Is the Sharp 8k Camera Better Than the GH5 in Low-Light?

In a previous video, Dave Altizer demonstrated the Sharp pre-production 8k camera. What many people really wanted to know was how this performed in low-light situations and its high ISO capabilities. 

The megapixel wars have commenced again it seems, and this time it's for video. We haven't even gotten used to 4k yet and camera manufacturers are producing 8k cameras already. In a recent video by Kinotika, Altizer demonstrates some of the low light capabilities of the Sharp 8k camera. Generally speaking, higher resolution cameras tend to offer worse low light performance at any given sensor size. Of course, this is not a rule and just a guideline as there are many cameras that offer better low-light performance even when resolutions have increased. What interesting to me is the fact that Altizer described the Sharp camera to potentially have better high ISO performance compared to the Panasonic GH5. If this is correct, we might have a very capable competitor. 

The video quality from the camera does look pretty good and Altizer has been kind enough to provide video clips for you to download and view. 

Personally, I'm really not bothered about 8k resolution video, it's mostly useless to me. I'm more interested in seeing features like higher frame rates at 4k, being able to shoot raw, significantly better dynamic range without the need of log profiles, and higher bit rates for color. Those features are actually useful and provide practical benefits. Simply increasing the resolution is easy and fun for the marketing department but is it really beneficial. 

In any case, who cares what I think, what are your thoughts on the new Sharp 8k camera?  

Log in or register to post comments

24 Comments

Spy Black's picture

Having a higher res relative to target viewing res is always handy in a few ways, like downsampled quality, especially with higher ISOs, but also handy for stabilization "breathing space", as well as greenscreen work in less than 4:4:4 depth, which pretty much everyone other than big productions have to deal with.

Usman Dawood's picture

444 is colour not resolution right?

Spy Black's picture

Yes, the accuracy of your key edges is dependent on color res, but you can mask ratty edges by downsampling. Well, ratty as compared to 4:4:4. :-) Modern keyers are cake compared to yesteryear, and you can eek some nice edges from modern keyers, but downsampling will just be the icing on the cake.

Rk K's picture

It has a bigger, 16:9 sensor, so it should have better low light.

Usman Dawood's picture

How much larger is the sensor compared to the GH5 I thought they were very similar. Also pixel size must be significantly smaller in the Sharp right?

Rk K's picture

Not by much, it only comes from the aspect ratio, but not insignificant either. Low light performance depends on the sensor size, sensor tech and aperture, not on the pixel size.

Usman Dawood's picture

Surely it works based on pixel size and not sensor size specifically. The reason why larger sensors are described as having better low light is that at any given resolution a larger sensor will have larger pixels allowing for more efficient light gathering abilities if all other factor remains the same.

Rk K's picture

Not how it works. And A7R3 for example has better iso performance than the A73. Pixel size doesn't matter with modern sensors.

Usman Dawood's picture

Actually, the a7 III is noticeably better than the a7R III when it comes to low light. Plenty of comparisons demonstrate this. Even in the DPreview studio comparison, you can clearly see the a7 III being about a stop better.

At 12800 on the a7 III the noise is about the same as the a7R III at 6400. Performance from the a7 III is about the same as the a9 in low light.

This is true even for video, in fact, it's especially true for video.

Rk K's picture

That's because you're looking at the comparison tool at 1:1, which is idiotic. You have to look at them scaled to the same output medium, because that's how images are used. Then the r3 is clearly better at very high isos. The a73 has better video because the r3 doesn't do a full sensor readout for that.

Usman Dawood's picture

People don’t look at images scaled they look at the image as a whole. Also the video is 4K for both and the a7 III is significantly better in low light.

What’s the point of having high resolution sensors if you’re going to scale images down. It’s idiotic to negate the resolution when comparing if you’re trying to determine that as the variable.

The a7S II is also much better in low light for video and both shoot the same resolution. I’m super 35 mode it is “full sensor” readout and even then the a7 III is noticeably better in low light.

Let’s explore your point specifically; what part of having a larger sensor provides better low light performance if pixel sizes are exactly the same?

Rk K's picture

"People don’t look at images scaled they look at the image as a whole." That's the entire point, dummy. You want to take the same photo with both cameras, so you should look at them adjusted to the same output medium (be it print or web), not zoomed in 1:1. Noone looks at images 1:1.

Super 35 is not full sensor readout, but a 1.5x crop. It's in the name! That's why the r3 is 1 stop worse in video iso performance. As for the s2, it's a specialty camera, quite a bit worse in normal iso ranges than the normal 73. It only gains a slight advantage in extreme cases (iso 100k+), but that's irrelevant, because both are unusable there.

Usman Dawood's picture

When I said "full sensor" (with the "") you knew what I meant, it was to suggest no line skipping. Both cameras in crop mode offer the same sensor size effectively and both have full pixel readout. Even still the a7 III is better at low light. Essentially same sensor size, same output resolution and both without line skipping but the a7III is still better. This is because pixel size makes the difference not sensor size. Once again I'm going to ask what part of having a larger sensor provides better low light performance if pixel sizes are exactly the same?

Also, don't make personal comments I thought we were having a polite discussion. It's childish at best.

Rk K's picture

The 73 is only better in FF mode (where it reads out the full sensor at 6k then downsamples). In crop mode the R3 is actually better, because it can do the 6k downsampling, while the 73 only reads 4k.

Usman Dawood's picture

That's not true have you looked at the comparisons or used both of the cameras?

The a73 is better in low light in both modes even with the a7R III having the advantage of downsampling from 6k.

You're wrong about the performance of these cameras in low light but in any case, we're moving away from the actual point. You still haven't answered my question

what part of having a larger sensor provides better low light performance if pixel sizes are exactly the same?

Rk K's picture

Yes, I looked. There's not much difference if any. What you should be more concerned about, in regards to your argument falling apart, is that the s2 isn't any better than either. It's actually worse in most cases than the 73 because it uses an older sensor.

Usman Dawood's picture

The a7 III about a stop better than the a7R III in video both in FF and crop mode.

I don't really have an argument here I'm posing questions and putting it to you based on some of my assumptions. You talked about sensor tech being a factor so you can't really compare current gen tech to previous gen tech. New sensors are BSI and the a7s don't have that tech.

If you compare the a7s II to any camera within that tech gen then the a7s was significantly better. I'm not suggesting that sensor tech is not a factor by any means I agree with you there but that's not what we're discussing here. I'm, saying that sensor size inherently is not a factor and that pixel size is the actual factor.

I've asked this question multiple times now but you've consistently evaded it.

what part of having a larger sensor provides better low light performance if pixel sizes are exactly the same? If you can't answer this question then your points are moot.

Rk K's picture

"The a7 III about a stop better than the a7R III in video both in FF and crop mode." Most certainly not, you'd have to be blind to believe that. It's 1 stop better in FF mode, for obvious reasons. It's about the same in crop mode.

BSI doesn't make that big of a difference, it doesn't actually make any noticeable difference if the sensor is FF and only 12mp...

I keep answering it, but you're too thick to realise. The content is scaled to the same output resolution, no matter what camera you shoot it with. That's how you should compare the cameras. Obviously, either the larger file gets downscaled, averaging the pixels and reducing the noise, or the smaller file upscaled, making the noise worse. You'll end up with roughly the same amount of noise and similar image quality, at high isos.

Alternatively, at lower isos, you can trade the extra detail for more noise reduction if you're using a higher res sensor.

Why do I have to keep explaining this? It's not some advanced, formal logic, a kid would understand it.

Usman Dawood's picture

You’re far too emotional to have a proper discussion with and I think think you have very little self control. You resort to childish name calling; it’s petty and laughable.

You haven’t answered anything. I switched discussions to video because output resolution is the same when comparing 4K. Your points make no sense so you go back and forth between subjects and points to try and make it sound like you know what you’re talking about.

You’re wrong about the Sony cameras and plenty of comparisons demonstrate this. There’s evidence to prove you wrong.

You offer nothing of value to this discussion and I don’t think you’re mature enough to continue this discussion.

Grow up.

Rk K's picture

I mean, you're way too stupid to have this discussion and change the subject when you don't understand something, so why do I even bother. Swiching the discussion to video makes no sense, because there's far more in camera processing going on with it, using different codecs, bitrates, etc. that make direct comparisons difficult, but you were the one bringing it up. Still, it proves my point (including the comparison between the r3 and the 73) just as well. This evidence you speak of does not exist. You were wrong about the dpreview scene, so you dropped it, you were wrong about the s2, so you dropped that as well and now you're wrong again. Grow up, admit it, maybe learn something...

Usman Dawood's picture

The a7S II is another great example

cameramanDop Shanghai Hong Kong's picture

At "ISO don't remember", anything looks great

T Scarb's picture

If pictures or video get any sharper it is going to hurt my eyes to look at it. Honestly I go into Bestbuy and some of the video they play on their top TVs makes me squint and go wow at the same time.

When is there too much detail.

Spy Black's picture

There's a lot of pseudo detail that you're seeing on those TVs on display, as their built-in sharpening is set pretty high. It's pretty easy to spot the ringing artifacts from over-sharpening on those sets at contrasting boundaries. It can give the image a lot of "crackle" and coarseness that can be off-putting.