I’VE MADE A MISTAKE! - 8-bit colour isn’t 128 colours, it’s 256. Kids, this is why you should ALWAYS check your work and not do what I do… but regardless, the same principle still applies 👌🏻 thanks for watching. Time to start proofing my work more 😂
I think your scene selection in woodland was less challenging that you'd get in some other situations. I'm thinking that smooth color or luminance gradients aren't often found in green foliage, but are sometimes found in cityscapes, sunsets over water etc., and then banding might be more obvious.
In general 8-bit's color looks muted compared to 10-bit...based on the difference in size you would think it would be much more noticeable..but it probably depends on how you're viewing the content e.g. Beautiful large OLED Panel vs tiny crappy phone 😂
The display really is where the differences will mostly be visible because those gaps between 8-bit and 10-bit are magnified. I think it’s also all comes down to the algorithm that’s compressing the data in which honestly any external monitor that converts 8-bit across an HDMI output to ProRes will be better than the internal solution in a camera. Regardless, one should simply learn Sonys color profiles and how to use them correctly for a particular scene in which brings me to the fact Sony made Cine and S-LOG variants for a reason. Most use S-LOG 2-3 improperly and should expose using a external monitor with a LUT applied. Otherwise, I think the formula is expose 2-stops over with S-LOG 2 regardless of what the zebras or internal reference states. I have no clue why Sony does not reference read/correctly internally.
If only banding is problem... theoretically that could be done by in post... by simply smoothing our sharp differences between colors. Not sure if that feature exists, but just saying it should be possible logically.
I feel like the differences were huge. You can really see it in the shot of the girls face, skin tones look more natural in cam A. Shadows in cam B have a blue/magenta tint like you said. Cam A overall feels more warm, soft, and has a more natural appealing look.
Same thing. I own a D3200, which is not a camera made for videography and it only does record 1080p@30 with 8Bit, but after seeing this maybe I should focus on mastering photography first before wanting a camera capable of 14Bit RAW. :)
One of the most important things is learning with what you have. But I won't deny, there's a few psychological elements to new gear... New video about that coming VERY Soon! Thanks for watching
@@fusseldieb i shoot with a dD5500! i invest in the motorized slider, weebill S with focus motor, lcd monitor, 35 1.8 and 50 1.8 lenses, variable ND Filter and a 150w led light with a 90cm softbox, its make a world apart! There are very good examples of what your camera is capable of here in YT with shooting video, in 5 years I have realized that what people like is how you tell the story, the editing. Don't fall for the silly marketing game that you use the newest camera to make art. 15 years ago people still produced magnificent cinematographic jewels with more basic cameras...
Most people don't even look at videos on a decent TV, much less a color calibrated one. They look at it on their phone. As a photographer, I may have a good color calibrated screen but experience has taught me that 99.9% of the people that will be looking at my photos won't. So in the end, how much does it really matter. These days, I just try the get the brightness levels high enough to be screen clearly and I no longer worry about the color because no one cares or can tell this difference anyway. Equipment manufacturers have to come up with new specs every year to convince people to buy new shit :)
The biggest difference I can see daily is backlight. Like person standing back to the window in sunny day. You can forget about window details while exposing the subject but... Is that really important or just telling the story and quicker workflow? Thats individual.
if you want to see spesifcly defferent between 8 bit and 10 bit, you should be shoot under or over exposure, from there we know that 10 bit better to save highlight and shadow.
Absolutely, I'd never normally go that heavy on a grade though haha! But some people do, and you've got to properly test the two. That's why I included both ungraded and graded
Lightbulb moment! Fantastic description of the difference between 8-bit & 10-bit. The staircase graphic was a stroke of genius. Made perfect sense. Thank you. Waiting for Black Friday to order my A7Siii, but its so hard to wait :-). One month, more or less, to go. Another wonderfully illuminating video Danny. Thank you for taking the time to share your knowledge in such a great way.
Glad it helped your understanding! And it’s a pleasure, I love making these videos as I learn as I make them. I’d never directly compared 8-bit and 10-bit until this, so it’s definitely eye openings as I make these videos. Thank you for watching!
@@DannyBligh yeah having the A7iii and the A7iv shows up a massive difference between the two. Time to sell the A7iii I think and upgrade that to the A7Siii. Thanks again and keep them coming.
Two of your videos just popped up and gave me exact answers that I was looking for - great channel man! Just picked up the Sony A7iv and finding the best settings. Cheers!
Even youtube already convert the video to 8bit, but i think we still can see which one is 8bit, which one is 10bit. I know you tried to match the both camera, but notice the detail (color of the tree, exposure from sky, skin color changing,...) I think it depend on situation, with the good lighting, slight color correction and grading will not make the big different. Ofc, with some project that you working with so much lighting, specialy with RGB lighting, 8bit and 10bit will make a big change. But i still prefer shooting with the highest quality you can, you dont know if you really need to fix it in post or not :D Good video btw, thank you man.
Nope. Couldn’t tell. Especially on a smart phone. I guess it really depends how the content is being consumed but I think the whole 10 bit vs 8 bit is bullshit. Just RU-vid “filmmakers” parroting specs like everyone else.
I'm using a 17" 1080p IPS monitor, watched it at 1080p and couldn't tell the difference. I thought the 8bit camera was 10bit. Really impressed me! As he said, YT compression makes it almost indistinguishable, unless you REALLY push your color grading to extreme levels. That means that for the average YT vlogger a 8Bit camera is completely reasonable, still.
An 8-bit camera is definitely reasonable. I always say in a lot of my camera comparison videos that there isn't really a right or wrong.. Nor is there a PERFECT camera that suits everybody. Gear is important but I think of 'important' being relative. If you are making a simple vlog, sure a phone will probably cut it. Could you make it better by shooting on a RED, absolutely! But it's down to what each user needs. There's some camera brands and cameras that I wouldn't ever use as I don't think they suit my needs, doesn't mean they aren't great cameras! Thanks for watching, both of you
Might be wrong, but I think you are missing the main point here. Yes, banding is an issue, and additional compression on RU-vid will make it worse anyway. But the main advantage of the 10 bit is how far can you push colors until they start to look all wonky. Therefore, there is a significant difference in possible end result.
Not missed the point at all. Colours fall apart in 8-bit far sooner than 10-bit, they where there's graded and ungraded clips, and the graded clips are heavily graded to try and break the image without being ridiculous. The reason 8-bit colours start to look what you call wonky is because of the lack of information, the lack of colours available, the shadows a lot of the time become super blocky and that's due to the banding which stretching the image causes. The main advantage is obviously as you say, how far you can push it. And it's the same with 12-bit and 14-bit. But the whole reason for this channel is to go further than what other people explain. I want everybody to go away with more understand of 10-bit is better because 10 is bigger than 8. A lot of RU-vid channels say 'WOW It's got 10-bit colour' and a lot of people I meet in person have no idea what 10-bit truly is, other than 'you can stretch the colour further'. So no, didn't miss the point. Just went deeper, way past the point. But appreciate the comment mate!
Viewing on an iPhone 14 pro at the default setting (480p), they looked indistinguishable. When I bumped the resolution to 2160p I could definitely see the magenta hue, but no way was any banding apparent, even when I magnified to 8x. I had to screenshot the image and zoom right in to one the gaps between tree branches and the sky to see the colour banding. At that magnification the bands in the colour gradations are visible on the left too, but on the right side they’re m ever so slightly more pronounced. That’s what told me the right hand side (B) was 8-bit. To be honest the magenta hue didn’t tip me off, though it should have as my a7RIV suffers from that, while my a7SIII doesn’t. I can’t wait to find a buyer for the RIV so I can splash out on the RV and the image from both my bodies will match right out of the box 🙂
Sony likely compresses 6K down to 4K to reduce most the banding and gradient issues, at least on older 8-bit Alpha cameras with color-profiles. Also, I’m sure it’s even more reduced if newer Alpha lines provide more than 100mbps (12MB/S) across the buss to SD cards. It’s odd that Sony only allows 100mbps when Panasonic allowed 200mbps on the the GH4 or higher. 100mbps is only 500KB per frame set at 24fps. 24MP is 1-2MB per image when compressed so Sony could have at least given us 200mbps for a full MB per frame of headroom.
I've never heard any viewer say I would have enjoyed the video, but I couldn't since it was shot at 8bit rather then 10bit. Only us camera nerds even care about these things.
Well, until recently, I also thought that this discussion is nonsense. Until my client wanted S-LOG 3, on a Sony 8 bit camera, cause that is how he is "used to color grade footage of"... I told him, not a good idea, on this camera (6300). With 8 bit, you really need to get the S-LOG 3 perfectly filmed in camera - talking bout over oder under exposure. He did not listen. I mean, I listened to him. I was "just" the DoP on this one. Well... after he corrected the temperature, added some contrast to his taste, NOISE reduction in Da Vinci, color space transform (to make the S-LOG 3 look "normal"), then added a film look LUT. . . You can imagine how many blocky stuff I have seen on this one. HORRIBLE IMAGE! Conclusion: 1. If you plan to DE-NOISE and film in S-LOG or any LOG, save up and buy a 10 bit camera! You will not regret it later, like I do now. . .
Unless, you film stuff around and just minor adjustments, or probably a lot of black and white footage with a lot of grain and "junkish", old school look, then keep the 8 bit around :)
Absolutely!! With the amount of info in an 8-bit image, you definitely have to get it pretty much perfect in camera. The whole ‘over exposing log’ is a lot more vital in 8bit to be able to get as much data in the shadows where with 10bit you have a lot more room for error but regardless, getting it right in camera and over exposing will always benefit you Thanks for watching
The camera also makes a big difference. After watching this, i went to test SLOG2 on my original ZV1 and it's terrible and reminds me of why i don't use it even at 100 Mbps. Too much noise and the minimum ISO of 1000 is not good.
First I was like clearly A is 10bit then I forced myself to change my mind XD no idea why, but in the and I guessed it wrong. But there is clearly visible difference. I don't understand how some people can't see it. Probably they just don't know what to look at. The explanation was great, I liked your video a lot, and I wanna work with 10bit :D I just learn color grading now days and I enjoy a lot, and sometimes I feel like I could do a better job with more information. But it is rare at the moment, also davinci asks money if I wanna color grade 10bit footage :D that's holding me back right now the most
Thank you! Glad yo hear that you liked it! For Sony cameras there’s definitely a colour shift between 8 and 10bit - mo idea why! But you can still colour grade 8-bit to a reasonable way. I’ve always found that 8-bit you have to get at the right exposure to get the most out of it. Practice with 8, get good and then you can invest in DaVinci to do 10bit ✌🏻 Thanks for watching
I’d go even further than that, I always recommend that people invest in the optics more than the camera bodies. You swap bodies every couple of years sometimes but the lenses stay with you for a while
Have to totally disagree on that. Will your lens give you 10-bit, 4k or better dynamic range? Just no.... A shitty camera won't be magicly saved by a good lens. And if for you the lens is more important because of the bokeh or because of more light.... a bit of ISO noise won't kill nobody ;) Peace!
I am gonna take a wild guess here and say B is 10 bit, at leat up to the image with the sign " This way", when I stopped and wrote this. Let's see if I am right! :) but at 1:30 looks like the camera A is the 10 bit...
can i tell the difference between 8 bit and 10? YES, absolutely when im grading my footage and pushing it in post...YES! oh im sorry, you meant can I tell from the images you just showed me.
The stairs was such a good way to explain it! I thought A was 8 bit, but changed my mind with the photo of the sign. I think the difference is so small most people wojldn't notice the difference, and I'm not sure the difference is worth another £1000 to upgrade to the a7iv.
It’s more down to how it’s captured and giving the the information to edit and effect it, not so much how it’s rendered. It’s a similar principal across many industries. For instance, a lot of audio in recording studios is recorded at 96Khz as the sample rate, allowing the system to capture anything from 1hz right through to 48Khz. The human ear can only hear between 20hz and 20Khz (and that’s being generous!) but it allows you to control and render it down to what we hear on iTunes 🤷🏻♂️😂 and at the end of it, we listen to it on our phones
@@DannyBligh So there is a difference if we record in 10bit despite the fact that YT will format us to 8bit, from the material I understood that there is not.
"8-bit footage has been around much longer..." Really, well, what was before DigitalBetacam? In the beginning, "digital" meant quality, not affordability.
B to me seems to have more dynamic range but it could just be the contrast is higher than Cam A - excited to see the outcome and awesome video = well put together and clear
I picked camera A for 10 bit only because I noticed slight difference in the shots with the leaves. I only recently have cameras that can shoot 10bit and some in raw, but honestly I don’t like to use those features as much because of the massive storage needed for them. I’ve been working with 8bit for years and know how to shoot without having to fix that much in post(where 10 bit and raw come in most useful)
I worked for a company making their social media content from phone footage. The footage was all 8-bit, but it was surprisingly gradable. 8-bit has more flexibility than people give it credit for. I grade and edit in resolve, so I had no issues as long as I used the tools that were colorspace aware. It could even withstand a CST to Arrilog c3.
Yes and no... 10-bit will always be better, if I had a choice I'd always go for 10-bit. However, if you're wanting to make a quick video and colour grading isn't on the agenda, 8-bit is more than acceptable. Thanks for watching!
subscribed! thanks danny that helped me to FINALLY understand the colour realm of camera science. that’s one i never really understood and it feels so good to level up in my knowledge thanks to the brilliant stair description that you used. did you use a ninja atmos recorder to bring your Siii up to 10 bit? thanks
That was a simple & therefore a powerfull eyeopener... Thanks a bunch 4 it... in my case it Was a notion that B was darker with a HDRish feel to it, while on the other hand A seemed to be more natural looking - I was leaning towards the 10 bit naturalness on the A, but it wasn't a sure bet, cause the HDR hype would suggest that it's actually B that's the picture world New Age :) Well the difference was noticable, but I think the winner would depend strongly on personal preference... One more THX 4 the video 👀
Got it right in the first 5 seconds honestly. Something about more color information makes it look less "cheap" digital look like mobile phone cameras do. 1080p 10bit or 12 bit 4:4:4/4:2:2 looks miles better to me the 8bit 4k or 8k video.
Well explained. With a Camera having no 10 bit option. Does picture profiles like slog Hlg etc gives better dynamic range or color (smooth color shift) compare to normal 8bit footage???
Any profile like log will give you better dynamic range. But you’ll see more banding if you use a narrower profile such as S-log 3. When I owned an A7iii I spent my time filming in cine 4, flatter than normal footage but only just. Also made it easier/faster to grade and correct
@@DannyBligh Thanks! So to be specific- apart from better Dynamic range does any Profile gives smooth color gradient (less color banding) in 8-bit, compared to a normal footage without any picture profile?
Hi, I am using Sony a7s II for almost 8 years now. I was thinking to change to a newer camera like fx30 or a7 iv for 10 bit and s-cinetone. Is it worth changing?
100% worth it. Just the ability to have 10bit but also the colours from the later Sony cameras have evolved so much since the A7Sii. On top of that, you have a huge amount of new features, better autofocus, 4k (with little to no crop on some cameras/modes). As for which to get, the fx30 is a fantastic camera from what i’ve seen. I don’t own one nor have I used one. But I have used an a7iv, i have a few videos on my channel about it and it’s a great camera
To be honest, it was hard to tell but I was able to tell at the end, slightly. Most people wont give a damn about that. I would probably record in 10 bits on a dark footage, it'll probably help when increasing shadows. Opinions?
I agree! We all get wrapped up in tech and specs. When you last watched a film did you think “jheeeez, this must be 14bit raw shot in 50fps in a full fr….” Because i certainly didn’t 😂 Agree with the shadows and gradient and fringing in the sky and between colours are a biggy between the two. But meh, i prefer 10 bit as i can stretch the colours more too, log is much easier to work with. But before I shot in log, i used 8bit 🤷🏻♂️
10 Bit is better for hard contrasts in sunny day, covered sky and diffuse light give no much advantages to 10 bits. In the past they recorded in studios each channel separately on expensive tape decks with high tape speed, spent hours mixing just to put it on noisy tapes or vinyl who had nice cracks like camp fire
😂😂 yup. Same as when I was at uni for music, we were always taught that you should record at 98khz sample rate, mix with particular things but you end up mixing it down to a 320kbps mp3 and played through an iphone speaker 😂