I'm also wondering if any HF-60 (or XA50/55) owners have updated to the latest firmware and checked out the new 'Custom Picture' feature, which includes Canon Log3 gamma. I'm particularly interested to know about noise levels when shooting in CLog3; do these camcorders really have the 'dynamic range' (exposure latitude) for it ?
The XF400/405 also get this feature with the latest firmware update. The HF-G50 gets a limited 'CP' feature set i.e. options to customize the 'Standard' (BT.709) picture profile, but does not get CLog3 and the EOS Std and Production Camera color matrices.
Clearly in adding this 'pro' feature (inherited from the Canon EOS cinema cameras and earlier XF300/305 models) Canon hope to extend the appeal to independant film/documentary makers who want more than out-of-the box 'looks' and greater latitude for stylized grading in-post.
Checking out the updated on-line user manuals though, I'm disappointed to find that Canon haven't seen fit to incorporate an in-camera 'videoscope' (waveform/parade), on any of the above models, to support this feature. They include a 'View Assistance' option when shooting in Clog3 that displays on the screen an 'approximation' of what the image would look like with Wide DR (BT.709) settings, but for tuning of custom picture settings, you really need scopes. Presumably they expect that people will use external monitors with videoscopes, or review test images in editing/grading software when creating custom picture profiles.
It's also of note that the GX10 doesn't get any firmware update, which I guess signals that Canon are phasing it out.
I've been eyeing the HF-G60 as a worthy upgrade to my HF-G40 for some time now. I nearly sprang for a GX10 when B&H were offering refurbished units for less than the HF-G60 earlier in the year. Glad I didn't now. This new 'Custom Picture' feature definitely sweetens the pot but I'd need to be convinced it is a viable option for shooting log.
Also worth mentioning that firmware brings improved auto-focus when zooming, on all models. Again, it would be nice to see that demonstrated in controlled tests. .
