-
funny codec named “None”
Drop any movie into Compresssor, and pick one of the Advanced Format Conversions.
Open Inspector for this job and go into Settings / Compression Type.
There is one choice called “None”. Choose that.
Now you can set “Depth” at 9 different descriptors from “Black and White” to “Millions of Colors +”.
And you can set “Quality” on a continuous scale from “Least” to “Best”.
Inspector shows what these settings do, more or less.Inspector shows the following pixel depth values for the “Depth” choices.
Black and White: 1
4 Grays: 34
4 Colors: 2
16 Grays: 36
16 Colors: 4
256 Grays: 40
256 Colors: 8
Millions of Colors: 24
Millions of Colors +: 32Obviously 4 grays is 2 bits. It seems Apple has coded the pixel depth for grayscale choices by adding 32 to the true value. OK, they’ve abuse of the term “pixel depth”. And for consistency they should call Black and White 33 bits deep.
You might be curious which 4 colors you get with that choice. It turns out that Compressor makes a 4 graytone image when you chose “4 Colors”, exactly the same as “4 Grays”.
“16 Colors” is however very interesting. If you set Quality on the low side, you will see about 10 colors. And you see about 4 gray tones plus black plus white. Those are the 16 colors. Quite weirdly, as a white title fades out it goes through 4 grays and then a light brown and then black. That light brown taking the place of the darkest grey must be a coding error. (Actually several of the 10 colors are wildly far from what they replace. This is not a serious attempt at 4-bit color.)
Using the same “16 Colors” setting with Quality set on the high side gives a completely different result. There appear to be hundreds of realistic colors and grays. If you look closer there are just the set of 16 but they are arranged in small clumps of identical pixels. So with codec “None”, the Quality control, which according to Inspector is spatial quality on a scale from 0 to 100, is actually put into the service of reproducing color. If you think about it, almost all colors could be reproduced using just red, green, blue, and black clumps of variable size. Millions of colors achieved with just 2 bit pixel depth, but lots of pixels. If this is cheating, then what is codec “None” doing?
Another oddity with the “16 Colors” setting is that the (spatial) quality slider yields exactly the same result (10 recognizable colors) between 0 and 49. At 50 it jumps to the pointillist color (looking like hundreds of colors), and this stays exactly the same to 100. Thus the continuous 0-100 scale is really a binary step. Whose idea was that?
The files for the two pictures have exactly the same size, because they have the same number of pixels and 4 bits for each. The number checks exactly. It’s a 30 sec PAL video: 30 * 25 * 720 * 576 * 4 = 1244160000 bits = 155520000 bytes = 148.3 MB. Codec “None” is 4:4:4.
Codec “none” seems to behave OK for 256 Grays, 256 Colors, and Millions of Colors. The latter is presumably 8 bits for each of Y, U, and V.
The great unknown is the 32 bit thingie at the end of the list. What does “Millions of Colors +”, more marketing than scientific language, mean? Does it mean more colors than “Millions of Colors” made, or does it mean the same colors “Millions of Colors” made and some non-color bonus?
There’s internet speculation that 32 bit color is simply 24 bit color with 8 bits of other stuff. I don’t buy this. Who knows how codec “None” understands 32 bit color?
-
20 Replies
-
Millions of Colors + refers to the images being an RGB + Alpha Channel. If you render as Millions of Colors, it’s just RGB no alpha information.
Richard Sanchez
North Hollywood, CA“We are the facilitators of our own creative evolution.” – Bill Hicks
-
[Dennis Couzin] “There’s internet speculation that 32 bit color is simply 24 bit color with 8 bits of other stuff. I don’t buy this. Who knows how codec “None” understands 32 bit color?”
The extra 8-bits is an alpha channel and that’s not speculation.
David Roth Weiss
Director/Editor
David Weiss Productions, Inc.
Los AngelesPOST-PRODUCTION WITHOUT THE USUAL INSANITY ™
A forum host of Creative COW’s Apple Final Cut Pro, Business & Marketing, and Indie Film & Documentary forums.
-
OK, if Apple means by “32 bit color” 24 bit R,G,B color plus an 8 bit alpha channel, I have some questions.
Doesn’t 10 bit uncompressed 4:2:2 video, which has 10 bits for each of Y, U, and V, produce better color (or more colors) than 32 bit color?
Alpha masks are only used in effects. Why is codec “None” offering 8 bits per pixel for alpha masking when almost no video has alpha masking? And why doesn’t codec “None” offer standard 1 bit alpha masking?
Video is normally coded Y,U,V, which makes subsampling so efficient. Perhaps codec “None” videos can’t be subsampled. Are all the other codec “None” depth settings coding R,G,B? This matters, since 24 bit depth distributed 8 to each of R,G,B yields different colors (though the same number) than 24 bit depth distributed 8 to each of Y,U,V does.I expected Apple codec “None” to be something fundamental. From the details in the original post, it seems something amateurish and screwy.
-
It’s 8-bit RGB, and it’s very old. It goes back to an age called the early 1990s when I believe YUV video didn’t even exist in a software-only video player. “None” is also a legacy AVI codec.
Why the fascination? Use the Apple Uncompressed codecs for YUV video. Use the AJA or Blackmagic Uncompressed codec for 10-bit RGB.
Sean
-
“Why is codec “None” offering 8 bits per pixel for alpha masking when almost no video has alpha masking”
If you’ve ever had to transfer footage to graphics house, or are integrating animation from a graphics house, you’ll work with video with embedded alpha channels all the time. The old method of doing it, would be to lay the graphic back to tape, and then lay the alpha channel (which will appear as a greyscale image) back to tape and use that as a luma travel matte. This is pretty standard workflow for sending footage and from graphics houses.
It isn’t just None that supports this, Apple Animation does, Avid DNxHD does, it would be nice if a new release of Apple Pro Res did.
Richard Sanchez
North Hollywood, CA“We are the facilitators of our own creative evolution.” – Bill Hicks
-
Thanks all, for the historical tidbits.
-
Nothing wrong with the old NONE, only that should be included among the “legacy codecs”.
Best codec information in the web;https://codecs.onerivermedia.com/
Rafael
-
Thanks Rafael Amador for the link to onerivermedia. So codec “None” has a respectable history. It was used with with large color depths. I wonder if the writers of Compressor added the several small color depth realizations of “None” as novelties. Someone certainly screwed up the 4-bit color realization, as detailed in my original post.
-
[Dennis Couzin] ” I wonder if the writers of Compressor added the several small color depth realizations of “None” as novelties.”
The bit depth options are not offered by Compressor, but by the very codec.
If you try to export NONE from QT, you will find the same options available.
This codec is quite old so haven’t take advantage of the compression techniques developed afterward.
Codecs like Animation or Planar RGB use compression schemes similat than “zip” or “tar” to reduce the files without lossing any information.
I don’t use NONE, but is good to heve it available.
Cheers,
rafael -
[Dennis Couzin] “Alpha masks are only used in effects. Why is codec “None” offering 8 bits per pixel for alpha masking when almost no video has alpha masking?”
so you would deny those that need it? This another one of those lame attempts at a discussion on how this or that needs to be changed because it does not affect you in any way shape or form. Do you Use the NONE codec- I doubt it, my guess is you would not have powerful enough storage to handle video with that compression.
NONE as a compression codec is a legacy carryover from the time when your machines were not fast enough to handle compression decompression cycles in realtime. This is why we have CODEC’s or compression / decompression engines that can handle this. This is also why as the needs of editing go up the hardware needed to get the best quality goes up to.
“Doesn’t 10 bit uncompressed 4:2:2 video, which has 10 bits for each of Y, U, and V, produce better color (or more colors) than 32 bit color? “
Part of the problem is that while the measurements sound like they are using the same metering for those measurements – they are not- the words are the same but the metrics for measurement are different
YUV video is more analogous to the ICE-LAB colorspace than the RGB space, and if you did actually know these color spaces you would understand that RGB info is indicating how many colors as Chroma that can be recorded while YUV bit depth is determined by levels of Luma that can be captured.
In the Planar and PRINT world RGB color is determined by the 8 bits ( 256 levels of grey) per channel x 3 channels to get 24bit color add an alpha channel and you get what is called 32 bits
IN VIDEO – the limit of a baseband video signal can currently be sent at more that 10 bits of data.
so for print that would be 10 x 3 channels (at 1024 levels of gray) without the ability for an alpha or the equivalent 30 bit info as it is calculated for print space.There is no workflow currently for video beyond 10bits, while frame based workflows and some tapeless cameras can capture data at higher bit depths, the limit for a “video” signal stands at what it is and I do not foresee it changing in my lifetime since humans do not have the visual acuity to see EVEN 1024 levels of gray at one time.
gary adcock
Studio37
HD & Film Consultation
Post and Production WorkflowsInside look at the IoHD
https://library.creativecow.net/articles/adcock_gary/AJAIOHD.php
Reply to this Discussion! Login or Sign Up