GH5 firmware 1.0

I'll keep this fairly brief, but I had the chance to use a (beta) GH5 with some firmware that had a much more sensible version number today. (Apparently the firmware arrived yesterday.) Here's what I have to add to my previous comments:

(1) It has (finally) gained a menu option for AF control.

Speed is as you'd think. Sensitivity is how much it tries to hang on to the current subject, or not.

(2) The video tracking in Locked-on sensitivity is impressive, even if someone walks in front of the camera.

(3) The video AF at short range is probably worse than before (which I hated). Do note this was indoors and so not that well lit. Changing from 5m to 1m it took 4 seconds to achieve focus, passing through correct focus and overshooting twice before getting it at the third attempt. This was repeatable. Anything under 3-4m wasn't good.

(4) I like the way the IBIS works with adapted lenses with no electronic connection. When you turn on the camera with such a lens attached it pops up a dialogue asking for the focal length. It also stabilised my 350mm "legacy" lens really well (that's 700mm FoV and very hard to use on a GH4).

(5) The menus have changed a lot, complete with dialogues explaining their function as you scroll down (which very annoyingly hide the next few menu entries so you don't know what's next, but can be turned off).

(6) Oh and on the subject of the menus, they go on forever, bring a book to read as you scroll down. (Plus some of the names aren't that intuitive.)

(7) BTW (not firmware related) shooting stills and comparing with my GH4 using the same lens there really wasn't that much of a difference in detail captured, which surprised me. Maybe the GH4 AA filter is really weak?

I'm leaning towards getting one, but I need to see that scaling the full 20MP sensor for video, rather than using a sensor pixel per video pixel (per GH4) doesn't lead to moire/aliasing effects (like Sony tend to) or a lack of sharpness (it quite amuses me that Sony A7rII users slag off the 5DIV for its crop ratio but then often shoot in Super-35 crop mode as using the full sensor is a lot less sharp).

Be sure I am jealous.

Is the viewfinder really that much better than G80 or EM1 II?

Franz Weber wrote:

Be sure I am jealous.

Is the viewfinder really that much better than G80 or EM1 II?

It's nice and a step-up from the GH4, I haven't looked through either of the others close enough to a GH5 to say for sure, but IIRC they aren't that different to the GH4 (all 2.3M dots).

Mikej12 wrote:

Thanks for the update Dr Jon.

Disappointed that the stills do not provide more detail than the GH4, I was hoping for an improvement in stills photography.

Thanks again,

Mike J

I feel it should be better, but I used the same lens on both cameras and the difference was small. Now obviously 20MP is a small step up from 16 and the lack of an AA filter will add at most a 10% improvement so maybe it's there if you look close enough. I just shot the same scene on both cameras at a bunch of apertures and compared the images. I'm sure more scientific results will follow and there might be some aspect of the scene that wasn't ideal, but I couldn't say what it might be.

Isn't there supposed to be a setting for 'moving object prediction' as I've seen in pictured menus? The hierarchy I've seen is 1. AF Sensitivity  2. AF Area Switching Sensitivity 3. Moving Object Prediction.

The menu you show looks very different and doesn't have the flexibility of what I've seen.

Ken Ross wrote:

Isn't there supposed to be a setting for 'moving object prediction' as I've seen in pictured menus? The hierarchy I've seen is 1. AF Sensitivity 2. AF Area Switching Sensitivity 3. Moving Object Prediction.

The menu you show looks very different and doesn't have the flexibility of what I've seen.

Didn't see that, just the ones in my first post, but the menus go on forever and I didn't photograph them again this time. The menus have quite a lot of changes from the ones of the betas, plus some are really poorly organised or have non-intuitive names.

Dr_Jon wrote:

I'll keep this fairly brief, but I had the chance to use a (beta) GH5 with some firmware that had a much more sensible version number today. (Apparently the firmware arrived yesterday.) Here's what I have to add to my previous comments:

(1) It has (finally) gained a menu option for AF control.

Speed is as you'd think. Sensitivity is how much it tries to hang on to the current subject, or not.

(2) The video tracking in Locked-on sensitivity is impressive, even if someone walks in front of the camera.

(3) The video AF at short range is probably worse than before (which I hated). Do note this was indoors and so not that well lit. Changing from 5m to 1m it took 4 seconds to achieve focus, passing through correct focus and overshooting twice before getting it at the third attempt. This was repeatable. Anything under 3-4m wasn't good.

Well, thats only very very small element. Better watch this Video first!

https://www.youtube.com/watch?v=_jfGrc0yYYw

B

Dr_Jon wrote:

Ken Ross wrote:

Isn't there supposed to be a setting for 'moving object prediction' as I've seen in pictured menus? The hierarchy I've seen is 1. AF Sensitivity 2. AF Area Switching Sensitivity 3. Moving Object Prediction.

The menu you show looks very different and doesn't have the flexibility of what I've seen.

Didn't see that, just the ones in my first post, but the menus go on forever and I didn't photograph them again this time. The menus have quite a lot of changes from the ones of the betas, plus some are really poorly organised or have non-intuitive names.

Yeah, you missed the more significant adjustments. You apparently in a different part of the menu.

Vesku wrote:

Dr_Jon wrote:

I'll keep this fairly brief, but I had the chance to use a (beta) GH5 with some firmware that had a much more sensible version number today. (Apparently the firmware arrived yesterday.)

Does the current FW show exposure info in video modes? It would be nice to see aperture, shutter and iso reading during video shooting.

I'm sorry, I really don't remember.

Timmi wrote:

Dr_Jon wrote:

I'll keep this fairly brief, but I had the chance to use a (beta) GH5 with some firmware that had a much more sensible version number today. (Apparently the firmware arrived yesterday.) Here's what I have to add to my previous comments:

(1) It has (finally) gained a menu option for AF control.

Speed is as you'd think. Sensitivity is how much it tries to hang on to the current subject, or not.

(2) The video tracking in Locked-on sensitivity is impressive, even if someone walks in front of the camera.

(3) The video AF at short range is probably worse than before (which I hated). Do note this was indoors and so not that well lit. Changing from 5m to 1m it took 4 seconds to achieve focus, passing through correct focus and overshooting twice before getting it at the third attempt. This was repeatable. Anything under 3-4m wasn't good.

Well, thats only very very small element. Better watch this Video first!

https://www.youtube.com/watch?v=_jfGrc0yYYw

B

There are no video-AF settings in that video I didn't mention, and I was only talking about video. It is a great video though and thanks for linking it.

I will say for locking onto a subject 90% is the sensitivity setting and you can get away with a lot on the speed setting in my experience.

Oh and the whole SET and ON stuff is such a Japanese Engineer way of doing things... alas... Plus it's still crap at short range

Dr_Jon wrote:

Timmi wrote:

Dr_Jon wrote:

I'll keep this fairly brief, but I had the chance to use a (beta) GH5 with some firmware that had a much more sensible version number today. (Apparently the firmware arrived yesterday.) Here's what I have to add to my previous comments:

(1) It has (finally) gained a menu option for AF control.

Speed is as you'd think. Sensitivity is how much it tries to hang on to the current subject, or not.

(2) The video tracking in Locked-on sensitivity is impressive, even if someone walks in front of the camera.

(3) The video AF at short range is probably worse than before (which I hated). Do note this was indoors and so not that well lit. Changing from 5m to 1m it took 4 seconds to achieve focus, passing through correct focus and overshooting twice before getting it at the third attempt. This was repeatable. Anything under 3-4m wasn't good.

Well, thats only very very small element. Better watch this Video first!

https://www.youtube.com/watch?v=_jfGrc0yYYw

B

There are no video-AF settings in that video I didn't mention, and I was only talking about video. It is a great video though and thanks for linking it.

I will say for locking onto a subject 90% is the sensitivity setting and you can get away with a lot on the speed setting in my experience.

Oh and the whole SET and ON stuff is such a Japanese Engineer way of doing things... alas... Plus it's still crap at short range

I think this is what you missed. Although I'm only posting 2 here, there are 4 shown:

Ken Ross wrote:

Dr_Jon wrote:

Timmi wrote:

Dr_Jon wrote:

I'll keep this fairly brief, but I had the chance to use a (beta) GH5 with some firmware that had a much more sensible version number today. (Apparently the firmware arrived yesterday.) Here's what I have to add to my previous comments:

(1) It has (finally) gained a menu option for AF control.

Speed is as you'd think. Sensitivity is how much it tries to hang on to the current subject, or not.

(2) The video tracking in Locked-on sensitivity is impressive, even if someone walks in front of the camera.

(3) The video AF at short range is probably worse than before (which I hated). Do note this was indoors and so not that well lit. Changing from 5m to 1m it took 4 seconds to achieve focus, passing through correct focus and overshooting twice before getting it at the third attempt. This was repeatable. Anything under 3-4m wasn't good.

Well, thats only very very small element. Better watch this Video first!

https://www.youtube.com/watch?v=_jfGrc0yYYw

B

There are no video-AF settings in that video I didn't mention, and I was only talking about video. It is a great video though and thanks for linking it.

I will say for locking onto a subject 90% is the sensitivity setting and you can get away with a lot on the speed setting in my experience.

Oh and the whole SET and ON stuff is such a Japanese Engineer way of doing things... alas... Plus it's still crap at short range

I think this is what you missed. Although I'm only posting 2 here, there are 4 shown:

I believe they are just for stills.

Regards

Pete Berry wrote:

Dr_Jon wrote:

Mikej12 wrote:

Thanks for the update Dr Jon.

Disappointed that the stills do not provide more detail than the GH4, I was hoping for an improvement in stills photography.

Thanks again,

Mike J

I feel it should be better, but I used the same lens on both cameras and the difference was small. Now obviously 20MP is a small step up from 16 and the lack of an AA filter will add at most a 10% improvement so maybe it's there if you look close enough. I just shot the same scene on both cameras at a bunch of apertures and compared the images. I'm sure more scientific results will follow and there might be some aspect of the scene that wasn't ideal, but I couldn't say what it might be.

The true linear res. gain from a 1.25X pixel increase alone is near 12% (sq. root of 1.25); and with the absent A-A filter, maybe 15-20% total expected. What lens, Jon?

12-35mm f2.8 mk I
Oh and I really think a missing AA filter is more 5-10%, e.g. https://www.lensrentals.com/blog/2015/06/canon-5ds-and-5ds-r-initial-resolution-tests/

Regards

Dr_Jon wrote:

Ken Ross wrote:

Dr_Jon wrote:

Timmi wrote:

Dr_Jon wrote:

I'll keep this fairly brief, but I had the chance to use a (beta) GH5 with some firmware that had a much more sensible version number today. (Apparently the firmware arrived yesterday.) Here's what I have to add to my previous comments:

(1) It has (finally) gained a menu option for AF control.

Speed is as you'd think. Sensitivity is how much it tries to hang on to the current subject, or not.

(2) The video tracking in Locked-on sensitivity is impressive, even if someone walks in front of the camera.

(3) The video AF at short range is probably worse than before (which I hated). Do note this was indoors and so not that well lit. Changing from 5m to 1m it took 4 seconds to achieve focus, passing through correct focus and overshooting twice before getting it at the third attempt. This was repeatable. Anything under 3-4m wasn't good.

Well, thats only very very small element. Better watch this Video first!

https://www.youtube.com/watch?v=_jfGrc0yYYw

B

There are no video-AF settings in that video I didn't mention, and I was only talking about video. It is a great video though and thanks for linking it.

I will say for locking onto a subject 90% is the sensitivity setting and you can get away with a lot on the speed setting in my experience.

Oh and the whole SET and ON stuff is such a Japanese Engineer way of doing things... alas... Plus it's still crap at short range

I think this is what you missed. Although I'm only posting 2 here, there are 4 shown:

I believe they are just for stills.

Regards

I was under the impression that was for video & stills, but I'm not 100%.

Pete Berry wrote:

Dr_Jon wrote:

Pete Berry wrote:

Dr_Jon wrote:

Mikej12 wrote:

Thanks for the update Dr Jon.

Disappointed that the stills do not provide more detail than the GH4, I was hoping for an improvement in stills photography.

Thanks again,

Mike J

I feel it should be better, but I used the same lens on both cameras and the difference was small. Now obviously 20MP is a small step up from 16 and the lack of an AA filter will add at most a 10% improvement so maybe it's there if you look close enough. I just shot the same scene on both cameras at a bunch of apertures and compared the images. I'm sure more scientific results will follow and there might be some aspect of the scene that wasn't ideal, but I couldn't say what it might be.

The true linear res. gain from a 1.25X pixel increase alone is near 12% (sq. root of 1.25); and with the absent A-A filter, maybe 15-20% total expected. What lens, Jon?

12-35mm f2.8 mk I
Oh and I really think a missing AA filter is more 5-10%, e.g. https://www.lensrentals.com/blog/2015/06/canon-5ds-and-5ds-r-initial-resolution-tests/

Regards

Well, the link mostly supports my point, with the sharpest of three lenses - 300mm/2.8 on a 50MP Canon 50DS compared to it's A-A filterless sister 50DS-R: "I was a bit surprised at the difference between the 5DS and the 5DS R in this test. The 5DS R version provided about a 10% increase in resolution over the 5DS." And this was in absolute optimum optical test bed conditions.

Minor confession, I may have mis-read your post. I'd have expected about a 10-15% total improvement and it seemed less. However more sunlight could change that.

Maybe you were thinking of the 22MP 5DIII he compared to the 50DS - about a 30% diff. w/ same lens?

OTOH, I will be very unhappy if my soon-to-arrive GH5 doesn't exceed my GH4 in relatively controlled shots using the elect. shutters, particularly when the GH5's 20MP are carefully downsampled to the GH4's 16MP.

Down-sampling will always lose some sharpness and downsizing a small amount can be worst-case. One of the reasons I haven't ordered a GH5 yet is I'm worried scaling the whole sensor to 4k will give aliasing/moire issues the GH4 doesn't have and any extra sharpness will get lost in the down-sides. (Like A7rII users find Super-35 4k, with no scaling, is sharper than full-frame 4k.) I'd really like to buy one, but would also really like to see it attempt my moire/aliasing test first. (BTW my only other main GH5 concern is I still find the video AF very poor at shorter ranges, although the IBIS may cancel that out, it's not like I'm unaccustomed to manual focus. I think I can live with no Dual-IS2 on the f2.8 zooms, as it's not like I currently have something better.)

Pete