Sony a7 IV
- Thread starter jonpais
- Start date
cyvideo
Veteran
BREATHING! Breathing, breathing, breathing. That's why I love to use manual cine designed glass. Very minimal breathing. Something still camera glass has never really had to be bothered with until lots of people started to use still camera glass for the moving image. Lack of breathing is especially important on zooms where breathing has to be controlled at all lengths. I trained in film on film glass and came to accept that that was how a good lens behaved. It didn't "zoom" when focused.
The fact that you have plenty of pixel real estate on the A7 iv I think lets you can get away with the slight crop the "de-breathing" function takes without taking too much of a hit in overall resolution.
Resolution, true resolution, well that's another thing altogether, isn't it? A 4096 sensor like the FS7 after a Demosaicing resolution loss of 20% is really only around 3.28K. Whereas the loss on the A7 iv with 7008-pixel sensor after Demosaicing still leaves you 5.6K resolution to be oversampled to UHD. What's the old saying "8K for the best 4K, 4K for the best 2K".
I only ever really shoot mainly 50p HD or 50p UHD/4K so the A7 iv crop would work for me with manual S35 glass where breathing isn't an issue. Even when using FF glass that doesn't breathe. I often used the Angenieux/Tokina FF zooms which exhibit virtually zero breathing. Even using them at times with a speed booster in APSC crop mode produced some very decent "breathless" results on previous FF sensors. Having the added benefit of an extra stop of light gain and an extra stop of shallower DOF when needed was an added side bonus. Shot a lot of interview material that way with great results.
The A7 iv after a 20% Demosaicing loss from its 4.6 crop mode (4672 pixels) still produces a sub 4K resolution of 3.73K. Having lived with the 3.2K "pseudo 4K" from the FS7 for a number of years I don't see that being a real concern for my work. On the rolling shutter side of things, I've lived with the 14msec rolling shutter of the FS7 with no great impact on my shooting so an APSC crop rolling shutter in 50/60p on an A7 iv of 12.8msec is something I can live with.
Well, the proof is in the pudding as they say. I'll find out sooner or later no doubt as the order is in for one.
Chris Young
Want good old lack of breathing? Maybe chase one of these down. If you do check the centring is correct. Old Angenieux designed Tokina 28-70mm F/2.6-2.8 or even the F/2.8. Marked differently for different markets but exactly the same lens. Either a Mark I or II. Don't touch any other models as they do not utilize the Angenieux parfocal breathless design. Speed boosted becomes an F2 100% parfocal, constant aperture lens in true cine-style. Nice colors, good bokeh and vintage flares if you like that kind of thing
Bassman2003
Veteran
I watched the CVP video and the color does not look that different but it is difficult to judge with the exposure difference. Maybe the exposure deficit can be fixed before production samples hit the streets. Sony gave pretty much everything us video folks wanted at this price point. If they can throw in the external RAW I think it would be a nice compliment to an A7sIII as the 2nd camera, not a total equal. The skin tones for the different cameras are all over the place in that video. I also noticed the huge detail in the A7IV compared to the R6 when they were both briefly on the screen at the same time with the color charts. The Sony probably has too much detail for flattering skin visuals. I look forward to seeing more footage and photo results.
Peter C.
Veteran
I haven't had the time to watch the reviews, but is it that surprising that the IV would have deficits compared to the A7S3? How could Sony sell a camera for less that was just as good? Otherwise there would be no reason to buy A7s3. Whether it's because of cheaper components or crippled by software does it matter? Logically speaking it would make more sense to limit or downgrade it by using less expensive internal components because Sony would be accomplishing two thing at once.
I was happy to see they finally added a flip screen and I liked what they did with the dial that makes it easier to switch to s&q and to use the top dial to control the exposure mode. I also like the full hdmi port.
So what are the cripples they put in? I expected lower res lcd and view finder and no 4k 60 or if it was included to be limited in some way.
I was happy to see they finally added a flip screen and I liked what they did with the dial that makes it easier to switch to s&q and to use the top dial to control the exposure mode. I also like the full hdmi port.
So what are the cripples they put in? I expected lower res lcd and view finder and no 4k 60 or if it was included to be limited in some way.
jonpais
Well-known member
As soon as ProRes RAW is ready, they'll update the Ninja V immediately.
It is not going to cannibalize sales of the a1 or the a7s III.
It's not clear whether the breathing compensation feature will work with PRR - we'll just have to wait and see. That would be a serious consideration to take into account for those of us who exclusively shoot RAW.
As far as ProRes RAW goes, on the a7s III, it''s 4.2K (4264 x 2408) resolution, which amounts to 23.7% higher resolution - which is visibly more detailed than the internal codec. It means there's a bit more leeway for cropping or stabilization than with the internal codec. De-noising ProRes RAW in post yields a much nicer, organic looking grain structure. Pretty much anyone who's seen a comparison between internal processing and de-noising in post agrees.
For HDR, internal XAVC-S I 4:2:2 should be avoided.
"The Y’CbCr encoding model is popular because it conceals subsampling artifacts vastly better than does RGB encoding. Sadly, while Y’CbCr works well in Rec 709, it doesn’t work very well for HDR. Because the Y’CbCr values are created from RGB values that have been gamma corrected, the luma and chroma values are not perfectly separate: subsampling causes minor shifts in both. This isn’t noticeable in Rec 709’s smaller color gamut, but it matters quite a lot in a large color gamut. Every process for scaling a wide color gamut image to fit into a smaller color gamut utilizes desaturation, and it’s not possible to desaturate Y’CbCr footage to that extent without seeing unwanted hue shifts.
My recommendation: always use RGB 4:4:4 codecs or capture raw when shooting for HDR, and avoid Y’CbCr 4:2:2 codecs. If a codec doesn’t specify that it is “4:4:4” then it uses Y’CbCr encoding, and should be avoided". -Art Adams
This paper by Dolby shows how 10-bit Y’CbCr is not only not great for displays but also that maximum color deviation of 10-bit Y’CbCr exceeds the acceptable threshold at every luminance level for theatrical presentation, making quantization distortions like banding and contouring visible. It damages the image at capture as well, affecting not only saturated blues and reds, but the all-important skin tones as well.
https://professional.dolby.com/sitea...paper_v071.pdf
It is not going to cannibalize sales of the a1 or the a7s III.
It's not clear whether the breathing compensation feature will work with PRR - we'll just have to wait and see. That would be a serious consideration to take into account for those of us who exclusively shoot RAW.
As far as ProRes RAW goes, on the a7s III, it''s 4.2K (4264 x 2408) resolution, which amounts to 23.7% higher resolution - which is visibly more detailed than the internal codec. It means there's a bit more leeway for cropping or stabilization than with the internal codec. De-noising ProRes RAW in post yields a much nicer, organic looking grain structure. Pretty much anyone who's seen a comparison between internal processing and de-noising in post agrees.
For HDR, internal XAVC-S I 4:2:2 should be avoided.
"The Y’CbCr encoding model is popular because it conceals subsampling artifacts vastly better than does RGB encoding. Sadly, while Y’CbCr works well in Rec 709, it doesn’t work very well for HDR. Because the Y’CbCr values are created from RGB values that have been gamma corrected, the luma and chroma values are not perfectly separate: subsampling causes minor shifts in both. This isn’t noticeable in Rec 709’s smaller color gamut, but it matters quite a lot in a large color gamut. Every process for scaling a wide color gamut image to fit into a smaller color gamut utilizes desaturation, and it’s not possible to desaturate Y’CbCr footage to that extent without seeing unwanted hue shifts.
My recommendation: always use RGB 4:4:4 codecs or capture raw when shooting for HDR, and avoid Y’CbCr 4:2:2 codecs. If a codec doesn’t specify that it is “4:4:4” then it uses Y’CbCr encoding, and should be avoided". -Art Adams
This paper by Dolby shows how 10-bit Y’CbCr is not only not great for displays but also that maximum color deviation of 10-bit Y’CbCr exceeds the acceptable threshold at every luminance level for theatrical presentation, making quantization distortions like banding and contouring visible. It damages the image at capture as well, affecting not only saturated blues and reds, but the all-important skin tones as well.
https://professional.dolby.com/sitea...paper_v071.pdf
Last edited:
jonpais
Well-known member
Why don't you do a teardown and let us know? :2vrolijk_08:
cyvideo
Veteran
Exactly. We all want it now and we want it for less and we want one camera to be the best at everything. I think a little bit of realism is called for here to balance out some of our flights of fancy on what we expect in a hybrid at that price point I see the A7 iv as being two fairly distinct cameras rolled into one. 1) a pretty decent 33 megapixel still camera and 2) a pretty decent APSC 50/60p 10-bit HD/UIHD 422 video camera.
Up against anything else from Sony in APSC, which are currently 420 8-bit cameras it has the fastest read out speed for rolling shutter at 12.8ms. The worst being the A6xxx series at 40ms @ 24p and 31ms @ 30p. The a7 iv full-width 4K video has a 26.5ms rolling shutter. Which is comparable with the other models in its class, which are mainly 20-24 megapixel cameras. Compared to the 24 megapixel Panasonic S1H it's very comparable as the S1H at maximum res and dynamic range runs out at 29.7ms. For a 33 megapixel camera I think that a 26.5m rolling shutter is to be expected and is in line with current camera development and expectations. If you want dramatically faster than that for FF video well then you have look at the A7S iii/FX3 which are peerless at their prices for video but pretty average for stills.
Obviously the video centric lower res sensors will display the best rolling shutter figures. The Sony FX6 (as well as the Sony a7S III) shows an 8.7ms rolling shutter for full-frame, 25fps UHD recording.
Its either go with one of the two video centric cameras above or go up-market to the A1. Which in 4K FF recording at 25 frames per second has a rolling shutter of 8.1ms with pixel binning pixel taking place. The A1 in APSC crop mode at 25fps has a rolling shutter of 11.1ms. Not dramatically different than the A7 iv at 12.8ms which surprises me as the A1 is running a stacked BSI sensor vs just a BSI sensor in the A7 iv.
A lot of us have good manual S35 and Sony E-Mount APSC glass. Some of us expensive hi performance manual PL glass. For those of us in that camp it's horses for courses for the $$$ and that's where the A7 iv drops in quite nicely.
If this is truly representative of the A7 iv's output I can only see it being another good addition to the line up.
OTH. Or possibly Nikon's leaked Z9? Where the video info is saying unlimited 8K at 60p with the worlds fastest scan rate and 120p continuous shooting. It never stops!
Chris Young
Last edited:
ahalpert
Major Contributor
I think it's really cool that they implemented breathing correction in-camera. There was a small discussion on dvxuser during the last year about why that had never been done.
I, too, am unconcerned about the impact on resolution. I think you've got more than enough for starters.
I'm always just a little wary of monkey business. You're not quite going to know whats going on with your focal length. You could lose a smidgeon of FOV at the wide-angle.
One potential issue i could see is: let's say the breathing is worse on a zoom at the wide end. So there's more of a crop there. And you want to execute a smooth zoom-in. But now the camera is cropping out slightly as you zoom in, in an unpredictable fashion. Maybe this would never happen, I dunno.
The other thing that comes to mind is that I used to use electronic image stabilization with the a7siii all the time on gimbal, which gives you a 1.1x crop. Then I got the 50 1.2 and realized that i thought it was even prettier without the crop because it has lovely edges. So now I use EIS with my 70-180 on gimbal but not my 50 or wider.
So, my guess is that if I had the a7iv that I would leave breathing correction turned off until I saw a shot that required it or I was using a lens with bad breathing. My current lenses aren't killing me with breathing although I'm sure they breathe.
But again, progress and innovation are good. More capabilities is a good thing
ahalpert
Major Contributor
this probably wouldn't even be that hard - you just need metadata for the focus distance and focal length and also have a lens profile for the software to work with. the only downside is that you'd be working with the 4K footage after it's already been downsampled.
jonpais
Well-known member
The metadata is in the files.
jonpais
Well-known member
How would I know? I don’t even have the camera.
ahalpert
Major Contributor
What do you mean? It detects the color temp and r/B g/m adjustment more accurately using auto white balance? Or that when you set a specific Kelvin color temperature, that the a7siii is more accurately balancing to the number you input while the a7iv is off by a margjn?
Tim Sewell
Well-known member
Mine rarely leaves my camera in general usage. A lovely lens.