Are Sony sensors ruining video with the 'Sony look'?

[PannySVHS]

@austinchimp I think, the GH1 has a beautiful colour palette. Best 709 of the Lumix cams imho. Would have loved a technical image quality of the GH3 or 4 combined with the colour of the GH1. Some gruesome grading out there for sure.:) GH1 footage looks effortless in the colour department. GX85 has some nifty colours up its sleeve. I always loved the graded image of the G6 and the G7. GX85 can be nice even without grading. The former two not so much. Now, S1 is another league of course. One of the best looking cams out there.

[noone]

19 hours ago, HockeyFan12 said:

By definition, any digital sensor (except Foveon) has 2.5 stops over 18% gray at base ISO, and that's not accounting for different channels clipping before others. Some digital cameras use 12.5% gray as middle gray to get 3 stops. The S1 has +6 stops at base 640 ISO. So I'm guessing what that actually is is base ISO (50-ish) with a digital push.

Of course, the read out for video is apparently 12 bit, which should limit the camera to 12 stops, but Panasonic claims 14, so I feel like we just don't have enough information (or I don't have enough technical knowledge) to make any conclusions based on specs alone. But I speculate 640 ISO V Log is base ISO with a digital push. This is why the dual native ISOs change with scene file or picture style or whatever it's called I suspect.

 

I just find it unusual.

The camera specs state a base ISO for stills anyway of 100 but ISO 50 has that big increase in DR over ISO 100 (14.5 VS 13.66 on DXO for print).    Drops to 12.07 at ISO 400 and then the second boost due to the dual gain sensor (12.58 at ISO 800).

Most cameras with extended ISO like the specs state 50 is are not much different from their base in terms of DR.

In any case, it would seem that for video, the S1 is mostly not going to get anywhere near 14.5 stops in general use though still be very good and better than the vast majority.

 

The dual ISO gives it that boost at ISO 

[BenEricson]

10 hours ago, austinchimp said:

One of the things I'm struck by when I watch old movies shot on film - from the 70s through to the 90s - is how natural and beautiful everything looks, before every shadow had to be dark blue and every skin tone had to be sunburnt orange or desaturated grey.

Yeah, those Rec709 grades with like 7 or 8 stops of dynamic range look so rich and vibrant. Film print style. Even 35mm film shot right now has a lot less punch than it used to. 

I always though the baked in profiles on the C300 Mkii or maybe the 1DC or 1DX look really really nice. Yes, you have to be more intentional with your exposing but the results are really nice.

Not all log footage is the same though. Flat film scans are incredibly easy to grade. You don't need any LUTs, just a standard correction will look really good and natural. 

[HockeyFan12]

55 minutes ago, noone said:

I just find it unusual.

Yeah, I don't have more than the slightest clue what's going on, but I think V Log 640 ISO is a lower ISO with a big digital push. Anyway the S1 dynamic range (not sure if you own one or not) is excellent.

20 minutes ago, BenEricson said:

Yeah, those Rec709 grades with like 7 or 8 stops of dynamic range look so rich and vibrant. Film print style. Even 35mm film shot right now has a lot less punch than it used to. 

I always though the baked in profiles on the C300 Mkii or maybe the 1DC or 1DX look really really nice. Yes, you have to be more intentional with your exposing but the results are really nice.

Not all log footage is the same though. Flat film scans are incredibly easy to grade. You don't need any LUTs, just a standard correction will look really good and natural. 

+1 

Also worth noting that the default Alexa LUTs Arri and Adobe distribute throw out about as much extra highlight detail as the Alexa has over other cameras. On a standard display you're going to struggle to have a 14 stop image that looks good. I miss t2i neutral more and more.

But it is weird, one knock against the S1 is the GHX series has better skin tones for some reason imo...

[BenEricson]

1 minute ago, HockeyFan12 said:

But it is weird, one knock against the S1 is the GHX series has better skin tones for some reason imo...

I think the color in general has always suffered as the light sensitivity gets better. Look at the original Black Magic Pocket, Digital Bolex, Arri Alexa. None of the those are lowlight cameras but in the right light they look really really good.

[HockeyFan12]

4 minutes ago, BenEricson said:

I think the color in general has always suffered as the light sensitivity gets better. Look at the original Black Magic Pocket, Digital Bolex, Arri Alexa. None of the those are lowlight cameras but in the right light they look really really good.

+1

I don't trust everything DXOMark publishes, their measurements indicate better color depth over time...

Whereas to my eye the color from my Digital Rebel XT is better than the color from my S1. Anyway it doesn't matter if they're right if I can't see it or make it work. But there are thinner bayer filter dyes apparently and I think it is leading to a "change" in color.

Oddly though I felt C300 Mk II was a step down for color from C300 Mk 1, F5 was a step down from F3, etc. but C300 Mk III looks like an improvement and Venice color is excellent, FX6 looks promising too but yet to work with it. So we might be heading back in the right direction.

[BenEricson]

1 hour ago, HockeyFan12 said:

Oddly though I felt C300 Mk II was a step down for color from C300 Mk 1, F5 was a step down from F3, etc. but C300 Mk III looks like an improvement and Venice color is excellent, FX6 looks promising too but yet to work with it. So we might be heading back in the right direction.

I own the C70 (Same Sensor.) I think the potential for nice color is definitely there. Really clean sensor. The color work flow Canon has provided is really terrible though. The LUTs are 6 years old and were built for the original C300 Mkii. Nothing has been updated. The LUT pack for C70 even includes LUTs for gamma profiles that no longer exist. (Clog2 / BT.709)

I think at this point, when a cell phone can shoot nice photos and videos, the camera manufactures really should be focusing on a solid post work flow that produces beautiful, consistent results or they will lose customers. 

[HockeyFan12]

3 minutes ago, BenEricson said:

I own the C70 (Same Sensor.) I think the potential for nice color is definitely there. Really clean sensor. The color work flow Canon has provided is really terrible though. The LUTs are 6 years old and were built for the original C300 Mkii. Nothing has been updated. The LUT pack for C70 even includes LUTs for gamma profiles that no longer exist. (Clog2 / BT.709)

I think at this point, when a cell phone can shoot nice photos and videos, the camera manufactures really should be focusing on a solid post work flow that produces beautiful, consistent results or they will lose customers. 

I like what I've seen from the C70 a lot so that is discouraging to read. I had a C200 for a minute but the Canon Log 2 image was too flat and the h264 image looked worse than my C100's so I never really warmed up to it. 

I like reading interviews with Roger Deakins and he has no idea how digital cameras work... but I get it. It's literally just iPhones and Alexas where the workflow is this simple. Log C has pretty good dynamic range to say the least but the default LUT adds a lot of contrast and clipping. So you can just go with that and it looks fine, or you can grade by eye and get some flexibility. But it's relatively simple. It's a little frightening to me that the simplest cameras around are iPhones and Alexa Classics.

 

[TomTheDP]

5 hours ago, noone said:

I just find it unusual.

The camera specs state a base ISO for stills anyway of 100 but ISO 50 has that big increase in DR over ISO 100 (14.5 VS 13.66 on DXO for print).    Drops to 12.07 at ISO 400 and then the second boost due to the dual gain sensor (12.58 at ISO 800).

Most cameras with extended ISO like the specs state 50 is are not much different from their base in terms of DR.

In any case, it would seem that for video, the S1 is mostly not going to get anywhere near 14.5 stops in general use though still be very good and better than the vast majority.

 

The dual ISO gives it that boost at ISO 

I think just depends on how you measure dynamic range. According to a more strict measurement that Arri uses the S1 gets more around 12-12.5 stops, while the Arri has 14. Which makes sense. With more liberal measurements that don't take into account how noisy certain stops are you get much higher numbers. 

[TomTheDP]

On 4/7/2021 at 3:24 PM, noone said:

Any idea what the minimum ISO would be for the RED cameras?  I see the S1 gets its 14.5 stops at ISO 50 but it takes quite a dip at ISO 100 (though still very good 13.66).      Are people shooting videos with the S1 at ISO 50?

Just curious.

Native ISO on the RED scarlet I own is 400 or 800 depends on who you ask. On the S1 for video its 640 and 4000 in vlog. I actually shot in 640 for the stills as well. I wasn't aware the native ISO for stills was 50. I would have needed a lot more light to shoot at 50 iso. 

[kye]

2 hours ago, HockeyFan12 said:

It's a little frightening to me that the simplest cameras around are iPhones and Alexa Classics.

I understand what you're saying, but would suggest that they are only simple to deal with in post because they've had the most work put into them to achieve the in-camera profiles.  

It is widely known that the ARRI CEO Glenn Kennel was an expert on film colour before he joined ARRI to help develop the Alexa.  Film was in development for decades with spectacular investment into its colour science prior to that, so to base the Alexa colour science on film was to stand on the shoulders of giants.

Glenns book is highly recommended and I learned more about the colour science of film from one chapter in it than from reading everything on the topic I could find online for years prior:

Also, Apple have put an enormous effort into the colour science of the iPhone, which has now been the most popular camera on earth for quite some time, according to Flickr stats anyway.

I have gone on several trips where I was shooting with the XC10 or GH5 and my wife was taking still images with her iPhone, and so I have dozens of instances where my wife and I were standing side-by-side at a vantage point and shooting the exact same scene at the exact same time.  Later on in post I tried replicating the colour from her iPhone shots with my footage and only then realised what a spectacular job that Apple have done with their colour science - the images are heavily processed with lots and lots of stuff going on in there.

and now that I have a BMMCC and my OG BMPCC is on its way, I will add that the footage from these cameras also grades absolutely beautifully straight-out-of-camera - they too (as well as Fairchild who made the sensor) did a great job on the colour science.  The P4K/P6K footage is radically different and doesn't share the same look at all.

[kye]

9 minutes ago, TomTheDP said:

I think just depends on how you measure dynamic range. According to a more strict measurement that Arri uses the S1 gets more around 12-12.5 stops, while the Arri has 14. Which makes sense. With more liberal measurements that don't take into account how noisy certain stops are you get much higher numbers. 

I saw this image from Cine-D that shows some of their tests and includes the Alexa - it shows that ARRI was conservative with their figures while most other manufacturers took sometimes wild liberties with the figures.

These numbers should be directly comparable to the other tests that they do, as the thresholds and methodologies should be the same.

[HockeyFan12]

34 minutes ago, TomTheDP said:

Native ISO on the RED scarlet I own is 400 or 800 depends on who you ask. On the S1 for video its 640 and 4000 in vlog. I actually shot in 640 for the stills as well. I wasn't aware the native ISO for stills was 50. I would have needed a lot more light to shoot at 50 iso. 

Some of the other picture profiles have different dual ISOs. I think 50 ISO for stills and 640 ISO for V LOG are exposing the sensor the same, but the curve that's applied in camera or when developing is very different.

I think Red is always exposing at the same ISO (except Gemini), and the ISO you rate it at is how the picture is developed. I wasn't a fan of Red's color but the newer development software looks great to me.

25 minutes ago, kye said:

I understand what you're saying, but would suggest that they are only simple to deal with in post because they've had the most work put into them to achieve the in-camera profiles.  

I think this is about right. A lot of it must be in-camera processing. The Alexa Plus draws >100 watts and Apple's latest iPhone is nearly as fast as my laptop. With older Reds, you need to develop on your computer because they're not fast enough to process the footage properly I believe. ArriRAW really doesn't look much better than Arri ProRes to me.

[TomTheDP]

2 hours ago, kye said:

I saw this image from Cine-D that shows some of their tests and includes the Alexa - it shows that ARRI was conservative with their figures while most other manufacturers took sometimes wild liberties with the figures.

Cine-D's testing can be useful but at the same time they often aren't accounting for noise reduction. Although in recent tests I've found they are a lot more detailed when it comes to accounting for things like noise reduction or lack thereof. Testing RAW is more consistent as there is no noise reduction with RAW. I wish they would have dynamic range readings for RED cameras. Would be interesting to see how they stack up. 

I am really surprised, if it's the case, that the RAW images don't have more dynamic range than the 10 bit video. I think 14 bit raw just has a lot of color and texture that 10 bit video doesn't. The ease of grading with raw controls probably makes a big difference too. Maybe the dynamic range seems higher as noise is less noticeable with stills. 

I think I need to get the Atomos Ninja V now. 

 

1 hour ago, HockeyFan12 said:

Some of the other picture profiles have different dual ISOs. I think 50 ISO for stills and 640 ISO for V LOG are exposing the sensor the same, but the curve that's applied in camera or when developing is very different.

I think Red is always exposing at the same ISO (except Gemini), and the ISO you rate it at is how the picture is developed. I wasn't a fan of Red's color but the newer development software looks great to me.

Yeah I'd probably need to do more comprehensive testing although my conclusion for now is the difference between stills and 10 bit video is not as much as I had thought. My second conclusion is a lot of my camera spec desires are just consumerism getting the best of me. 

 

2 hours ago, kye said:

I understand what you're saying, but would suggest that they are only simple to deal with in post because they've had the most work put into them to achieve the in-camera profiles.  

It is widely known that the ARRI CEO Glenn Kennel was an expert on film colour before he joined ARRI to help develop the Alexa.  Film was in development for decades with spectacular investment into its colour science prior to that, so to base the Alexa colour science on film was to stand on the shoulders of giants.

It makes sense as well with Arri having a background in film cameras. The Alexa was meant to be something that would replicate and replace film. Germans are also great and everything haha. 

RED didn't have a background in film and neither does Sony.

Canon and Fujifilm both have a background with film photography and are also both pretty well regarded for their color science. 

[osmanovic]

3 hours ago, kye said:

and now that I have a BMMCC and my OG BMPCC is on its way, I will add that the footage from these cameras also grades absolutely beautifully straight-out-of-camera - they too (as well as Fairchild who made the sensor) did a great job on the colour science.  The P4K/P6K footage is radically different and doesn't share the same look at all.

Agree. The fairchild sensor is very good. The different look on the new Blackmagic cameras, is a big reason why we're passing on it.

[kye]

8 hours ago, HockeyFan12 said:

I think this is about right. A lot of it must be in-camera processing. The Alexa Plus draws >100 watts and Apple's latest iPhone is nearly as fast as my laptop. With older Reds, you need to develop on your computer because they're not fast enough to process the footage properly I believe. ArriRAW really doesn't look much better than Arri ProRes to me.

You raise an interesting point about the Prores vs RAW and I don't think I ever got to the bottom of it.  With the Prores they can put whatever processing into the camera that they want (and manufacturers certainly do) but technically the RAW should be straight off the sensor.  Of course, in the instance of the Alexa, it isn't straight off the sensor due to the dual-gain architecture which combines two readouts to get (IIRC) higher dynamic range and bit-depth, so there is definitely processing there, although the output is uncompressed.  Perhaps they are applying colour science processing at this point as well, I'm not sure.

The reason that this question is more than just an academic curiosity is that if they are not applying colour science processing to their RAW, then at least some of the magic of the image is in their conversion LUTs, which we all have access to and could choose to grade under if we chose to (and some do).

7 hours ago, TomTheDP said:

Cine-D's testing can be useful but at the same time they often aren't accounting for noise reduction. Although in recent tests I've found they are a lot more detailed when it comes to accounting for things like noise reduction or lack thereof. Testing RAW is more consistent as there is no noise reduction with RAW. I wish they would have dynamic range readings for RED cameras. Would be interesting to see how they stack up. 

I am really surprised, if it's the case, that the RAW images don't have more dynamic range than the 10 bit video. I think 14 bit raw just has a lot of color and texture that 10 bit video doesn't. The ease of grading with raw controls probably makes a big difference too. Maybe the dynamic range seems higher as noise is less noticeable with stills. 

I think I need to get the Atomos Ninja V now.

...........<snip>

It makes sense as well with Arri having a background in film cameras. The Alexa was meant to be something that would replicate and replace film. Germans are also great and everything haha. 

RED didn't have a background in film and neither does Sony.

Canon and Fujifilm both have a background with film photography and are also both pretty well regarded for their color science. 

Yes, testing DR involves working your way through various processing, if you can't get a straight RAW signal.  I'm assuming that they would have tested the RAW Alexa footage but they haven't published the charts so who knows.

Bit depth and DR are related, but do not need to correlate.

For example, I could have a bit-depth of 2 bits and have a DR of 1000 stops.  In this example I would say 0 for anything not in direct sun, 1 for anything in direct sun that wasn't the sun, 2 for the sun itself, and only hit 3 if a nearby supernova occurred (gotta protect those highlights!!).  Obviously this would have so much banding that it would be ridiculous, but it's possible.  Manufacturers don't want to push things too far otherwise they risk this, but you can push it if you wanted to.

6 hours ago, osmanovic said:

Agree. The fairchild sensor is very good. The different look on the new Blackmagic cameras, is a big reason why we're passing on it.

You're not the only ones, I hear this a lot, especially in the OG BMPCC forums / groups.

[kye]

9 hours ago, HockeyFan12 said:

I think this is about right. A lot of it must be in-camera processing. The Alexa Plus draws >100 watts and Apple's latest iPhone is nearly as fast as my laptop. With older Reds, you need to develop on your computer because they're not fast enough to process the footage properly I believe. ArriRAW really doesn't look much better than Arri ProRes to me.

 

41 minutes ago, kye said:

You raise an interesting point about the Prores vs RAW and I don't think I ever got to the bottom of it.  With the Prores they can put whatever processing into the camera that they want (and manufacturers certainly do) but technically the RAW should be straight off the sensor.  Of course, in the instance of the Alexa, it isn't straight off the sensor due to the dual-gain architecture which combines two readouts to get (IIRC) higher dynamic range and bit-depth, so there is definitely processing there, although the output is uncompressed.  Perhaps they are applying colour science processing at this point as well, I'm not sure.

The reason that this question is more than just an academic curiosity is that if they are not applying colour science processing to their RAW, then at least some of the magic of the image is in their conversion LUTs, which we all have access to and could choose to grade under if we chose to (and some do).

To elaborate on what I said before, if they are applying different colour science to the Prores and to the RAW then you'll have to use a different conversion LUT, but I think there is only one LUT for ARRI, right?  If so, either the magic is in the LUT or in the camera and being applied to the RAW.  I suspect the latter, as it's the only good way to keep it far from prying eyes and people who would steal it.

[austinchimp]

13 hours ago, kye said:

and now that I have a BMMCC and my OG BMPCC is on its way, I will add that the footage from these cameras also grades absolutely beautifully straight-out-of-camera - they too (as well as Fairchild who made the sensor) did a great job on the colour science.  The P4K/P6K footage is radically different and doesn't share the same look at all.

I have a theory that the new Sony sensors have a lot of inbuilt processing done on the chip itself or in the accompanying electronics, which you can see in the new BMPCC4k cameras. It feels a lot more sharpened and as if there's processing or noise reduction in there. The newer BM pocket 4k and 6k cameras look very similar to the GH5s and Z-Cam cameras for me. A nice image, but not with the same magic or feel.

Some have speculated that it's about the jump up to 4k, but I feel like there's something else going on there.

[TomTheDP]

2 hours ago, austinchimp said:

I have a theory that the new Sony sensors have a lot of inbuilt processing done on the chip itself or in the accompanying electronics, which you can see in the new BMPCC4k cameras. It feels a lot more sharpened and as if there's processing or noise reduction in there. The newer BM pocket 4k and 6k cameras look very similar to the GH5s and Z-Cam cameras for me. A nice image, but not with the same magic or feel.

Some have speculated that it's about the jump up to 4k, but I feel like there's something else going on there.

Blackmagic RAW has artificial sharpening going on and NR as does the GH5/S. I had heard some people preferred the CDNG image on the P4k that was present before the firmware update. Would be interesting to see a comparison between the two. I think the Zcam image is more organic with recent firmware that allows for noise reduction to be turned off as well as sharpening. Originally there was only a low setting on both. 

[kye]

12 hours ago, austinchimp said:

I have a theory that the new Sony sensors have a lot of inbuilt processing done on the chip itself or in the accompanying electronics, which you can see in the new BMPCC4k cameras. It feels a lot more sharpened and as if there's processing or noise reduction in there. The newer BM pocket 4k and 6k cameras look very similar to the GH5s and Z-Cam cameras for me. A nice image, but not with the same magic or feel.

Some have speculated that it's about the jump up to 4k, but I feel like there's something else going on there.

It would be nice to see some analysis of that.  Considering that images are just bunches of numbers, we can analyse them in almost any way that you can imagine, but we do basically no analysis whatsoever, but instead just fight about our preferred religion manufacturer online...

It's quite sad really.

10 hours ago, TomTheDP said:

Blackmagic RAW has artificial sharpening going on and NR as does the GH5/S. I had heard some people preferred the CDNG image on the P4k that was present before the firmware update. Would be interesting to see a comparison between the two. I think the Zcam image is more organic with recent firmware that allows for noise reduction to be turned off as well as sharpening. Originally there was only a low setting on both. 

What a fascinating thing - that BM RAW has processing on the RAW output..  that almost completely defeats the purpose of RAW!  There's a way to work out how to un-do it in post, but it's a PITA to do.

Sharpening isn't always bad, but it can always be added more in post, so manufacturers should only add it sparingly.  I've heard that the Alexa adds a small amount of sharpening to the Prores, as the compression smooths some detail so they add a little back in to match the original look.

In a normal camera they add sharpening, then compress the image with h264/5, which I think is the killer combo.  Sharpening and blurring are mathematical opposites, so with the right algorithm you should be able to reverse the sharpening in post with blurring, but because there is a compression in-between so the information gets lost and by the time you blur enough to get rid of the edge sharpening then the image is blurry as hell.