GH4 Panasonic GH5 coming in 2017

Getting conflicting info on whether the speedbooster XL works or not with full frame EF lenses. It seems though that the speedbooster ultra(non-XL) works.

Also, i have not seen mentions of dynamic range(but then again a lot of these pre-orders dont seem to have V-log). Any info?

From the DPreview stuff i saw, footage reminds me a lot of the FS7 but with more noise when bumped to 3200iso+(still amazing for a m43 sensor!)
 
The stuff from DPReview look good. I feel like the slow motion stuff (60p conformed to 24p) looked a bit odd, perhaps it was a weird shutter setting as it looked stuttery. As a wedding videographer, I believe I could easily work with this camera in terms of the noise at 6400 ISO. I'm excited to take my first step in the GH arena with the GH5, wish I didn't have to wait to start testing with it to see if it's a good fit for me.
 
ISO 6400 looks years better than GH4 ISO 3200 and maybe even better than ISO 1600 on the GH4? ISO 3200 sort of has an ISO 800-1200 look to my eye in comparison to GH4, which is great, and ISO 1600 looks sharp and clean.

All in all, great news for live event and documentary work.

Re: the slow motion, they appear to have done it wrong and just used speed control in Premiere to slow the 60fps version, rather than using VFR mode, or conforming footage.
 
Some horrible horrible denoising...

http://www.imaging-resource.com/news/2017/02/14/panasonic-gh5-beta-first-shots

GH5 vs A6500 @iso6400
 
Last edited:
brudney said:
Some horrible horrible denoising...

http://www.imaging-resource.com/news/2017/02/14/panasonic-gh5-beta-first-shots
Click to expand...

Yeah :( Checking the higher ISO's it does appear there is some very strong noise reduction - however this is stills.

Will await some proper side by sides from GH4/GH5 straight from the camera with -5 for NR

Downloading the footage from DPR - the noise is there, but not mashed at all - and even 1600 has a nice look about it - fine grain. 3200 looks useable but 6400 is still a mess still by the looks of it.
 
TheTrickster said:
Yeah :( Checking the higher ISO's it does appear there is some very strong noise reduction - however this is stills.
Will await some proper side by sides from GH4/GH5 straight from the camera with -5 for NR
Click to expand...

They have two sample JPEG images for each ISO level on this page. One is at the default noise reduction level (labeled NR6), the other has less noise reduction (labeled NR1):
http://www.imaging-resource.com/PRODS/panasonic-gh5/panasonic-gh5A7.HTM

There are also RAW files available for download, although I'm not sure if existing RAW processors can interpret them.

I believe they also have the same test shots from the GH4 somewhere on the site as well. Of course the still performance may not translate exactly to video as JPEG and H.264 compression are different, and I'm sure there may be other processing differences. Not to mention this is still all with pre-production firmware.
 
brudney said:
Some horrible horrible denoising...
Click to expand...

The GH5 default NR settings are just set too high.

The GH4 default NR settings are notorious with both stills and video, where details evaporate from your image when the NR is set too high, and I expect this is the case for the GH5 too.
 
philiplipetz said:
look at the same Imagining Resource shots done with NR set lower. They are much better.

http://www.imaging-resource.com/PRODS/panasonic-gh5/GH5hSLI06400NR1.HTM
Click to expand...

I downloaded the raw images to look at what the sensor can really handle. It is the only real way to judge what the sensor itself can do. ISO6400 has noise. Removing the noise with raw processing helps but there is a loss. ISO3200 cleans up a lot although there are still artifacts. ISO1600 looks really good even without NR. Still has visible noise but if delivering HD it mostly goes away. What I'm still very impressed with is how good the color and contrast is at the higher ISO levels. Typically a lot more factor of image quality suffer outside of just noise and on the GH5 if one is ok with the noise every other aspect of ISO1600 and ISO3200 on the GH5 seems usable.

A more sophisticated NR like NeatVideo might have much better results. Haven't tested that yet.
 
Thomas Smet said:
I downloaded the raw images to look at what the sensor can really handle. It is the only real way to judge what the sensor itself can do. ISO6400 has noise. Removing the noise with raw processing helps but there is a loss. ISO3200 cleans up a lot although there are still artifacts. ISO1600 looks really good even without NR. Still has visible noise but if delivering HD it mostly goes away. What I'm still very impressed with is how good the color and contrast is at the higher ISO levels. Typically a lot more factor of image quality suffer outside of just noise and on the GH5 if one is ok with the noise every other aspect of ISO1600 and ISO3200 on the GH5 seems usable.

A more sophisticated NR like NeatVideo might have much better results. Haven't tested that yet.
Click to expand...

RAW results are irrelevant for video GH5 shooters
 
philiplipetz said:
RAW results are irrelevant for video GH5 shooters
Click to expand...

jpeg is even more irrelevant. What the raw images provide however is the ability to see what the sensor is capable of without any other odd outside factors. You can also take the raw samples and convert them to other formats like a video. The jpeg is already damaged way beyond what the h264 encoder in the camera is doing.
 
philiplipetz said:
I disagree. While the encoding is different from video the NR procesing of the RAW signal is the same.
Click to expand...

No it tells you how the noise looks with a freeze frame and not moving video. Very different concepts. Do you also have any proof that the processing is exactly the same between jpeg and H264 video? There is a lot more going on when shooting video vs a still image and the processor is working much harder on other forms of processing.

The reason I look at the raw samples is that they tell you what the sensor is capable of in a controlled environment without any outside influence. I tend to shoot video with a -5NR or no in camera noise reduction at all and the raw samples will be closer to that than the standard settings used in these tests. I would never in a million years use a +1 or higher NR on the GH4. In this case the raw samples are a much better representation to how I shoot video vs jpegs with horrible baked in NR.
 
Thomas Smet said:
No it tells you how the noise looks with a freeze frame and not moving video. Very different concepts. Do you also have any proof that the processing is exactly the same between jpeg and H264 video? There is a lot more going on when shooting video vs a still image and the processor is working much harder on other forms of processing.

The reason I look at the raw samples is that they tell you what the sensor is capable of in a controlled environment without any outside influence. I tend to shoot video with a -5NR or no in camera noise reduction at all and the raw samples will be closer to that than the standard settings used in these tests. I would never in a million years use a +1 or higher NR on the GH4. In this case the raw samples are a much better representation to how I shoot video vs jpegs with horrible baked in NR.
Click to expand...

I agree. Hopefully we will see an option to turn Noisereduction and Sharpening off completely in video, just like raw. Am not impressed by the video samples right now, because they look videoish, thanks to the heavy processing going on... :(
 
Thomas Smet said:
No it tells you how the noise looks with a freeze frame and not moving video. Very different concepts. Do you also have any proof that the processing is exactly the same between jpeg and H264 video? There is a lot more going on when shooting video vs a still image and the processor is working much harder on other forms of processing.

The reason I look at the raw samples is that they tell you what the sensor is capable of in a controlled environment without any outside influence. I tend to shoot video with a -5NR or no in camera noise reduction at all and the raw samples will be closer to that than the standard settings used in these tests. I would never in a million years use a +1 or higher NR on the GH4. In this case the raw samples are a much better representation to how I shoot video vs jpegs with horrible baked in NR.
Click to expand...

Panasonic has stated definitively that the GH5's video noise-reduction is both temporal and spacial, so it is most definitely different than the NR applied to JPEG's.
 
joe12south said:
Panasonic has stated definitively that the GH5's video noise-reduction is both temporal and spacial, so it is most definitely different than the NR applied to JPEG's.
Click to expand...
If it is similar to how the UX90, UX180, and DVX200 work, you should have about 10 levels of adjustability, and each level makes a really big difference. Spatial noise reduction seems to be much more prevalent on the higher settings, and perhaps eliminated entirely on the lowest settings.
 
You must log in or register to reply here.
Back
Top