Cosimo Bullo
Veteran
It's amazingly good, I'd say. At 200 mbps it's more than an equivilent of a 4k crop, especially at 60p.
GH4 i.Dynamic
- Thread starter bruceallen
- Start date
It's amazingly good, I'd say. At 200 mbps it's more than an equivilent of a 4k crop, especially at 60p.
Camerapilot
Active member
What model of speed booster are you using?
Wow thats very interesting! So its way better than the regular 1080p i presume? Please post some comparisons
hfr 96fps, regular 1080, ex tele, 4k.
I'm using the MFT-to-F Mount model.
You can hope, but when was the last time Panasonic made any real improvements to a camera via a firmware update?
Has anyone made any more tests on the GH4's Dynamic Range? No one has done tests that show anything better than eight or so stops. Im considering moving from the GH3 to the GH4 but probably wont bother if there is no improvement on DR.
bruceallen
Well-known member
Sadly no. Macgregor and my tests vs the F35 were done with stock Cine-D settings.
I do believe that if you use -5 Contrast, -5 Sharpening and at least i.Dynamic Std, you could gain another stop or so.
But I haven't tested it due to being busy at work. Anyone in the Los Angeles area is welcome to borrow a GH4 if they want to do a scientific test!
Throw in Neat Video and you gain another stop (but so does the F35).
Bruce Allen
www.boacinema.com
Cosimo Bullo
Veteran
We're getting 200mbps without a hack. No, that's not the problem. This issues with 1080 on the GH4 are hardware based, I think, pixel binning/selecting, etc.
snowblindfilm
Carbonite Member
We will not see real capabilities of this camera, until recorded in 10bit to an external recorder. The difference between internal 8bit 420 and external 10bit 422 should be night and day in the post. From what I have seen, this camera equals C300 in IQ, at a fraction of the cost. Add 2.5K for Shogun and SSD and it will be remarkable combination.
Now, if anyone can tell me if you can record 96fps to an external recorder, or is it internal recording option only?
Now, if anyone can tell me if you can record 96fps to an external recorder, or is it internal recording option only?
I picked up my GH4 at UPS yesterday and transported it to a set where I unboxed it, popped in my GH3 battery, fiddled with the menu settings and began shooting another director's short film in 4k.
I have not done a scientific DR test, but based on yesterday's shooting, I can say the DR is improved. I have been a heavy GH user since the GH1, so I instinctively know when highlights are due to blow out on those cameras. Shooting in Cinelike-D yesterday at 800 ISO, my feeling is that there's another 2/3 to 1 full stop more dynamic range than on the GH3. That's significant.
I used my GH4 on a location scout today and discovered something about iDynamic by mistake.
Somehow, I had iDynamic set to HIGH when I thought it was OFF. I wondered why I could see so much in tha shadows in a dark warehouse at 200 ISO! When I got home and viewed the footage on my iMac, I noticed the 200 ISO footage looked like it was shot at 3200 or 6400! Just horrible colored noise in the shadows and midtones. Then I realized I had iDynamic set to HIGH.
So I repeated the test in a similar environment at 200 ISO and 400 ISO. With iDynamic set to OFF and iDynamic set to HIGH. At both ISOs, with iDynamic OFF, shadows are dark but noise free. With iDynamic on HIGH, the shadows and midtones have horrible, horrible colored noise.
I did some color correction on the clips with iDynamic OFF and found that if I lifted the exposure to the level of the iDynamic HIGH clips, the noise looked cleaner and less saturated.
Also, other clips that had a more narrow range of brightness that I shot while accidentally on HIGH didn't have as much noise, which leads me to believe iDynamic adjusts itself as necessary based on the contrast in the scene. This makes it unusable for me doing narrative work, where I need to keep the key-to-fill ratio consistent.
So, I am not convinced iDynamic is a good thing. At the very least, don't set it on HIGH.
Somehow, I had iDynamic set to HIGH when I thought it was OFF. I wondered why I could see so much in tha shadows in a dark warehouse at 200 ISO! When I got home and viewed the footage on my iMac, I noticed the 200 ISO footage looked like it was shot at 3200 or 6400! Just horrible colored noise in the shadows and midtones. Then I realized I had iDynamic set to HIGH.
So I repeated the test in a similar environment at 200 ISO and 400 ISO. With iDynamic set to OFF and iDynamic set to HIGH. At both ISOs, with iDynamic OFF, shadows are dark but noise free. With iDynamic on HIGH, the shadows and midtones have horrible, horrible colored noise.
I did some color correction on the clips with iDynamic OFF and found that if I lifted the exposure to the level of the iDynamic HIGH clips, the noise looked cleaner and less saturated.
Also, other clips that had a more narrow range of brightness that I shot while accidentally on HIGH didn't have as much noise, which leads me to believe iDynamic adjusts itself as necessary based on the contrast in the scene. This makes it unusable for me doing narrative work, where I need to keep the key-to-fill ratio consistent.
So, I am not convinced iDynamic is a good thing. At the very least, don't set it on HIGH.
bruceallen
Well-known member
Interesting! Question is: if you did Shadows +5, would you have gotten the same noise as with i.Dynamic?
I definitely do think that the contrast settings available in camera exceed the sensor's actual ability.
If you do something that brings up the shadows a lot, you might not find a ton of usable data. Just a bunch of noise... and then you're wasting bits on noise if you plan to darken that stuff down in post anyway. On the other hand, if you plan to use Neat Video etc they might be useful?
More tests would be awesoe. And agree it's always better to expose correctly... as usual
Bruce Allen
www.boacinema.com
I definitely do think that the contrast settings available in camera exceed the sensor's actual ability.
If you do something that brings up the shadows a lot, you might not find a ton of usable data. Just a bunch of noise... and then you're wasting bits on noise if you plan to darken that stuff down in post anyway. On the other hand, if you plan to use Neat Video etc they might be useful?
More tests would be awesoe. And agree it's always better to expose correctly... as usual
Bruce Allen
www.boacinema.com
Bruce, I tend to like my video pretty contrasty. I don't often run and gun in low light, but am more often in situations where I have some lighting or bounce material at my disposal. So lots of details in the shadows isn't something I obsess over. I tend to be more concerned about blown highlights.
But what concerns me about iDynamic is that it seems to change its strength based on the dynamics in the scene. So, as the scene gets more contrasty, iDynamic lifts the shadows more aggressively to compensate. That means it could change in the middle of a shot, and I just don't like that.
My custom Shadow/Highlight setting on the Gh4 is Shadows- 0 and Highlights minus 4. In other words, I leave Shadows alone. That way, if I'm at 400 ISO, say, I know I'm getting noise levels in the shadows appropriate to that ISO.
I'm shooting with Cinelike D currently -3,-5,-4,-2, 0.
But after this current project, I may go back to using Natural like I did with the GH3. Or I may explore Portrait. My rudimentary tests showed that Cine-D doesn't offer much more dynamic range than the other profiles on the camera.
Problem is, I haven't had enough time to test the GH4. I just jumped right in to shooting a short film with it, and I'm discovering things as I go!
But what concerns me about iDynamic is that it seems to change its strength based on the dynamics in the scene. So, as the scene gets more contrasty, iDynamic lifts the shadows more aggressively to compensate. That means it could change in the middle of a shot, and I just don't like that.
My custom Shadow/Highlight setting on the Gh4 is Shadows- 0 and Highlights minus 4. In other words, I leave Shadows alone. That way, if I'm at 400 ISO, say, I know I'm getting noise levels in the shadows appropriate to that ISO.
I'm shooting with Cinelike D currently -3,-5,-4,-2, 0.
But after this current project, I may go back to using Natural like I did with the GH3. Or I may explore Portrait. My rudimentary tests showed that Cine-D doesn't offer much more dynamic range than the other profiles on the camera.
Problem is, I haven't had enough time to test the GH4. I just jumped right in to shooting a short film with it, and I'm discovering things as I go!
bruceallen
Well-known member
Thanks for sharing your settings, dvbrother! I'll have to give them a try!
I agree that if i.Dynamic adjusts per scene then it is a major issue. If anyone can find proof of it, I will amend my thread's first post to avoid leading others astray
I have been trying to trick my cam into getting i.Dynamic to misbehave but so far no luck. I am on Low most of the time though.
If anyone can record it being naughty, that would be great, of course.
Yes, I 100% agree that Cinelike-D doesn't offer a huge dynamic range increase over the other modes. Furthermore, once you start playing with contrast, highlights / shadows, etc... most modes can be so low contrast that they bump into the noise floor.
I was just playing with Natural today... it appears to be pretty much Portrait with less saturation, which is awesome as a big problem I was having was that I couldn't turn saturation down enough.
Comparing Natural and Cine-D, I think that Natural even allows for less saturation than Cine-D. And Cine-D has really weird color bands in the highlights for me (or at least, it saturates the highlights to the point where the distortions in the color science become more obvious).
Bruce Allen
www.boacinema.com
I agree that if i.Dynamic adjusts per scene then it is a major issue. If anyone can find proof of it, I will amend my thread's first post to avoid leading others astray
I have been trying to trick my cam into getting i.Dynamic to misbehave but so far no luck. I am on Low most of the time though.
If anyone can record it being naughty, that would be great, of course.
Yes, I 100% agree that Cinelike-D doesn't offer a huge dynamic range increase over the other modes. Furthermore, once you start playing with contrast, highlights / shadows, etc... most modes can be so low contrast that they bump into the noise floor.
I was just playing with Natural today... it appears to be pretty much Portrait with less saturation, which is awesome as a big problem I was having was that I couldn't turn saturation down enough.
Comparing Natural and Cine-D, I think that Natural even allows for less saturation than Cine-D. And Cine-D has really weird color bands in the highlights for me (or at least, it saturates the highlights to the point where the distortions in the color science become more obvious).
Bruce Allen
www.boacinema.com
Thanks for sharing your settings, dvbrother! I'll have to give them a try!
I agree that if i.Dynamic adjusts per scene then it is a major issue. If anyone can find proof of it, I will amend my thread's first post to avoid leading others astray
I have been trying to trick my cam into getting i.Dynamic to misbehave but so far no luck. I am on Low most of the time though.
If anyone can record it being naughty, that would be great, of course.
Yes, I 100% agree that Cinelike-D doesn't offer a huge dynamic range increase over the other modes. Furthermore, once you start playing with contrast, highlights / shadows, etc... most modes can be so low contrast that they bump into the noise floor.
I was just playing with Natural today... it appears to be pretty much Portrait with less saturation, which is awesome as a big problem I was having was that I couldn't turn saturation down enough.
Comparing Natural and Cine-D, I think that Natural even allows for less saturation than Cine-D. And Cine-D has really weird color bands in the highlights for me (or at least, it saturates the highlights to the point where the distortions in the color science become more obvious).
Bruce Allen
www.boacinema.com
Yeah, I kind of settled on standard for my outdoor shots. Works pretty well. Really don't make any adjustments except for knocking down highlights just a bit on the curve. (-3, I think)
In regards to dynamic range and boosting shadows -- there is no free lunch. You could lift shadows a number of ways, but at some point it'll cause noise... just have to be aware of your situation. I have found, though, that you can boost shadows pretty effectively in post with 4k... so that may be the best option at times.
bruceallen
Well-known member
Interesting! Are you really finding anything extra in the highlights? Every time I try to get info from there, I am disappointed but maybe I need to test more.
In regards to dynamic range and boosting shadows -- there is no free lunch. You could lift shadows a number of ways, but at some point it'll cause noise... just have to be aware of your situation. I have found, though, that you can boost shadows pretty effectively in post with 4k... so that may be the best option at times.[/QUOTE]
BTW, I just read the DPReview pages on i.Dynamic and Scene modes... it does appear that Natural does give a flatter image. It also appears that turning i.Dynamic up makes the tonal response more linear(!).
Maybe Natural with -5 Contrast is enough. Anyway, I'm having fun! I'm ordering some more Stouffer wedges (still can't find my old one) and will test.
GH3:
http://www.dpreview.com/reviews/panasonic-lumix-dmc-gh3/15
GH2:
http://www.dpreview.com/reviews/PanasonicDMCGH2/12
Bruce Allen
www.boacinema.com
Bruce, regarding those Stouffer wedges, I found a page for them (http://www.stouffer.net), but which model is the right one to get for film/video application? Also, if I'm understanding it correctly, they only seem to cover about 6 or 7 stops depending on model, so to test the full dynamic range of the camera do you just take multiple shots at different exposures?
bruceallen
Well-known member
Shenan, the T4110 is the one you want as it has the most dynamic range. Well, get the C version if you want them to actually calibrate and check it.
You can also get the DSC Labs Xyla but that is thousands of dollars.
The T4110 has a fine camera test heritage, eg Barry Green has been known to use it:
http://www.dvxuser.com/V6/showthread.php?187624-GH1-meet-the-Stouffer-4110C-dynamic-range-test
And RED has been tested on it here:
http://bealecorner.org/red/test-080108/page2.html
Heck, I even used it to test a HV20 back in the day.
The main issue with the Stouffer is that there is a little bit of cross-contamination of light which lessens DR. You want to have no light other than whatever is backlighting the Stouffer in the room as well.
Only with the very best cameras (F35, Alexa, Dragon) would the dynamic range of the Stouffer be too little (in which case, yes, you need a chart with more DR).
I have actually ordered 8 Stouffer charts so that I can put various gels in front of them to see what is happening with color across the dynamic range spectrum in one shot easily.
Bruce Allen
www.boacinema.com
Awesome, thanks Bruce! I was looking at the reflective ones, which only seem to go about 6 to 8 stops. 13 and 2/3 stops then.
Barry's post is also informative, but I still have a couple of questions:
What is the standard procedure for lighting this thing, aside from what you already mentioned with having the backlight source only? Put it against a lightbox, softbox, diffusion, or something else? I would imagine you would want the most even lighting and the most accurate white balance possible, and also probably mask everything except for the chart itself so you can see only the steps on the waveform.
Also, Barry mentions using a waveform monitor. I guess I would need a field monitor with a waveform monitor, or is there some other way around that? 100% zebras on the clearest box? Histogram somehow? I hope I'm not derailing this thread too much with my questions. If you have any links to articles on how to do all this, I would appreciate it.
Thanks!