Loss of quality from super 16 to video

Robert Edge · February 13, 2005

I am planning to start a project in super 16 and complete the project in either standard or high definition video. I have a question on which I am receiving conflicting advice. What information will I have on the film that I will lose due to the transfer? In particular, will I lose contrast latitude? My understanding is that one of the reasons people shoot film is to maintain contrast latitude when the film is televised, but someone has suggested that this is wrong. If there is a loss in latitude, does this mean that when we are shooting the film we have to gear the lighting to the contrast latitude of video? Thanks for help on this question.

Alvin Pingol · February 13, 2005

You will lose latitude. Once the image is on video, you will not be able to digitally "recover" any footage that may be a few stops underexposed, and most certainly won't be able to do anything to overexposed footage. However, during telecine you will, allowing you to make adjustments while still maintaining quality, since these adjustments use information from the film itself - so make sure you get a proper transfer. I may even go as far as to say you don't have any latitude once on video, since there's no amount of under or overexposure that will still produce an acceptable image when corrected. (Well, maybe one or two stops under, but that's it!)

You will also lose color information and dynamic range, for the unlimited bit depth and high contrast ratios of analoge film must now be quantized into (not always, but commonly) 8-bit digital representations, which is not nearly enough to capture all of the information on film.

Depending on the final format, throw in some chroma subsampling and high compression ratios, and you lose even more.

If I were shooting on S16, I wouldn't waste the money on going straight to SD. It has been said that S16 has similar "resolution" to current HD formats, so an HD transfer would be logical. Moreover, one can make a great SD copy from an HD master, but not vice-versa!

(You can search the forum for more discussion re: latitude and dynamic range. They are not synonymous)

David Mullen ASC · February 13, 2005

Color negative has a wider exposure range than video can handle -- and that's a GOOD thing because you have the ability to adjust the image in the transfer and optimize it for video presentation. Plus real life has an even broader exposure range and what's nice about shooting on color negative and transferring to video -- versus shooting on video -- is that the film sort of "compresses" the exposure range of reality into the narrower one for film, and then you sort of compress it again for video to an even narrower range. This means you have a lot of flexibility when transferring color negative to video. You definitely do not have to light it flatter -- if anything, you have to light flatter if you were going to make a print off of the negative, whereas if you transfer from the negative, you have more information to play with.

Alessandro Machi · February 13, 2005

You will lose latitude. Once the image is on video, you will not be able to digitally "recover" any footage that may be a few stops underexposed, and most certainly won't be able to do anything to overexposed footage. However, during telecine you will, allowing you to make adjustments while still maintaining quality, since these adjustments use information from the film itself - so make sure you get a proper transfer. I may even go as far as to say you don't have any latitude once on video, since there's no amount of under or overexposure that will still produce an acceptable image when corrected. (Well, maybe one or two stops under, but that's it!)

You will also lose color information and dynamic range, for the unlimited bit depth and high contrast ratios of analoge film must now be quantized into (not always, but commonly) 8-bit digital representations, which is not nearly enough to capture all of the information on film.

Depending on the final format, throw in some chroma subsampling and high compression ratios, and you lose even more.

If I were shooting on S16, I wouldn't waste the money on going straight to SD. It has been said that S16 has similar "resolution" to current HD formats, so an HD transfer would be logical. Moreover, one can make a great SD copy from an HD master, but not vice-versa!

(You can search the forum for more discussion re: latitude and dynamic range. They are not synonymous)

<{POST_SNAPBACK}>

I have re-color corrected 50 year old 8mm film after it was one-light transferred to video. It's surprising what information is actually there but buried in the lower end blacks, even after it's been transferred to video. I've readjusted and brought back the lower end detail whether the film transfer was on VHS or BetaCam SP.

One of the advantages of higher end analog editing (such as BetaCam SP), is it has very very mild compression and as a result holds information in the blacks better than many "affordable" NLE codecs that seem to fall apart in the very low end of the black scale.

-------------------------

SD can look perfectly fine on HD, it can look "improved" on HD as long as you don't change the aspect ratio.

Alvin Pingol · February 13, 2005

SD... can look "improved" on HD
<{POST_SNAPBACK}>

Unless the SD footage is lightly, but preferably un-, compressed, doesn't suffer from extreme edge enhancement, and is as sharp as can be for SD, I imagine an upconvert would only emphasize the shortcomings of the lower resolution format. For example, the level of detail computer monitors can resolve make all DVDs appear more compressed than usual, and also a little soft - things that are hidden by SD monitors, or television sets.

Personally, I don't like the way SD looks on HD TV sets. I'd much rather watch that same SD signal on an SD monitor versus on an HD TV of the same size.