[Graham Jones]

Hmmm…

I decided to do a very quick down and dirty test — low lighting in my office to exaggerate the grain, tried capturing 1080p30, 1080p24, 1080i60 by HDMI (ProRes HQ), and recording at the same time to the camera at 24 Mbit AVCHD to import with Log and Transfer.

Notes:
– The camera appears to Final Cut as 1080i60 regardless of which internal mode the camera is set at, whether using HDMI or Log and Transfer. I notice the 24p-style motion when recorded in that mode, but not true progressive frames. Bummer. This may not be a FCP issue — when I check the HDMI Status in the camera menus, it says 1080i60 and has no other options.

– Though the HDMI-captured image does have significantly more colour and more detail in moving areas, it also has significantly more noise. This doesn’t make sense to me — unless it is an issue of the HDMI more accurately capturing the color noise of the grain.

– Because of this, in low light the non-HDMI actually looks better.

– That said, in my broadcast color grading experience I know that even very high end pro cameras look terrible when there is gain noise in low light. I am eager to see how these tests might differ in full studio light. If I don’t have to deal with the grain, I think the HDMI improvements will be worth it.

My light kit will ship in another week or two — I’ll report back when I have more tests to share.

Thanks,
G.

[Graham Jones]

Hmmm…

I decided to do a very quick down and dirty test — low lighting in my office to exaggerate the grain, tried capturing 1080p30, 1080p24, 1080i60 by HDMI (ProRes HQ), and recording at the same time to the camera at 24 Mbit AVCHD to import with Log and Transfer.

Notes:
– The camera appears to Final Cut as 1080i60 regardless of which internal mode the camera is set at, whether using HDMI or Log and Transfer. I notice the 24p-style motion when recorded in that mode, but not true progressive frames. Bummer. This may not be a FCP issue — when I check the HDMI Status in the camera menus, it says 1080i60 and has no other options.

– Though the HDMI-captured image does have significantly more colour and more detail in moving areas, it also has significantly more noise. This doesn’t make sense to me — unless it is an issue of the HDMI more accurately capturing the color noise of the grain.

– Because of this, in low light the non-HDMI actually looks better.

– That said, in my broadcast color grading experience I know that even very high end pro cameras look terrible when there is gain noise in low light. I am eager to see how these tests might differ in full studio light. If I don’t have to deal with the grain, I think the HDMI improvements will be worth it.

My light kit will ship in another week or two — I’ll report back when I have more tests to share.

Thanks,
G.

[Graham Jones]

Hi,

unfortunately I only have CS3, which didn’t come with a Mac version of OnLocation.

Do you know of any better Mac-based Decklink-compatible capture tools, that would be better than FCP?

Thanks,
Graham.

[Graham Jones]

Hi Joshua,

thanks for your response. I was thinking I would record to the camera’s hard drive as a backup anyway, so probably wouldn’t have to worry about the demo mode.

Since I have the Decklink Studio and camera already, I’ll run some tests to see if there are any issues.

Thanks,
Graham.

[Graham Jones]

Oops, just realized that my Decklink Studio maxes out at 4:2:2 — you need the Extreme for 4:4:4. So — I guess the only question is HQ vs non-HQ.

I guess I’ll just have to experiment to see whether non-HQ will do what I need.

Thanks,
G.

[Graham Jones]

Hi Dave,

thanks for your response. I forgot to mention that because I do a lot of HD broadcast editing and color grading, I already have massive RAIDs at home — two RAID arrays totaling 16 TB, on two separate quad Mac Pro machines. So storage isn’t an issue. 😉

Both Prores HQ and Prores are 10-bit 4:2:2 — they only differ on the data rate. You’re right in suggesting I should experiment with non-HQ, as what I’m really after is the 4:2:2, and the higher data rate of HQ may be just overkill.

I may be wrong, but I think the HDMI only carries 4:2:2 — so I think Prores 4444 wouldn’t be worth it. Plus, for my purposes, the 4:2:2 will do fine.

Yeah, I have the lighting under control. 4 years of film school taught me enough of that. 😉

Thanks,
G.

[Graham Jones]

Hi,

well, I think part of my problem was just needing a restart. When I came in this morning the AJA control panel worked perfectly, and finally the “Crop” setting worked as intended.

Then I followed your advice and used the LH DVCProHD 1080i/29.97 Easy Setup — and it works like a charm! It even lets me continue to render in ProRes HQ, and downconvert beautifully to SD 4:3.

Many thanks for your help — it was in the nick of time.

– Graham.

[Graham Jones]

Hi,

I was referring to the Sequence Settings window where I chose the “Load Sequence Settings” button on the lower left, and indicated the Kona LH’s preset for DVCProHD 1080i60 — so yes, sequence preset.

This is from memory — I am at home and do not have the editing computer in front of me.

Thanks,
Graham.

[Graham Jones]

An important distinction I forgot to mention — because of luma shift issues upon render, we were forced to render our 1080i60 sequence as ProRes HQ (there was a noticeable and reproducible luma boost upon render when keeping the sequence settings at DVCProHD 1080i60).

Most of my 4:3 downconvert testing was done with ProRes-rendered sequence. However, I did do a test with DVCProHD 1080i60 as the sequence preset and it did not seem to fix the downconverting issue.

Just thought this was worth mentioning.

Thanks,
Graham.

[Graham Jones]

Hi,

sorry if I was not more clear… I did try exactly what you suggest, and the KonaLH control panel was not responding to my indication to “Crop” — whether FCP was open or not. It would always letterbox no matter what the Crop/Letterbox popup said. Also, from the HD sequence the downconversion to letterboxed SD was jittery.

I do not have the machine in front of me right now… at some point tomorrow I will go over your FAQ item, comparing with my test sequence and KonaLH control panel, to see if there’s anything I missed.

Thanks,
Graham.