Is R3D 4:4:4 or 4:2:2?

[Adrian Jebef]

Absolutely not, the Alexa has a nice output image but in terms of pure color rendition the Epic wipes the floor with it. Including skin tones. However with more information comes greater responsibility and the need for greater understanding of the material and processes required to control and manipulate said information. If you like your images spoon fed to you the Alexa/C300 etc. are a good fit. If you want full access to what is really there the only choice is Epic.

My second contention is that saying a GRBG Bayer filter array is 4:2:2 or 100%/50%/50% is like saying that a 4 cylinder is 30% less powerful than a 6 cylinder. Despite the fact that the 4 cylinder is a tuned Subie STI and the 6 cylinder is a Camry. Debayering is a computational act, it uses real, interpretive and calculated information to create color channels that measure much higher than the 4:2:2 or 100/50/50 you imply because no color is absolute, it is made up of components and those can be calculated to a much higher degree of accuracy. Measured chroma resolution here is the important thing to consider because that is the chroma resolution actually available to you when you debayer and process your footage. Is in absolute terms the color resolution lower than the luma? Yes, but by only a small margin and at least at 5K you'd be getting around 4K or more, again tested, usable, measured resolution, not theoretical numbers based on a sub-sampling method that doesn't even apply to RGB. Remember YUV is in and of it's self a chroma subsampling method.
Simply by staying in the RGB space you've saved significant color information from being thrown down the drain.

Put into YUV terms if the Alexa produces a pure 4:4:4 image (It doesn't but I'll go with it) then the Epic produces a 20:16:16 image, which would you choose?

Evin:

It has been my experience that ALEXA and F35 footage has "truer" skin tone reproduction than Epic footage. To me true color reproduction should allow for what is seen by eye on the set to be translated to what is seen on the screen. In order for this to happen a whole lot of technical crap has to happen first. There are steps to the process. Light energy needs to be captured. That energy needs to sampled and then converted (analog to digital) to a numerical value representing the intensity of that energy. This is then scaled into the degree of accuracy of the system: 8-bit 256 values, 10-bit 1024 values, 12-bit 4096 values, and so forth. At this point it is RAW and it can be recorded. This is what the Epic does and what the R3D file represents. But is is not an image. It is not what was seen on set. These numbers need to be crunched, algorithized, transformed, debayered, reborn, interpreted. Debayering is a computational act. It is math. And it does use "real" (discreet), sampled information but it also uses interpretive and calculated (non-discreet) information to turn the math into an image. My experience has been that the accuracy of this process is relative to how the image was discreetly sampled at the sensor. Math can be re-calculated to get slightly better results but math cannot add discreet values that were never present in the first place.

The Epic is an excellent 4K black and white camera. But it's color reproduction I find lacking. The idea that the end user must take on more responsibility and a greater understanding of RED's "mysterious" image-making process is simply silly. Yes, RAW image data can be computationally changed an infinite number of times without (theoretically) destroying the underlining data. But RAW image data cannot "magically" produce chroma resolution that wasn't there (sampled) in the first place. RGB, YUV, XYZ, or ABC do not really matter as they are simply quantitative spaces in which the RAW data will eventually inhabit. Just because I have a 16-bit RAW file or a 10-bit RGB image does not mean that I have all of the luma and chroma information that was originally present in the scene at the time of capture.

I have no interest in which camera system provides bigger and better technical results. My interest is in the system that provides results that are as close to what I see by eye.

BTW: if we say the ALEXA is 4:4:4 at 2K/HD and the EPIC is 20:16:16 at 4K then wouldn't your logic lead us to declare the EPIC is 4:2:2 at 2K/HD?

 

[M Most]

Absolutely not, the Alexa has a nice output image but in terms of pure color rendition the Epic wipes the floor with it. Including skin tones....

An opinion. One that you are, of course, entitled to. But also one that is not widely shared by the great majority of cinematographers that I know and work with. Epic has many qualities and many reasons to recommend it, especially for features. But at least to this point, what you're claiming is generally not one of them. I would go so far as to say that Alexa has succeeded in its intended market (primarily television) despite Epic's technically superior specs. There are two primary reasons for that: simplicity, and the images themselves. On paper, Red should be winning hands down. But in reality, it's all about the images, particularly in television, which doesn't need the resolution that Red delivers, at least not to this point. The images are what have sold Alexa to the cameramen who have selected it. The simplicity of delivery is what sold it to the producers.

I'm very optimistic that continued refinement of the color matrixing will always lead to "better" (always in the eye of the beholder......) Red images, thus minimizing the first perceived advantage. But if Red really wants to gain a larger foothold in television, they should at least consider integrating HD recording, as both Alexa and F65 do (well, F65 uses a dockable module...but it's built by Sony, uses their codec, and will be updated in lock step with the camera..). That would go a long way to eliminating the perceived complications caused by having multiple color matrices, multiple gamma curves, multiple shooting formats, internal settings, external settings, FLUT, Alchemy, and a host of other settings all available to be used - or abused, as is often the case. Third party solutions are nice, and useful, but it's not the same, especially in that television market that I'm referring to. You don't see a whole lot of Codexes attached to Alexas shooting television shows...

 

[Graeme Nattress]

But RAW image data cannot "magically" produce chroma resolution that wasn't there (sampled) in the first place

Agreed - so why are you arguing that 1440 chroma samples across an image is better than 2560?

Graeme

 

[Gavin Greenwalt]

I feel like the question is getting semantically ignored while we all know exactly what he's asking.

Does the green channel have more resolution after debayer? Yes. In my experience blue and red are softer after debayer.

Is it half the resolution? No. Making up bullshit numbers off the top of my head it feels more like (3.8)3.2)3)

Does any camera deliver "true" 4:4:4? No.
Does any other camera deliver more red and blue resolution from what I've seen? No.
Does redcode above about 5:1 start to really eat into the blue resolution and sharpness? Yes.
Does film deliver 4:4:4 as meaning "full resolution in every channel"? No. The blue channel in every film scan I've seen is also softer just like RED.
Does Bayer pattern sometimes introduce 'odd' mosaic patterns? Yes.
Are said patterns noticeable at 4k or after supersampling to 1080p/2k? Not since like build 14.

Should 4:4:4, interlacing, pull downs, non-square pixel aspect ratios and tape all be relegated to the dustbin of history? Yes, please for the love of God yes.

 

[Graeme Nattress]

Pretty good summary Gavin - I'll take you up on one point

"Does any camera deliver "true" 4:4:4? No. " and just add in ",not at it's native sensor resolution, or not without excessive aliasing" but your overall message is well taken.

Graeme

 

[Adrian Jebef]

Agreed - so why are you arguing that 1440 chroma samples across an image is better than 2560?

Graeme

G:

I'm not arguing that at all. The number of samples is only relevant to the intended target resolution. And the way these chroma samples are handled within the debayering process equates to real, perceivable, on-the-screen color rendition.

 

[Adrian Jebef]

I feel like the question is getting semantically ignored while we all know exactly what he's asking.

Does the green channel have more resolution after debayer? Yes. In my experience blue and red are softer after debayer.

Is it half the resolution? No. Making up bullshit numbers off the top of my head it feels more like (3.8)3.2)3)

Does any camera deliver "true" 4:4:4? No.
Does any other camera deliver more red and blue resolution from what I've seen? No.
Does redcode above about 5:1 start to really eat into the blue resolution and sharpness? Yes.
Does film deliver 4:4:4 as meaning "full resolution in every channel"? No. The blue channel in every film scan I've seen is also softer just like RED.
Does Bayer pattern sometimes introduce 'odd' mosaic patterns? Yes.
Are said patterns noticeable at 4k or after supersampling to 1080p/2k? Not since like build 14.

Should 4:4:4, interlacing, pull downs, non-square pixel aspect ratios and tape all be relegated to the dustbin of history? Yes, please for the love of God yes.

Gavin:

You're getting close but I'm actually not interested in technical resolution whether it be luma or chroma. I'm interested in color accuracy, rendition, and most importantly beautifully natural skin tones. I constantly have issues with how the EPIC records and processes color. Especially skin tones. I find it very disingenuous that some among us would prefer to beat the 2-is-better-than-1 drum no matter what. More resolution (technical sharpness) does not always deliver superior results over fidelity (perceived sharpness).

 

[Graeme Nattress]

G:

I'm not arguing that at all. The number of samples is only relevant to the intended target resolution. And the way these chroma samples are handled within the debayering process equates to real, perceivable, on-the-screen color rendition.

You're not making a coherent argument.

Graeme

 

[Graeme Nattress]

Gavin:

You're getting close but I'm actually not interested in technical resolution whether it be luma or chroma. I'm interested in color accuracy, rendition, and most importantly beautifully natural skin tones. I constantly have issues with how the EPIC records and processes color. Especially skin tones. I find it very disingenuous that some among us would prefer to beat the 2-is-better-than-1 drum no matter what. More resolution (technical sharpness) does not always deliver superior results over fidelity (perceived sharpness).

Look Adrain, you've jumped in here and your comments are all over the place, and then you come out with "Actually the ARRI Alexa utilizes a 3K+ Bayer sensor which downscales to 2K/HD for proper 4:4:4 color resolution. " which you know is absolutely factually wrong. From there it's gone downhill rapidly.

The original poster asked what no-doubt they thought was a simple enough question, but in reality the answer is a hell of lot more nuanced than they anticipated. The more nuanced answers have been accurately given as part of this thread, but you're constantly wanting de-rail the discussion along your own lines and that doesn't help us out here. We've had some very good contributions to the discussion, but when you say: "but I'm actually not interested in technical resolution whether it be luma or chroma" you've basically just said you're wanting to discuss something entirely different to what this thread is about.

Graeme

 

[David Mullen ASC]

Skintone reproduction is as much an art as it is a science and it is highly subjective -- film emulsion designers have spent nearly a century trying to get skintones to look pleasant as much as they have spent trying to accurately reproduce a grey scale or a color chart. In digital photography, it is also an evolving art and science. Some of it is technical, you need enough bit depth, color separation, accurate spectral response, etc. (so as cameras improve, we've seen an improvement in the naturalness of skin) but some of it is also an eyeball thing... and it goes all the way from the design of the sensor to the colorist at the end of the process, not to mention the display technology. We'll never hear the end of this debate, not as long as color photography exists.

But for many reasons, I feel that the skintone reproduction of Epic images is quite nice and a visible improvement from early Red One images. Just as with early film D.I.'s, I think the complaints about caucasian skin looking blandly band-aid pink/tan have become increasingly rare as the science of digital color-correction improves, and as digital cameras improve.

 

[Stuart English]

BTW: if we say the ALEXA is 4:4:4 at 2K/HD and the EPIC is 20:16:16 at 4K then wouldn't your logic lead us to declare the EPIC is 4:2:2 at 2K/HD?

No.

If you believe that Alexa has better skin tone reproduction then best to just leave the commentary at that, but the 4:2:2 / 4:4:4 numbers you keep throwing around are just wrong.

 

[Graeme Nattress]

That's another post I can agree with David. Also, we're improving our colour decoding here, and are always working with DP's and colourists to make the image the camera produces better. I've certainly learned a lot these past 5yrs or so, but it doesn't stop here - it keeps going and getting better.

Graeme

 

[Eric Lange]

Given the dynamic range of “real life” v.s. the dynamic range of even a notional 4:4:4 “system’s” capability to record real life color what is accurate color rendition of a scene anyway?

I have worked with very high res large plate film scans as well as spectrophotometers and I still ask this question.

Year on year I find Graeme’s argument of “nice looking images” to be ever more persuasive... And I seem to be gaining more of an appreciation for the technical "art" (like what David is saying) behind this deeper thinking and understanding. It seems RED’s latest software offerings seem or might be offering “easier” “out of the box” color?... (If we are allowed to say that? In the MF wolrd that is also a contoversial issue)

E

 

[Evin Grant]

BTW: if we say the ALEXA is 4:4:4 at 2K/HD and the EPIC is 20:16:16 at 4K then wouldn't your logic lead us to declare the EPIC is 4:2:2 at 2K/HD?

No by that math it would be 5:4:4 or 4:3.2:3.2.

An opinion. One that you are, of course, entitled to. But also one that is not widely shared by the great majority of cinematographers that I know and work with.

Mike, I work with many of the same DPs you do and of course we've even worked together on a show before, but I disagree that the reason that Alexa's have dominated TV is because of the image they record. I believe that that it's the image they display on set that is far more responsible for this, that and the aforementioned simplicity (Hopefully RC3/RG3 will address this). What I've found is that if you compare the skin tone rendition of the Prores file of the Alexa to one from an Epic R3D on the same viewing device you can match them identically if you like, but you can also take the Epic image places I think the Prores file is just not capable of. The Alexa also has some significant mixed light issues I don't see handled as well as the Epics colorimetry. There will always be differences of opinion as there will always be bias, Arri is a long time, trusted motion picture Co. and loyalty as well as being able to present a beautiful monitor image to an impatient producer is what is really responsibly for it's success in TV.

 

[Adrian Jebef]

No.

If you believe that Alexa has better skin tone reproduction then best to just leave the commentary at that, but the 4:2:2 / 4:4:4 numbers you keep throwing around are just wrong.

How so? Perhaps you can suggest a simple definition of 4:4:4 ?

 

[Stuart English]

... but I disagree that the reason that Alexa's have dominated TV is because of the image they record. I believe that it's the image they display on set that is far more responsible for this, that and the aforementioned simplicity (Hopefully RC3/RG3 will address this). What I've found is that if you compare the skin tone rendition of the Prores file of the Alexa to one from an Epic R3D on the same viewing device you can match them identically if you like, but you can also take the Epic image places I think the Prores file is just not capable of.

We would concede that Alexa has provided a nicer default image from its HD-SDI output than either RED ONE or EPIC. We've always looked at our HD-SDI output as more of a video tap than an end point because the .R3D file has the full gamut and precision of the sensor stored in it which the HD-SDI output just can't carry through rigorous post production.

However algorithms advance, so it would be a good idea to re-evaluate the HD-SDI default output of EPIC under Redgamma 3 / Redcolor 3 ...

 

[M Most]

What I've found is that if you compare the skin tone rendition of the Prores file of the Alexa to one from an Epic R3D on the same viewing device you can match them identically if you like, but you can also take the Epic image places I think the Prores file is just not capable of.

This conversation was really about skin tones, and how they compare. Being able to "take an image places" has little to do with properly photographed and properly reproduced skin tones. And - speaking as a colorist - I would also contend that you can't take an Epic ***image*** anywhere that you can't take images from other cameras. What you can do is change pre-debayer settings to get different results, but at that point you're dealing with how the original sensor data is being interpreted, not with an RGB image. The result is, to many, the same thing, but to me it's not.

I agree that you can come relatively close to matching skin tones across both cameras. But there is a lot more to skin reproduction than simply hue and saturation. There is continuous tone reproduction across the grey scale, particularly at the extremes, that contributes significantly to the final look of a human face on screen. Intangibles such as "waxy" skin highlights, and overly red lowlights are common problems with digital cameras that the Alexa really seems to have addressed. You really don't see that looking at the monitors on Video Village, and you really don't see it as a DIT because you're not dealing with those kind of nuances that are required when you're turning out the final product and have the proper conditions and tools with which to judge them. Most of the cameramen I talk to who like the Alexa images talk about the "creamy" skin tones, and they're really commenting on exactly what I just referred to. And their judgements are being made based on viewing those images in grading rooms, under proper conditions. There are a lot of factors in play here, not just the ability to get the same color using various tweaks and secondary isolations. Some of these factors are optical as well as electronic, but they all combine to yield a result.

The Alexa also has some significant mixed light issues I don't see handled as well as the Epics colorimetry.

Personally, I haven't found that to be the case, at least not based on the files I've encountered. Of course, I'm not the one on the camera setting up metadata to deal with that. :smilewinkgrin:

There will always be differences of opinion as there will always be bias, Arri is a long time, trusted motion picture Co. and loyalty as well as being able to present a beautiful monitor image to an impatient producer is what is really responsibly for it's success in TV.

I don't completely disagree, but I still stand by what I said earlier. Alexa hit a sweet spot for television production on a number of levels and succeeded because of a combination of those factors. I think we can basically agree on that.....

 

[Gavin Greenwalt]

How so? Perhaps you can suggest a simple definition of 4:4:4 ?

I think the simplest definition someone wants is "1920x1080 resolution in all 3 colors." RED has enough sensor sites to satisfy that requirement for 4:4:4.

If you view 4:4:4 as a fancy way of saying 1:1:1 (e.g. as a ratio) then no the ratio is not 1:1 for all channels of resolution @ 4k. Once you supersample though down to HD (and compare to Alexa) it's *at least* 1:1.

And none of that has anything to do with skintones. That's a colorimetry question not a resolution question. If it were a resolution question then SD skintones from RED would be better than the same image in HD--which isn't the case. In fact I would say the opposite is true. At HD you can dither the color and get perceptual color accuracy beyond that of the same bit-depth in standard definition. I still see agree that skintones out of RED are more frustrating than out of some other cameras and film in particular--hopefully something that Redcolor 3 will resolve.

 

[M Most]

We would concede that Alexa has provided a nicer default image from its HD-SDI output than either RED ONE or EPIC. We've always looked at our HD-SDI output as more of a video tap than an end point because the .R3D file has the full gamut and precision of the sensor stored in it which the HD-SDI output just can't carry through rigorous post production.
.

Television programs generally don't go through "rigorous" post production, and now that Next Element is no more, I don't really know of any Red shows being finished directly from the R3D files. If I'm wrong on that, please feel free to correct me.

Television and features are very different beasts, with very different requirements, conditions, post schedules, and post production approaches. The advantages of Red are very clear in features, as they largely relate to resolution and its ability to hold up under very large screen viewing conditions, as well as flexibility when the original files are made available, which is more common on features and more necessary due to the use of "look creation" techniques. They're not so clear in television, where speed and simplicity are the orders of the day, and where the loop is not nearly as closed in terms of people doing what they're supposed to do all along the post path.

 

[Gavin Greenwalt]

*Opinion hat*: I think historically RED has been most concerned with color *accuracy* to the detriment of color subjectivity. A slavish reproduction of linear gamma and photometric color accuracy has resulted in a less pleasing image. If you look at film it's extremely non-linear and its color is very unpredictable... but often more pleasing in my opinion.

Something I would like to see going forward is for the REDGamma improvements to be passed on to REDLogFilm. While film is encoded in LOG color space, if you apply an inverse log LUT it doesn't become perfectly linear... it's stilly highly unlinear--it's got that nice roll-off and toe which is baked in. What would be best is something like REDGamma 3 -> .4545 gamma -> log with a LUT to convert to 'true linear' for FX and then re-apply the shadow and highlight curves. Then we could apply FX in linear space and grade in linear space but reapply REDGamma3 on top of everything when we're done (including 100% cg plates).