What is RAW?
- Thread starter kodjo
- Start date
AndrewGentle
Well-known member
No, RAW doesn't mean uncompressed. What RAW endeavours to do is to record the data coming from the sensor. This data comes from further up the signal processing chain which presents us with a viewable image than usual video formats. The advantage to using RAW is that you can make changes to adjustments to the image which would otherwise be burned in by the camera's signal processing. White balance, ISO, noise reduction and sharpening are all examples of attributes which may be adjusted in post and performed by a computer which has far more processing power than could be included in a camera, thereby allowing it to use more effective algorithms. RAW data is usually compressed, sometimes losslessly but often with a small amount of loss.
AndrewGentle
Well-known member
No, XDCAM is not a raw codec. This means that you need to nail your whitebalance, gain, sharpness e.t.c. on set because you don't have the flexibility to modify them in post production.
AD Tom
Active member
The difference is that RAW-Material must be "interpretated" in your computer. That means the settings like Kelvin, ISO, Color Space, etc. must be set manually AFTER recording. This gives you a much higher flexibility in changing exposure, colors, etc. with your material (and also a little bit more work). The RED-RAW codec is one (or probably THE one) RAW-Codec for digital filmmaking.
Hope this helps
Thomas.
Hope this helps
Thomas.
Ted Ramasola
Veteran
The Red one uses its version of raw. Rated at 320 asa. Also the Ikonoskop.
I'm not sure if the viper does , i think its uncompressed. Not sure though. I'm sure about the red.
The files are bigger. Not impossibly big, just bigger. It gives you lotsa a wiggle room in post.
Handling red footage in Red cineX is easy though, even in moderately powered PCs you can do a grade then export the clips to a format that your NLE can ingest.
There's ARRI RAW/REDCODE/IKONOSKOP records to DNG sequences. I'm not sure if either Otis's camera (drama) or Si-2K shoot RAW too, but there are many cameras out there, yet none at all as affordable as RED one.
The one thing everyone's waiting for someone to do is to standardize a RAW format, with unified colorspace, cause every company has its own RAW encoding, and I read that this was a problem. I believe it was, 'The film Industry Is Broken' on Prolost.com I'll post a link soon, but at a lousy cafe in beirut with 5.6K connection. I know it's 2010 its embarrassing.
But that was a good article, from what I gather (just a feeling didn't read this) is that DNG is as close as a standard as it gets I think. By Adobe. Stands for Digital Negative, which cleverly describes it.
It's the same still you get off your DSLR that makes 'Camera RAW' pop up. I'm not proficient with RAW at all, still new, and it's intimidating (I know I should just do it I am saving up for an A-Cam and still mess with R3ds from shoots but still need to read more) but that's when you demosaic it, turning into let's say a TIFF. Crazy flexibility, crazy boundaries for aggressive grading. I mean artiste aggressive. I don't like that the RED one is big, but you just don't have the heart to open any other format after a bout with RAW.
Shortens the distance between the idea and its realization if you ask me.
I wonder what RAW stands for.
If I got any of that wrong, I apologize, hoping to be corrected
cheers all
The one thing everyone's waiting for someone to do is to standardize a RAW format, with unified colorspace, cause every company has its own RAW encoding, and I read that this was a problem. I believe it was, 'The film Industry Is Broken' on Prolost.com I'll post a link soon, but at a lousy cafe in beirut with 5.6K connection. I know it's 2010 its embarrassing.
But that was a good article, from what I gather (just a feeling didn't read this) is that DNG is as close as a standard as it gets I think. By Adobe. Stands for Digital Negative, which cleverly describes it.
It's the same still you get off your DSLR that makes 'Camera RAW' pop up. I'm not proficient with RAW at all, still new, and it's intimidating (I know I should just do it I am saving up for an A-Cam and still mess with R3ds from shoots but still need to read more) but that's when you demosaic it, turning into let's say a TIFF. Crazy flexibility, crazy boundaries for aggressive grading. I mean artiste aggressive. I don't like that the RED one is big, but you just don't have the heart to open any other format after a bout with RAW.
Shortens the distance between the idea and its realization if you ask me.
I wonder what RAW stands for.
If I got any of that wrong, I apologize, hoping to be corrected
cheers allThe important feature of raw is that it records the actual per pixel linear exposure data values with minimal or no processing. The only variables that affect this data are exposure settings: aperture, shutter speed or angle, frame rates and light levels in the scene. Like a film negative, these recorded values have to be processed to produce a viewable image.
A video camera processes sensor data in camera and spits out a finished encoded image.
A raw camera records the sensor data without processing and all image processing is applied in post. Image properties such as ISO speed, white balance, color balance, black levels, contrast,brightness, gamma, color space, and final encoding format, are all defined from the raw data in the processing software. These settings can be changed at any time by going back to the original raw file, as it is never changed. All values are just meta data stored with the file until it is encoded and saved to the final image format.
It is raw recording that has given digital still photography the quality and flexibility to largely replace film as an acquisition medium. Now the same revolution is just beginning to happen for digital motion cameras.
The main issues with recording raw motion at high resolutions are sensor and A/D conversion speed, data bandwidths, high data rates and storage space required.
To give you an idea, I have a RED R3D raw 4k resolution (4 times HD resolution) file I downloaded to play with and learn the work flow. It is a 40 second clip and a 1.5 gigabyte file. Typical video camera file formats run up to 50 megabits/sec data rates. At its lowest quality setting at 24 fps 4k, the data rate for a Red 1 is about 225 megabits/sec.
A video camera processes sensor data in camera and spits out a finished encoded image.
A raw camera records the sensor data without processing and all image processing is applied in post. Image properties such as ISO speed, white balance, color balance, black levels, contrast,brightness, gamma, color space, and final encoding format, are all defined from the raw data in the processing software. These settings can be changed at any time by going back to the original raw file, as it is never changed. All values are just meta data stored with the file until it is encoded and saved to the final image format.
It is raw recording that has given digital still photography the quality and flexibility to largely replace film as an acquisition medium. Now the same revolution is just beginning to happen for digital motion cameras.
The main issues with recording raw motion at high resolutions are sensor and A/D conversion speed, data bandwidths, high data rates and storage space required.
To give you an idea, I have a RED R3D raw 4k resolution (4 times HD resolution) file I downloaded to play with and learn the work flow. It is a 40 second clip and a 1.5 gigabyte file. Typical video camera file formats run up to 50 megabits/sec data rates. At its lowest quality setting at 24 fps 4k, the data rate for a Red 1 is about 225 megabits/sec.
Last edited:
AndrewGentle
Well-known member
Zacatac
Well-known member
well now, the difference between RAW and H.264 is HUGE
G
G.P.
Guest
I don't think you are fully understanding what RAW is... simple explanation is this:
RAW is the sensor data being recorded before anything else touches the image within the camera.
Their are very few cameras in the world that have this ability. Most of them don't have the capability to do this on the actually camera itself, and is only an option for a tethered recording option.
This isn't a type of codec, because there is no "codec" really used. RAW is just the sensors data, and it usually takes dedicated software to debayer and decode the image information.
RAW is in the middle of being standardized to a general set, but this is not really going to last long because of the advancement of technology is too fast for it to be fully implemented. Pretty much every major maufacturer of a camera that can have an option to record RAW uses its own way of getting that information.
D1 and D2 tape is not a RAW recording. RAW is the basic 0 and 1 from the sensor before anything is done to the image you can't even view the image unless it is decoded properly to show you what your "negative" looks like.
RAW is the sensor data being recorded before anything else touches the image within the camera.
Their are very few cameras in the world that have this ability. Most of them don't have the capability to do this on the actually camera itself, and is only an option for a tethered recording option.
This isn't a type of codec, because there is no "codec" really used. RAW is just the sensors data, and it usually takes dedicated software to debayer and decode the image information.
RAW is in the middle of being standardized to a general set, but this is not really going to last long because of the advancement of technology is too fast for it to be fully implemented. Pretty much every major maufacturer of a camera that can have an option to record RAW uses its own way of getting that information.
D1 and D2 tape is not a RAW recording. RAW is the basic 0 and 1 from the sensor before anything is done to the image you can't even view the image unless it is decoded properly to show you what your "negative" looks like.
Ian-T
Veteran
Go buy a DSLR....take a RAW picture and then another but with the typical compressed JPeg (or whatever it's called). You can manipulate the RAW every which way while you will break the other easily. It's like manipulating film itself. That's what a lot of folks want with video...RAW images. No more compression artifacts...or muted colors. The images will be so much richer and detailed. Most seem to think that would be the death of film.
I know what it means, but why is it always in caps? Genuinely, no clue why it's in caps all the time. Nobody writes raw. They write RAW.
Is it cause it is awesome?
AndrewGentle
Well-known member
I think it's to differentiate between the recording format and the generic term for uncut footage. For example; 'I'm uploading some raw footage,' and, 'I'm uploading some RAW footage.'
Even redcode isn't really RAW as a photographer understands it, since it undergoes lossy compression, whereas the only compression true RAW undergoes is very mild lossless.
To put some figures to it, my 550D outputs on average a 24.5 MB RAW file. Stringing together 25 of these per second (PAL) gives 612.5MB per second or 4900 Mbps . Compare this with the HVX's 100 Mbps DVCProHD or Canon's 44Mbps H264.
Some fairly heavy data pushing is needed for this sort of data, and it isn't even overcranked, just bog-standard 25 fps.
Dave
To put some figures to it, my 550D outputs on average a 24.5 MB RAW file. Stringing together 25 of these per second (PAL) gives 612.5MB per second or 4900 Mbps . Compare this with the HVX's 100 Mbps DVCProHD or Canon's 44Mbps H264.
Some fairly heavy data pushing is needed for this sort of data, and it isn't even overcranked, just bog-standard 25 fps.
Dave