Jump to content

Color - SOOC vs. LUTs/Grading


SRV1981
 Share

Recommended Posts

2 hours ago, Clark Nikolai said:

Interesting. I've been working in REC709 for the sole reason is that I'm just not ready to jump in to REC2020 and the learning curve that it might have. I also might need a new monitor or something. At some point I'll learn all about it as it seems to have huge advantages over REC709 and I want to have them. (Another thing that I want to one day learn about is ACES, another big scary thing.)

My D16 has a colour space that is like HLG called Bolex Wide Gamut. It keeps the full dynamic range of the image but isn't flat looking like Log would be so is watchable even before using a LUT or something. Here's a white paper about it.

https://digitalbolex.com/wp-content/uploads/2016/05/Bolex_Log_Bolex_Gamut_TechSum.pdf

 

 

Below is an approximate comparison of transfer curve for various Log and Sony's three versions of HLG.

From the Bolex PDF above:

image.png.91bc094fbfd25ddc255645991ee2ebb8.png

The three versions of HLG that Sony cameras support (from https://xtremestuff.net/sony-and-hybrid-log-gamma-hlg/ ) - HLG3 (upper curve) is the closest to the Rec. 2100 standard :

Sony-HLG.png?x31412

A chart of various Log formats from https://www.dpreview.com/forums/post/64243940 (note the 3 EV spread in the high EV limit between the various curves). The highest DR curves look to be C-Log2, S-Log3 and V-Log - but of course they do that by being flatter in the important mid-range area i.e. fewer levels per EV, increasing the chance of banding if pushed too far in post. Note this chart is -10 to +10 EV vs. IRE 0 to 110%, the other two are -8 to +8 EV vs. 10-bit digital code values.

62f1ebaa6a2141a2a5cee25751f8871c

Link to comment
Share on other sites

EOSHD Pro Color 5 for Sony cameras EOSHD Z LOG for Nikon CamerasEOSHD C-LOG and Film Profiles for All Canon DSLRs
7 minutes ago, ac6000cw said:

The highest DR curves look to be C-Log2, S-Log3 and V-Log - but of course they do that by being flatter in the important mid-range area i.e. fewer levels per EV, increasing the chance of banding if pushed too far in post. Note this chart is -10 to +10 EV vs. IRE 0 to 110%, the other two are -8 to +8 EV vs. 10-bit digital code values.

I read that clog2 has slightly more DR than clog3 but that in clog2 the NR is turned down/off and the shadows are way noisier than clog3.  How do you guys see that trade-off?  worth it? or would you have rather reduce noise in post.

Link to comment
Share on other sites

If you're looking for SOOC color, you probably want a decent amount of SOOC noise reduction.

If you're working on a project which will have reasonable post-production work done, you want the least (or no) noise reduction in camera.  Tools like the denoiser in Resolve, Neat Video, and Topaz Video AI give a better result as well as giving you a lot more control of the trade-offs of noise vs lost details and plastic/wax skin.

Link to comment
Share on other sites

1 hour ago, eatstoomuchjam said:

If you're looking for SOOC color, you probably want a decent amount of SOOC noise reduction.

If you're working on a project which will have reasonable post-production work done, you want the least (or no) noise reduction in camera.  Tools like the denoiser in Resolve, Neat Video, and Topaz Video AI give a better result as well as giving you a lot more control of the trade-offs of noise vs lost details and plastic/wax skin.

How would you categorize some of the main cameras discussed here ?

Link to comment
Share on other sites

7 hours ago, SRV1981 said:

makes sense when you decouple that from the original intent of the thread - minimal processing/grading or even SOOC. 

Which is it?  Minimal processing / grading? or SOOC?

They're VERY VERY different!

Just for arguments sake, let's definite "minimal processing/grading" of an image as being limited to:

  • 5 minutes of adjustments applied to the whole project
  • 30s of adjustments per clip
  • only using basic operations that can be done in PP/FCPX/Resolve using the standard tools

This definition, if you were to make a 5 minute edit, with an average shot length of 3 seconds, then that's less than an hour of colour grading for the whole project.

The differences that that effort can make will completely overwhelm any minor differences that different cameras have.  If you were to say that the best SOOC colour was 5/10 and the average was 3.5/10, then the graded images you can do in that time with those rules will easily be 9/10.

You might think that an hour sounds like a long time, but it's nothing compared to how long it will take you to edit something anyway.  Casey Neistat did his daily vlogs, which were usually between 5-10+ minutes each, and took 5-9 HOURS to edit.  This might sound like a lot, but he was an experienced editor even before he did his 800+ daily vlogs, and he also mostly knew what the film was about etc, so he wasn't filming without a plan.
I've heard other YT film-makers (where the result is a film and not a vlog or whatever) say that they spent 30 HOURS even just colour grading their 20 minute film!

The other thing that you might not be considering is that shooting for great looking images SOOC will require you to either have boring flat and lifeless looking images, or you need to crank up the saturation and contrast in-camera (as someone mentioned earlier in the thread) but this requires you to shoot really really carefully to ensure that all images are shot with exactly the right exposure and the right light levels and contrast levels.

One thing I find in colour grading is that different images require very different levels of contrast etc to look coherent together - you might have one shot with something bright or dark in the background and then the next doesn't have it - so in order to look coherent you need to adjust the contrast between the two images slightly.  Black and shadow levels is another thing that you want to try and get relatively consistent between the shots in the edit.

Shooting meticulously like this will take a lot more time during shooting than to just move a few sliders in post - you have to setup each shot, ideally to adjust the level of contrast and saturation between each setup to get a coherent look, watching your levels and histograms etc.  This would add 30s or more to each shot before you hit record.  If you're recording anything except a controlled environment where everyone is waiting for the camera to be ready then you'll miss all the good stuff.

It's fun to talk about technical things, sure, but these things aren't independent of the rest of the process.  Your question may as well begin with "Let's imagine we're in a parallel universe where instead of cameras being for making videos, they're really....." 

Link to comment
Share on other sites

Here's my recommendation for SOOC shooting - Sony AX100.  As Dave says "Sony AX100 looks better than your camera".

Just look at the nice contrast, saturation, and above all...  skintones!  and in mixed lighting no less!

Good luck getting that with a "better" camera - they all have far too much DR to give you a punchy image from their 709 LUTs or profiles.

Link to comment
Share on other sites

I've been mentioning this before,that I found the GH5 much easier to grade than the S series. It has 709 as colour space vs V-Gamut on the S cameras, also two to 2.5 stops less dynamic range. For lit scenarios I was very fine with the full curve and gamut as contrast and color was much more stylized and designed. The GH5 in MK II alteration is still a camera I wanna check out with its creamy high resolution image. It's more classic and organic compared to the crispier image of the MK I. I have only seen one person ever with the MKII. It has 10bit in 60p 4K as well. And from the Mark I the beloved S16 crop or optional 2/3" mode via 2x digizoom. Is the latter without heavy aliasing and moiré on the GH5? @kye

Link to comment
Share on other sites

7 hours ago, kye said:

Here's my recommendation for SOOC shooting - Sony AX100.  As Dave says "Sony AX100 looks better than your camera".

Just look at the nice contrast, saturation, and above all...  skintones!  and in mixed lighting no less!

Good luck getting that with a "better" camera - they all have far too much DR to give you a punchy image from their 709 LUTs or profiles.

I want one!

Link to comment
Share on other sites

9 hours ago, kye said:

You might think that an hour sounds like a long time, but it's nothing compared to how long it will take you to edit something anyway.  Casey Neistat did his daily vlogs, which were usually between 5-10+ minutes each, and took 5-9 HOURS to edit.  This might sound like a lot, but he was an experienced editor even before he did his 800+ daily vlogs, and he also mostly knew what the film was about etc, so he wasn't filming without a plan.

Based on my own experience of putting together 10-20 minute YT videos, I'd agree with that. I don't usually do the whole editing process in one go - more often it's broken into 1-2 hour sessions and from start to finish it spans a week or more.

My usual process to create something for YT is:

  • Decide on the topic/subject for the movie (based on what content I've already got, usually from an event or place I've visited, recently or in the past).
  • Choose the clips that might be included.
  • Decide on a 'flow' for the movie - is it going to be basically chronological or geographical?
  • Put the clips that might fit into the flow into the editor media bin.
  • Put the clips onto the timeline in the right order. There might be 30 or more.
  • Do a very rough cut and play it. Delete or trim the clips to improve the flow and get the overall length within the target (it's almost always too long to start with). Repeat the cut & trim cycle until happy.
  • Go through the timeline and adjust brightness/contrast/saturation/colour as necessary (this can be very time consuming!). If a clip is proving really awkward to adjust, decide if it's essential to the flow. If not, could it be discarded or replaced with different one?
  • Do a similar adjustment process for the audio content (equally or more important as the picture content).
  • For the wobbly clips (which for old content will be quite a lot of them!) try to stabilise them - another time consuming activity!
  • Add titles to tell the viewer more about what they're looking at. This often involves some research to find the information.
  • Add timeline markers and generate a list of YouTube chapter points from them.
  • Render out the movie and play it through to review it. Fix the issues found (there will be some...).
  • Render it out again and upload it to YT as a 'private' video.
  • Add a description, the chapter list, a suitable thumbnail and the 'end screen' stuff in the YT 'Studio'.
  • Check it plays and looks OK on YT.
  • Make it 'public' on YT and hope al least a few people watch it for more than 10 seconds...

The above also needs cups of coffee and maybe wine too 😉

SRV1981 - next time you watch something on YT, just think about how much creative work and time has probably gone into it.

 

 

Link to comment
Share on other sites

30 minutes ago, PannySVHS said:

I've been mentioning this before,that I found the GH5 much easier to grade than the S series. It has 709 as colour space vs V-Gamut on the S cameras, also two to 2.5 stops less dynamic range. For lit scenarios I was very fine with the full curve and gamut as contrast and color was much more stylized and designed. The GH5 in MK II alteration is still a camera I wanna check out with its creamy high resolution image. It's more classic and organic compared to the crispier image of the MK I. I have only seen one person ever with the MKII. It has 10bit in 60p 4K as well. And from the Mark I the beloved S16 crop or optional 2/3" mode via 2x digizoom. Is the latter without heavy aliasing and moiré on the GH5? @kye

I'm not familiar with the GH5v2 but Panasonic was (at that time) updating cameras with all the user-feedback, and your description was certainly things that the community was wanting.

I definitely agree that one of the main challenges is taking a clip that was shot in LOG and has 10-14 stops of DR in it, and somehow stuffing that into Rec709 which has just over 5 stops of DR.  This obviously manifests in having to crush or severely compress various areas of the luminance range, but it also means that the source material can have colours that are dramatically more saturated than Rec709 can contain and you'll need to work out how to contain those too.

Once you have enough DR to shoot the scenes you need to shoot, having more is actually a liability rather than a feature.

2 minutes ago, ac6000cw said:

Based on my own experience of putting together 10-20 minute YT videos, I'd agree with that. I don't usually do the whole editing process in one go - more often it's broken into 1-2 hour sessions and from start to finish it spans a week or more.

My usual process to create something for YT is:

  • Decide on the topic/subject for the movie (based on what content I've already got, usually from an event or place I've visited, recently or in the past).
  • Choose the clips that might be included.
  • Decide on a 'flow' for the movie - is it going to be basically chronological or geographical?
  • Put the clips that might fit into the flow into the editor media bin.
  • Put the clips onto the timeline in the right order. There might be 30 or more.
  • Do a very rough cut and play it. Delete or trim the clips to improve the flow and get the overall length within the target (it's almost always too long to start with). Repeat the cut & trim cycle until happy.
  • Go through the timeline and adjust brightness/contrast/saturation/colour as necessary (this can be very time consuming!). If a clip is proving really awkward to adjust, decide if it's essential to the flow. If not, could it be discarded or replaced with different one?
  • Do a similar adjustment process for the audio content (equally or more important as the picture content).
  • For the wobbly clips (which for old content will be quite a lot of them!) try to stabilise them - another time consuming activity!
  • Add titles to tell the viewer more about what they're looking at. This often involves some research to find the information.
  • Add timeline markers and generate a list of YouTube chapter points from them.
  • Render out the movie and play it through to review it. Fix the issues found (there will be some...).
  • Render it out again and upload it to YT as a 'private' video.
  • Add a description, the chapter list, a suitable thumbnail and the 'end screen' stuff in the YT 'Studio'.
  • Check it plays and looks OK on YT.
  • Make it 'public' on YT and hope al least a few people watch it for more than 10 seconds...

The above also needs cups of coffee and maybe wine too 😉

SRV1981 - next time you watch something on YT, just think about how much creative work and time has probably gone into it.

I co-produced a 5-min short with my sister a long time ago, and we estimated that all up it had 10,000 person-hours in it.

But enough of this blasphemous film-making talk - we should go back to talking about camera colour profiles like film-making doesn't exist!

Link to comment
Share on other sites

1 hour ago, kye said:

I co-produced a 5-min short with my sister a long time ago, and we estimated that all up it had 10,000 person-hours in it.

Not at all surprised - doing that is far more complex and creative than the simple documentary stuff I normally produce.

Link to comment
Share on other sites

1 hour ago, kye said:

Once you have enough DR to shoot the scenes you need to shoot, having more is actually a liability rather than a feature.

I agree.

1 hour ago, kye said:

I definitely agree that one of the main challenges is taking a clip that was shot in LOG and has 10-14 stops of DR in it, and somehow stuffing that into Rec709 which has just over 5 stops of DR.  This obviously manifests in having to crush or severely compress various areas of the luminance range, but it also means that the source material can have colours that are dramatically more saturated than Rec709 can contain and you'll need to work out how to contain those too.

I often apply an S shaped contrast curve, compressing the highs downwards and lows upwards and expanding the mid-range to increase contrast. Balancing the compression and expansion (and the inflection points) to get it look nice is the tricky part of course....

And then there's the accuracy or otherwise of the Rec 2020 to Rec. 709 colour conversion - I think every HLG to Rec 709 conversion LUT I've tried has a different take on this...

Link to comment
Share on other sites

1 hour ago, kye said:

enough of this blasphemous film-making talk - we should go back to talking about camera colour profiles like film-making doesn't exist!

I just realized - I did come to the wrong forum. I just assumed this was a wider community than what actually is. Many view themselves as filmmakers. I do not, I’m sure there’s other forums/places to chat just about equipment etc and not ruffle weathers of those who see themselves as “filmmakers”. 

Link to comment
Share on other sites

11 minutes ago, SRV1981 said:

I just realized - I did come to the wrong forum. I just assumed this was a wider community than what actually is. Many view themselves as filmmakers. I do not, I’m sure there’s other forums/places to chat just about equipment etc and not ruffle weathers of those who see themselves as “filmmakers”. 

Don't take forum discussions and comments quite so seriously - Kye discusses equipment too. I think this forum has always been biased towards that side of things. But the equipment is used as creative tools, so both it and the creativity are very much linked together. But if you're mainly interested in the equipment that's fine as far as I'm concerned - probably most of the threads on the forum are related to equipment and tools.

Link to comment
Share on other sites

20 minutes ago, SRV1981 said:

I just realized - I did come to the wrong forum. I just assumed this was a wider community than what actually is. Many view themselves as filmmakers. I do not, I’m sure there’s other forums/places to chat just about equipment etc and not ruffle weathers of those who see themselves as “filmmakers”. 

Your questions have already been answered here. You're simply just trolling at this point, and not very well at that. Not sure what exactly is is you ARE looking for..

Link to comment
Share on other sites

36 minutes ago, ac6000cw said:

I agree.

I often apply an S shaped contrast curve, compressing the highs downwards and lows upwards and expanding the mid-range to increase contrast. Balancing the compression and expansion (and the inflection points) to get it look nice is the tricky part of course....

And then there's the accuracy or otherwise of the Rec 2020 to Rec. 709 colour conversion - I think every HLG to Rec 709 conversion LUT I've tried has a different take on this...

What software are you grading in? and what's your pipeline?

I'm in Resolve, so tend to use Colour Space Transforms, which gives a whole other way to look at things because everything is on the table...  the ARRI LUTs, Print Film Emulations, etc etc.

27 minutes ago, SRV1981 said:

I just realized - I did come to the wrong forum. I just assumed this was a wider community than what actually is. Many view themselves as filmmakers. I do not, I’m sure there’s other forums/places to chat just about equipment etc and not ruffle weathers of those who see themselves as “filmmakers”. 

Lots of discussions about equipment occur here, but it's basically impossible to discuss cameras without doing it in the context that they're for a particular thing.  

Otherwise you could be talking about how the GH1 is better than the Z8 because it makes a better paperweight, or how an AE-1 is the best camera ever because when you mount it on a pole in your cornfield the sun glints off the mirror and keeps the crows away.

Without context there is no basis to discuss anything.

Link to comment
Share on other sites

14 minutes ago, ghostwind said:

Your questions have already been answered here. You're simply just trolling at this point, and not very well at that. Not sure what exactly is is you ARE looking for..

Not really, I've asked basic questions for discussion purposes but the elitism and condescension as your post demonstrates is strong here - i'd rather just find a place with more amateur enthusiasts rather than certain folks who think highly of themselves

Link to comment
Share on other sites

On 4/14/2024 at 3:00 PM, SRV1981 said:

Is it fair to say some cameras produce normative or more pleasing color to most but if using log, you can get similar color/image from most cameras equally?  
 

1. if you wanted a personal camera with fast turn around what brands are you usually happy with color wise?

 

2. when deciding for more professional or bigger projects, how do you decide what system/log system to get?

 

seems canon and then Fuji reign when SOOC is discussed and it’s more nuanced for the latter.

Literally a non-controversial, non-trolling, basic question that you can partake in politely or not. 

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

  • EOSHD Pro Color 5 for All Sony cameras
    EOSHD C-LOG and Film Profiles for All Canon DSLRs
    EOSHD Dynamic Range Enhancer for H.264/H.265
×
×
  • Create New...