Video Bit Depth & Chroma Sub Sampling Basics

[@oc2phish07]

FINALLY! The best video I have ever seen explaining what these terms mean. Really good stuff Michael, I have been struggling to get my head around it.

[@lhd7105]

Sky is showering brain candy!! Michael thank you for the super crystal clear explanation and tying it so clearly to "So what?" or how the more information makes it easier for colorists. Keep up the amazing work, how can only 14K people have seen this???

[@alrestivo2746]

I finally looked up the word maven, it is the definition of you.

[@lallamallamarada]

That was a nice explanation of Sub Sampling. Thank you so much.

[@johnharte1221]

This is a fantastic explanation! Thank you for making this easy to understand.

[@sacredmurals]

Wonderful explanation!! Thank you!!!!❤❤❤

[@svbow6347]

So glad you talked about the Z6 and Atomos. The Z6 is and will be amazing, thanks Nikon.

[@JOAOAUGUSTODIASABREU]

So much valuable information Michael, Thank you. You're channel rocks! and who has disliked this (or any of your videos and tech inputs) has the intestines attached to the brain!

[@johnbhome2]

Great info and I learned something new today! Thank You!!

[@MMaven]

That’s what I’m going for! Just want to give my viewers a well rounded foundation of technical information!

[@town3]

Nice work Michael like always.

[@andreasmoyseos5980]

What a great explanation. Instant suscriber!

[@LaurenceBakerAdobe]

Great video. Not everyone's cup of tea as a subject. But if you want to understand digital imaging properly this is the video.

[@shoaibasghar_h]

what about 10-bit 4:2:0 - can you still do some decent colour grading with this ?

[@randydean23]

I gotta hand it to you Michael, you have been killing it with your content lately. I haven't had a chance to read through the comments, but I'm sure you know about Atomos announcing that the new Nikon Z series firmware will not only allow for 10 bit 4-2-2 and of course Pro-Res Raw, but now they're saying 12 bit is going to be available...in the case of the Z6 in a sub 2000 dollar camera?!?!? It has me trying to get in the current Nikon trade-up offer to get folks into their new mirror-less system. I had planned to pick one up because I shoot a lot of video (some professionally and then I use youtube for recreational purposes mostly) but hadn't been in a rush because - outside of being full frame - the Z's weren't nearly as attractive for stills as my current Nikon DSLR kit. But in any case I do believe that Nikon's position as being a stills first company for so long have allowed them a latitude not afforded to Canon, Sony, or Panasonic and that is not having to worry about protecting their $10,000 cinema cameras as they don't happen have a presence there. I for one am all for it and believe this is healthy competition on a front which I believe more and more to be the most likely to produce substantive innovations going forward, but that's just an opinion. In any case, thanks again for the content Cheers

[@MMaven]

Thank you Randy! Its definitely an exciting time as a filmmaker!

[@oren.kirschenbaum]

Great vid, Michael, as always. Thanks!

[@wadesmith666]

Great vid- cheers

[@shayuppim]

Thank you very helpful information

[@sbaaaa76]

very helpful information

[@MrBaracas]

So if you had to chose between these? 8 bit 4:2:2, or 10 bit 4:2:0?

[@dialectixemcee2428]

dope....thanks....but now my question is: I'm shooting on a T5i with no option Im aware of a to change bit...is this bit you're talking about "bit rate" like when exporting video in premiere? If so, what does making the file "16 bits" do if my camera may not shoot in that? Sorry Im new to video relatively I been getting into it lately more but I started learning photography fist

[@MMaven]

Dialectix Emcee you are shooting on 8 bit 4:2:0

[@dialectixemcee2428]

@@MMaven thank you. So I can save time and file size by exporting in 8 bit vs 16 in premier haha

I always get confused on how bit depth affects luminance, I mean when you have a 32 bit linear image or video on after effects or photshop you can pretty much color it like if it was a raw file. I thought bit depth was only about banding but you can also get more dynamic range? I don't understand it completely. Thank you!

[@lnz3704]

Black magic pocket can do cheap raw

[@briandecasa5444]

Very informative! Thank you! Question - if ny main medium is Facebook and KZbin should I just use the h.264 8 bit 420 on the XT3 then if I dont do extensive color grading?

[@MMaven]

Yes! Absolutely

[@briandecasa5444]

@@MMaven so those comparison videos on KZbin that show more banding on 8 bit does that mean they pushed the color too much for the difference to be noticeable vs 10 bit?

[@briandecasa5444]

Do you mind covering bit rate and All I vs long GOP next video?

[@timmeier4136]

Does this mean that a full HD 4:2:0 picture has the same color information as an 4:4:4 quarter full HD picture? so 960*540?

[@DeyvsonMoutinhoCaliman]

I think it's strange that everything is 8bits, even monitors, because games have been said to be using 32bits colors for many years now. Some games let you choose the quality to improve performance, and if you choose like 16bits the colors become horrible and 8bits would be like a cartoon. At least in Dark Souls it gets horrible if you decrease the color bits, but it's not the case in some games where if you change the color bits it makes little to no perceptible difference. I'm surely getting something wrong here, they must measure it in games in a different way

[@MMaven]

There are many variables in the final output, even your graphics card matters, and to maximize everything you have to have all the right links, including monitors.

[@Enlightenment_2025]

@@MMaven, is there 16 bit video even in computer graphics? I never heard of it.

[@mediaconsumer9195]

You totally mixed up the things: 32 bits for games means overall bit depth of one pixel, but not a bit depth per channel that this video is about. I.e. 8 bit for red channel + 8 bit for green channel + 8 bit for blue channel + 8 bit for alpha channel = 32 bits. 8 bits for game graphics means that you have some sort of predefined palette, 256 colors for everything. And that 8 bit pixel value is not the color itself, but just an index of the actual color in that pallete. The best example of 8 bit color thing is GIF image format. The best example of 32 bit thing is PNG format. JPEG is 24 bit format, without the alpha channel.

[@MMaven]

​@@Enlightenment_2025 I was answering about the discrepancy about colors being horrible. Not the 32 bit colors.

[@mediaconsumer9195]

@@MMaven also, there is another one thing in bit depth: TV range (16-235) vs full RGB range (0-255). Sometimes this difference can lead to bug in software (especially in mobile apps like Filmic pro), which can lead to problems with video dynamic range. For example, check the latest Gerald Undone's video about S-Log on Sony a7 III

[@LavikkaPhotography]

Good video.

[@abyhamouda3813]

Thank you

[@xyzimagenes]

Interesting, thanks!

[@TMM6900]

What about a computer monitor to display 10 bits 422?

[@MMaven]

Yes! I covered this!

[@Camilamscine]

I dont understand why 422 is always related to 10bit and 420 to 8bit. Couldnt i record in 8bit with 422 compression? Im trying to understand why, can someone help?

[@MMaven]

Camila Mendes Santos There are cameras that record 8bit 422. Sony cameras can do it through HDMI in the Canon 5D Mark four does it internally

[@Mostly_straight]

8 bit footage sucks when grading it

[@winniduesseldorf]

Reat Infos THANK YOU !

[@batfan1939]

I think ten-bit color should be the next "big thing" for screens moving forward, rather than resolution. I've seen several videos talking about how, especially with dark frames, there are just not enough colors to cover a 3,000-plus pixel display width all the way across for something like a gradient. That's a legitimate problem with current standards, and I'd love to see companies address it. Would also be an excuse to create an image standard with a fourth pixel (extending cyan, not yellow), further increasing the number of possible colors!

[@darkftforce]

really very informative video and great efforts by Sir Michael The Maven. very few people who are Real Photographers put out something like this which is very different from day to day almost simillar topics discussed by every tom dick and harry of photography. thanks.

[@MMaven]

This is my goal, so thank you!

[@svbow6347]

You don't talk about another issue or image quality. Talk about 10 bit 422 that compressed smaller file vs 8 bit 420 uncompressed larger file. The GH5 has 10bit 422 that compressed and sharpened. Give me 8 bit with better color, that not compressed or compromised.

[@ratheist5356]

Medium crossover

[@hipstershep]

2 to the power of 8 = 256

[@jamesshort6856]

256 total values, however this includes 0, so its a 0->255 scale