Sony a7 III discussion

[Inazuma]

14 hours ago, scotchtape said:

Look at the mountains top right @ 8s in.  (Location scouting for personal timelapse project)

No grade

Can't wait for Ronin-S (hopefully it's good)

Just to be sure it's the camera, have you turned off the IBIS and checked if the problem still occurs?

[Don Kotlos]

There is no way for IBIS to create that artifact since it cannot adjust the YAW/PITCH of the sensor.

IBIS can only compensate for YAW/PITCH movements by X/Y shifts because with medium to long lenses a YAW/PITCH movement can be approximated as an X/Y shift. 

That type of transformation that can be seen in that video posted from scotchtape can only arise from an imbalanced/faulty gimbal or digital stabilization.  

[scotchtape]

48 minutes ago, Don Kotlos said:

There is no way for IBIS to create that artifact since it cannot adjust the YAW/PITCH of the sensor.

IBIS can only compensate for YAW/PITCH movements by X/Y shifts because with medium to long lenses a YAW/PITCH movement can be approximated as an X/Y shift. 

That type of transformation that can be seen in that video posted from scotchtape can only arise from an imbalanced/faulty gimbal or digital stabilization.  

Did you mean the opposite? 

IBIS can be implemented in many ways, you know the Sony has 5 axis including the yaw/pitch...

I do think the overloaded gimbal combined with the IBIS made that artifact, so I will test with a bigger gimbal later, however it does show how wonky things can get with UWA and IBIS on the Sony (full frame too), which makes sense considering the size of e-mount, the limited distances it has, and the distortion of UWA lenses at the edges.

If you don't think that this effect is caused by IBIS you can check this video below, at one point there is a comparison between the EMII and GH5 while the guy is walking (I'm pretty sure without a gimbal just handholding), you can see similar type artifacts on the EMII side, video should be linked @ 5:43, and this is on M43 sized sensor.  

This type of artifact pretty much has to be due to IBIS/OIS because it has to do with the direction of the sensor shifting in relation to the lens.  Without OIS/IBIS the sensor stays in position relative to the lens, so you would see a normal perspective change.

Whether it's better than shaky cam or not is subjective.

 

[Don Kotlos]

48 minutes ago, scotchtape said:

Did you mean the opposite? 

IBIS can be implemented in many ways, you know the Sony has 5 axis including the yaw/pitch...

No I did not. While IBIS can compensate to some degree for yaw/pitch, it cannot change the yaw/pitch of the sensor. The yaw and pitch corrections are really only X/Y IBIS movements. E-M1ii has also digital stabilization thats how you get these artifacts.

[Trek of Joy]

18 hours ago, Robert Collins said:

Even more idiotic is that the cameras reset the clip names (back to 0001) with each SD card (and this cant be changed unlike with stills where the number can be sequential.)

That's always frustrating too. I travel and shoot a lot of stills and video, after downloading footage/stills I reformat my cards. I keep my stuff organized according to what country I'm visiting, and in all my Sony-shoot folders there are a number of sub folders with each day's content. It makes file management a royal PITA. Every other camera I've owned kept the sequential numbering, so I could just fill a folder with new content from reformatted cards. I went to a couple dozen countries with the XT2's and each country has an easily organized folder. I did the same with the a7r2/s2 and every folder is a mess.

People have been complaining about file management with Sony's for years. They finally updated the menu's (I didn't say fixed because IMO they're still a mess), it seems like this would be relatively simple in comparison to address.

Chris

[Don Kotlos]

I know it is not a solution for everyone, but I use Lightroom to copy all my footage (stills/clips) and with the "rename during import" function you can adjust the naming as you wish. I use the date + sequential numbering. 

[scotchtape]

1 hour ago, Don Kotlos said:

No I did not. While IBIS can compensate to some degree for yaw/pitch, it cannot change the yaw/pitch of the sensor. The yaw and pitch corrections are really only X/Y IBIS movements. E-M1ii has also digital stabilization thats how you get these artifacts.

So you're saying all the diagrams from Sony showing yaw and pitch correction from ibis on the sensor are all misinformation? Is there a source for this? They literally market it correcting (obviously constrained by physical spave) yaw pitch and roll, as well as xy movement. 

[Don Kotlos]

29 minutes ago, scotchtape said:

So you're saying all the diagrams from Sony showing yaw and pitch correction from ibis on the sensor are all misinformation? Is there a source for this? They literally market it correcting (obviously constrained by physical spave) yaw pitch and roll, as well as xy movement. 

While implied, they never showed the sensor tilting. Their claim is that it can correct for yaw/pitch which it can but not very well. 

Here is a small discussion: https://diglloyd.com/blog/2015/20150823_0800-SonyA7R_II-IBIS.html

But you can do the experiment yourself as well. Take the lens off and observe the light reflections off the sensor while you tilt the camera when it is on or off. If it was actually changing the yaw/pitch of the sensor you wouldn't notice the light reflections changing since the angle should remain the same (stabilized). What you will see is that the reflections change the same way when the camera is off or on.

You can even do another test to see how with longer focal lengths, tilts of the camera affect the x/y shift more (since their trying to compensate for yaw/pitch).

 

[scotchtape]

Thank you for providing that link it explains things nicely. That answers how the ibis works for Sony (no tilting).

I do think the ibis ifeature is what is causing the artifacting, as the other video I posted shows.  With ibis off you don't see that type of artifacting vs having it on.

The blog post you referenced also explains the artifacting caused by the ibis implementation as it applies a transformation to the image attempting to correct for tilt and yaw.  You said ibis  can't generate that artifcat because it doesn't tilt the sensor, but the implementation is that the image is digitally transformed instead of having the sensor tilted. This transformation is especially noticeable with uwa lenses (also noted in the blog post).

If you mean the physical ibis implementation cannot generate that artifact then that might be true, however the implementation of ibis includes the digital transformation neccesarily, so turning on ibis will generate those artifacts depending on the camera movement, which would casually mean it is the ibis causing the warping.

[Don Kotlos]

1 hour ago, scotchtape said:

The blog post you referenced also explains the artifacting caused by the ibis implementation as it applies a transformation to the image attempting to correct for tilt and yaw. 

No, the only issue with correcting yaw/pitch movement with x/y shifts is that for wide angle lenses the focus plane changes and thus you get corners that are out of focus.

With Sony there is no digital stabilization, and with the E-M1II there is an optional extra digital stabilization which can be turned off. 

[scotchtape]

If the IBIS is not causing the warping, please explain how the gimbal is causing the warping.  The gimbal controls where the camera is pointed.  How would the gimbal cause the edges to warp, when the center remains in the same position?  If the gimbal were to move the camera's direction, the center would certainly shift too. 

It sounds like you're saying if I have IBIS on and get the warping, it is not from the IBIS.  But if I turn off the IBIS and get no warping, it is still not the IBIS, even though the only time the warping appears is with IBIS on...

[Xavier Plagaro Mussard]

I have the Sony A7 III in my mind. If someone has it and have found big problem, please tell me. And a couple of basic questions: does it overheat?? Can you record a couple of hours straight??  Thanks!

[Geoff CB]

31 minutes ago, Xavier Plágaro Mussard said:

I have the Sony A7 III in my mind. If someone has it and have found big problem, please tell me. And a couple of basic questions: does it overheat?? Can you record a couple of hours straight??  Thanks!

Does not overheat, unfortunately it still has the 30 minute record limit. 

[Don Kotlos]

6 hours ago, scotchtape said:

If the IBIS is not causing the warping, please explain how the gimbal is causing the warping.  The gimbal controls where the camera is pointed.  How would the gimbal cause the edges to warp, when the center remains in the same position?  If the gimbal were to move the camera's direction, the center would certainly shift too. 

It sounds like you're saying if I have IBIS on and get the warping, it is not from the IBIS.  But if I turn off the IBIS and get no warping, it is still not the IBIS, even though the only time the warping appears is with IBIS on...

My guess is that because the gimbal was unbalanced, it had some residual vibrating motion with a pivot point in front of the camera. That would cause the camera to move in both X & Y and change the YAW/PITCH. You can imaging it as the movement space belonging on the surface of a small sphere. The IBIS of the camera would compensate for the X,Y shifts, but the uncorrected yaw/pitch could bring out this warping. It might be that the rolling shutter & lens distortions exaggerate this effect. While it is very hard for this movement (surface of the sphere) to occur in any natural circumstances, the combination of IBIS with an unbalanced gimbal could do this. 

 

[Xavier Plagaro Mussard]

1 hour ago, Geoff CB said:

Does not overheat, unfortunately it still has the 30 minute record limit. 

So I assume you can rehit record every 30 minutes for a couple of hours, no problem?

I would love the sensor of this Sony and the codecs from the GH5!!! ;-)

[BasiliskFilm]

2 hours ago, Xavier Plágaro Mussard said:

I have the Sony A7 III in my mind. If someone has it and have found big problem, please tell me. And a couple of basic questions: does it overheat?? Can you record a couple of hours straight??  Thanks!

Like most hybrid cameras, recording over 30 mins is deliberately disabled to keep it taxed differently from a video camera. I imagine the work-around is to use an external recorder, and a SSD drive is probably  a more cost-effective way of capturing that amount of footage anyway. The camera might need extra power too.

[jonpais]

6 minutes ago, BasiliskFilm said:

Like most hybrid cameras, recording over 30 mins is deliberately disabled to keep it taxed differently from a video camera. I imagine the work-around is to use an external recorder, and a SSD drive is probably  a more cost-effective way of capturing that amount of footage anyway. The camera might need extra power too.

Then how can Panasonic offer unlimited recording?

[BTM_Pix]

16 minutes ago, jonpais said:

Then how can Panasonic offer unlimited recording?

I guess they must have paid the additional tariff in the EU at least?

There is a way around the 30 minute limit on the other A7/A7R/A7S cameras with the OpenMemories tweak but unfortunately the A7Riii and A7iii don't have the PlayMemories app stuff in it so thats not possible on those and I'm guessing on future Sony cameras either.

[Inazuma]

Still haven't shot properly with this camera yet but did a quick test of the s&q mode. Love it. Footage is quite sharp and relatively free of digital artefacts (sharpening/moire/aliasing). 

 

[Don Kotlos]

Yeah I like the S&Q mode as well. It is very easy to create a nice output, slomo or timelapses. Just keep in mind that if you need the highest quality then the normal 120/1080p uses ~100mbps whereas the S&Q mode uses ~60mbps.