Why are 4k cameras sharper than some best 1080p cameras like F3?

zlfan

Active member
Even though F3 has a 12.9 mp sensor (according to Alan Roberts' test).
The same goes C300 original, AF100, etc. They all have 8-12 mp sensor.
By the way, AF100's sensor is 12 mp, F3's is 12.9 mp, C300/C100s' are 9 mp. However, C300 is the sharpest, F3 is close to AF100 as being soft. So, mp does not determine the sharpness here, olpf kicks in for sure.

Is it possible that 4k cameras with a 8-12 mp sensor have thinner olpf? If so, does this mean 4k cameras may have more alias and moire than those "old best 1080p" cameras?

If olpf is the same across the board, theoretically, only a camera with a sensor of 20 mp or higher is truly significantly sharper than a camera with a sensor of 8-12 mp, right?

If so, FS7, FS5, F5, F55, FX6 are actually not sharper than F3, FS100, FS700, C300, AF100, etc. FX9 is truly sharper. Right?
 
Last edited:
That is a complicated topic because it varies greatly on the camera manufacturer and the camera model and what you're comparing it to. But some 4K cameras definitely have more aliasing and moire than the best 1080p cameras.

But regardless of how many MPs or which OLPF, the quality may be limited in other ways, both in HD and 4K.

In general, sharper for most people means "clearer" and less blurry and more detail, and many great 4K cameras utilize their additional resolution well, especially in wider shots where you may really see the difference if you're watching the high-resolution, minimally compressed file.
 
On my 4k F5, I tested using the same lens, the following modes: 4k, 2k in 4k scan (down sampling), 2k in 2k full scan (pixel binning), 2k in 2k center scan (s16). 4K is as sharp as 2k downsampling, sharper than 2k pixel binning and 2k S16.
So, I think 4k to 2k downsampling may not affect the sharpness or resolution much. I agree with you that this is camera and manufacturer specific.
 
Having great downsampling is when many people say they don't see much of a difference between that mode and a higher-resolution mode on their cameras.

I swear, until this day, Blackmagic's 1080p is just as good as the 4K (in many of the models)...they do downsampling so well.

I think this is more because the 4K just isn't as good as it could be. The F5 is awesome, but it was never known to be a great 4K camera.
 
Having great downsampling is when many people say they don't see much of a difference between that mode and a higher-resolution mode on their cameras.

I swear, until this day, Blackmagic's 1080p is just as good as the 4K (in many of the models)...they do downsampling so well.

I think this is more because the 4K just isn't as good as it could be. The F5 is awesome, but it was never known to be a great 4K camera.

Maybe you are right about F5, since F5 has several OLPFs for different modes. As F5 was only a 2k camera when introduced, 4k upgrade is just buying an activation key. After upgrading to 4k, or more accurately, activating 4k recording, F5 may need a 4k specific OLPF accordingly. I doubt anyone did that to his F5.
 
Last edited:
Having great downsampling is when many people say they don't see much of a difference between that mode and a higher-resolution mode on their cameras.

I swear, until this day, Blackmagic's 1080p is just as good as the 4K (in many of the models)...they do downsampling so well.

I think this is more because the 4K just isn't as good as it could be. The F5 is awesome, but it was never known to be a great 4K camera.

I am interested in your experience on BM cameras.
If BMPCC 6k or pro or G2 or BMPCC 4k, have similar high quality 6k/4k/1080p, this is great news. As at 1080p, you get up to 240 fps.
F5's 2k 240 fps is not good, it is either 2k pixel binning or 2k s16 center, significantly softer than its 4k or 2k downsampling modes.
 
IIRC, Mitch Gross once said a 5.7K sensor was a sweet spot (EVA1) or enough to see the benefit of downsampling to 4K, so also keep that in mind when you're evaluating your F5.

And other cameras which have 4K from only a 4K or 4K-ish chip; maybe not quite enough to see the true benefit of 4000 pixels.

___

As far as BM, the slow-motion - depending on the framerate - could be windowed so it's not as good as 24p. I don't remember when the cropping starts, but the quality also gets softer.

Personally, I have never seen 240 frames per second look impressive from anything besides a Phantom. Maybe passable from some systems, but that's uncharted territory still, even in only 1080p, IMO.
 
You are right. Seems great 4k cameras are actually 6k to 8k sensor down sampled to 4k. 4k sensor 4k cameras may be lack lusting.
Why is Phantom better than all others? Maybe the sensor read out is much faster? Or shutter design?
 
There are only a few but I would think so, but I also think some of the newer 4K just from 4K-ish chips look good as well.

Different aspects of a camera's image-making pipeline have improved and contribute to the final result.

___

I'm not sure exactly why but I imagine capturing that many frames with decent picture quality is for a specialized industry, one that Vision Research is in. Those cameras are extremely expensive, lol.
 
Maybe the newer 4k-ish sensors are new designs, like dgo for greater dr, and faster read out, and back luminating, etc? So even the cameras are 4k resolution, higher dr, better iq at 60p or 120p, make the final footage looking better?
 
def could be, I don't see why not...every year most electronics seem better, even by a little
 
BTW the opposite is often true with modern 4k cameras. 1080p is less sharp on 4k camera. How can this be? Well many of these cameras shoot 4k natively and line skip to obtain 1080 yielding a poor quality. You have film in 4k and export from your editor 1080.

I have a camcorder that films in 4k and HD but there is no obvious difference once exported as 1080. This could be because the camera codec is highly compressed and further compressed in export from the editor and Youtube compress it again. I've also discovered the camcorder crops in on the already small 1" sensor when shooting in 4k. Every camera does things differently so one can't make definitive proclamation.

Heck we have the FX6 is soft thread...but I think in that case the user was shooting RAW and didn't realize the cameras applies internal sharpening so when using RAW you need to add sharpening in post. When comparing cameras one could appear to sharper but maybe that's a result of more internal sharpening.

If you follow what is the current cinematic trend they look down on sharpness and add Promist filters to soften. Moral of the story you along with every techy camera nerd will never be happy with the image quality.
 
Last edited:
Don't let your eyes fool you...all close-ups/tighter shots (resolution matters much less) plus the lighting and grading makes them better...and they are stills, you might see them in motion and think differently, or maybe not :)
 
Don't let your eyes fool you...all close-ups/tighter shots (resolution matters much less) plus the lighting and grading makes them better...and they are stills, you might see them in motion and think differently, or maybe not :)

Yeh, you are right. Closeup. I was very cautious when watching camera companies' promo short films, but not on his work. Hahaha.
 
Has anyone taken off F3's OLPF? This may elevate F3 to the current 4k cameras' resolution levels.
 
Very rarely does anyone do that with any camera unless it's easily removable (like a feature).
 
And RED sold different ones so it was part of their business plan (although I don't doubt they also did it after receiving real industry feedback from directors/DPs, etc).
 
Im in the have a proper OPLF camp.

having worked for fine weave fabric companies moiree can ruin your shoot.

Failing to be in this converation is false detail

One should observe a chart blending into softness just below the resolution so a 4k chip should be filtered to 3.9k

While non oplf camera may look sharp.. they can create garbage that must be rejeected.

6k aquired a 6k with a 5.9k oplf is probably the 'proper way' to aquire 4k

--

Compression is also not noted. It seemed to be that the 600mbs 444 sr codec from th F35 whopped the '4k' of the FS7.. but that was not scientific.
 
Back
Top