image OPTICALLY.

Think of SQUEEZE MODE as shooting a 4:3 image optically then blowing it up digitally in camera. The camera did use all of the CCD's to capture the 4:3 image, but now must throw away the top and bottom as it STRETCHES from top and bottom to the point where the portions of the screen that would contain the balck letterbox bars are no longer visisble, -- just like letterbox --. Letterbox blowups have slightly higher resolution because they have a cleaner starting point and your handy dandy computer is better at digitallly anamorphing it than your camera is.

The difference in letterboxed footage and squeezed footage (330-350 lines of rez) is not as drastic as the difference between these two options and the anamorphic adaptor (500 lines of rez). Like Barry said (of course) the difference between letternbox and anamorph is smaller. But the difference between those and Anamorphic Adaptor is big.

Supposedly the way to go for big screen is the anamorphic lens. And yet the DP that won the Cinematography Award at Sundance on the DVX opted not to use it - after testing - because she really like the leica lens and didn't want to put another piece of glass in front of it and because it's really hard to work with.

If you continue to shoot in squeeze you will make your editor's life easier in that the footage will match, but I think mostly because he'll have less to do in post; working with photo zoom pro is tedious.

If there are a lot of panormaic exteriors or super wide shots of the set or location where you might need every bit of resolution you can get, you might consider shooting 4:3 for those "production value shots" (fixing the aspect ratio in post) and using Squeeze the rest of the time.

No question you will make everyone's life easier if you stick with squeeze. I think going to Anamorphic Lens is not an option now becuase it wil out rez the other footage. Unless you happen to have shot only close ups, in which case you might be able to use it for wider production value shots ... Barry?

Real question is: is the smaller resolution difference in letterbox and squeeze worth switching gears for midstream? ... Barry?

Also is this a documentary or, and intimate tale, or an action extravaganza. If it's a documentary, then I think the above questions are mute. Stick with the way you started.

 

[Jack Daniel Stanley]

Software

Software

So if my theory is correct, are there software programs I can push the footage through which will bring a more "film look" ie. saturated and soft later? Thoughts?

Lots of people like Magic Bullet. I definately use it. I really like Graem Nattress's film effetcs however:
http://nattress.com
It's super cheap ($100) and you can download it and start using it right away. Plus the customer service is amazing. If you have questions, you are not emailing some customer service rep, you will get a reply from the software designer and CEO himself, Mr. Nattress, usually within the day. He is also a member of DVXuser.com
Add Shake into the mix and you can compete with Peter Jackson and Lucas.

What I hear from, most people is get a quality image and stylize it in post. This is definately the way I work, and yet the DP for Novemeber, the film that won the Sundance Cinematography award, who has severakl stylized nearly monochome color setions in her film achieved her look by white balancing and using filters on set. She says that color always looks forced in post and hated the few sections they had to re-color and move around for narrative reasons in post.

I think real hardcore seasoned DP's get what they can in camera when time allows- optical is generally better than digital - but for the rest of us there is color correction, magic bullet and Graeme Nattresses G-film effects, and they are excellent and flexible solutions.

 

[Barry_Green]

Squeeze is worse than letterbox - period. Although I understand people use it and get great reults.

Squeeze is for a different purpose than letterbox. Squeeze is for making 16:9 footage. Letterbox is for making 4:3 footage that looks like it's 16:9. It's not that one is "better" or "worse" than the other; it's that they're for different purposes.

Think of SQUEEZE MODE as shooting a 4:3 image optically then blowing it up digitally in camera. The camera did use all of the CCD's to capture the 4:3 image, but now must throw away the top and bottom as it STRETCHES from top and bottom to the point where the portions of the screen that would contain the balck letterbox bars are no longer visisble, -- just like letterbox --. Letterbox blowups have slightly higher resolution because they have a cleaner starting point and your handy dandy computer is better at digitallly anamorphing it than your camera is.

This is an accurate description. For film blowup, you may get slightly better performance from letterbox mode because both will need to be greatly magnified, but letterbox has a cleaner "starting point". For a 16x9 DVD though, I think squeeze is certainly a reasonable mode to use.

Supposedly the way to go for big screen is the anamorphic lens. And yet the DP that won the Cinematography Award at Sundance on the DVX opted not to use it - after testing - because she really like the leica lens and didn't want to put another piece of glass in front of it and because it's really hard to work with.

For ultimate resolution, you can't beat the anamorphic adapter. But it is definitely a lot more hassle, and a lot more limiting, than not using it.

No question you will make everyone's life easier if you stick with squeeze. I think going to Anamorphic Lens is not an option now becuase it wil out rez the other footage. Unless you happen to have shot only close ups, in which case you might be able to use it for wider production value shots ... Barry?

That's one place where the anamorphic adapter's advantages can easily be seen, is wide shots. Squeeze works excellently on close-ups, and the anamorphic performs at its best during wide shots. Outdoors wide shots let you zoom out and stop down the iris, two things the anamorphic needs in order to deliver sharp images. Closeups usually require you to zoom in and interior shots require a more-open iris; zooming in and opening up are a recipe for lousy shots with the anamorphic.

So yes, if you wanted to mix squeeze and anamorphic, that would be the way to do it -- use the anamorphic for wide shots, and the squeeze mode for your closeups.

Real question is: is the smaller resolution difference in letterbox and squeeze worth switching gears for midstream? ... Barry?

If you've already been shooting with squeeze, there's no reason not to continue. Just MAKE SURE you're using THIN line detail!

 

[Jack Daniel Stanley]

Squeeze is for a different purpose than letterbox. Squeeze is for making 16:9 footage. Letterbox is for making 4:3 footage that looks like it's 16:9. It's not that one is "better" or "worse" than the other; it's that they're for different purposes.

I meant that letterbox was higher resolution than squeeze, as per your book, and that uprezzing from letterbox would look better than squeeze for a big screen TV or for film blowup because:

...letterbox has a cleaner "starting point". For a 16x9 DVD though, I think squeeze is certainly a reasonable mode to use.

Would the difference not be noticeable a big wide screen TV, or a hi def monitor?
I'm planning on using disjecta's work flow - uprezzing 4:3 to letterbox in post for anything that "matters" (festival submission) and then outputting it DVD, but using squeeze for home movies and if I'm pressed for time in post (like this weekend I'm doing a 72 hour film contest) then outputting to DVD. Does photozoom pro not buy enough resolution to make it worth it for DVD output?

...If you've already been shooting with squeeze, there's no reason not to continue. Just MAKE SURE you're using THIN line detail!

Should you also use thin line detail if you are going to use djecta's magic and uprezzing to 16X9 in post?