GH5 VS URSA MINI 4.6K Downloadable RAW and Vlog

Thanks for posting the footage samples. Here is a quick grade of both in Resolve. I wanted to see how much DR could be recovered from the two shots and how they held up to pushing color. A couple of observations:
1. The URSA sample was overexposed by about 2 stops so the sky is clipped. Skies are clipped in both examples, but the URSA raw could hold the full range with correct exposure.
2. The GH5 has an OLPF, the Ursa does not. Personally I prefer a well matched OLPF, but the GH5 codec does not hold up well to sharpening.
3. It doesn't show up so much after web compression, but I could pull a much broader more subtle range of shading and colors from the URSA as one would expect with DNG raw footage. Still like my Bolex better for color gamut though.
4. The Ursa shading allowed more HDR like depth and dimension to the footage. Tree trunks looked round and there was a more 3D sense of depth perspective. The GH5 seemed flat by comparison.
5. I used color managed workflows for both files in Resolve for color space conversion. I used BMD4.6k film 3 to REC709 for the URSA and Vlog to REC709 for the GH5. Both worked well as a starting point for the grades.

 
Last edited:
This is pretty much the nightmare scenario for cameras. If you expose for the highlights the forest looks like a semi night scene. If you expose for the forest the highlights look like they are on fire. Looking at the original video I would say the GH5 has holding up better in terms of dynamic range. The URSA was clipping the highlights a bit more. Overall this shows how far we have to go with digital dynamic range and cameras. The grade by Razz shows some improvement but both had blue highlights which means the program is basically painting in to make up for the sensor deficits.

I would like to see the new Panasonic and Canon cameras in the same setting to see how the 15 stop DR claims hold up. But in the end, the $2,000 GH5 is a great value.
 
Bassman2003 said:
This is pretty much the nightmare scenario for cameras. If you expose for the highlights the forest looks like a semi night scene. If you expose for the forest the highlights look like they are on fire. Looking at the original video I would say the GH5 has holding up better in terms of dynamic range. The URSA was clipping the highlights a bit more. Overall this shows how far we have to go with digital dynamic range and cameras. The grade by Razz shows some improvement but both had blue highlights which means the program is basically painting in to make up for the sensor deficits.

I would like to see the new Panasonic and Canon cameras in the same setting to see how the 15 stop DR claims hold up. But in the end, the $2,000 GH5 is a great value.
Click to expand...

The Ursa has the DR to handle this beautifully without clipping. I had to back the exposure down 3 stops in Resolve to get into a normal value range for the raw footage. Way overexposed. The GH5 was much closer to optimum exposure, so highlights were not quite as burned. But there is much less room to manipulate the image without breaking something.
 
Razz16mm said:
The Ursa has the DR to handle this beautifully without clipping. I had to back the exposure down 3 stops in Resolve to get into a normal value range for the raw footage. Way overexposed. The GH5 was much closer to optimum exposure, so highlights were not quite as burned. But there is much less room to manipulate the image without breaking something.
Click to expand...

NEWS FLASH: "Camera at less than half the cost that records h264 has less usable dynamic range than much more expensive camera that records raw."
 
joe12south said:
NEWS FLASH: "Camera at less than half the cost that records h264 has less usable dynamic range than much more expensive camera that records raw."
Click to expand...

True, but the GH5 still looks very good if not as refined. You just have to nail your exposure and settings when you shoot to get the best results. It stood up to the grading process better than I thought it would.
 
whoa, the GH5 footage looks crappy.
Like some crazy sharpening and even worse watercolor effect going on while panning...

Used sharpening an nr at 0 or worse?
-5/-5 definitely isn't that bad and with some filtering in post and a Diffusion filter applied in front of the lens even quite usable - though still not Blackmagic Ursa 4.6 or even Mini like...
 
Razz16mm said:
True, but the GH5 still looks very good if not as refined. You just have to nail your exposure and settings when you shoot to get the best results. It stood up to the grading process better than I thought it would.
Click to expand...
I agree, the GH5 (V-log L, 10 bit) grades fine.
The BM was overexposed.
 
You must log in or register to reply here.
Back
Top