10 bit vs 8 bit uncompressed

Would anyone care to explain to me the difference between these 2 forms of compression. I always use 8 bit, never had issues but just want to know the difference. thanks.

The end user will almost always be seeing 8-bit, and your video sources are
probably 8-bit as well, unless you have a very nice camera. 10-bit video
offers more detailed shades of color. For example, if there is a very dark
area in your image, 8-bit might just show black, whereas 10-bit could show
some detail. Of course, if you deliver your final result in 8-bit, then you'll
need to somehow brighten those areas to bring them into the 8-bit range.
It's about the amount of data you have available to represent your image.
Let me try a metaphor. Bit depth is like image size. So, for example, let's
say you shoot a video at 1080 and you are delivering it at 720. It makes
sense to stay in 1080 until the very end, because all your details will be clearer,
and the approximations more accurate as you process. Then, at the final
output stage you can make the compromises inherent in reducing the size.
If you have 10-bit footage (i.e. DigiBeta), and/or you're planning to do a lot of color correction,
effects, and gradients go with 10-bit. You'll get better results.
On the other hand, if you're a beginner working with basic 8-bit footage,
stick to 8-bit. It will work just fine for now.

Similar Messages

How to Export UYVY 4:2:2 8-bit Uncompressed AVI?

I've been doing a lot of testing in Premiere Pro CS5 (on Windows) but cannot figure out how to export video to a generic 4:2:2 8-bit uncompressed AVI file. Here's some background - I've captured some video to an uncompressed Matrox AVI file (via an MXO2 I/O Device). Now I've done some simple cutting and need to get the video off Premiere's timeline (and into another yet-to-be-determined multimedia application). I don't want to lose any quality - hence my requirement for 4:2:2 uncompressed AVI. Also, the file needs to be a standard/generic AVI that doesn't require any special CODECs installed.
In Premiere's File > Export > Media window, there seems to be 2 ways of getting uncompressed 4:2:2 AVIs:
Option 1:
Format: Uncompressed Microsoft AVI
Video Codec: UYVY
Quality: 100
Width: 720
Height: 576
Frame Rate: 25
Field Type: Upper First
Aspect: PAL (1.0940)
Audio Codec: Uncompressed
Sample Rate: 48000 Hz
Channels: Stereo
Sample Type: 16 bit
Option 2:
Format: Microsoft AVI
Video Codec: Uncompressed UYVY 422 8bit
Quality: (Greyed out at 100)
Width: 720
Height: 576
Frame Rate: 25
Field Type: Upper First
Aspect: PAL (1.0940)
Audio Codec: Uncompressed
Sample Rate: 48000 Hz
Channels: Stereo
Sample Type: 16 bit
Audio Interleave: 0
Option 1 is great except that the audio codec detected is "0xfffe extensible wave format" (according to GSpot). I can play the video on practically any machine including something as old as Windows Media Player 9 running on Windows 2000! The audio plays back fine as well in Windows Media Player. However, audio does not play (not detected correctly) in VLC media player, or in Sony Vegas-type software. FYI, the video codec detected (by GSpot) is "UYVY (packed 4:2:2)".
Option 2, on the other hand, is quite the contrary. The video does not play in any version of Windows Media Player (regardless of operating system), but the audio plays fine though. The audio codec detected (according to GSpot) is the standard "PCM audio" format and the video codec is again "UYVY (packed 4:2:2)". In VLC media player, both the video and audio play fine. But in other editing software such as Edius, the video is not detected and only the audio is loaded onto the timeline.
What I would like is a generic 4:2:2 uncompressed AVI file which has audio that works in anything. Is there some setting/parameter that I have missed? Can I customise Option 1 with "PCM audio" rather than "0xfffe extensible wave format"? I want to avoid "Video Codec: None" if at all possible given that the file size will be significantly larger for absolutely no quality gain (I presume this is uncompressed 4:4:4?).
Any suggestions would be most welcome.

Hi Jim, I too am confused!
Why should the video codec for Option 1 be V210? Isn't this reserved for 10-bit video?
http://www.fourcc.org/yuv.php#V210
http://help.adobe.com/en_US/premierepro/cs/using/WS398C372B-B748-4ef6-BA0D-396D2E360707.ht ml
I would have thought that the UYVY option was the correct choice for 8-bit video.
http://www.fourcc.org/yuv.php#UYVY
Videos exported using Option 2 definately do not play correctly in Windows Media Player - I get no video, just audio. I have tried many different computers running different versions of Windows and Windows Media Player. I'm running Premiere CS5 v5.03 if that makes any difference.
I don't understand why videos from Option 1 play fine in Windows Media Player even though the systems don't have AJA hardware installed (or any video I/O device) - they are just plain vanilla office PCs

QT won't read 10 bits uncompressed HD file

I think i read somewhere that QT on windows does not support 10 bits uncompressed ( only 8 bits ), anyone knows if there a plug-in i can d/l in order for it to work ??
My file is 10 bits uncompressed 1920x1080, it plays fine on a mac. Thanks

That part was fine but I cannot get the drivers to load from my install disk (10.4.8).
The Mac OS X 10.4 disks don't contain Boot Camp drivers. Insert a Mac OS X 10.5 installation disk and get the drivers from it.
(42155)

Gamma Shift After Rendering 8 Bit Uncompressed

I'm getting a really bad black level shift after I render in Final Cut 5. I'm using the final cut 8 bit uncompressed codec, which is what the footage was captured at. Before I render, the preview quality looks fine, but once I render it out, it looks awful, the blacks get a lot brighter, to the point where there's no detail at all. I've changed all my video output settings in the render settings to all the settings possible, and nothing works. I use an AJA iola for video output/capture, but I know it's not that becuase I hooked my sysem up through a normal firewire setup and it does the same thing. It seems to do it everytime I render, and it doesn't matter what kind of fiter or change I'm making to the video. Any help would be greatly appreciated. Thanks.
2gig dual core G5, OS X 10.4.5, Final Cut Pro 5, AJA iola

ProRes is not a long GOP structure (it is not HDV). It is an all "i" frame with compression while maintaining a 4:2:2 color space. Its getting very good reviews including it's ability to work in Color and other grading/finishing options.
Just for your own piece of mind, why not get a segment in both formats and do some testing. That's the only way you'll know for sure the best workflow for your interests.
Good luck,
x

Saving a mov file as a 24 bit uncompressed avi

I have a MOV file that I created in LiveType. I need to convert it to a 24 bit uncompressed avi - I tried in Compressor (v3.0.5) but couldn't find a way. I tried in Quicktime Pro and saved it as an uncompressed 36bit avi, but don't know how to reduce it town to 24bit -- is it just reducing the colours to thousands or something?
Argh!

Why do you need to make it to that specification?
Tell us what you are trying to do -what is the recipient going to do with the file?
Maybe we can help you find another way to do this that will be acceptable to your client.

Worth the effort? 8 bit AVCHD 1080p 4:2:0 versus 1080i 10 bit uncompressed

Hello,
I am i homing in on a final MacBook Pro versus Mac Pro decision to enable both real-time capture (argues for portable) and final cut - color - motion work flow.
i have a new sony HXR 5NU that provides me with 25 Mbps 1080p AVCHD 8 bit 4:2:0 from the camera or 10 bit uncompressed 1080i 4:2:2 via SDI or HDMI. The latter would involve a MacBook (not lugging a 45 pound Mac Pro around on the set) and an external MOTU HD Express HDMI or or a Matrox O2 LE.
The key question is if the extra information is significant in terms of final look, constructing quality masks, etc. I believe the answer is yes as it relates to simply comparing 8 bit 4:2:0 versus 10 bit uncompressed 4:2:2. Seems the answer is murkier with the trade-off of 1080i (the uncompressed signal) versus 1080p from AVCHD.
I do not know the internal workings of the camera, it is possible that it is 1080i off of the sensor electronics and a pull down is happening with the avchd encoder which means that it is not really a trade-off.
Any opinions?
Otherwise I just focus on getting the best out of my camcorder's avchd output using a waveform monitor and go for a Mac Pro in Post.
Any thoughts on this?
Wayne
Portland

Thoughts or opinions? Remember, you asked for it.
You haven't done the math.
You do seem to realize there might be some kind of quality difference between 4:2:0 AVC and 10 bit uncompressed, which is a good thing, but haven't dealt with the larger issues.
First off, while FCP might deal with AVC without too much fuss, it is not an approved codec for use with COLOR, and you will be dealing with a transcode at some point.
If you are dealing with some kind of sub-cinema format camera, which is producing AVC, then the practical reality is that its sensors are indulging in a serious amount of voodoo, and if your project is going to involve anything like a lot of greenscreen, then you are also buying yourself a certain amount of downstream pain, since the bandwidth that an actual 10-bit 4:4:4 camera produces is what you might really want. But that brings us to the real math.
Your first paragraph contains the phrase:
MacBook Pro versus Mac Pro decision to enable both real-time capture
Uh.... I guess you'd also be looking for some other kind of AJA/Io or whatever interface to work with the MacBook Pro, since it only supports USB and FireWire natively, neither of which will support 10-bit Uncompressed, or some other kind of slot34 device. I expect you really meant "on-set" capture, rather than real-time, since the only other way into a MacBook would be a P2-type card interface. In addition, real time Uncompressed 1920x1080 is very demanding no matter whether it is 24/30 frame, 8- or 10-bit. I'd be surprised if any Mac laptop could hack it. Its harder to do than HDV.
If you are recording on tape, then you can bet that the recorded format is segmented, which is a special, non-temporally related form of interlace, that allows a progressive frame to be recorded on a format that is maxed out at 1080i. Pull down is a term reserved for use with generating redundant frames to bridge a rate difference -- most notably "24" fps to "30" fps, which you don't mention. But which would be another significant hurdle.
I used to have this handy-dandy chart that listed all the bitrates and storage requirements for uncompressed formats SD, HD, 2K, etc., but it isn't in my field of view at the moment. 10-bit Uncompressed 1920x1080 at 29.97 takes up about a TeraByte per hour... do you have that much storage for field work? It isn't RED. It also requires a transfer rate in the Gigabit range, so in a practical sense it really requires a fibre-channel interface.
I'm kind of thinking you will eventually wind up with a ProRes workflow. You could get away with a laptop and AJA interface doing that.
jPo

8-bit uncompressed on a powerbook and G-Raid?

Has anyone tried to use their powerbook with a G-Tech G-raid using 8-bit uncompressed? I just bought a couple of G-Raids, a fw800 card for my G4. It works great on my G4, but I plugged in the drive to my laptop, and tried to edit the same project and the laptop couldn't keep up?
I ran some benchmarking tests, as suggested by tech support at GTech, and I got very low numbers when connected to the fw800 port on the laptop.
I've edited dv res material plenty of times on the laptop using lacie drives, and its always worked well. Anybody had any luck with 8-bit and GRaid on a powerbook?
System specs of powerbook are below.
Thanks,
Chris

Hi Chris - here's the way I rig mine when I've got 8 bit uncompressed stuff to do. fwiw - I've only ever had to rig it for short form stuff so not too sure about how well it would survive long form, anyways ...
pb 17 1.67 - 1.5 RAM
fw 800 Dual channel raid 0 -
LaCie fw 800 cardbus - 1 drive
pb fw 800 port - 1 drive
I get 3 layers of 8 bit uncompressed playing in RT.
That's grey bars. Additional layers up to 4 it'll go light green (for a total of 7), yellow after that.
2 layers of DVCPro 50 grey bars - light green for additional layers up to 5. Yellow after that.
And dv ? - I get 1 layer, then light green for the others, same as a normal drive. I guess because of dv compression.
Benchmarks for this - QuickbenchX - gives Sustained 100 mb read of around 115 and around 80 for write.
You can daisy chain also if you want - Quickbench will give you around 85ish for Read.
Not bad for a single processor G4 !!!

How do I convert titles from DV to 10 bit uncompressed?

I made titles using text in FCP working in a DV timeline. Now I need it in 10 bit uncompressed. Do I have to remake the titles in a 10 bit uncompressed timeline or can I simply move the title clips to a 10 bit uncompressed timeline and re-render, without losing any quality?
Thanks,
Mike

You can drop your sequence into a sequence set for 10 bit uncompressed. Everything will have to be re-rendered. You will not loose quality, but neither will you gain quality. You can never, ever gain quality. Plus you will need a RAID to be able to edit 10 bit uncompressed video, and it is going to take up an enormous amount of hard drive space.
So why go through all of this?

ProRes render times vs 8-bit Uncompressed

We have a project being brought in to us with the footage as Pro Res 422
Our editor thinks that perhaps it may be beneficial to media manage to 8-bit Uncompressed and do the edit and then re-link to the original Pro Res material because she thinks that it may save rendering time.
There will be quite a bit of colour grading, vignettes, masks and multi-layers etc, that will almost certainly give us red render bars.
Final delivery will be to SD and HD and as it happens the originator of the material can supply us media managed 8-bit Uncompressed clips anyway so we don't have to worry about wasted time there.
It's really to discover which codec is most efficient with respect to rendering.
I'd initially thought about the data rates that each requires and thought perhaps that might give a clue - but then the whole HDV issues of long render times made me re-think.
I'd love to do tests but the client will be there from day one and I'd love to be able to suggest a work flow straight away. (Searches on the forum havn't come up with a comparison - hence this request...)
We're using G-Raid drives that seem to handle both codecs quite happily.
Oodles of thanks for your collective wisdom.
LEE

ProRes is not a long GOP structure (it is not HDV). It is an all "i" frame with compression while maintaining a 4:2:2 color space. Its getting very good reviews including it's ability to work in Color and other grading/finishing options.
Just for your own piece of mind, why not get a segment in both formats and do some testing. That's the only way you'll know for sure the best workflow for your interests.
Good luck,
x

Exporting SD 4:3 to 8 bit uncompressed for DigiBeta

I need to deliver an 8 bit uncompressed file 1920 X1080 of an SD 4:3 18 minute video to a lab for transfer to DigiBeta. The film includes titles, still images, SD and HD video. The DigiBeta will then be delivered to TV station, where they'll take care of readying for broadcast, captioning, etc. I've only done this once before with trial and error, but for Digi Beta theater projection. I had to fiddle with settings to lose some artifacting (cost a couple trips to lab). I thought I took good notes, but not sure if thorough enough.
First, Sequence Settings: NTSC DV/DVCPRO 720X 480 Field Dominance set to "None" (this was the last change the last time that fixed the artifacts that I thought deinterlacing would take care of)
Render YUV
Setting when exporting:
Uncompressed 8-bit, Quality medium (no other option) 1920 x 1080 HD, letterbox checked "preserve aspect ratio, checked deinterlace.
Sound Linear PCM 48 kHz 24 bit stereo, Little Endian checked.
IMac OS 10.5.8 2.4 gHz 4 GB Ram Final Cut Academic 6.0
Need any more information?
I'd like to get this right the first time.
Am I on the right track?
Thanks in advance,
Dennis

Broadcast is about bit-depth and Digibeta is a 10bit 4:2:2 format and supports high bit rate audio. HDcam also has a 10bit 4:2:2 hence why they asked for that too. In lamens terms it means it holds a huge amount of colour information. Imagine a stripe, one side is black and the other pure white and in the middle is grey scale- with 10bit the scale is really smooth, whereas 8bit is noticably chunky in comparison- espeically near the blacks.
DVD is a compressed format and to make the file size fit onto a DVD they basically (in lames terms) strip away parts of the grey/colour scale so that is perceptually like the original, when in fact it is missing lots of the colour/grey range. But the audience won't know, unless they are viewing it on a screen capable of showing the range / resolution. You'll notice the lack of shadow detail in DVD's and if a broadcaster were to use it, it would be extremely blocky, grany saturated etc, because the broadcasters will encode/compress it too so they need the best quality.

Problem with 8-bit uncompressed quicktime file

I have exported my FCP video using the 8-bit uncompressed setting considering I initially captured the content uncompressed via a Blackmagic Decklink card. The problem is, regardless of the settings I have chosen for the uncompressed export, the quicktime file always has horizontal lines going across it when I watch it.
It almost seems like this could be an interlace issue, but I don't really know why that would be the case. I imported from DigiBeta that had telecined 16mm footage on it, but the footage should be at 29.97 when it is imported. I'm sure I'm an idiot and there's a simple answer to this, so if anyone has any ideas, please let me know.
Thanks.
PowerMac G5 2 x 2.0 GHz, 2.5GB RAM, 1 TB Disk Mac OS X (10.4.4)

If the exported movie plays fine in FCP and but not in the QT player, I would assume everything is okay and it's just the QT player. A couple of tests you can try:
1. Make sure the Quicktime player is set to actual size in the view menu.
2. Open up the movie properties in QT player. I have version 7 and can do this by going to the window menu, then show movie properties. If I select the video track and then the visual settings tab, there is a high quality preference checked. If I have it checked, I can see visual lines in the credits of an DV movie I have. if I have it unchecked, the horizontal lines dissapear. There are other settings there that affect the playout of the film, but shouldn't affect your final format.
QuickTime Player is probably showing you the highest quality of image and taking an interlaced video and playing it progressive on your computer monitor.
If you are going back to tape, you'll be using FCP to go back out to tape and shouldn't have those line issues.
If you are compressing for DVD, you shouldn't have those issues.
If you find a more detailed explanation for this, post back so I can know it.

DV PAL vs 8 Bit UNCOMPRESSED

I'm currently editing a TV show thats shot on DVCAM and is being ingest SDI via a KONJA card as 8 Bit Uncompressed in FCP.
What i'd like to know is if I instead ingested the DVCAM tapes in its native format DV, will there be a quality difference as opposed to ingesting them as 8 Bit uncompressed?
The show is Real Estate so its simple cuts with minor graphics, Name straps, Summery of the property & price & agents usually give me photos that I add, and I add a little motion to spice them up. The only time FCP need to render is any graphics added to the sequence.
I understand 8 Bit Uncmpressed is 20 mb/s and DV is only something like 4 mb/s. If my native footage is DVCAM then wouldn't it make sense to keep it DV and then just mastter the final to DigiBETA?

Steve is absolutely, 100-percent right, but there's still going to be a layer of loss in going from computer graphics to video. I only mention it because it just hit a client of mine yesterday.
Client wanted large title cards on black, with screen-filling red type. No problem, right? Well lo and behold, the titles ended up looking jaggy +even though+ they were going into a ProRes 422 HQ 1920x1080 timeline.
Why? Because of the conversion from 4:4:4 to 4:2:2. Video (with a few exceptions) only uses half as much horizontal color resolution as black-and-white resolution. So bright red type on a solid black background will end up looking like it's low-res, even though it's not. It's actually the perfect resolution +for video.+
A television is not a computer screen. There are limits to what you can do, limits that depend on technology choices made decades ago. The reason we all get paid the big bucks is because we've developed the patience to live within those limits.

Newbie question about 10 bit Uncompressed Video

Hi guys
Am I correct in understanding that Uncompressed formats are only for digitizing analog video?
Like, with DV or HDV or HD 1080p/ P2 etc.. I would never need to use any Uncompressed codecs.. is that right??
Basically my question is: when would I use one of these Uncompressed codecs?
Thanks!

Animation for interchange between platforms is no compression, and with alpha if millions+ is chosen. Basically these files would be exported from 10 bit files at our shop. Image sequences would be generated the same way unless they were going out to a grading station or VFX in which case they would be 16bit tiff or DPX files.
The main reason for not passing 10 bit video is gamma interpretation between platforms.
we have had success with AJA 10 bit uncompressed being passed to PC shops with them using the AJA codec on their systems. Most of these transfers are aftereffects based projects, captured via Kona3 in FCP automatic ducked to AE and passed with media and project. They come back the same way in an AE project. we can then control which layers go where depending on their use and workstation.

Best Settings to import Native 525 interlaced 29.97 AJA codec 10 bit uncompressed masters in CS5.

This is the note from the post house. "These are all 525 interlaced masters and digitized as that. So the ingest was native 525 29.97 AJA codec 10 bit uncompressed. Your timeline will have to match these settings. As with all telecine at that time, there was no 23.98 (24P) transfers. Hence all film is 3:2 pulldown and should be constant cadence although it might not be."
I have tried everything I can think of and burned test clips to DVD's as well. The best I have gotten was jumping over to progressive because the interlacing is a mess otherwise. These clips are all .mov. 720 X 486, when I drag them on the timeline it makes them progressive but the titles still jump. This is a turn key PC based system. THey came to me in .mov format.
Any Ideas? Thanks!

I recently had to convert some clips from a very old Matrox system, with clips delivered in the Matrox codec at 720x486. First, in the Bin, check the clip Properties and look at the field order, should be Lower First. Use Modify > Interpret if needed to correct. Sequence settings should be 720x486, NOT 720x480.
For export, DVD is 720x480, so you would need to use CROP settings in AME to remove the 6 extra lines. Which ones? Trial and error. I ran a series of short test clips (same clip exported multiple ways), trying Top and Bottom settings of 6,0 and 5,1 and 4,2 and so forth, naming clips accordingly to keep track of results. These clips were dropped into a 720x480 sequence and reviewed to see which looked the best to determine final output settings.
I hope you find this helpful.
Jeff Pulera
Safe Harbor Computers

10 Bit Uncompressed .mov problem

I've just been given a file that's PAL 10 Bit Uncompressed. When I import into FCP it plays as plain white with audio.
Can anybody help?

That codec is often associated with AJA or Decklink cards. Try the codec link here and see if it helps: http://www.aja.com/ajashare/AJA_SoftwareCodec.zip

HD 10 bit uncompressed Playback problems

Hi
I have some problems with HD 10 bit uncompressed material
I captured HD material with Blackmagic HD extreme
and I can't play it in Final Cut Pro without rendering.
In our company we have 5 (Mac Pro 2x2.8 Quad-Core Intel Xeon)
with Final Cut Pro 6.0.4
On two of our systems I can play it without rendering
but the other 3 need rendering and annoys me!
It are the same systems with the same soft & hardware
Can somebody help me
thanks

If the machines are similar with the same version of software, copy the FCP project from a machine that doesn't require rendering to a machine that does. Now, does the moved project require rendering?
If not, then the sequence settings are indeed different somewhere.
If it does, then maybe your preferences no longer match, i.e. RT Playback.

10 bit vs 8 bit uncompressed

Similar Messages

Maybe you are looking for