Highlight Rolloff and DR - fx3/a7s3 sensor vs. A7iv sensor

[SRV1981]

As title asks - which is better of the two? I can’t find good tests comparing. Many of the new bodies sport these two sensors so I’d be curious which one produces the superior image for video. 

[kye]

In digital, highlights don't roll-off - they clip faster than a Ferrari without breaks.  The roll-off is applied as part of the processing that occurs after the image is read from the sensor.

In terms of DR, you have to find a source that has tested both with the same methodology.  Luckily, CineD is one such source:

https://www.cined.com/sony-a7-iv-lab-test-rolling-shutter-dynamic-range-and-latitude/

https://www.cined.com/sony-a7s-iii-lab-test-does-it-live-up-to-the-hype/

 

[The Dancing Babamef]

you asked like the sensor in A7S3 and FX3 would be different. They're not.

[Al Dolega]

2 hours ago, The Dancing Babamef said:

you asked like the sensor in A7S3 and FX3 would be different. They're not.

They pretty clearly didn't- "fx3/a7s3 sensor vs. A7iv sensor" obviously acknowledges the FX3 and A7sIII are the same.

[SRV1981]

6 hours ago, kye said:

In digital, highlights don't roll-off - they clip faster than a Ferrari without breaks.  The roll-off is applied as part of the processing that occurs after the image is read from the sensor.

In terms of DR, you have to find a source that has tested both with the same methodology.  Luckily, CineD is one such source:

https://www.cined.com/sony-a7-iv-lab-test-rolling-shutter-dynamic-range-and-latitude/

https://www.cined.com/sony-a7s-iii-lab-test-does-it-live-up-to-the-hype/

 

Thanks it seems the A7iv has more viewable DR than fx3 due to A7iv being 7k and fx3 is 4k but has internal noise reduction that can’t be bypassed 

 

 

[kye]

14 minutes ago, SRV1981 said:

Thanks it seems the A7iv has more viewable DR than fx3 due to A7iv being 7k and fx3 is 4k but has internal noise reduction that can’t be bypassed 

I think comparisons like this are a journey into very tricky territory.

I'm not saying that DR doesn't matter, or that the A74 doesn't have more than the FX3, but I wonder what situations this will really make a meaningful difference in.

12.9 vs 12.4 isn't that much of a difference, especially once you've got this much DR already.  I like to think about this in terms of "what shots can you get with one and not the other?".  I have spoken a lot about DR in my work, but that's because I tend to regularly be shooting people in front of a sunset, and I want the persons face and also not to clip the sunset behind them.  I think I could get that shot with either of these cameras.

Another shot is of people sitting by a fire, and wanting to show the surroundings as well as not clip the fire (white fire looks odd but you can clip little bits of it as they end up looking like fire highlights).  This isn't a shot I do so I'm not sure if this needs more DR than the A73.

You also need the skill to pull the shadows out in post.  I don't mean just raising the shadows slider either, I mean having the skill to match a shot with (let's say) 12.9 stops of DR with the previous and next shots which are likely to have dramatically less DR.  Imagine that the surrounding shots are of people sitting around the fire - they might have half the DR of the previous shot.  When you have the contrast / curves / LGG wheels / etc cranked one way for one shot, then cranked the other way for the next shot, will all the colours remain similar?  Will the contrast seem natural?  

Colour grading software is getting better but did you know that the normal colour grading controls were developed with algorithms that were simple for computers to do (to reduce processor requirements) rather than looking good?  This often doesn't matter when you're only adjusting things a little bit, but large changes will ruthlessly reveal the weaknesses.  Things like the Resolve HDR Palette were designed to be perceptually uniform, but are you grading in Resolve using this panel?

You also have to consider the other implications of a choice between the FX3 and A74, for example.  Comparing DR is great but it's an "everything else being equal" comparison, but of course, everything else isn't equal.  The extra DR might come at the cost of something else that is more meaningful to what you're doing.  Film-making doing anything other than shooting still-life images on a tripod inside (where everything is controlled and you have all the time in the world) is an exercise in having less time and attention than you'd want to, so you're always trying to work out what is most important and paying attention to that.  I don't know about you, but I routinely look at the footage I have shot and see all kinds of things that I should have done differently - but then I have to remind myself that I was walking down a staircase holding the camera with one hand and the railing with the other and trying not be steady and use the ninja walk while keeping the right focus distance to my subject and keep the nice framing and also not hit my head etc etc etc.  I didn't have mental capacity to do anything else, and if I had paid attention to that other thing then maybe I would have screwed up the shot entirely, etc.

You have to look at the total situation and review what the requirements are for you in your situation and think through what implications you'll have of swapping from one camera to another and the effect on your viewers as they are watching the final edit.  Everything else other than how your viewers feel is a means to an end.

[kye]

It's also worth pointing out that the film you posted could have been shot with almost any camera just by controlling the ratio on set with lighting.  I'm not being dramatic here - you could literally shoot that with the GH1 or my GF3 (of course the resolution and bitrates would suffer..) but the contrast could be replicated.

[SRV1981]

2 hours ago, kye said:

I like to think about this in terms of "what shots can you get with one and not the other?"

So how do we go about answering this when comparing the a7iv sensor (a7c2) versus the fx3 sensor (a7s3/zve1)

[eatstoomuchjam]

You could check the cined lab tests for both.  If you do, the a7 iv seems to have better DR (though much worse rolling shutter).

[SRV1981]

6 hours ago, eatstoomuchjam said:

You could check the cined lab tests for both.  If you do, the a7 iv seems to have better DR (though much worse rolling shutter).

I’ll check, thanks! RS doesn’t affect me for what I do or care about and DR is more important to me. 
 

 

am I seeing better noise in A7iv sensor here?

 

[independent]

This is why the Canon R5c can test relative poorly in raw yet approach very close to 13 stops with some noise reduction applied (in camera) and downsampled from 8k, which Gerald showed this in his review.

Of course, dynamic range isn’t the same as highlight roll off, in which the R5c is better than the FX3/A7s3 because it allocates more stops in the highlights. But it comes at a cost of a lack of latitude in the shadows, like the Panasonic s5iix.

The sigma fp and red Komodo are the opposite, so you have to really protect the highlights. 

None of this really matters if you understand your camera and exposure strategy. 

The Sonys and Canon c70 are balanced in their latitude, which can be better for doc work and run & gun.
 

[SRV1981]

57 minutes ago, independent said:

This is why the Canon R5c can test relative poorly in raw yet approach very close to 13 stops with some noise reduction applied (in camera) and downsampled from 8k, which Gerald showed this in his review.

Of course, dynamic range isn’t the same as highlight roll off, in which the R5c is better than the FX3/A7s3 because it allocates more stops in the highlights. But it comes at a cost of a lack of latitude in the shadows, like the Panasonic s5iix.

The sigma fp and red Komodo are the opposite, so you have to really protect the highlights. 

None of this really matters if you understand your camera and exposure strategy. 

The Sonys and Canon c70 are balanced in their latitude, which can be better for doc work and run & gun.
 

Thanks for the knowledge - so how would you advise someone considering a camera - video first photo second when the a7c2 sports the a7iv sensor and the zve1 sports the fx3 sensor? 

[eatstoomuchjam]

I'd say that if there is a specific feature that one has which the person cares about (higher frame rate, etc), then go for it.  If not, the majority of differences between those cameras will become negligible pretty quickly once you start using them.

I try to make sure to go rewatch Reverie every so often.  It's a great reminder that the 5D Mark II, released in 2008, could look pretty good in the hands of a Vincent Laforet.  Is the quality equal to even an entry-level modern mirrorless camera?  Not really.  Does it look better than about 40% of short films being made by local filmmakers in my area?  Yes.

So nobody can tell you which camera is the best one for you.  That's a personal decision - but FWIW, between the two cameras that you're looking at, I don't think either one is the WRONG choice, especially if you're just getting started.
 

 

[kye]

20 hours ago, SRV1981 said:

So how do we go about answering this when comparing the a7iv sensor (a7c2) versus the fx3 sensor (a7s3/zve1)

I'd suggest testing this for yourself.

Find the situation / situations where your existing setup doesn't have enough DR, and use a camera in stills mode to measure the brightest part of the image and the darkest part.  I'd suggest using zebras.

For example, if you open up your camera lens to f2.8 and set a 1/10 shutter and ISO 3200 and this sets off the zebras then you just adjust those parameters down until the highlights don't set off the zebras and then you'll have your number of stops.

This might sound a bit fiddly, and it is, but it will give you a definite answer.  The alternative is spending thousands of dollars based on a review / vlog from some camera bro with mediocre technical knowledge and unknown commercial interests on YT.

[eatstoomuchjam]

Is there an existing camera?  Sorry if I missed that.  If so, which is it?  That could also be useful to know.

[SRV1981]

7 hours ago, eatstoomuchjam said:

Is there an existing camera?  Sorry if I missed that.  If so, which is it?  That could also be useful to know.

I’m just generally curious and interested in the topic. Currently using an iPhone 14 and Ricoh GriiiX

may dip into mirrorless for video and so find the topic fascinating 

[SRV1981]

https://www.photonstophotos.net/Charts/PDR.htm
 

seems a7iv may have better DR in lower ISO and they’re even in higher 

[eatstoomuchjam]

If you haven't used a mirrorless ILC for video before, either camera is more than enough to get started.  I'll go back to saying that if you care about a specific feature on one of them, that's probably the one to choose.  If not, neither of these cameras will be what holds you back.  The differences in dynamic range between the cameras in most practical shooting situations are unlikely to be severe. 

The A7 IV has some of the best test chart performance for DR of any camera in the industry, similar to my C70 and my GFX 100 II (in 4K with DR boost turned on) - and they are both fantastic.  The other one tests about half to a full stop worse?  If your exposure is way off or the scene has too much contrast, there's a good chance that the difference won't be enough to save your image.  Using similar exposures on my C70 (best DR of any camera under $10k) and my R5 (comes in at like #30 or #40 on cined's charts), the visibilities in DR are different in the form of highlight/shadow details, but it's FAR from night/day.

[TomTheDP]

With FX3 you have the optinon to shoot RAW which will give you a bit more dynamic range. Testing shows the FX3 and Sony A1 have about 1 stop more of usable dynamic range which can be seen in latitude tests. 

Outside of that the FX3, A1, and A7IV all perform very similiar. The A1 has the best internal codec, least processing. FX3 has too much noise reduction but it can be avoided by shooting raw externally. 
 

[eatstoomuchjam]

8 hours ago, TomTheDP said:

With FX3 you have the optinon to shoot RAW which will give you a bit more dynamic range. Testing shows the FX3 and Sony A1 have about 1 stop more of usable dynamic range which can be seen in latitude tests. 

 

Keep in mind from context (quoted below) that the OP is potentially not talking about the actual fx3 and more likely is talking about the zve1 and is comparing it to the fx3/a7s3 due to similar sensor.  The zve1 doesn't shoot raw.  Also, if it's worth mention that recording raw on any of these cameras requires an external recorder which will add $300+ to the price.
Also also, that the OP doesn't currently have a video camera or experience (so raw might be questionable choice) 😃
 

On 3/19/2024 at 9:49 PM, SRV1981 said:

So how do we go about answering this when comparing the a7iv sensor (a7c2) versus the fx3 sensor (a7s3/zve1)