Thomas Smet
Veteran
Good points. The A7III only shoots 60p @1080 and 8bit. Yeah it's showing its age, but it's still attractive even at it's current price. I assume they'll maintain its price until the release of the A7IV then offer a discount to clear inventory before they discontinue it. S5 has better IBIS on a gimbal its moot. I think GH5 still has the best IBIS.
As far as the 10bit color, I was watching video comparing the two and surprisingly there was almost no difference between 10 and 8bit. Best explanation was with the high internal compression you end up losing much of the gained color. The effects of bit rate on color bit depth has been brought up in other threads. You can't make conclusions on specs alone.
10bit is one of those things you only need it when you need it. 8bit can look great and if you don't need to push a grade very hard then 8bit is likely more than enough. Where 10bit becomes important is when shooting a situation where you would get banding in the shot. Once an image has banding it will always have banding. Its kind of like bitrates in compression. 150mbps IPB is likely more than enough for 90% of shooting situations. Its that 10% of an explosion of surf waves or whatever that can break the codec. 10bit is kind of the same thing. Most of the time there are enough colors to reproduce the scene in 8bit. Its those rare situations where it is not enough where 10bit can help and render a clean gradient.
Grading today is very sophisticated with 32bit float which creates in between samples during grading. As long as there are no visual gaps then 32bit float should be able to fill in new samples. Unlike upscaling where the math is very complex to recreate details, 32bit float wants to create a more subtle smooth gradient which is easy for math to do. So a lot of online videos comparing 8bit and 10bit are inaccurate when they try to push a grade and say they look the same. Of course they do because visually the shot looked fine in both 8bit and 10bit anyway. The 32bit float could easily fill in the values as the grade was pushed for both.
Online examples need to shoot something that would normally fall apart with 8bit.
10bit like dynamic range, extra bit depth extra audio bit depth and so forth is all about overhead in a professional environment. Its to have more than needed so there is room to adjust or to make sure something subtle isn't lost. We like more DR so we don't lose the subtle details in the highs and/or lows. We like intraframe formats vs 150mbps IPB to make sure we don't get macro blocking on complex scenes. We like 10bit to make sure we never get any kind of banding in any shot.