X-T3 vs A7III vs EOS R vs Z7 vs Pocket 4K - Video Quality Compared!

[noone]

4 hours ago, thebrothersthre3 said:

Thank you for the info, that makes complete sense now. Seems Sony is one of the more accurate ones. I wonder if they are consistent across all cameras though. 

My experience is that Sony IS probably the most accurate and it is pretty much across all (most) of their cameras.

That said, I don't think most manufacturers are "fudging" and I think it is just a way for DXO to match them up (maybe a couple do seem to be almost a stop different to others Fuji for instance at least historically).

I always like to use the dynamic range charts on DXO for a(rough) comparison as that has  the DR and both measured and manufacturer ISOs.

[Kisaha]

Max did a chart a couple of years ago and the most accurate was Samsung NX, while m43 were the least. I am at work now, when I am going back to my base I could find the chart somewhere.

[Yurolov]

Where are all the Halloween jokes for the Nikon guys. Come on, we are better than this. Something funky going on there. 

[TurboRat]

12 hours ago, Robert Collins said:

For lowlight, these results are pretty much as expected

The A7iii is full frame supersampled from 6k to 4k, the BMPCC is M43 (-2 stops), EOS-R FF sensor heavily cropped (-2 stops), XT-3 APSC (-1 stop) and the Z7 is high resolution sensor with pixel binning (a comparison with a Z6 would be fairer.)

Indeed looks like A7iii is better in low light but from comparisons (with light lol) , skin color and sharpness looks better on the EOS R. I dont own any of those cameras, just based on clips from these guys:

.  

[wolf33d]

10 hours ago, thebrothersthre3 said:

Z7 looks bright but doesn't hold much detail. 

My comment was ironic. The Z7 looks like a zombie with makeup 

[noone]

12 hours ago, Kisaha said:

Max did a chart a couple of years ago and the most accurate was Samsung NX, while m43 were the least. I am at work now, when I am going back to my base I could find the chart somewhere.

Out of curiosity, I just looked at a heap of cameras from most manufacturers and compared the various manufacturers settings on DXO versus the measured ISOs.

Things are all over the place.

Fuji does not get scored these days (because of the X trans thing I think).

I looked at five different ISOs and things vary at each settings but there were a few things that surprised me.

I think Sony and Samsung WERE probably overall the closest to measured standards* but the recent Sony cameras have dropped back a fraction so they are similar to the others.      Samsung was closer to the measured amounts too but they stopped making cameras around the same time as Sony dropped a bit.     (It probably goes something like Sony, Samsung, Samsung Sony).

If I had looked at different ISO settings I might have seen something slightly different.

It does seem that at least for the thirty odd cameras I looked at and at the five ISOs I looked at the closest to measured might be the original A7s and then the NX1.        At some ISOs the Samsung is closer and at others the Sony.       The Samsung NX300 is also close and then maybe the Sony A3000 (I looked at the usual suspects and some cameras I have owned and some I am interested in and some historic ones). EDIT A7s is closer to measured over more ISOs in their common range than the NX1.

At some ISOs, some cameras that are overall quite a way from the measured ISO are better than others.      For instance, at ISO 100, the cameras closest to 100 are the Canon 7D and 7Dii (both 94) and then the Panasonic GH4 (88) but they all drop away a lot above ISO 100.        Despite the Canon 7d pair being so close to measured ISO at ISO 100, the other Canons I looked at are all further away than most APSC or larger sensor cameras.    Panasonic GX7 is also close but at ISO 125 instead of 100 (108 measured).

The Olys ARE quite a way behind (EM1 ii measures 83 but at ISO 200).

All this really means NOTHING but I was curious and find it interesting.

* Fuji isn't tested these days and does have a reputation for being further away from measured ISO but the camera I found to be closest overall was actually the Fuji S5 (and was closer to measured ISO at all three I looked at including 100 (95).

Again, this means NOTHING and if I looked at other cameras and other ISOs things might be different.

Oh and DXO is a great guide but it isn't Gospel ( I do love their RAW conversion software though).

[Kisaha]

@noone wow! you took it to another level!

I found the segment from the Max's video. Obviously is a couple of years old, but the trends should be similar.

I would like to know the P4K's measurements. The only thing keeping me from buying the P4K is the lack of real world experiences (as of now) and the fear of the small sensor. My preferable one is an APS-C.

[Robert Collins]

Typically the older the camera the 'closer' the iso is to the measured iso. As each generation goes by the manufacturers 'cheat' more so the 'improvement' in iso 'looks' better. Take the E-M1 measured iso at 6400 is 3870 - the EM1 ii measured iso at 6400 is 2797!!

And look at the big 'shift' in iso of the GH series between GH1 and GH5?

[webrunner5]

6 hours ago, Kisaha said:

@noone

I would like to know the P4K's measurements. The only thing keeping me from buying the P4K is the lack of real world experiences (as of now) and the fear of the small sensor. My preferable one is an APS-C.

 

Yeah but if you put on a Speed Booster XL you are nearly FF like Andrew says. 1.2 crop. A 24mm ends up being a 28.8. A 16mm a 19.2. Not that big of deal. And if you use one of those wireless control things that are cheap now for focus and one also for aperture you have a near FF Raw beast for less than 2000 bucks. A poor man's Arri LF. 

Sure not something you are going to hang off your neck. But take off the wireless thing's and put on a small, light prime and away you go. The PK4 is a game changer.

[DBounce]

17 minutes ago, webrunner5 said:

Yeah but if you put a Speed Booster XL you are nearly FF like Andrew says. 1.2 crop. A 24mm ends up being a 28.8. A 16mm a 19.2. Not that big of deal. And if you use one of those wireless control things that are cheap now for focus and one also for aperture you have a near FF Raw beast for less than 2000 bucks. A poor man's Arri LF. 

Sure not something you are going to hang off your neck. But take off the wireless thing's and put on a small, light prime and away you go. The PK4 is a game changer.

Speedboosters do not change depth of field. 

[webrunner5]

28 minutes ago, DBounce said:

Speedboosters do not change depth of field. 

F stops determines DoF.

[Shirozina]

4 minutes ago, DBounce said:

Speedboosters do not change depth of field. 

They do and they don't. On a zoom where you maintain camera to subject distance and zoom to maintain the same FOV they do. On a fixed focal length lens where the FOV changes and you change your camera to subject distance to compensate then no.

[webrunner5]

30 minutes ago, webrunner5 said:

F stops determines DoF.

But then there is the equivalence thingy, but that gets sort of crazy. Some truth to it and not.

From this EOSHD link down the page.

 

[thephoenix]

 

[Mako Sports]

43 minutes ago, thephoenix said:

 

Already posted that.. 

[noone]

F stop AND focal length AND distance to subject determines depth of field.

You can have infinite DOF with a f0.95 lens and you can be out of focus at f8.

Speed Boosters and other focal reducers do NOT turn APSC cameras into FF ones, they do not affect the sensor at all.

All they do is change the lens, just like adding a tele converter does so you change the angle of view so it matches what it would be if it was on a FF camera and the give a boost in speed as well of around a stop so you can use a lower ISO on the boosted camera against the FF camera for similar result.

If there were native lenses the same as the speed boosted lenses it wouldn't be confusing would it?

Equivalence is real and any differences are down to things not being EXACTLY matching (I would almost never try and get things the same between sensor sizes and I have owned ILCs with sensors of most sizes -Pentax Q, M43, 1.5x APSC, 1.6x APSC, FF and MF with film).

With regards the cameras closest to ISO 6400 measured VS manufacturer above, that might be the A7s at ISO 5407 measured but again, those are all for RAW stills and video might be very different and why I doubt the P4k will be measured by DXOmark.

 

[webrunner5]

10 minutes ago, noone said:

Equivalence is real and any differences are down to things not being EXACTLY matching (I would almost never try and get things the same between sensor sizes and I have owned ILCs with sensors of moat sizes (Pentax Q, M43, 1.5x APSC, 1.6x APSC, FF and MF with film).

Equivalence is NOT real.

https://***URL removed***/forums/post/57745283

[noone]

20 minutes ago, webrunner5 said:

Equivalence is NOT real.

https://***URL removed***/forums/post/57745283

Seriously?      Look at the post by "Great Bustard" in particular   https://***URL removed***/forums/thread/4002825?page=3#forum-post-57746271 but others as well.

Unless someone ever does a set up to match EXACTLY, there will always be some small differences but in the real world it is close enough for those that need to match things though again, it isn't something that matters much to me as I use whatever I have to get the shot and don't try to shoot something so it matches what another camera/lens combination would get.        As long as I can get the DOF I need using the ISO, focal length and shutter speed I have available, I am happy (with enough pixels).

[webrunner5]

7 minutes ago, noone said:

Seriously?      Just look at the third post in that thread by "dennis" (as well as a few others "Great Bustard" in particular).

Unless someone ever does a set up to match EXACTLY, there will always be some small differences but in the real world it is close enough for those that need to match things though again, it isn't something that matters much to me as I use whatever I have to get the shot and don't try to shoot something so it matches what another camera/lens combination would get.        As long as I can get the DOF I need using the ISO, focal length and shutter speed I have available, I am happy (with enough pixels).

I can't find the article that really proves it. I am not home at my computer. I am not arguing about your thought about the three things that make up DoF. I am arguing that when it comes to DoF FF versus m4/3 as an example it is pretty much bullshit. FF is Not really any better at  Shallow DoF than m4/3 is.

[noone]

2 minutes ago, webrunner5 said:

I can't find the article that really proves it. I am not home at my computer. I am not arguing about your thought about the three things that make up DoF. I am arguing that when it comes to DoF FF versus m4/3 as an example it is pretty much bullshit. FF is Not really any better at  Shallow DoF than m4/3 is.

FF isn't any better at shallow DOF than M43 and equivalence doesn't say it is in theory.

It does work out that way practically though because of what is available (not because of equivalence).

The same applies with FF against MF.      MF though has far fewer options and except for some rare expensive aerial photography lenses, MF lenses are not that fast generally.

I have a lovely old FD 24 1.4 L that I miss using on my A7s while it is (still) broken.      To match the DOF on M43, I would need a 12mm f0.7 (and it would then not be an EXACT match so there would be differences).     How many 12mm f0.7 M43 lenses are there?

Same with my old FD 85 1.2 L, for equivalence purposes I would need an M43 42.5 f0.6 aprox lens and to get equivalence with a lens like a FF 50mm f0.95 lens well that one technically you can't use equivalency as you would need a M43 25mm lens slightly faster than f0.5 which is as fast as you can use in air I believe (though stand to be corrected) though a 25mm M43 f0.5 lens would be close-     Any of THOSE available?