I'm a newb to 10 bit footage in vegas pro. I just got a Sony ZV-E10 that records 10 bit footage, and I want to take advantage of it
 

What are the camera settings to record the 10bit footage?

I'm a newb to 10 bit footage in vegas pro. I just got a Sony ZV-E10 that records 10 bit footage, and I want to take advantage of it
 

What are the camera settings to record the 10bit footage?

The ZV-E10 (Mark 2) records in 10bit 422

I'm a newb to 10 bit footage in vegas pro. I just got a Sony ZV-E10 that records 10 bit footage, and I want to take advantage of it
 

What are the camera settings to record the 10bit footage?

I'm also curious, 'cos E10 cannot record in 10bit, nor output.

Maybe E10 Mark II?

I'm a newb to 10 bit footage in vegas pro. I just got a Sony ZV-E10 that records 10 bit footage, and I want to take advantage of it
 

What are the camera settings to record the 10bit footage?

I'm also curious, 'cos E10 cannot record in 10bit, nor output.

Maybe E10 Mark II?

Yes the mark 2

off: i'm interested in this cam. Mind you share some short 10bit footage so we can see the advantages?

10 bit 422 HEVC only decodes with Intel GPUs. 420 decodes with any GPU for better performance without proxy files.

Try 32 bit full with view transform off as a high bit alternative to 8 bit full. I assume you are shooting log, otherwise you may benefit less from 10-bit files (they hold up better to extreme color correction).

ACES is an option if your camera settings match one of the media color spaces like Slog3 sgamut3.cine. In that case use the sRGB view transform.

No need for special render settings, the point is to use the extra data precision so the image doesn't fall apart going from log to Rec709. 8-bit is fine for viewing.

Can you elaborate on the "420 decodes with any gpu for better performance without proxy files".

Essentially, I'm getting a new camera. Wanting to go from 8 bit to 10 bit. Let's say I record in 10 bit, but only record in 4.20. Are you saying it decodes so well that it is even faster/better than having a proxy file? I know in the past I could have a very weak intel machine, and 4k footage was no problem with a proxy file being used. I've been able to edit off of all thin and light laptops, with 16GB of RAM, at times even with only 8GB of RAM, mostly no problem at all, all because of proxies. Will I really get the same performance WITHOUT proxies from recording in 420?

The one kicker to this whole issue: I'm actually using an ARM device. I'm using the Snapdragon X Elite, which is supposed to be supported SOON, and ARM was shown running Vegas Pro 22 at an event a few months before the release of VP 22. But it's still not officially out for arm, so basically it takes me 2 hours to render a video. Even though I'm using "emulation" I have been able to easily edit videos through proxies,

But now I'm upgrading to 10 bit on the ZV-E10 (Mark 2), I'm open to shooting 4.22 or 4.20. I'm just trying to understand what settings I need to use in Vegas to ENSURE I'm properly taking advantage of 10 bit. And while I'm at it I'm trying to learn more because I'm very ignorant. I don't plan to use HDR.

I should add that Davinci Resolve is a pain (feels much more bulky than Vegas) but it does FULLY support ARM today. So I'm on the fence about switching if I can't get Vegas working well. Yes it takes 2 hours to render a video, but I can do that while I'm doing other things. The key is being able to quickly edit with proxies. OR not need proxies.

Also, I need to make sure whatever proxies I make, I need to make sure those are 10 bit proxies too. That's a whole different thing I can't figure out with vegas.

off: i'm interested in this cam. Mind you share some short 10bit footage so we can see the advantages?

That's going to take some time. Many of the advantages are found in color grading and editing. But I haven't even figured out how to edit any footage yet. But it's for my Youtube Channel (Android Digest) so hoping to improve some things.

i dont ask for color graded or edited footage. Instead, 10bit videos straight from the cam.

"I haven't even figured out how to edit any footage yet. "

That's why i ask for footage, for giving you editing tips :)

10 bit 422 HEVC only decodes with Intel GPUs. 420 decodes with any GPU for better performance without proxy files.

Try 32 bit full with view transform off as a high bit alternative to 8 bit full. I assume you are shooting log, otherwise you may benefit less from 10-bit files (they hold up better to extreme color correction).

ACES is an option if your camera settings match one of the media color spaces like Slog3 sgamut3.cine. In that case use the sRGB view transform.

No need for special render settings, the point is to use the extra data precision so the image doesn't fall apart going from log to Rec709. 8-bit is fine for viewing.

Can you elaborate on the "420 decodes with any gpu for better performance without proxy files".

Essentially, I'm getting a new camera. Wanting to go from 8 bit to 10 bit. Let's say I record in 10 bit, but only record in 4.20. Are you saying it decodes so well that it is even faster/better than having a proxy file? I know in the past I could have a very weak intel machine, and 4k footage was no problem with a proxy file being used. I've been able to edit off of all thin and light laptops, with 16GB of RAM, at times even with only 8GB of RAM, mostly no problem at all, all because of proxies. Will I really get the same performance WITHOUT proxies from recording in 420?

Yes, you can shoot 10-bit 420 HEVC or AVC and likely play back at full speed with no proxy file now that VEGAS has better decoders. With NVDEC I'm getting 100+ FPS with a 2080 GPU and some formats. 422 requires the CPU to decode- it may be able to play back at full speed if the CPU is fast.

The one kicker to this whole issue: I'm actually using an ARM device. I'm using the Snapdragon X Elite, which is supposed to be supported SOON, and ARM was shown running Vegas Pro 22 at an event a few months before the release of VP 22. But it's still not officially out for arm, so basically it takes me 2 hours to render a video. Even though I'm using "emulation" I have been able to easily edit videos through proxies,

When native VEGAS support comes it may get way faster.

But now I'm upgrading to 10 bit on the ZV-E10 (Mark 2), I'm open to shooting 4.22 or 4.20. I'm just trying to understand what settings I need to use in Vegas to ENSURE I'm properly taking advantage of 10 bit. And while I'm at it I'm trying to learn more because I'm very ignorant. I don't plan to use HDR.

The settings I described are what I'd use. Edit in 8-bit full, color and export with 32-bit full.

I should add that Davinci Resolve is a pain (feels much more bulky than Vegas) but it does FULLY support ARM today. So I'm on the fence about switching if I can't get Vegas working well. Yes it takes 2 hours to render a video, but I can do that while I'm doing other things. The key is being able to quickly edit with proxies. OR not need proxies.

 

Also, I need to make sure whatever proxies I make, I need to make sure those are 10 bit proxies too. That's a whole different thing I can't figure out with vegas.

You don't need or want 10-bit proxies, you use the originals for coloring (best/full).

i dont ask for color graded or edited footage. Instead, 10bit videos straight from the cam.

"I haven't even figured out how to edit any footage yet. "

That's why i ask for footage, for giving you editing tips :)

@relaxvideo Look at the duck videos, the interesting ones to download are 1,2,4, they are rec709, but you can simulate that it's log if you're interested in Vegas performance with a color transform plus grade.

Vegas plays 1 and 2 poorly for me, 32bit project with color grade. These are the HEVC variants

@RogerS if you or others with 422 10bit GPU decoders could look at duck videos 1 and 2 and report back performance 32bit project, and could compare it to upcoming release2, if you're beta testing release2 then that makes things awkward.

https://www.photographyblog.com/reviews/sony_zv_e10_ii_review

You don't need or want 10-bit proxies, you use the originals for coloring (best/full).

Most pro editors have 10 and 12bit proxy options, but to the OP, Vegas's proxy is 8bit

I don't think there's any difference between this Sony footage and that from similar cameras so I've tested it all already. But I agree, no point in testing until the next release is out (apparently was supposed to be released 9/3 according to Hub so it's imminent).

What's the benefit of a high-bit proxy? In those workflows people do color grading on the proxy file not the original? Or is it for fine tuning masks and compositing that benefits from greater precision and being able to play back footage in real-time?

@RogerS Re: High bit proxies. That's one of the advantages VegasDerek mentioned with the Prores proxies, it's very important for caching - alpha channels, green screens, color grading. As Vegas doesn't currently have automated caching the 10/12?bit proxy will give better results than 8bit, require less switching back between original and proxy.

These files don't contain P frames, they're all I and B frame, with a high GOP, 48 and 96 but as you say, Sony's with recent firmware should all be like that with HEVC.

Thanks, that makes sense to me. I look forward to the next step in VEGAS's evolution.

@AndroidDigest difference between pixel format and color format in laymen terms :) (if you're still interested)

Pixel format is the technical representation of color. Your eye has 3 different cell types that can differentiate color. They are most sensitive at blue, green and orange. So your software uses a 3 component pixel format to edit images or video frames called RGB (Red/Green/Blue). It's the usual and most accurate representation to calculate effects and the like.

There is a 4th type of cells that can "only" differentiate brightness levels. But this type is way more sensitive to light then the 3 types the human eye uses for color perception. And the eye has way more of this 4th type (ca. 120 million) compared to the 7-8 million color sensitive cells. So for storing or transmitting images a technical trick is used to convert this accurate RGB format in something called YUV. The Y stands for a brightness/luminance component, U and V are used for color. As you are much more sensitive to luminace than color some color information can then be dropped without any visual disadvantage (you can measure it, but you won't see it most of the time).

That are these subsampling terms used:
- 4:2:0 means a block of 2x2 pixel uses 4 luminance values, but only one pair of uv color information
- 4:2:2 means a block of 2x1 pixel (2 neighbour pixel) have 2 luminance values but share a uv color information
- 4:4:4 is most accurate and every pixel has its own luminance and uv color values

So YUV 4:4:4 is as a accurate as RGB, YUV 4:2:2 and 4:2:0 drop some color information. But in a way you usually cannot see. Especially with the footage a camera can record. That's the pixel format.

The color format kind of describes what is meant with "Red"(and "Green"/"Blue") Is it more purple? Or more orange? And is R=0.5 really half as red as R=1.0? Or is there another "transmission" curve?

And the color format describes, how much of the possible colors your eye can see can be expressed in RGB values between 0.0 and 1.0. That's hard to explain, but think of it as a subset of all possible colors.

So it is basically a translation of this technical RGB values to real colors (of a certain wave length) and depending on the color format a RGB value of (0.5, 1, 0) may represent a different real color/different wave length of the light.

I hope that's not too confusing... 🙂

With that said, the typical GPU can usually decode 4:2.0 formats with hardware acceleration, 4:2:2 and 4:4:4 is typically decoded much slower in software. So as you won't see the difference from 4.2:0 to 4:2.2 I'd recommend using a 4:2:0 format for the ease of handling in Vegas with a high resolution and high bit depth.

For the resolution you may want to experiment as your camera lenses have a optical resolution as well. So the preset with the highest resolution only makes sense, if the optical resolution of the lense system can catch up with it.

And the bit depth is simply, how many levels of red (and green/blue) can be distinguished...

I'm a newb to 10 bit footage in vegas pro. I just got a Sony ZV-E10 that records 10 bit footage, and I want to take advantage of it, and edit it. I plan to use proxies anyway.

My question: Do I need to change the color format? Would it help me at all? If so, what Pixel Format should I use? And what aces info should I use? I'll be recording in that 10 bit 422 format.

If you can help me set up my "project settings" to fix this, that would help. Thank you

 

PS: If you explain the difference between "Pixel Format" and the color format, that helps, but please do it in laymen's terms. The biggest thing is I want the very best project settings to work with 10 bit, and if I need to change any render settings I will.

Please post your Vegas and MediaInfo Properties for assistance on your first four questions.

I can promise that there is quite a bit one would need to know about your source footage to give you workable advice about setting your Project and Render Properties.

https://www.vegascreativesoftware.info/us/forum/faq-how-to-post-mediainfo-and-vegas-pro-file-properties--104561/

There is good advice in the previous replies to your last question; however, tuck that away for a while until you learn the vocabulary; also, having a working knowledge of Algebra 1 won't hurt even a little bit.

https://www.vegascreativesoftware.info/us/forum/speaking-good-video-a-beginner-s-guide--104463/

First thing you need to ask yourself... what is my target format? Most target formats are 8 bit... DVD, Blu-Ray (except 4K Blu-Ray), broadcast TV, etc. are all 8 bit. The only delivery format that is 10 bit is HDR. Youtube, Facebook, Instagram, Tiktok all support HDR video. You may notice that some videos you open in those platforms appear much brighter and more vibrant on your phone than others, those are the HDR videos.

There are some advantages to color grading 10 bit video over 8 bit, as well as 4:2:2 over 4:2:0, so that's why, even if you are targeting 8 bit, it's better to shoot in those formats. You can push them farther when it comes to changing the color or brightness levels than you can 8 bit 4:2:0 video. You also get cleaner chroma key when doing green screen when you have that extra color information to work with.

It isn't as simple as just "10 bit or not", though... you also have to take gamma curve into account. Are you shooting a Log format? Log formats use a logarithmic curve to their gamma, which creates a very desaturated, low contrast look to the video, but it actually captures more of the sensor's dynamic range in the process of doing this. The end result is video that must be color corrected (often by using LUTs or ACES view transforms to do the conversion) back to "normal looking" video, and the coloring the video to your liking from there. I suggest looking both of those things up on youtube if you want to learn more about what they are and how they work.

Ideally, you'd want to shoot in a log format with your 10 bit video. Sony cameras use what is called "SLog", and there's different variants of that, like "Slog, Slog2, Slog3" etc.

To color grade Slog Video in VEGAS Pro, you want to import the video, then right click on it in the media pool, and select "properties". Go into the color space options on the following menu, and select the "Slog" type that you recorded in, so "Slog2" for example. You'll set that up in your camera's settings menu when recording, so remember what type you select.

If you don't record to "Slog" and just do regular 10 bit, then "Rec709" will be the color space you want to select, but again, I recommend using SLog to get the most out of your footage.

After setting the color space, edit your video as you would with 8 bit project settings (the default). This edits faster than 32 bit, so we'll switch that later. Once the editing is complete, you want to go into project settings, and set the project to 32 bit (full range) mode. Then set the view transform to your target format. If you want to create an HDR video, you'll turn HDR mode on at this point... just keep in mind you must have a HDR monitor connected to VEGAS in order to see what you are doing, otherwise you'll be flying blind and could end up with much different results than what you see on your screen. If you are targeting standard 8 bit video as your export format, then leave HDR mode off, and set your view transform to either sRGB or Rec709.

As long as you've set your color space properly, you should notice that your video no longer looks low contrast and desaturated (assuming you shot it in SLog), VEGAS automatically converted it to "normal looking" video for you. Now you can use the color grading panel to adjust the image to look however you want it to. Get it looking how you like, and then render it out from there.

There is a 4th type of cells that can "only" differentiate brightness levels. But this type is way more sensitive to light then the 3 types the human eye uses for color perception. And the eye has way more of this 4th type (ca. 120 million) compared to the 7-8 million color sensitive cells.

@VEGASMichael

Can you point me to some peer-reviewed resources for this information?

Thanks.

@VEGASMichael

Can you point me to some peer-reviewed resources for this information?

Thanks.

@mark-y He's talking about rods and cones, rods make up 95% of photo receptors - brightness levels, night vision, the cones , short, medium and long wave receptors send the signals to your brain for processing into colors. There's also another photo receptor that's sensitive to blue light that controls the production of melatonin, and that's why phones/monitors have night mode that filters blue light, these cells are not used for vision.

@AndroidDigest Here's a pretty comprehensive review of the Sony ZV-E10 II that includes 5 downloadable video clips at the end.

https://www.photographyblog.com/reviews/sony_zv_e10_ii_review

Here's a summary of the media info of the 5 video clips:

If those clips are representative, it looks like the camera's top write speed to media is 200 mbps. Given that the sensor pixel geometry matches up to 4:2:0 encoding before in-camera upscaling, and that the camera's 4k transfer rate to media is capped, I'd recommend hevc 4:2:0 media capture to get the most real 10-bit sensor data out of the camera. Unless you can record some other way at a higher bit-rate, like via sdi or hdmi.

^This would depend upon the camera's native resolution, if it is capturing a much higher native pixel resolution, and downsampling from there to say, 4K, then that changes things. In this case, we're going from 26 megapixels down to 8 for 4K. At 26MP, you have 3 times more pixels than you need for 4K, so even if the bayer pattern has far more green pixels than red or blue, then you're still getting a lot more than 4:2:0 upon downconverting that to 8mp (depending on how that conversion is done in-camera, there are enough physical pixels of each color there to end up with a higher than 4:2:0 end-result).

However, 4:2:0/4:2:2/4:4:4 has to do with YUV compression chroma sub-sampling of the image after capture, not physical pixel characteristics.

As for 200mbps, I record at 150 and still get a very noticeable improvement when recording in 10 bit over 8 bit, and some beautiful HDR end results, so that should be just fine.

 

However, 4:2:0/4:2:2/4:4:4 has to do with YUV compression chroma sub-sampling of the image after capture, not physical pixel characteristics.

@fr0sty Howard won't ever change his mind about this, but for the education of others read why it's possible to have real 422 color from a single sensor and why this isn't a camera conspiracy.

https://cinematography.com/index.php?/forums/topic/72071-is-422-from-a-4k-sensor-real-4k-422/