Really great work Josh! Not being able to color correct without compromise defeats the purpose of RAW video. Thankfully, the colors look great straight out of Nikon. Probably best to stick with NLOG or Rec709 in camera. Appreciate you taking the time to do all those tests.
Quite the deep dive in this. Thank you! Been using the 4k mode without really noticeable issues lately, but im wondering if ETTR helps that, as I tend to expose as bright as I can without clipping.
I've been using the Z9 since release along with my RED V Raptor, I do mostly Extreme Sports such as Skateboarding and Big Wave Surfing, lots of challenging environments. This is what I use, if I want 8k the 8k 60fps is absolutely fantastic from the Z9 and matches my RED V Raptor 8k almost perfectly with minimal effort in Resolve. If I am using 4k which to be honest 80% of everything I do is then I use the Z9 H265 4k 60 with oversampling turned on and just get the exposure pretty good in camera and it produces from my experience using nearly all of the top tier cameras/cinema cameras one of the best looking 4k videos on the market today, it is an absolute joy to use. The 4.1k RAW 24fps in the Z9/Z8 is just not worth it in any scenario. The H265 120fps is also another absolute gem and when compared to my RED 4k 120fps footage it looks practically identical which is a massive bonus, being 4:2:0 affects nothing for me.
...so in depth! Thank you for putting in the work!! There were a lot of strange but somewhat expected results here. Yet, I'm hopeful that if Nikon could focus on making a video first camera with some R&D help from Red, they will iron out these kinks and could deliver a truly unique prosumer cinema camera in the near future! Heres hoping...
FINALLY... spent too much time trying to figure out the Nikon Z8 422 10bit formats to basically come to the same conclusion, its only via HDMI output that it will give what most Sony shooters/editors want which is 422 10bit H265 (they will even take 422 10bit H.264 since may have used it for years) so for second shooters and video guys that want to give that format its a bummer the camera doesn't shoot it internally. Shoots both higher and lower but not that damn one! ...apprecite the deep dive and keep up the great work.
@@Tmk969 if you have the H.265 enabled via purchase or feature pack, then yes the Ninja V (in my case) will record 10bit 422 H.265 footage via HDMI from the Nikon Z8 in my most used case of 4K 24fps. I will point you to the Nikon Z8 Reference manual, (reference, not user manual) which is available online, and page 297 shows output bit depth for resolution and frame rates. Hope that helps.
@@nicolascarle8210 yes, have seen that, just wanted to ask someone who used or have atomos :) for me h265 420 is also big disappointment, we need bring that problem to the Nikon somehow
@@Tmk969 agreed, how Nikon missed this one is a head scratcher. Good part is with the Ninja V recording the 422 h265 my 1TB drive on the Ninja gets ridiculous amounts of recording time, paired with a few V-mount batteries I can record long format talks etc for hours and hours without heat or battery issues at all.
You are the only creator (that I've watched - and I've watched a ton) that has actually come out and said that the h265 footage is 420 vs 422. Everyone else that I've seen says otherwise but one day as I was looking in premiere's properties section for a few clips that I took said it was 420 (extended) and I was so confused. I always thought the footage didn't compare to my fx3 footage and I couldn't figure out why I was struggling so much with my z8. I thought I was going crazy and it was only after going back and forth with Nikon that they eventually confirmed it was 420. Thanks for the detailed review as always!
@@Tmk969 I swear I tried searching everywhere. Guess I didn’t search enough but Nikon didn’t even know why it said 420. It was only after a few exchanges that their experts said 422 is only with raw or prores hq 🤷🏻♂️ i was most confused because other reviewers kept saying h265 was 422
@@__eugenelee true :) i think other reviewers just didn't used it much or tested more heavy and deep as this video author. I found it long ago and i'm tellin it everywhere, i wrote about it under another video of this author (Josh). I hate it, because h265 would be mostly used for me.
@@Tmk969 same! It’s such a shame because it would make this way more flexible and I’m honestly surprised it isn’t 422… wonder if it’s a limit on the sensor or if it can be updated via firmware. 🙏🏻🙏🏻🙏🏻
So Highlights Recovery is not available with N-RAW on Camera Raw page of Davinci Resolve? I think with Cinema DNG from any camera that toggle is availbe there. Please try converting N-RAW to DNG to check if thar toggle appears or not in resolve.
I wouldn't get too hung up on the lack of 422 using h.265. Yes, other cameras have the option but have you ever done a side-by-side comparison recording 422 vs 420 using the A7SIII for example? I have and I never could see any difference whatsoever even pushing colors to extremes. Of course there's a huge difference between 8 bit and 10 bit, but any kind of meaningful differences between 420 and 422 are not noticeable... at least with Sony cameras using SLOG3. Thanks for the video, very well done.
The 4k Raw option was the most attractive to me as a R5C owner, turns out it's not without compromises, hopefully fixable with firmware updates. I'm still holding out for something with similar specs that offers open gate.
Great video, thanks the deep dive. Except the bitrate, what's the advantage of using something like a h265 422, over the ProResHQ ? I get the point of using Davinci (I use it also) but I think it's worth comparing the N-RAW against the ProResRaw, for the "non-destructive" part. It can also be a great option for Final Cut or Premiere users.
I did some tests with 4k Prores and 4k H.265 and the H.265 files are slightly softer. So keep that in mind. With a proper All-I H.265 file you should technically get better color and a significantly reduced file size, but with this Long GOP compression it tends to soften the image up a little. Would have been cool if the camera had an All-I H.264 4:2:2 format at least, since I'm not sure anyone does ALL-I H.265 yet. This is why Sony does this with H.264, it eliminates the need for Prores. So it's probably better to shoot 8k H.265 then sample down to 4k to something other than Long GOP to retain that detail, this would also make that 4:2:0 subsampling a non-issue or at least less of one. Or just stick with Prores. Would also like to see ProresRAW comparisons as I'm on Premiere and don't want to pay for Resolve Studio to load 8k NRAW.
Had a lot of picture degradation on Z9 at 4k 25p raw and went to 8k. Thanks for confirming. Also, I thought only prores HQ and h26? was oversampled up to 30 p…NOT raw formats. Might have changed? Have done a lot of grading with the nraw 8k and found it extremely flexible. Your point about not being able to change the color temp to match because the slider didnt facilitate is shallow. There is an abundace of data to enable a perfect match, though it requires some work. Otherwise, the depth of your methodology and analysis was superb. Thank you.
You're welcome. As I said in the video, Nikon told me that the 4.1K raw mode is oversampled in 24/25/30 fps. I have been in touch with one of their engineers. Yes the N-RAW is very flexible, but there were two things I was trying to show with the white balance examples. First is the difference between how you can change the image in raw vs ProRes. The other one is the fact that the raw isn't really raw since changing the white balance is destructive, which it shouldn't be. Thanks for watching!
Very interesting as someone who primarily uses the 4.1K N-Raw for almost a year to date on my Z8, I’ve never had any of the issues you described or show in your video ?🤔 For the past year I’ve gotten consistently good footage. I do use neat video for noise removal. I do use the Nikon lut to grade, but I never use it at full strength,lol. I only use Nikon native lenses?? I use Davinci Studio 18.5 version And I edit on a 60inch tv screen, so I’m always pixel peeping,lol. I have def shot Scenes in “low light” situations and while the footage is noisy and hard to clean up in those scenarios , no color shift that I’ve noticed? Maybe before adding noise reduction? But to be fair I only focus on the finished results, so I’ve never noticed.
You're welcome! I got in the habit of doing that a little while ago because people would watch one of my videos a year or two after I posted it and complain how things were different. And yeah, the firmware can make a big difference. v2.01 made a lot of changes to the performance of the Z8.
I read or heard from some official nikon source, and i really wish i could find it, that the oversampling is only effective when shooting 4K30 and higher, not 4K24.
In the 4.1K raw yes that is true. I state this in the video. You can do an oversampled 4K60 in the non raw mode if you turn the "Extended Oversampling" mode on.
Very interesting, nice work. This helps me to get the best video quality from the Z8. One thing I wonder is whether the problem in N RAW 4K 24 fps remains constant even for 25,30,50 fps? thanks for this video
You're welcome. Although I didn't test those other frame rates in 4.1K. I would assume 24/25/30 act the same. And I would assume 50/60/100/120 act the same.
16:9 is not what we wanted ( cropped vertically). We want open gate mode, as Panasonic or Blackmagic do it. Of Nikon is serious about filmmaking, they should start paying attention.
So it's looking like the best workflow for H.265 is to shoot 8k, then downsample to 4k in a timeline while upsampling to 4:2:2 or 4:4:4. You'll have to make sure resolve is set up properly to do this on export. This technically should work, although it's maybe not as perfect as 4:2:2 4k but probably very close. So this would give you poor mans 4k 4:2:2. This would also give you a little more detail from the slightly softer Long GOP H.265 8k files.
I had to re-do a lot of tests and update a lot of the video because there was a recent firmware update that fixed a lot of the issues that I found. Thanks for watching!
As I said several times in the video, the H.265 is 10-bit 420. Page 191: download.nikonimglib.com/archive6/vnbui00jEg3u068C2mQ470eyQ539/Z8RG_(En)04.pdf You can get 422 out of over HDMI, which is what you might be seeing in a different section.
@@NickDelDuca With firmware v2.01, I didn’t see any difference with the Extended Oversampling mode on or off with the 4.1K raw modes in any frame rate. I did test this but omitted it in my video because it didn’t have an effect.
@@Josh_Sattin Its so weird that issue happens. Its only in 4k24 and 4k30 right? Once you go to 4k60 it's fine? Those are the 2 framerates I film in the most. Really hope Nikon can sort it out with firmware. They released a new video recently about how to handle nraw in resolve and it was definitely something guys from RED probably showed them how to do. So I have hope.
Why are you messing around with the clip data? Use the Color Managed workflow. Resolve automatically use the RAW metadata to apply a color space transform. Choose Rec.709 and you have a base “look” to work from. Non-RAW files you will have to specify the input color space at the clip level if you’re using a mix of camera sources.
Of course the compressed raw is compromised raw. You don't see that kind of damage in RED raw because its 16 bit file. Even in 14 bit vs. 12 bit the difference would be noticeable in heavy WB changes or shadow lift.
Just want to add, I believe the sensors read data that is converted to 12 bit based on the log curve associated with the format, so SDR vs N-LOG will in fact result in different gamma values, meaning, you will in fact have more information in highlights with N-LOG than SDR. The actual sensor for the Z9 reads out at 14 BIT, which is a shame, since the A7SIII does 16 bit RAW recording.
Raw data should not be affected by which curve is selected in camera. It's raw, not log, SDR, etc. When I asked Nikon about this, they told me that it had to do with how exposure is measured while shooting. This should not affect what is recorded in raw. N-Log vs SDR is just metadata telling the NLE what to automatically apply, but you can change it.
@@Josh_Sattin thanks Josh, a little disappointing in that sense, since it really makes no difference. Interesting how the very low ISO in N-LOG affect the curve. I noticed far less noise in shadows with ISO Lo 2.0
Lowering the ISO below the base and raising the exposure with aperture or adding light will shift more dynamic range to the shadows in most cameras: ua-cam.com/video/AWzqLE64TzI/v-deo.html
The best in detail analysis video about the Z8. You are doing a great job, better than the biggest reviewers out there.
Thank you!!
You're welcome. Thank you for the kind words and for checking out the video!
Love how detailed you go! Finding the noise reduction that the camera does is CRAZY! And testing all that.
This is like a nikon z8 Bible!
This feels like a dissertation 😂 you had a question, hypothesized, tested it and summarized... Cool
Thanks for watching Morty!
Holy shit man this video is so damn helpful for the Nikon Z8 users.
I hope so! Thanks for watching Brandon!
Amazing! We use the Z9 for video, so this is a great addition to the knowledge on how this works. Much appreciated.
You're welcome. Thanks for watching!
Really great work Josh! Not being able to color correct without compromise defeats the purpose of RAW video. Thankfully, the colors look great straight out of Nikon. Probably best to stick with NLOG or Rec709 in camera. Appreciate you taking the time to do all those tests.
Thanks for watching!
Thank you! Very useful information. Much appreciated.
You're welcome. Thanks for watching!
Would love to see open gate and anamorphic mode in Z8. Maybe with the next firmware update Nikon?
Quite the deep dive in this. Thank you! Been using the 4k mode without really noticeable issues lately, but im wondering if ETTR helps that, as I tend to expose as bright as I can without clipping.
I've been using the Z9 since release along with my RED V Raptor, I do mostly Extreme Sports such as Skateboarding and Big Wave Surfing, lots of challenging environments. This is what I use, if I want 8k the 8k 60fps is absolutely fantastic from the Z9 and matches my RED V Raptor 8k almost perfectly with minimal effort in Resolve. If I am using 4k which to be honest 80% of everything I do is then I use the Z9 H265 4k 60 with oversampling turned on and just get the exposure pretty good in camera and it produces from my experience using nearly all of the top tier cameras/cinema cameras one of the best looking 4k videos on the market today, it is an absolute joy to use. The 4.1k RAW 24fps in the Z9/Z8 is just not worth it in any scenario. The H265 120fps is also another absolute gem and when compared to my RED 4k 120fps footage it looks practically identical which is a massive bonus, being 4:2:0 affects nothing for me.
...so in depth! Thank you for putting in the work!!
There were a lot of strange but somewhat expected results here. Yet, I'm hopeful that if Nikon could focus on making a video first camera with some R&D help from Red, they will iron out these kinks and could deliver a truly unique prosumer cinema camera in the near future! Heres hoping...
You're welcome. I had heard from so many people about N-RAW in the Z8 and Z9 so I needed to fully test it out. Thanks for watching!
thanks for the test and explanation, hope they noticed this issue and update next firmware
You're welcome. Thanks for watching!
Great work, thank you
You’re welcome. Thanks for watching!
Great video, Josh! So informative. Hopefully that 4.1k 24p gets fixed and agreed, I would love to see h265 422 in it.
Thanks Tyler! Yeah, I hope they fix this stuff!
FINALLY... spent too much time trying to figure out the Nikon Z8 422 10bit formats to basically come to the same conclusion, its only via HDMI output that it will give what most Sony shooters/editors want which is 422 10bit H265 (they will even take 422 10bit H.264 since may have used it for years) so for second shooters and video guys that want to give that format its a bummer the camera doesn't shoot it internally. Shoots both higher and lower but not that damn one! ...apprecite the deep dive and keep up the great work.
but you can record it with atomos, yes?
@@Tmk969 if you have the H.265 enabled via purchase or feature pack, then yes the Ninja V (in my case) will record 10bit 422 H.265 footage via HDMI from the Nikon Z8 in my most used case of 4K 24fps. I will point you to the Nikon Z8 Reference manual, (reference, not user manual) which is available online, and page 297 shows output bit depth for resolution and frame rates. Hope that helps.
@@nicolascarle8210 yes, have seen that, just wanted to ask someone who used or have atomos :) for me h265 420 is also big disappointment, we need bring that problem to the Nikon somehow
@@Tmk969 agreed, how Nikon missed this one is a head scratcher. Good part is with the Ninja V recording the 422 h265 my 1TB drive on the Ninja gets ridiculous amounts of recording time, paired with a few V-mount batteries I can record long format talks etc for hours and hours without heat or battery issues at all.
Very thorough. Eye opening about the RAW being destructive and kinda all over the place. What's the point?
Thanks for all the details man, I wish they give redcode raw in future😁
You're welcome. I can't wait to see what Nikon does with RED's IP and tech.
Wonderful review. May I know whats your recommendation for the CFExpress Card(s) for 8K60p n-raw recording?
Great job!
Thanks!
You are the only creator (that I've watched - and I've watched a ton) that has actually come out and said that the h265 footage is 420 vs 422. Everyone else that I've seen says otherwise but one day as I was looking in premiere's properties section for a few clips that I took said it was 420 (extended) and I was so confused. I always thought the footage didn't compare to my fx3 footage and I couldn't figure out why I was struggling so much with my z8. I thought I was going crazy and it was only after going back and forth with Nikon that they eventually confirmed it was 420. Thanks for the detailed review as always!
You are very welcome. Sometimes you just have to read the manual!
it's written in manual, also its was talked in Nikon z filmmakers groups
@@Tmk969 I swear I tried searching everywhere. Guess I didn’t search enough but Nikon didn’t even know why it said 420. It was only after a few exchanges that their experts said 422 is only with raw or prores hq 🤷🏻♂️ i was most confused because other reviewers kept saying h265 was 422
@@__eugenelee true :) i think other reviewers just didn't used it much or tested more heavy and deep as this video author. I found it long ago and i'm tellin it everywhere, i wrote about it under another video of this author (Josh). I hate it, because h265 would be mostly used for me.
@@Tmk969 same! It’s such a shame because it would make this way more flexible and I’m honestly surprised it isn’t 422… wonder if it’s a limit on the sensor or if it can be updated via firmware. 🙏🏻🙏🏻🙏🏻
Wahouuuu, Great job. Thanks
You're welcome. Thanks for watching!
BM Ursa 12K has the same sensor tech kind of when you go up with fps and down with resolution it do not crop...and the new ursa cine as well...
So Highlights Recovery is not available with N-RAW on Camera Raw page of Davinci Resolve?
I think with Cinema DNG from any camera that toggle is availbe there.
Please try converting N-RAW to DNG to check if thar toggle appears or not in resolve.
Also, did anyone compare the 25P to 24P or should I get to it?
I wouldn't get too hung up on the lack of 422 using h.265. Yes, other cameras have the option but have you ever done a side-by-side comparison recording 422 vs 420 using the A7SIII for example? I have and I never could see any difference whatsoever even pushing colors to extremes. Of course there's a huge difference between 8 bit and 10 bit, but any kind of meaningful differences between 420 and 422 are not noticeable... at least with Sony cameras using SLOG3. Thanks for the video, very well done.
The 4k Raw option was the most attractive to me as a R5C owner, turns out it's not without compromises, hopefully fixable with firmware updates. I'm still holding out for something with similar specs that offers open gate.
I'm waiting for the Panasonic S1H2, but the Z8 is very tempting
Great video, thanks the deep dive.
Except the bitrate, what's the advantage of using something like a h265 422, over the ProResHQ ?
I get the point of using Davinci (I use it also) but I think it's worth comparing the N-RAW against the ProResRaw, for the "non-destructive" part.
It can also be a great option for Final Cut or Premiere users.
Yeah the main advantage of a H.265 codec is the bitrate and thus overall data usage.
I did some tests with 4k Prores and 4k H.265 and the H.265 files are slightly softer. So keep that in mind. With a proper All-I H.265 file you should technically get better color and a significantly reduced file size, but with this Long GOP compression it tends to soften the image up a little.
Would have been cool if the camera had an All-I H.264 4:2:2 format at least, since I'm not sure anyone does ALL-I H.265 yet. This is why Sony does this with H.264, it eliminates the need for Prores.
So it's probably better to shoot 8k H.265 then sample down to 4k to something other than Long GOP to retain that detail, this would also make that 4:2:0 subsampling a non-issue or at least less of one. Or just stick with Prores.
Would also like to see ProresRAW comparisons as I'm on Premiere and don't want to pay for Resolve Studio to load 8k NRAW.
Did you use normal quality for testing 4.1k NRAW? Normal has about one stop impact on the dynamic range.
Had a lot of picture degradation on Z9 at 4k 25p raw and went to 8k. Thanks for confirming. Also, I thought only prores HQ and h26? was oversampled up to 30 p…NOT raw formats. Might have changed? Have done a lot of grading with the nraw 8k and found it extremely flexible. Your point about not being able to change the color temp to match because the slider didnt facilitate is shallow. There is an abundace of data to enable a perfect match, though it requires some work. Otherwise, the depth of your methodology and analysis was superb. Thank you.
You're welcome. As I said in the video, Nikon told me that the 4.1K raw mode is oversampled in 24/25/30 fps. I have been in touch with one of their engineers. Yes the N-RAW is very flexible, but there were two things I was trying to show with the white balance examples. First is the difference between how you can change the image in raw vs ProRes. The other one is the fact that the raw isn't really raw since changing the white balance is destructive, which it shouldn't be. Thanks for watching!
Very interesting as someone who primarily uses the 4.1K N-Raw for almost a year to date on my Z8, I’ve never had any of the issues you described or show in your video ?🤔 For the past year I’ve gotten consistently good footage. I do use neat video for noise removal. I do use the Nikon lut to grade, but I never use it at full strength,lol. I only use Nikon native lenses?? I use Davinci Studio 18.5 version And I edit on a 60inch tv screen, so I’m always pixel peeping,lol. I have def shot Scenes in “low light” situations and while the footage is noisy and hard to clean up in those scenarios , no color shift that I’ve noticed? Maybe before adding noise reduction? But to be fair I only focus on the finished results, so I’ve never noticed.
I wonder how the HDMI out videos compare
And one more thing, did you have the Extended Oversampling option on or off in your video settings. I wonder what impact this has?
I had it turned off, but I tried it on and off. It had no effect on the raw files with the v2.01 firmware.
Thanks for stating the firmware, every review out there fails to mention this. The firmware version makes a big difference in how it behaves.
You're welcome! I got in the habit of doing that a little while ago because people would watch one of my videos a year or two after I posted it and complain how things were different. And yeah, the firmware can make a big difference. v2.01 made a lot of changes to the performance of the Z8.
It is a pity that H.265 10 bit has only 4.2.0 (
I read or heard from some official nikon source, and i really wish i could find it, that the oversampling is only effective when shooting 4K30 and higher, not 4K24.
In the 4.1K raw yes that is true. I state this in the video. You can do an oversampled 4K60 in the non raw mode if you turn the "Extended Oversampling" mode on.
Very interesting, nice work.
This helps me to get the best video quality from the Z8.
One thing I wonder is whether the problem in N RAW 4K 24 fps remains constant even for 25,30,50 fps?
thanks for this video
You're welcome. Although I didn't test those other frame rates in 4.1K. I would assume 24/25/30 act the same. And I would assume 50/60/100/120 act the same.
@@Josh_Sattin I tested 24p vs 30p in 4.1k NRAW. It acts exactly the same indeed. Same weird color cast in shadows.
16:9 is not what we wanted ( cropped vertically).
We want open gate mode, as Panasonic or Blackmagic do it. Of Nikon is serious about filmmaking, they should start paying attention.
hi, which camera was used to shoot the review in the studio
RED Komodo and Sigma 18-35.
So it's looking like the best workflow for H.265 is to shoot 8k, then downsample to 4k in a timeline while upsampling to 4:2:2 or 4:4:4. You'll have to make sure resolve is set up properly to do this on export. This technically should work, although it's maybe not as perfect as 4:2:2 4k but probably very close.
So this would give you poor mans 4k 4:2:2. This would also give you a little more detail from the slightly softer Long GOP H.265 8k files.
Is this video a reupload from something last week? Could’ve swore I watched this exact video lol
I had to re-do a lot of tests and update a lot of the video because there was a recent firmware update that fixed a lot of the issues that I found. Thanks for watching!
So h.265 is 422 color or 420 color? The manual says both in two different places so it’s confusing…
As I said several times in the video, the H.265 is 10-bit 420. Page 191: download.nikonimglib.com/archive6/vnbui00jEg3u068C2mQ470eyQ539/Z8RG_(En)04.pdf
You can get 422 out of over HDMI, which is what you might be seeing in a different section.
I am wondering if there is a difference between the NRAW 4.1K 24p with and without Extended Oversampling enabled.
I tested this. There actually was a difference on the v2.0 firmware but there is no difference in the v2.01 firmware.
@@Josh_Sattin Does it fix the black clipping issue one way or the other? I'm kinda sad about that mode it was going to be my go to 😞
@@NickDelDuca With firmware v2.01, I didn’t see any difference with the Extended Oversampling mode on or off with the 4.1K raw modes in any frame rate. I did test this but omitted it in my video because it didn’t have an effect.
@@Josh_Sattin Its so weird that issue happens. Its only in 4k24 and 4k30 right? Once you go to 4k60 it's fine? Those are the 2 framerates I film in the most. Really hope Nikon can sort it out with firmware. They released a new video recently about how to handle nraw in resolve and it was definitely something guys from RED probably showed them how to do. So I have hope.
@@NickDelDuca Which issue? The color shift issue in 4.1K24? If so, yes, the issue is not there in 4.1K60/120. I show that in this video.
Why are you messing around with the clip data? Use the Color Managed workflow. Resolve automatically use the RAW metadata to apply a color space transform. Choose Rec.709 and you have a base “look” to work from. Non-RAW files you will have to specify the input color space at the clip level if you’re using a mix of camera sources.
whats the memory card you use in z8 ?????
geni.us/TDSMH
Of course the compressed raw is compromised raw. You don't see that kind of damage in RED raw because its 16 bit file. Even in 14 bit vs. 12 bit the difference would be noticeable in heavy WB changes or shadow lift.
RED's R3D codec is compressed raw.
At 4.1K is line skipped raw video. You simply can't over sample with raw video codecs.
I agree this is weird but a Nikon engineer told me that the 4.1K raw mode is oversampled in 24/25/30 fps.
Just want to add, I believe the sensors read data that is converted to 12 bit based on the log curve associated with the format, so SDR vs N-LOG will in fact result in different gamma values, meaning, you will in fact have more information in highlights with N-LOG than SDR. The actual sensor for the Z9 reads out at 14 BIT, which is a shame, since the A7SIII does 16 bit RAW recording.
Raw data should not be affected by which curve is selected in camera. It's raw, not log, SDR, etc. When I asked Nikon about this, they told me that it had to do with how exposure is measured while shooting. This should not affect what is recorded in raw. N-Log vs SDR is just metadata telling the NLE what to automatically apply, but you can change it.
@@Josh_Sattin thanks Josh, a little disappointing in that sense, since it really makes no difference. Interesting how the very low ISO in N-LOG affect the curve. I noticed far less noise in shadows with ISO Lo 2.0
Lowering the ISO below the base and raising the exposure with aperture or adding light will shift more dynamic range to the shadows in most cameras: ua-cam.com/video/AWzqLE64TzI/v-deo.html
Its not oversampling at 4k
I agree this is weird but a Nikon engineer told me that the 4.1K raw mode is oversampled in 24/25/30 fps.
Yes it is