Just a question... I know DV is compressed video but i was wondering, If i have a lap top on had while shooting, and have the firewire plugged into the laptop capturing in real time to my hard drive, would this by-pass compression (Probably not... but i'm curious).

nope.

sadly, no. the compression is needed both to fit enough video on a tape, AND because EVEN firewire can't transfer THAT much data at once. an uncompressed dv size frame would be several megabytes.

That's too bad... I thought I was onto something.

uncompressed video requires special cameras(the new "consumer" HD camera works), and special computer capture devices and raid drives. The problem with uncompressed, or partially compressed video on a laptop is that the hard drive cannot write the information at the speed it is being brought in, (over hundreds of megabytes a second) also, a laptop's hard drive would fill up after a few minutes of video were captured in an uncompressed form.

Yes its true - and all of the points made here hit it head on. Heres a few issues -

1: the internals of DV cams compress it to 4:1:1 before it gets to either the tape or the firewire jacks.

2: Like mentioned above, firewire cant handle the throughput of HDV. The new firewire format i believe is built for it.

3: Even if it could, the data rate of 720p HDV is about 19Mbps, and 1080i is 25 Mbps. Good luck getting a laptop to capture it without drops. But for the record you can capture direct to a laptop if it is well equipped and skip tapes completely. Theres even software designed for it, its called DV Rack. They adverise in DV all the time. Apparently it allows for some real time footage analysis too. Swizzle.

The 4:1:1 isn't really DV compression, it's standard RCA and coaxial video signal for the channels of information.

DV compresses the signal 5:1, i think. DigiBeta compresses it 2:1. I think HD tape is the only thing that captures 1:1.

okay ... misinformation:
the new "consumer" HD cameras are highly compressed. they use mpeg2 which is 4:2:0 color space. Panasonic has a new HDV codec that can go over firewire, but it's still a compressed signal. For reference, uncompressed 601 standard def is 25 megabytes and 10 bit uncompressed HD is 125 megabytes.
4:1:1 is the DV compression and it has nothing to do with RCA (the connector and the company) or coaxial video signal. But, 5:1 is also the compression for DV-25 (i.e. miniDV). What? You're talking apples and oranges. 4:1:1 is how the color space is compressed (4 being luma "Y" and the ones are Red minus Y and Blue minus Y). So, that is the compression. But, if you compare the data rate of 601 uncompressed video (about 24 MB/s) to the data rate of DV-25 (just under 4 MB), you get your 5:1 compression ratio. (yes, I know the math does add perfectly and I don't know why)
I know I link him a lot, but if you check out www.adamwilt.com , it gives a excellent description of all the different formats, and the difference between the various color space compression formats.

Well put joren. Thanks for the clarification.

So joren, I was talking about signal compression with a telecine technician just last week as i was transferring stuff, and as far as i know he knows his stuff.

So we were talking about colors, how the reds looked way better on the print thant he transfer, and i asked if with new HD chips used in the machines, if that would change any of that. So he told me why it might change, but not much:

He explained that the reason standard video uses 4:1:1 for the color rendering is because that is all the electrical cable could handle as it was being sent to the monitor, meaning that RCA and coaxial.

I think it was something like: 4 layers of luminosity at 8 bits, plus one layer of red minus y and one layer of blue minus y, which totaled 48-bits of color information.

(Which incidentally is considered the high end of consumer photo scanners (48 bit color rendition. i think photoshop only reads up to 48 bits anyway)

He then said that cables like S-Video and component could handle 4:2:2, meaning 64-bits of information, because at that end, they are used to transfer digital data instead of electrical data such as RCA and coaxial cables. Meaning they can transfer more.

HD signals, broadcast and HD tape when you're recording off an HD camera (don't know about the prosumer stuff) is recorded at 4:4:4, giving you 96-bits of color information. And that is why it looks amazing to our eyes at home. And that is the standard that the U.S. will be switching to.

He said that 4:4:4 monitors and decks can read 4:1:1 or 4:2:2 signals, since they have less information, but can't improve on the quality of the signal since it can't just make up the lost bits.

The transfer machines use an 8:8:8 color profile to read the negative during the transfer, but they only do so internally. Meaning the conrtoller, the transfer machine, and the software doing the color correction all read a 192-bit color profile, but, since the highest quality signal we can record today is HD, and that is a 4:4:4 profile, you will always lose some information, and so the colors will never look as vibrant or as good as the negative until video moves on to 8:8:8 or larger.

Is that correct or incorrect?


So as I understand it:

Color profiles are a part of the Data transfer rate, but the data transfer rate also includes resolution, timecode, blackburst, and eh, probalby some other stuff. When talking about 601 uncompressed or miniDV compression, it is talking about ALL the information on the tape.

When talking about color profiles: 4:1:1 is broadcast standard for now, 4:2:2 is DigiBeta and DVD, 4:4:4 is HD and also the highest we can record at the moment.

That sound about right?

See, these sites sort of explain it in terms of video signal bandwidth:

http://www.videohelp.com/glossary?0#4:2:2,%204:4:4%20and%204:4:4:4

http://www.tvtechnology.com/features/Tech-Corner/f-RH-4.2.2-07.10.02.shtml

See? 4:2:2 was introduced with component video signals, because 4:2:2 is more information than a composite (RCA) video signal can send

What about these new hard drives? At http://www.focusinfo.com/products/firestore/firestore.htm any input? has anybody used them? seems like a good idea to me.

It doesn't really have to do with the harddrive, it's your camera.

The camera will always record at whatever compression it is programmed to record to the format it is made for (miniDV, Beta, HD). The harddrive will record whatever signal it recieves from the camera or deck. So it will record a miniDV signal just as well as it will record an HD signal.

A harddrive is just recording the information, and doesn't really need understand it or interpret it, since they only record bits. Your program has to be able to read and interpret it, and it will only interpret/read what it is programmed to read (some programs can't read HD signals or 601 uncompressed video)

I think we need to clear up a few things.

First of all, there's a difference between color space and compression. Color space is simply a reference to frequency of sample. Bit-rate is a reference to the fidelity of that sample. The correlating numbers are independent of eachother. Something with a 4:1:1 color space could be designed to sample at 16-bits.

In other words, the quality of color in something sampled at a 4:1:1 ratio can be the same as something sampled at 4:4:4. The human eye percieves value differences much better than color differences. This is why delivery on 4:1:1 is fairly acceptable.

For a multitude of reasons we in the world of cinema like to modify our imagery in post, which is why we like to keep a bunch of information around that might otherwise be imperceptable to the average human viewer (And one of the reasons why film is still the gold standard).

quote:

---

Originally posted by RFranco:
HD signals, broadcast and HD tape when you're recording off an HD camera (don't know about the prosumer stuff) is recorded at 4:4:4, giving you 96-bits of color information. And that is why it looks amazing to our eyes at home. And that is the standard that the U.S. will be switching to.

---

I think you may have some bad information. First of all, you can't really throw around the term HD like you would SD(SMPTE's definition was even changed in favor of ambiguity). While there is a broadcast standard, the world has yet to really embrace any set of numbers as really being HD (In terms of resolution, or color). Regardless, the only HD systems currently offering 4:4:4 color space are HDCam-SR and raw data recorded straight to hard-drive (Sound familiar Diego? They have to use fiber-optic cable, though). This only represents a drop in the bucket of today's HD workflow. Not only that, but many HD systems are 10-bit, 12-bit or even 16-bit, so the numbers don't add up.

Sorry if this is more confusion than clarity.

Nota "16-bit 'o honey" Mono

Actually, NotaMono that makes sense, about the different HD. I think it's just that our school whenever our teachers talk bout HD they talk about whatever the highest end is since they're all learning that stuff right now in L.A.

Almost all our teachers hang out at Panavision down there, and Panavision is really pushing the HD systems and having their optics work with the cameras.

So here's another question someone might know the answer to:

This prosumer HD, is it called HD because of the high resolution it records at, but lacking any of the other aspects of HD such as the high contrast ratios and color difference? Is that why they're able to make it so small and cheap?


And I thought the FCC already decided on what the specs for HD were going to be for the U.S. and that's why they wanted to change broadcast standards by like 2012 or something like that. Lemme go find articles.