F5: Sony Raw Viewer vs. Blackmagic Resolve

Admit One

Admit One

Active member
I recently shot a short 4K raw test clip with the F5, and opened it in both Resolve and the Raw Viewer. Here is a tight crop of the results:

cap_1.6.3.jpg

Quite different results, particularly in the right hot edge of the caps.

Anyone else notice similar differences?

Cheers.
 
Did you process the Resolve clip with ACES?

Sony has been saying that Raw viewer is preferable owing to some tweaks Blackmagic needs to do with Resolve. But if you didn't process with ACES, try running that test again and comparing all 3.

Ned Soltz
 
ACES is definitely the way to go with the raw footage.

Interested in seeing more of your tests.

Ned
 
Admit One said:
And I though Resolve was the way to go, until this. Unless they straighten this out, they're out...

Cheers.
Click to expand...

Wait, so the top shot is Resolve?

Been working a lot with ACES/Resolve, haven't seen anything like this EXCEPT with old camera firmware, but that was XAVC material, not raw.

Please forgive, you had the IDT and ODT in Resolve set correctly?

You should contact one of the Resolve team about this, maybe on the CC board. They really care about this stuff, and they will never see it over here on this forum.

[edit: I just noticed you said F5, not 55. I wonder if the more Rec709 targeted color filter array on the F5 sensor is part of why this is happening. I would bet lunch that the ACES IDT for Sony raw is designed for the sensor CFA on the F65 and 55]
 
Hmm... looks like some form of chromatic aberration correction to me. Maybe look into whether the raw viewer is automatically applying chromatic aberration correction, and whether Resolve offers a similar feature?
 
Nate Weaver said:
Please forgive, you had the IDT and ODT in Resolve set correctly?
Click to expand...
Yes.

[edit: I just noticed you said F5, not 55. I wonder if the more Rec709 targeted color filter array on the F5 sensor is part of why this is happening. I would bet lunch that the ACES IDT for Sony raw is designed for the sensor CFA on the F65 and 55]
Click to expand...
Could be, but that seems a streatch. The pixels on the F5 and F55 are all in the same place.

Cheers.
 
Grug said:
Hmm... looks like some form of chromatic aberration correction to me. Maybe look into whether the raw viewer is automatically applying chromatic aberration correction, and whether Resolve offers a similar feature?
Click to expand...
That's what I thought at first. To do that, it would seem the Raw Viewer would need to know the lens characteristics. But it's not just purple fringing. The lights above show multicolored aberrations.

Cheers.
 
Admit One said:
That's what I thought at first. To do that, it would seem the Raw Viewer would need to know the lens characteristics. But it's not just purple fringing. The lights above show multicolored aberrations.

Cheers.
Click to expand...

I see cyan and green fringing in the lights above, purple on the brim of the cap. All three of those tend to correctable (to an extent) with chromatic aberration correction tools (in still photography programs at least) without the need for specific lens characteristics.

I can't think of another style of correction tool that would address those aberrations in the way they appear to have been handled by the raw viewer.
 
Grug said:
I see cyan and green fringing in the lights above, purple on the brim of the cap. All three of those tend to correctable (to an extent) with chromatic aberration correction tools (in still photography programs at least) without the need for specific lens characteristics.

I can't think of another style of correction tool that would address those aberrations in the way they appear to have been handled by the raw viewer.
Click to expand...
Well, it seems that either a different (not so good) de-Bayer algorithm is being used by Resolve or lens correction is being applied by Sony. I'm not sure which.

I've ordered Vegas 12, and will see if it produces similar results. Unfortunately, at the moment, none of my other raw processing programs can decode Sony's raw format.

Cheers.
 
Admit One said:
Well, it seems that either a different (not so good) de-Bayer algorithm is being used by Resolve or lens correction is being applied by Sony. I'm not sure which.

I've ordered Vegas 12, and will see if it produces similar results. Unfortunately, at the moment, none of my other raw processing programs can decode Sony's raw format.

Cheers.
Click to expand...

Well happy hunting! Keep us updated if you find out the explanation.
 
Admit One said:
Well, it seems that either a different (not so good) de-Bayer algorithm is being used by Resolve or lens correction is being applied by Sony. I'm not sure which.

I've ordered Vegas 12, and will see if it produces similar results. Unfortunately, at the moment, none of my other raw processing programs can decode Sony's raw format.

Cheers.
Click to expand...

I wish you would post it on the Resolve forum. They fix things VERY fast as soon as they know about it, and about 4 people from Blackmagic including the lead engineer read the forum daily.
 
Rad. Peter Chamberlain won't sleep on it. I saw it in the BM forum.

I just realized this is a more extreme version of something I saw a couple months ago. Raw only, not on XAVC. It didn't have the color fringes, it was more of a colorless CA-ish bloom with a 1pixel hard edge, so abrupt it was clearly software generated.

I do now think you're right. It's not just F5 raw. 55 too. Sorry I discounted ya.

I wish I knew why Juan Salvo doesn't like ACES though. I get really great results out of it with 55 material. Red, not so much.
 
You must log in or register to reply here.
Back
Top