Can't open a 10-bit 4:2:2 recording

Since my camera offers 4K 10-bit 4:2:2 recording using .MOV format, I thought I'd experiment with it and compare with my usual 8-bit 4:2:0 recording. I'm finding, however, that the free version of Resolve will not open it.

In Resolve, Project Setting is set to 10-bit (I guess by default as I have always used 8-bit video). Are there any other settings that would allow me to open the 10-bit file? The file details are:

Frame width: 3840
Frame height: 2160
Data rate: 149942 kbps
Total bit rate: 151479 kbps
Frame rate: 29.97 fps

Thanks.

We know the container format is MOV, but what about the formats of the video and audio streams inside it?
A tool like mediainfo is useful for examining files.

I'm finding, however, that the free version of Resolve will not open it.

Pretty sure there's your problem.

Resolve free on Windows doesn't support 10 bit h.264/5

https://documents.blackmagicdesign.com/ ... 3705210000

your options are;

Use a Mac
Buy a $300 studio key
transcode everything to DNxHR HQX with shutter encoder (free)



I would request you add your support for my feature request to add error messages to resolve for this. You are going to be one of multiple posts this week with the same confusion.

viewtopic.php?f=33&t=199305&hilit=+error

I had done a quick search but somehow didn't come across similar topics. In any case, it didn't take long for this subject to go above my head. I was only curious, as a hobbyist, to compare 10-bit and 8-bit videos but I'll just continue using 8-bit unless and until a compelling need for 10-bit comes up.

Thanks for all the responses.

it didn't take long for this subject to go above my head

Welcome to the Wonderful World of Video! Don't feel bad. Nobody is born knowing this stuff and I'd wager most of us have learned the Inner Mysteries by tripping over them.

Welcome to the Wonderful World of Video! Don't feel bad. Nobody is born knowing this stuff and I'd wager most of us have learned the Inner Mysteries by tripping over them.

Thanks. Here I thought I would advance my understanding a little by learning about and testing bits/pixel and using a different chroma subsampling but I guess I was overly optimistic.

We know the container format is MOV, but what about the formats of the video and audio streams inside it?

What more is there to the video format besides the file properties that I posted and what impact would it have on the ability to open it in Resolve?

Hi,
You can use the free "Shutter Encoder" to encode the file as Cineform 10bit mov (at the highest quality you have storage for) and the free version of Resolve can open it. I find I can edit smoother with this intermediate format than h.264.
Sorry, I just saw ZRGARDNE posted a similar response.
10 bit colour is useful for chromakeying and for heavy colour grading.

And for HDR.
For editing and mild corrections in SDR you don’t need it. Try to get as close to the picture you like as you can with the camera settings and proper exposure, and 8 bit will suffice.

We know the container format is MOV, but what about the formats of the video and audio streams inside it?

What more is there to the video format besides the file properties that I posted and what impact would it have on the ability to open it in Resolve?

The container is just the outside package. Inside it might be 264 or 265.

265 4.2.2 can be decoded by a few things. 264 4.2.2 is more problematic.

264 is older and compresses less. Its mainly used for 4k or lower .

Unless your HW can decode either the playback will be on your CPU and that can really slow things down.

264 is older and compresses less. Its mainly used for 4k or lower .

I would not recommend using H.264 for 4K video.

Preferably, I would avoid using H.264 altogether unless absolutely necessary. While H.264 (also known as AVC) is still widely used, it is less efficient compared to newer codecs like H.265 (HEVC) or AV1, especially for high-resolution content. These newer codecs offer better compression and quality for 4K video.

People tend to get stuck with older solutions and forgetting that newer and better options are available.

The problem is people can only use what the camera companies provide.

On my camera(s) the 8 bit option uses 264. 10 bit gets HEVC.

The problem is people can only use what the camera companies provide.

On my camera(s) the 8 bit option uses 264. 10 bit gets HEVC.

If you have a camera that supports 10 bit why would you ever want to use 8 bit?

The OP doesn't have studio. Can't open 10 bit.

You may need to deliver 8 bit.

The problem is people can only use what the camera companies provide.

On my camera(s) the 8 bit option uses 264. 10 bit gets HEVC.

If you have a camera that supports 10 bit why would you ever want to use 8 bit?

Because not all cameras with 10 bit recording will record 10 bit in all frame rates. Also they may record 10 bit in h264 4:2:2 at some frame rates and of course 10 bit 4:2:0 in HEVC at other frames rates. Depends what is needed for editing/grading and output.

We know the container format is MOV, but what about the formats of the video and audio streams inside it?

What more is there to the video format besides the file properties that I posted and what impact would it have on the ability to open it in Resolve?

At the risk of over-simplifying the whole mess:
A media file is a container for video and audio information. Containers come in various formats and are usually known by their file extension.
Inside the container, the video and audio streams each have their own set of formats, bit and sample rates, and bit depths.
Container: MOV, MP4, etc
__ Video stream: DNxHD, H.264, H.265, etc
__ Audio stream: AAC, WAV, etc.

Resolve can be fussy about what combinations of formats it will accept or output. On Linux, for example, it does not like AAC audio. The Supported Codecs List that ZRGARDNE posted above shows the various combinations Resolve currently uses:
https://documents.blackmagicdesign.com/SupportNotes/DaVinci_Resolve_19_Supported_Codec_List.pdf

For further reading:
https://drive.google.com/file/d/1N91KnI ... sp=sharing

So I recorded several short videos of different scenes in both 8-bit and 10-bit and followed the advice here to use Shutter Encoder to transcode to DNxHR HQX. I could then open the (now much bigger) 10-bit file in Resolve. I was hard-pressed to find any noticeable differences between 8- and 10-bit until I used a greatly exaggerated curve tool that brought out some subtle banding in the sky in the 8-bit video when viewed at 100%. I conclude that for my modest purposes I will be sticking with 8-bit for now. But it was an interesting learning experience.

That’s what is to be expected.
But to make sure, also test an upload to YT ( if your work is intended for it).
Their compression can make it look worse.

I conclude that for my modest purposes I will be sticking with 8-bit for now. But it was an interesting learning experience.

It matters when things aren't perfect.

If I go out to the beach on a sunny day the skies and clouds are quite a few stops from the darkest part of the scene. It's very easy to slightly mess up the exposure . With 10 bit you'd have a chance of recovering things.

Obviously if it's clipped it's clipped but at times you aren't totally gone.

Eight bit is like shooting JPGs. If you get it perfect in camera great. But the further you are from perfect the harder it'll be to fix things.

Obviously if it's clipped it's clipped but at times you aren't totally gone.

Which applies to some Sony cameras with 'Superwhite'.

I conclude that for my modest purposes I will be sticking with 8-bit for now. But it was an interesting learning experience.

In my experience (which might be wrong as I am far from experienced like many here are) 8-bit will hold up _IF_ I have recorded one of Rec709 variations _AND_ don't push the file much. However _IF_ I record in SLog3 _AND/OR_ need to push file around 10-bit 4:2:2 will make a huge difference, difference between make it or break it, especially if I am trying green screen. In turn I don't bother recording with anything but 10-bit 4:2:2.

I share that experience. S-log is poison for 8 bit, while minor tweaks of Rec 709 can be acceptable. The general principle is getting as close as possible to your intended look by exposure, white balance and lighting and apply only minor tweaks if you only have 8 bit. Don't try to lift the shadows, in particular.
Sony has some relatively nice profiles, which can also be helpful, like Cine 4.

Canon even has the free Picture Style Editor to create your own looks, which will be applied by the camera before conversion to 8 bit. It hasn't been updated in a while, so I don't know if it still works, since we have no Canon camera around any more:
https://www.canon-europe.com/support/co ... =EN&os=all

I share that experience. S-log is poison for 8 bit, while minor tweaks of Rec 709 can be acceptable.

I have no experience with other log profiles besides Slog2 and 3 but would it be safe generalization to say 8-bit is a poison for any log curve, whether it is Sony's or some other manufacturer's?

I use Vlog L on my GH5S 8 bit and happy. I am sure it cannot be pushed too much but the colours using Vlog are much better than the other picture profiles.

If I have to recommend, I will buy the studio version.....

Isn't 8 bit log just a left over from when hybrid / ML cameras didn't tend to support 10 bit?

Yes.
And in the case of Sony it was developed for their professional "Cine Alta" line, and then they moved it to hybrids. So, the owners could brag "I shot with the same profile like the big boys and girls."
But S-log is much too flat for an 8 bit camera, and with most regular scenes it'll only use about half of the 8 bits.

I do not push my shots much and my main reason to use VlogL on the GH5S in 8bit is colour. All other picture profile do not reproduce red very well on the Panasonics. So I do not use it to recover dynamic range. It works well to match the GH5S and the GH6.

This is link to Panasonic . A little old I think as it does not include latest GH6 or G9 2 etc which are 10 bit full Vlog.

https://www.panasonic.com/uk/consumer/cameras-camcorders/lumix-expert-advice-learn/technique-technology/what-is-v-log.html

To each their own, but insisting on 8-bit log encoding when you have a camera that supports 10-bit is, in my opinion, quite ludicrous. Log encoding compresses a high dynamic range into a limited signal, and 8-bit simply doesn’t provide enough precision to retain the smooth tonal gradations needed, especially in highlights. Using 8-bit log almost guarantees banding issues, limits the flexibility of color grading, and often amplifies compression artifacts.

To each their own, but insisting on 8-bit log encoding when you have a camera that supports 10-bit is, in my opinion, quite ludicrous. Log encoding compresses a high dynamic range into a limited signal, and 8-bit simply doesn’t provide enough precision to retain the smooth tonal gradations needed, especially in highlights. Using 8-bit log almost guarantees banding issues, limits the flexibility of color grading, and often amplifies compression artifacts.

I shoot at 59.94fps and the GH5S will only shoot at 8bit for that frame rate. Temporal motion is more important to me. I see no difference to 10 bit on my GH6 . If anything the GH5S has a better image in the environments I shoot in the theatre !! I shoot 3840x2160 but edit for 1920x1080 project for output for DVD and Bluray. I accept things may be different if project was 3840x2160 in the outdoors with sky etc. But that is not the case for me. Comparing GH5S 8 bit and GH6 10bit the GH6 has more noise even using faster lens than GH5S. They do work well together with GH6 10 bit 4:2:0 full stage and GH5S mid framing and I use AX100 XAVC-S for closeups of course also 8bit !! Project is colour managed and they all match easily this way. None need to be pushed too much just a little level adjustment .

To each their own, but insisting on 8-bit log encoding when you have a camera that supports 10-bit is, in my opinion, quite ludicrous.

Vlog L and many other 'log' encodings are designed specifically for 8bit acquisition and/or telecine style grading. They do not have simple linearizable formulaic transfer functions but rather custom tone curves that try to optimize sensor capable dynamic range vs quantization. There is nothing wrong with shooting 8bit VLog L or DLog M or many other flavors of 'flat' 'log like' profiles for quicker turnarounds or simplistic post needs. These days it is a lot simpler to properly manage real log encodings but not everyone understands the subject well enough or have the needs to work like that.

To each their own, but insisting on 8-bit log encoding when you have a camera that supports 10-bit is, in my opinion, quite ludicrous.

Vlog L and many other 'log' encodings are designed specifically for 8bit acquisition and/or telecine style grading. They do not have simple linearizable formulaic transfer functions but rather custom tone curves that try to optimize sensor capable dynamic range vs quantization. There is nothing wrong with shooting 8bit VLog L or DLog M or many other flavors of 'flat' 'log like' profiles for quicker turnarounds or simplistic post needs.

I’m stepping out of this discussion; enjoy your "optimized quantization"!