Rendered at 17:57:10 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
Spiwux 12 hours ago [-]
I don't understand this post. It doesn't explain the concrete problems with grading log footage. And you can convert log footage to REC709 in a way that would give you the same result as if you'd just recorded in REC709 to begin with. At least for most established camera brands. Maybe this post is referring to grading more fringe brands that have bad/undocumented color science or don't offer proper conversion paths?
Either way it does not explain the concrete issues of recording in log. If you have a strong title like "Stop doing X" you better have some damn good arguments if you think you can tell other people what to do.
rickdeckard 9 hours ago [-]
He's a colorist, stating that many videographers shoot log incorrectly and decide to defer the work to post-production, creating unnecessary complexity and bad source-material.
The plea is for filmmakers to stop following the hype, use what they are experienced with and take ownership of exposure again ("Many times a videographer will shoot log incorrectly – generally underexposed – expecting to ‘fix it in post’").
Because in his experience log doesn't add substantial value, it creates lower-quality base-material with substantial post-production work.
It makes sense. Creative work comes from decisions being taken, log means that the videographer is not deciding but is deferring the task to post-production, creating a unnecessary burden (and possibly diluting his own creativity).
whizzter 8 hours ago [-]
I can kinda see "both sides of it".
Technically, the issue if I'm getting is that the camera physically (or in teh encoding) will be better at capturing within a certain range that you adjusted for, having parts above or below that isn't capture properly and if material is in those above or below ranges the output after grading will not have a good range of values (ie possibly introducing banding).
Underexposing probably feels like a "safe fix" rather than adjusting properly all the time since with more limited ranges it'd be easy to "wash out" with overexposure (irreparably removing things due to clipping), and with budgets cut it's probably easy to just skimp on a lot of these "details" if the system "is capable".
So it's budgets/speed/creativity at shooting time vs material that might just not be good to work with in the end. (Also as he mentions some of these systems aren't as sensitive, probably leaves a situation where some people who's used the better systems just don't realize that it's a problem coming to others).
Edit: Noticed some other comment mentioning that these "log" encodings are 10bits, yeah that's going to leave a bad trail of numeric resolution if you've exposed badly, it'll probably be workable for a huge range but it's not enough by far to recover really badly exposed stuff.
rickdeckard 8 hours ago [-]
I see the rationale for the other side. But it's the "safe fix" as you said, to create maximum room for a decision-maker in post.
The way I read it, the point is especially for experienced videographers to stop playing it safe but "making the call".
The author implies that many of them now "shoot for post-production" which basically leaves alot of their skill dormant to just create much more tuning/decision-work down the line.
From the perspective of a colorist it makes alot of sense to me. The shot and its "message" should be decided on-set already, deciding ALL of this in post is not creating as much value as people think, it rather dilutes the work as a whole.
Or as Orson Welles said: "The enemy of art is the absence of limitations"
yokoprime 11 hours ago [-]
I think the message was "stop trying to fix things in post", but I too am still I bit bewildered as to why it's an issue.
poulpy123 6 hours ago [-]
I'm not in the field so I can say I'm right, but it's quite common that the "I will fix later" mindset end in in worse result even if it's technically doable because of lack of time, lack of money, lack of interest or more importantly, lack of intent
Spiwux 11 hours ago [-]
Upon re-skimming the post with your comment in mind I think you are right. The main issue (from the author's point of view) seems to be: You're more likely to record bad footage in log because you think you can just fix it in post.
That's certainly an interesting take to have (and it could've been compressed to a single paragraph)
simoncion 8 hours ago [-]
> (and it could've been compressed to a single paragraph)
It seems to me that it was. The essay's opening paragraph is:
I may ruffle some feathers with [the "Stop Filming in Log" title]; but after years as a colorist, log-encoded clips can be an unnecessary challenge. Using camera log profiles is often a crutch for location and studio set-ups. Or it’s simply used because the internet tells you log = film. That’s just not true. Many times a videographer will shoot log incorrectly – generally underexposed – expecting to ‘fix it in post’. The same issues apply to shooting in camera raw or HDR.
fps-hero 7 hours ago [-]
I think it’s necessary to separate the art from the science.
A CCD captures light with a fixed exposure time and analog + digital gain. This gives you a raw, or in photographer speak, RAW frame. The issue being that the required bandwidth to store RAW frames of video is simply too high to be practical.
So, you need some form of compression. RED is a great example of seeing the problem and solving it. Now the issue is compressed RAW video is a patent minefield, and just recorded video equally so.
And so, we are resorting to tricks such as “log raw encoding”, or “just get the videographer to capture it right in the first place” just to get around the limitations of the storage format.
Honestly capturing images and displaying them accurately is such a mess that it’s a miracle that we have more than 8-bits of colour, and that was hard enough as it is. From a RAW Bayer frame to what you see on a monitor preview display is multiple layers of white balance, gain, equalisation, linear to gamma, and whatever your monitor decides to do with the digital signal it’s fed.
rickdeckard 9 hours ago [-]
yep, also my takeaway from the article, and I can see the point.
It's kinda like always shooting photos in RAW, always 4-5 photos of every scene, but then never taking the decision/work which of them is representing the scene you were trying to capture.
simondotau 8 hours ago [-]
Shooting RAW makes perfect sense when you’re both photographer and editor. It lets me defer technical and creative decisions until later, freeing me to focus on capturing the moment. Anything technology can do to reduce the number of variables I need to think about at the time of capture is a win imho.
What I’d really like to see in future cameras is subject-aware shutter speed selection, where the camera intelligently sets a minimum shutter speed based on what’s happening in the scene. I’d also love real-time feedback on image quality: the camera analyses a shot within a second of capture and give a subtle beep or chime if, for example, neither eye is in focus or the subject is blurred by motion. The less time I spend worrying about the technical details, the more attention I can give to the photograph itself.
rickdeckard 8 hours ago [-]
> "Shooting RAW makes perfect sense when you’re both photographer and editor."
Exactly. And that article describes a scale where both tasks are assigned to different roles, with the videographer mainly working to create more room for later decision-making (than practically needed).
in my (exaggerated) simile to casual RAW photography, the photographer defers technical and creative decisions to a later step that then never happens, so he actually never conveys his message of that captured moment.
In this context the article makes most sense to me.
"You are there, you know the equipment, you know what you want to capture, go and do it."
KaiserPro 9 hours ago [-]
Hmm unless something has changed, rec.709 is 8bit, so yeah if you're good you can just shoot in rec.709
But if you want to play with the colour, loosing 4bits of colour precision means you aint recovering whats in the blacks. (log supposedly allows better detail recovery in blacks, but that was for film, I don't know about digital anymore)
ikatson 8 hours ago [-]
Disclaimer: I'm not a video pro, have nothing to do with that industry.
"rec.709 is 8bit" sounded wrong, so a quick check suggests it's either 8 or 10 bit [1], where 10 bit formats may be used during production, and 8 bits for delivery [source: gemini].
Ahhh this must have changed after I left the VFX buisness.
rickdeckard 9 hours ago [-]
> "And you can convert log footage to REC709 in a way that would give you the same result as if you'd just recorded in REC709 to begin with"
I think beside of this extra-step in post (which comes with its own set of parameters again), that's part of his point:
The videographer didn't expose for a specific look in a scene, he didn't decide at all and postponed to post-production. This is not adding value, it's creating work.
"Many times a videographer will shoot log incorrectly – generally underexposed – expecting to ‘fix it in post’. The same issues apply to shooting in camera raw or HDR."
ryandamm 10 hours ago [-]
I think what the author is trying to say is that directors of photography don’t know how to expose an image properly when they are targeting Rec709 output but shooting and monitoring in a different color space.
But this is a nontechnical article for a nontechnical audience.
buildbot 12 hours ago [-]
Color science/color management seems like a complete dark art; super interesting post. The rest of the blog is full of cool posts about film production in general.
It’s also a field chock full of misinformation. A couple of years ago I made a short film and when I was looking to color grade it I found many apparently well respected resources that offered outright contradictory information on what should be straight up technical topics. It was very surprising to me.
I’m not sure I really buy the original article here. The reason I shot in log is because it recorded a wider color gamut than otherwise and skipped in camera processing. Doesn’t mean I didn’t try to nail the exposure in the first place, just that I had more latitude to make adjustments in post without as much image quality loss.
t0bia_s 10 hours ago [-]
Does your audience look at your graded footages in wide gammut displays?
buildbot 1 hours ago [-]
These days a lot of people probably do between phones and decent OLED TVs.
enneff 10 hours ago [-]
No, and I don’t distribute it in HDR. The point of shooting that way is I have more latitude for adjustments in post.
t0bia_s 3 hours ago [-]
Sure but either you create unrealistic scene where nothing is clipped but it lack a depth or you need wide gamut monitor. So technically, unless your content is created to cinema or HDR streaming service, log is pointless and you need properly lit scene with standard profile.
LoganDark 10 hours ago [-]
Yeah, I'm shocked by how many people get it wrong. Linux is catching up to macOS, but Windows may never get proper color management.
lambdaone 10 hours ago [-]
The author is being intentionally provocative, as they admit in their first sentence. I can see what they are trying to get at, which is that film-makers should not abuse HDR and set their exposure intentionally for viewer comfort, rather than trying to fix everything in post, but shooting Rec. 709, an old colour standard with wonky primaries derived from old-fashioned CRT TV phosphors, seems like a really peculiar way of going about it. "Make it look like DVD" should not be the ideal of film-making.
prmoustache 6 hours ago [-]
> "Make it look like DVD" should not be the ideal of film-making.
I don't know, it really depends what is the target audience.
I don't know the technical details but I found out that many modern TV shows have a lots of scenes with what I would call the netflix/youtube influencer's look with a very dark room and various small lights in the background that just don't work well if you don't have a very good screen.
And a lot of people are watching these productions from devices of varying qualities.
It is a bit like criticizing the loudness war and overuse of compressors in music production. Sure it makes for a huge loss of details when listening in your living room with a good quality equipment and in a quiet place. But most people listen to music in hostile environment: in a vehicle, in a car, in badly insulated rooms/apprtments with noisy neighbors and traffic, with shitty bluetooth speakers, etc.
Sometimes you can target for the beat equipmemt, sometimes you should focus on making stuff work on what people use.
As a vfx supervisor with strong color science knowledge (I have designed the color pipeline for more than one studio) and written many specs for the color workflow on feature films… this is a very frustrating thread to read. A major problem with color science is that the terms are poorly defined, many are overloaded, and half the challenge of having a constructive conversation is to be sure you are all speaking the same language. There aren’t many people who understand digital color well, so the majority of info on the web is wrong or confusing. You have the spend a long time to get to the point where color no longer has any mystery to you.
Having said all that. Shoot log always. But if you are viewing log on set while you are shooting and composing your picture, you are doing it horribly wrong. You need to choose a viewing lut and look at your pictures through that.
_aavaa_ 6 hours ago [-]
What resources do you recommend for learning about color science properly?
rob74 8 hours ago [-]
TIL that "log" has more meanings than (1) a piece of wood or (2) some sort of periodic data recording, be it a ship's log or a server log. In this case it's short for "logarithmic profile" (https://en.wikipedia.org/wiki/Log_profile).
bede 8 hours ago [-]
Presumably this choice should be dictated more than anything by the capture device. ARRI colour discussed in the article is famously good / critically acclaimed, and a great starting point as suggested. For many cameras however—especially lower end pro and consumer gear—shooting log is the best escape from an awfully oversaturated default colour profile.
wyager 12 hours ago [-]
The point of shooting in a log format isn't to make the footage look a certain way; it's to reduce the degree to which quantization/clipping throws away information you want to have in post.
If you film with a standardized transfer function like HLG or S-Log3, you should be able to apply the inverse OETF in your grading software and work in whatever color space you want (linear, ACEScct, etc.). If you want, you can just compress to rec.709 there! There's really no need to squash to rec.709 in the camera.
kstrauser 11 hours ago [-]
I know zero about video but a decent amount about digital audio. What you said sounds, by analogy, like we don’t use 24-bit masters because we need that dynamic range (we absolutely do not) but so that we have enough extra bits to adjust the dynamic range without introducing a bunch of noise into the mixing process.
Is that about right?
Melatonic 11 hours ago [-]
Sort of. The problem is most cameras record in H264 and now (finally) sometimes higher than 8 bit depth.
That's like using a low quality audio codec as your master.
Recording in log squeezes the maximum amount of usable latitude into an otherwise not ideal codec.
KaiserPro 8 hours ago [-]
Kinda.
Its a hack to get the result of 24bit.
24bit means that you have way more dynamic range than you can possibly want. log is a bit more like slapping a compressor in to capture something like a snare without needing the full 24bit. You set the compressor to open right at the start but then close off as it get louder.
log film space means that the darkest darks get way more precision, falling off exponentially as they get brighter and brighter.
t0bia_s 10 hours ago [-]
Not good analogy. Log in video means more noise in shadows because lowest ISO is higher. Does 24bit sound need a more gain that increasing noise?
Melatonic 10 hours ago [-]
Log theoretically doesn't mean that but it does often implement like that. Usually it shoots at the real base iso of the sensor and doesn't allow changing (or for dual base gain sensors they often have a second log setting for low light with the second "base" iso)
t0bia_s 3 hours ago [-]
On Sony or Fujifilm that I used, ISO <800 (or 1000) is disabled and can be changed higher only. Never fixed.
Melatonic 2 hours ago [-]
That's because on those sensors that's the "real" iso of the sensor. Below that is just it doing some ADC mumbo jumbo.
t0bia_s 2 hours ago [-]
Not true. ISO depending on log. F-log has ISO 800, F-log2 has ISO 1000, same sensor.
loki-ai 11 hours ago [-]
ironically the post shows how tricky color management can be, where him gets multiple facts wrong. just go after some proper technical literature, ignore youtube influencers on the topic and continue shooting in log
zuInnp 12 hours ago [-]
The whole “how to take a video” bubble in the internet is rather confusing. You find so many hints and tutorials and advice about how to shoot and color correct … it is pretty confusing.
Last year I learned about the correct position of the log to rec 709 transformation, which makes sense looking now at it, but which also means I did it wrong for some years
KaiserPro 8 hours ago [-]
Depends on what you mean by correct though.
Unless you are sending your stuff for broadcast, which has a clear and rigid delivery spec, then frankly it all about what looks good.
Most displays are only really capable of displaying 8bits of srgb. even then its suspect. Manufactures will put colour tints/transformations/LUTs to stuff look bright and colour rich, even though the underlying footage/screen is sup standard.
At the end of the day, can you make your output look good on a TV and or phone? if you can, thats video, well done. Ignore the purists unless you want to get a very specific feel of colour that requires what they are selling.
KaiserPro 9 hours ago [-]
Strap in children, you're about to get a stream of consciousness from someone who was there in the old days.
Unless I have missed something re.709 is 8bit. rec 7.09 is the colourspace that most "full HD" screens can sorta reproduce in an "accurate" way (spoiler, most of them dont, but thats another issue. )
The author talks about cineon which is/was a 10 bit log format meant to record film "properly". From what I remember, and how it was explained to me was: we want the detail in the blacks, because thats where we are keen to make changes. I never was involved in actual film processing so I don't know what the frequency response of film is, all I know was that for the longest time it had greater dynamic range than digital(depending on the stock you were using).
Now what the author misses out in the talk about RED cameras is that actually their colour was a bit dodgy. Mind you, so were a lot of the other cameras at the time. (a lot of people were filming on canon 5dmk3, and then a whole range of other weird cameras)
With film, you'd send it to a DI lab to get scanned. You'd sit with the colourist, who's team would have conformed the base grade to something normal, and then the colourist would do magic according to what you wanted (assuming you had enough headroom)
For film, you'd ever print out to film again, so keep that 10/12bit colour, or dump out to DCP, which was, I seem to recall jpeg2000 (much like RED....)
Now for most of us, we don;t have that kind of cash, which, I suppose is the point of the article.
So, advice:
1) don't shoot with RED. They are bastard cameras, some of them are colourblind (see the hobbit)
2) turn zebras on in your camera (they give you a visual warning of over/under exposed areas in your shot)
3) use the live histogram. If you have massive peaks somewhere, you need to adjust.
4) light properly. film looks like film, mostly because of the lighting allowing you to grade the way you do.
5) shoot a macbeth chart in your scene, that way you can use the auto tools to make a base grade.
there is nothing wrong with log. Out of the camera it looks flat and a bit grey. find the macbeth chart, use the tooling to capture the output and set a base grade. Then drag a LUT in to make it look the way it should (there are loads of premade LUTs now, so go and grab a few and try them out. )
Digital cameras are wonderful, and with a bit of skill allow you to shoot sumptuous looking stuff without an entire team of set and in post. Log is fine, its a reasonable way to store the 14+ bits of colour that mp4/h265 cant really render. just make sure you have davinci or what ever adobe has set up to receive and correct the colour
whizzter 8 hours ago [-]
Ugh, if the "log" he's talking about is only 10 bits then I can definetly understand his issues, I thought we'd gotten to atleast something like 16bit capture/storage or more before using log scales, I can definetly see the problem then.
KaiserPro 8 hours ago [-]
so Cineon was deffo 10bit, but that was in the 90s. I do remember some stuff being 10bit log, but that was a rarity. most was 16bit log even in 2007.
but I am pretty sure that everything is at least 12bits, even on the cheap stuff.
klodolph 12 hours ago [-]
Yeah. I’m not really a professional here but I would say sensitometry is something of an interest of mine. Take what I’m saying with the appropriate grain of salt.
Your camera has some sensor which records some range of values. The range is a lot wider than your monitor can reproduce and generally follows some curve up to a maximum value, where the highlights clip.
Log is something of a fantasy in that it represents an idealized, mathematical idea of what your camera does, but it does not really correspond to what the sensor does or what the monitor displays. You can convert your video to log, or from log, but the log data itself is only really useful as an intermediate step when post-processing. I don’t think you need to capture in log. I would just go for some “regular” wide like Canon’s Wide DR.
YMMV, I am a little old-school about these things and there is a lot of stuff that I prefer to do in-camera. (Be warned that while doing things in post gives you flexibility to undo your work and change your mind, there are a lot of things you can’t reasonably do in post… even things that seem basic like white balance are better to do with lighting or filters, and if you don’t understand the sensitometry then you can easily end up with footage that you can’t rescue in post.)
davidgay 12 hours ago [-]
The one thing worth considering here is that the noise is intrinsically (short of a handy laser illuminated scene IIUC) proportional to the sqrt of the illumination level. So we should recored in sqrt for better coding efficiency/range :)
weinzierl 11 hours ago [-]
"there are a lot of things you can’t reasonably do in post"
I have my own list of these things.
Mine is a bit general, including for example the anti-aliasing filter, and polarizers. I'm curious about yours. Would you mind to share?
t0bia_s 10 hours ago [-]
Properly lit scene. If you have huge contrast in scene, you cannot reproduce it on common display with attitude of having everything unclipped.
weinzierl 9 hours ago [-]
This is probably the most important one.
112233 11 hours ago [-]
I am even less of a professional and this whole field looks exactly like audiophile snake oil peddling. "log" makes total sense because codec cannot encode full precision sensor values (otherwise, just save that!) Any floating point is "log". As long as image carries color space and the mapping LUT, you can convert it into a linear space, which is only thing that makes sense for most operations. Instead, people are taking these "shmony super log8 plus ultra" files, and looking at them directly! they say "log looks flat", they adjust contrast and curves on it, ... How is that making any sense? What "look"? Why is reversing any such LUTs not done automatically? Super weird
KaiserPro 8 hours ago [-]
log _was_ a compromise. It allowed one to do film scanning at the right level between precision and storage.
cineon, which is where you could argue that log came from, was a 10bit format to represent something that could have easily been 16bit. The compromise was that you kept the precision in the blacks for recovery, but with the ability to actually handle the files on hardware at the time. (bear in mind that realtime 10bit colour playback at 2/4k required a rack full of harddrives up until at least 2010 (see baselight 8 hardware))
I would hope there are 14 and 16 bit formats that come out of the camera that allow you to have some sort of linearish recording. Now that sensors are actually capable of capturing a dynamic range greater than top end film stock. However light isn't linear, so log is actually closer to how we see than linear is.
What ever it is, there are many levels of hidden LUTs to get the camera to record in a way that its meant to. what ever you are doing involves fiddling someone else LUTs
jiggawatts 11 hours ago [-]
Video formats historically did not include gamut information, made wild assumptions, etc.
So when log formats were introduced, you had to know the color space and undo the mapping semi-manually!
DaVinci Resolve — a colourist tool ffs — just got automatic color space detection a couple of versions ago!
Compared to still image photography where color management is the default, videography is insane.
LoganDark 10 hours ago [-]
The crux of the article appears to be that it's too much effort. If you enjoy the effort, like me, nothing has changed.
jiggawatts 10 hours ago [-]
> Understand that HDR video is not the same thing as HDR photography. They use different science and apply different technology to the sensor image.
This is where the author lost all credibility.
I have a Nikon Z8 "hybrid" camera that can take breathtaking 4K HDR footage or HDR still images.
I can easily develop both the stills and video from the same sensor, the same lenses, and interleave them in the same HDR video and have the two be nigh indistinguishable. It just looks like pressing the "pause button" on the video.
The differences -- if any -- are largely down to historical or stylistic choices, and can be freely changed depending on one's preference.
At the end of the day, they're just... lookup tables.
KaiserPro 8 hours ago [-]
HDR is a thing thats related, but not the same as this.
Its also complicated by the issue that HDR photography is all about compressing many more bits than 8 into an 8 bit image.
HDR video is a way to represent physically brighter output than 8bit on a screen.
Capturing HDR isn't that hard anymore, especially for stills. I remember seeing a light probe camera that could capture an entire film set, where we could adjust the exposure in post to the point where I could see the filament of the tungsten bulb, and the bloke eating a sandwich in the dark.
dofm 9 hours ago [-]
The author lost no credibility.
I assumed from that sentence that the author was trying to make the point to an audience more experienced with stills that video HDR is not like multiple-merged-exposure HDR, which is historically what the term HDR has generally been taken to mean in the stills world. Basically they are saying “put that interpretation out of your mind”.
Of course, in the hybrid mirrorless world, it’s identical to a single-shot HDR.
jiggawatts 8 hours ago [-]
He talks specifically about cameras that take a single-shot HDR image: "Yes, there are camera and displays capable of some level of HDR".
Except that "some level" in even my previous Nikon D800 camera has been a 14-bit RAW with 13.5 stops of dynamic range, which I got back in 2012!
That puts it somewhere between the VESA DisplayHDR 600 and 1000 standards, which is plenty HDR.
I've recently been going through my old photos from the 2010s using the latest Adobe Lightroom, which can natively process RAW images into HDR output, and they look spectacular.
It's not cheating, it's not multiple exposure, it's not local contrast. The results look like you paused a Netflix 4K HDR show.
dofm 7 hours ago [-]
You don't need to talk to me about DSLRs and dynamic range — I have owned many DSLRs dating back to one of the earliest (the Fuji S2 Pro — the dynamic range king of its day)
But nevertheless the point I think he is making to the reader is "don't think of this as what you colloquially know as HDR from stills". A surprisingly large number of non-photographers think "HDR" means those gaudy things people used to make from exposure blends.
(I only ever had one proper application for an exposure-blended HDR: a holiday home room that had a spectacular view that I needed to hint at in an interior shot)
jiggawatts 7 hours ago [-]
It's possible that he meant old-school "fake" local contrast adaptation HDR, but the way it was phrased implies "new HDR-capable cameras" are somehow not the same as HDR video... which is only technically true in the most pedantic way imaginable.
ikatson 7 hours ago [-]
It is a bit apples-to-oranges to compare camera stops to DisplayHDR.
Camera stops are about relative contrast ratios, whereas DisplayHDR is a luminance-first standard where the 600 and 1000 mandate absolute peak output nits [1].
If HDR were defined purely by contrast ratio, an OLED display peaking at just 100 nits would technically be "infinitely HDR" simply because its black level is zero.
I mean they probably do save in different capture formats depending on what camera you're using. But theoretically with a high quality lossless or lossless compressed in either they would be comparable
There's also I suppose the process of making the HDR - with photos sometimes your combing many more into a single HDR frame with theoretically as much latitude as you want. Video your limited probably with two exposures
Either way it does not explain the concrete issues of recording in log. If you have a strong title like "Stop doing X" you better have some damn good arguments if you think you can tell other people what to do.
The plea is for filmmakers to stop following the hype, use what they are experienced with and take ownership of exposure again ("Many times a videographer will shoot log incorrectly – generally underexposed – expecting to ‘fix it in post’").
Because in his experience log doesn't add substantial value, it creates lower-quality base-material with substantial post-production work.
It makes sense. Creative work comes from decisions being taken, log means that the videographer is not deciding but is deferring the task to post-production, creating a unnecessary burden (and possibly diluting his own creativity).
Technically, the issue if I'm getting is that the camera physically (or in teh encoding) will be better at capturing within a certain range that you adjusted for, having parts above or below that isn't capture properly and if material is in those above or below ranges the output after grading will not have a good range of values (ie possibly introducing banding).
Underexposing probably feels like a "safe fix" rather than adjusting properly all the time since with more limited ranges it'd be easy to "wash out" with overexposure (irreparably removing things due to clipping), and with budgets cut it's probably easy to just skimp on a lot of these "details" if the system "is capable".
So it's budgets/speed/creativity at shooting time vs material that might just not be good to work with in the end. (Also as he mentions some of these systems aren't as sensitive, probably leaves a situation where some people who's used the better systems just don't realize that it's a problem coming to others).
Edit: Noticed some other comment mentioning that these "log" encodings are 10bits, yeah that's going to leave a bad trail of numeric resolution if you've exposed badly, it'll probably be workable for a huge range but it's not enough by far to recover really badly exposed stuff.
The way I read it, the point is especially for experienced videographers to stop playing it safe but "making the call".
The author implies that many of them now "shoot for post-production" which basically leaves alot of their skill dormant to just create much more tuning/decision-work down the line.
From the perspective of a colorist it makes alot of sense to me. The shot and its "message" should be decided on-set already, deciding ALL of this in post is not creating as much value as people think, it rather dilutes the work as a whole.
Or as Orson Welles said: "The enemy of art is the absence of limitations"
That's certainly an interesting take to have (and it could've been compressed to a single paragraph)
It seems to me that it was. The essay's opening paragraph is:
A CCD captures light with a fixed exposure time and analog + digital gain. This gives you a raw, or in photographer speak, RAW frame. The issue being that the required bandwidth to store RAW frames of video is simply too high to be practical.
So, you need some form of compression. RED is a great example of seeing the problem and solving it. Now the issue is compressed RAW video is a patent minefield, and just recorded video equally so.
And so, we are resorting to tricks such as “log raw encoding”, or “just get the videographer to capture it right in the first place” just to get around the limitations of the storage format.
Honestly capturing images and displaying them accurately is such a mess that it’s a miracle that we have more than 8-bits of colour, and that was hard enough as it is. From a RAW Bayer frame to what you see on a monitor preview display is multiple layers of white balance, gain, equalisation, linear to gamma, and whatever your monitor decides to do with the digital signal it’s fed.
It's kinda like always shooting photos in RAW, always 4-5 photos of every scene, but then never taking the decision/work which of them is representing the scene you were trying to capture.
What I’d really like to see in future cameras is subject-aware shutter speed selection, where the camera intelligently sets a minimum shutter speed based on what’s happening in the scene. I’d also love real-time feedback on image quality: the camera analyses a shot within a second of capture and give a subtle beep or chime if, for example, neither eye is in focus or the subject is blurred by motion. The less time I spend worrying about the technical details, the more attention I can give to the photograph itself.
Exactly. And that article describes a scale where both tasks are assigned to different roles, with the videographer mainly working to create more room for later decision-making (than practically needed).
in my (exaggerated) simile to casual RAW photography, the photographer defers technical and creative decisions to a later step that then never happens, so he actually never conveys his message of that captured moment.
In this context the article makes most sense to me.
"You are there, you know the equipment, you know what you want to capture, go and do it."
But if you want to play with the colour, loosing 4bits of colour precision means you aint recovering whats in the blacks. (log supposedly allows better detail recovery in blacks, but that was for film, I don't know about digital anymore)
"rec.709 is 8bit" sounded wrong, so a quick check suggests it's either 8 or 10 bit [1], where 10 bit formats may be used during production, and 8 bits for delivery [source: gemini].
[1] https://glenwing.github.io/docs/ITU-R-BT.709-6.pdf - page 5, quantization levels
I think beside of this extra-step in post (which comes with its own set of parameters again), that's part of his point:
The videographer didn't expose for a specific look in a scene, he didn't decide at all and postponed to post-production. This is not adding value, it's creating work.
"Many times a videographer will shoot log incorrectly – generally underexposed – expecting to ‘fix it in post’. The same issues apply to shooting in camera raw or HDR."
But this is a nontechnical article for a nontechnical audience.
For example, https://digitalfilms.wordpress.com/2025/08/15/the-film-look-...
It's mostly oriented towards painters and similar artists, but I find it fascinating as a non-artist.
Especially his color wheel Wednesday videos makes you appreciate just how much weird theories were about, and still linger around.
[1]: https://www.youtube.com/channel/UCXGRyrSH8Ak31H7QGpPlm9w
I’m not sure I really buy the original article here. The reason I shot in log is because it recorded a wider color gamut than otherwise and skipped in camera processing. Doesn’t mean I didn’t try to nail the exposure in the first place, just that I had more latitude to make adjustments in post without as much image quality loss.
I don't know, it really depends what is the target audience.
I don't know the technical details but I found out that many modern TV shows have a lots of scenes with what I would call the netflix/youtube influencer's look with a very dark room and various small lights in the background that just don't work well if you don't have a very good screen.
And a lot of people are watching these productions from devices of varying qualities.
It is a bit like criticizing the loudness war and overuse of compressors in music production. Sure it makes for a huge loss of details when listening in your living room with a good quality equipment and in a quiet place. But most people listen to music in hostile environment: in a vehicle, in a car, in badly insulated rooms/apprtments with noisy neighbors and traffic, with shitty bluetooth speakers, etc.
Sometimes you can target for the beat equipmemt, sometimes you should focus on making stuff work on what people use.
It is the same when writing software and accounting for the still large amount of computers used globally with a screen resolution below 1080p. https://gs.statcounter.com/screen-resolution-stats/desktop/w...
Having said all that. Shoot log always. But if you are viewing log on set while you are shooting and composing your picture, you are doing it horribly wrong. You need to choose a viewing lut and look at your pictures through that.
If you film with a standardized transfer function like HLG or S-Log3, you should be able to apply the inverse OETF in your grading software and work in whatever color space you want (linear, ACEScct, etc.). If you want, you can just compress to rec.709 there! There's really no need to squash to rec.709 in the camera.
Is that about right?
That's like using a low quality audio codec as your master.
Recording in log squeezes the maximum amount of usable latitude into an otherwise not ideal codec.
Its a hack to get the result of 24bit.
24bit means that you have way more dynamic range than you can possibly want. log is a bit more like slapping a compressor in to capture something like a snare without needing the full 24bit. You set the compressor to open right at the start but then close off as it get louder.
log film space means that the darkest darks get way more precision, falling off exponentially as they get brighter and brighter.
Unless you are sending your stuff for broadcast, which has a clear and rigid delivery spec, then frankly it all about what looks good.
Most displays are only really capable of displaying 8bits of srgb. even then its suspect. Manufactures will put colour tints/transformations/LUTs to stuff look bright and colour rich, even though the underlying footage/screen is sup standard.
At the end of the day, can you make your output look good on a TV and or phone? if you can, thats video, well done. Ignore the purists unless you want to get a very specific feel of colour that requires what they are selling.
Unless I have missed something re.709 is 8bit. rec 7.09 is the colourspace that most "full HD" screens can sorta reproduce in an "accurate" way (spoiler, most of them dont, but thats another issue. )
The author talks about cineon which is/was a 10 bit log format meant to record film "properly". From what I remember, and how it was explained to me was: we want the detail in the blacks, because thats where we are keen to make changes. I never was involved in actual film processing so I don't know what the frequency response of film is, all I know was that for the longest time it had greater dynamic range than digital(depending on the stock you were using).
Now what the author misses out in the talk about RED cameras is that actually their colour was a bit dodgy. Mind you, so were a lot of the other cameras at the time. (a lot of people were filming on canon 5dmk3, and then a whole range of other weird cameras)
With film, you'd send it to a DI lab to get scanned. You'd sit with the colourist, who's team would have conformed the base grade to something normal, and then the colourist would do magic according to what you wanted (assuming you had enough headroom)
For film, you'd ever print out to film again, so keep that 10/12bit colour, or dump out to DCP, which was, I seem to recall jpeg2000 (much like RED....)
Now for most of us, we don;t have that kind of cash, which, I suppose is the point of the article.
So, advice:
1) don't shoot with RED. They are bastard cameras, some of them are colourblind (see the hobbit)
2) turn zebras on in your camera (they give you a visual warning of over/under exposed areas in your shot)
3) use the live histogram. If you have massive peaks somewhere, you need to adjust.
4) light properly. film looks like film, mostly because of the lighting allowing you to grade the way you do.
5) shoot a macbeth chart in your scene, that way you can use the auto tools to make a base grade.
there is nothing wrong with log. Out of the camera it looks flat and a bit grey. find the macbeth chart, use the tooling to capture the output and set a base grade. Then drag a LUT in to make it look the way it should (there are loads of premade LUTs now, so go and grab a few and try them out. )
Digital cameras are wonderful, and with a bit of skill allow you to shoot sumptuous looking stuff without an entire team of set and in post. Log is fine, its a reasonable way to store the 14+ bits of colour that mp4/h265 cant really render. just make sure you have davinci or what ever adobe has set up to receive and correct the colour
but I am pretty sure that everything is at least 12bits, even on the cheap stuff.
Your camera has some sensor which records some range of values. The range is a lot wider than your monitor can reproduce and generally follows some curve up to a maximum value, where the highlights clip.
Log is something of a fantasy in that it represents an idealized, mathematical idea of what your camera does, but it does not really correspond to what the sensor does or what the monitor displays. You can convert your video to log, or from log, but the log data itself is only really useful as an intermediate step when post-processing. I don’t think you need to capture in log. I would just go for some “regular” wide like Canon’s Wide DR.
YMMV, I am a little old-school about these things and there is a lot of stuff that I prefer to do in-camera. (Be warned that while doing things in post gives you flexibility to undo your work and change your mind, there are a lot of things you can’t reasonably do in post… even things that seem basic like white balance are better to do with lighting or filters, and if you don’t understand the sensitometry then you can easily end up with footage that you can’t rescue in post.)
I have my own list of these things. Mine is a bit general, including for example the anti-aliasing filter, and polarizers. I'm curious about yours. Would you mind to share?
cineon, which is where you could argue that log came from, was a 10bit format to represent something that could have easily been 16bit. The compromise was that you kept the precision in the blacks for recovery, but with the ability to actually handle the files on hardware at the time. (bear in mind that realtime 10bit colour playback at 2/4k required a rack full of harddrives up until at least 2010 (see baselight 8 hardware))
I would hope there are 14 and 16 bit formats that come out of the camera that allow you to have some sort of linearish recording. Now that sensors are actually capable of capturing a dynamic range greater than top end film stock. However light isn't linear, so log is actually closer to how we see than linear is.
What ever it is, there are many levels of hidden LUTs to get the camera to record in a way that its meant to. what ever you are doing involves fiddling someone else LUTs
So when log formats were introduced, you had to know the color space and undo the mapping semi-manually!
DaVinci Resolve — a colourist tool ffs — just got automatic color space detection a couple of versions ago!
Compared to still image photography where color management is the default, videography is insane.
This is where the author lost all credibility.
I have a Nikon Z8 "hybrid" camera that can take breathtaking 4K HDR footage or HDR still images.
I can easily develop both the stills and video from the same sensor, the same lenses, and interleave them in the same HDR video and have the two be nigh indistinguishable. It just looks like pressing the "pause button" on the video.
The differences -- if any -- are largely down to historical or stylistic choices, and can be freely changed depending on one's preference.
At the end of the day, they're just... lookup tables.
Its also complicated by the issue that HDR photography is all about compressing many more bits than 8 into an 8 bit image.
HDR video is a way to represent physically brighter output than 8bit on a screen.
Capturing HDR isn't that hard anymore, especially for stills. I remember seeing a light probe camera that could capture an entire film set, where we could adjust the exposure in post to the point where I could see the filament of the tungsten bulb, and the bloke eating a sandwich in the dark.
I assumed from that sentence that the author was trying to make the point to an audience more experienced with stills that video HDR is not like multiple-merged-exposure HDR, which is historically what the term HDR has generally been taken to mean in the stills world. Basically they are saying “put that interpretation out of your mind”.
Of course, in the hybrid mirrorless world, it’s identical to a single-shot HDR.
Except that "some level" in even my previous Nikon D800 camera has been a 14-bit RAW with 13.5 stops of dynamic range, which I got back in 2012!
That puts it somewhere between the VESA DisplayHDR 600 and 1000 standards, which is plenty HDR.
I've recently been going through my old photos from the 2010s using the latest Adobe Lightroom, which can natively process RAW images into HDR output, and they look spectacular.
It's not cheating, it's not multiple exposure, it's not local contrast. The results look like you paused a Netflix 4K HDR show.
But nevertheless the point I think he is making to the reader is "don't think of this as what you colloquially know as HDR from stills". A surprisingly large number of non-photographers think "HDR" means those gaudy things people used to make from exposure blends.
(I only ever had one proper application for an exposure-blended HDR: a holiday home room that had a spectacular view that I needed to hint at in an interior shot)
Camera stops are about relative contrast ratios, whereas DisplayHDR is a luminance-first standard where the 600 and 1000 mandate absolute peak output nits [1].
If HDR were defined purely by contrast ratio, an OLED display peaking at just 100 nits would technically be "infinitely HDR" simply because its black level is zero.
[1] https://displayhdr.org/performance-criteria/
There's also I suppose the process of making the HDR - with photos sometimes your combing many more into a single HDR frame with theoretically as much latitude as you want. Video your limited probably with two exposures