IPP2 benefit in grading?

[Christoph Tilley]

Hi folks,

just upgraded and tried 6.4 with the new IPP2 today and I really like it.
But now I'm curious: is IPP2 enhancing the way my weapon reads sensor data or is it just a more advanced way of previewing Helium?

In other words: if I keep shooting in RG4/DC2 and use Log3G10/RWG as a starting point for grading my R3Ds in Davinci Resolve, will there be any noticeable difference to IPP2 shot footage?

Thanks for clearing this up for me!
Best,
Chris

 

[Gavin Greenwalt]

IPP2 won't affect the RAW image except in so far as you base any exposure or lighting decisions on the Dragoncolor/REDganma look on your monitor.

 

[Björn Benckert]

There will be no difference. You are shoting raw in both cases and end up with the same result from there. So the ipp2 in camera is just togive a preview of it while shooting and you get your r3ds metadata set right from the get go.

 

[Graeme Nattress]

IPP2 does not effect the raw R3D data recorded (other than in respect to metadata to describe the in-camera IPP2 settings).

Graeme

 

[Aleksandr Frayman]

Hello! There is some nuance. Can I do ipp2 without rec 709? I love the LOG space. I also noticed that the skin tone is not uniform and spotty. Compared to Dragoncolor 2 and Red logfilm

 

[Aleksandr Frayman]

IPP2 does not effect the raw R3D data recorded (other than in respect to metadata to describe the in-camera IPP2 settings).

Graeme

Hello! There is some nuance. Can I do ipp2 without rec 709? I love the LOG space. I also noticed that the skin tone is not uniform and spotty. Compared to Dragoncolor 2 and Red logfilm

 

[Stacey Spears]

Hello! There is some nuance. Can I do ipp2 without rec 709? I love the LOG space. I also noticed that the skin tone is not uniform and spotty. Compared to Dragoncolor 2 and Red logfilm

709 is deliver format. You can target any color space you like. Does not have to be 709. Can be P3, 2020, XYZ, its up to you.

 

[Anton Esteban]

One thing to note, something Graeme mentioned on the Facebook Epic W page, is Redcine X is using a new algorithm that makes slightly sharper images and has better highlight retention in the processed IPP2 workflow vs the 3D Luts provided on the RED site. This doesn't mean your footage will miraculously bring back blown highlights, crushed shadows, or out of focus footage but it is a more precise and direct implementation of the new contrast, color, and highlight controls (Graeme am I doing justice to what you so patiently explained to me). I can't wait until this new SDK is finally available in other programs. This new IPP2 is a game changer, especially the new highlight rolloff controls.

 

[Graeme Nattress]

One thing to note, something Graeme mentioned on the Facebook Epic W page, is Redcine X is using a new algorithm that makes slightly sharper images and has better highlight retention in the processed IPP2 workflow vs the 3D Luts provided on the RED site. This doesn't mean your footage will miraculously bring back blown highlights, crushed shadows, or out of focus footage but it is a more precise and direct implementation of the new contrast, color, and highlight controls (Graeme am I doing justice to what you so patiently explained to me). I can't wait until this new SDK is finally available in other programs. This new IPP2 is a game changer, especially the new highlight rolloff controls.

Specifically it's the processing from raw data to RWG/Log3G10 that gets the extra benefit in REDCine-X. If you process like that an apply the LUTs, the LUTs will not add to or remove that benefit.

Graeme

 

[Anton Esteban]

Wow this is even better than I understood it.

 

[Andrew R Lewis]

IPP2 does not effect the raw R3D data recorded (other than in respect to metadata to describe the in-camera IPP2 settings).

Graeme

So, capturing in camera using IPP2 is nice, but for the moment won't add any benefit when moving to Davinci Resolve for the grade? Or, will Resolve read the IPP2 metadata captured and stored in the RMD just as natively as it would the legacy pipeline? Am I off or ahead of myself?

 

[Graeme Nattress]

So, capturing in camera using IPP2 is nice, but for the moment won't add any benefit when moving to Davinci Resolve for the grade? Or, will Resolve read the IPP2 metadata captured and stored in the RMD just as natively as it would the legacy pipeline? Am I off or ahead of myself?

At the moment, setting IPP2 in camera gives you benefits for monitoring.

Graeme

 

[Neil Abeynayake]

One thing to note, something Graeme mentioned on the Facebook Epic W page, is Redcine X is using a new algorithm that makes slightly sharper images and has better highlight retention in the processed IPP2 workflow vs the 3D Luts provided on the RED site. This doesn't mean your footage will miraculously bring back blown highlights, crushed shadows, or out of focus footage but it is a more precise and direct implementation of the new contrast, color, and highlight controls (Graeme am I doing justice to what you so patiently explained to me). I can't wait until this new SDK is finally available in other programs. This new IPP2 is a game changer, especially the new highlight rolloff controls.

Is there such thing as "Facebook Epic-W page?"

 

[Aaron Lochert]

Is there such thing as "Facebook Epic-W page?"

Yes.

https://www.facebook.com/groups/1816671215236474/

 

[Alain M]

Wow... So for now is a benefit we ll get only in REDCine X Graeme Until other platforms update?

 

[Björn Benckert]

Wow... So for now is a benefit we ll get only in REDCine X Graeme Until other platforms update?

You can use RWGRGB in other apps than RCXP but as I understand the curves doing the conversions to viewing colorspaces are mathematical curves in RCXP and in other apps you would need to use the IPP2 luts. There is a difference in between the two as I understand as conversion is "cleaner" than lut´s but the difference is rather slim as long as the lut got high resolution.

 

[Graeme Nattress]

You can use RWGRGB in other apps than RCXP but as I understand the curves doing the conversions to viewing colorspaces are mathematical curves in RCXP and in other apps you would need to use the IPP2 luts. There is a difference in between the two as I understand as conversion is "cleaner" than lut´s but the difference is rather slim as long as the lut got high resolution.

There's even more to it than that. In RC-X IPP2 beta you're getting a sharper demosaic and more highlight extension on clipped highlights.

Graeme

 

[Mike P.]

There's even more to it than that. In RC-X IPP2 beta you're getting a sharper demosaic and more highlight extension on clipped highlights.

Graeme

Will this be coming to the SDK? Depending on how much of an image improvement it is, it'd be kinda hard to take advantage of it in the real-world if you have to send everything through RCXp for rendering...

 

[Graeme Nattress]

Will this be coming to the SDK? Depending on how much of an image improvement it is, it'd be kinda hard to take advantage of it in the real-world if you have to send everything through RCXp for rendering...

RC-X sits on top of the SDK, so this kind of functionality does indeed make its way to the SDK.

Graeme

 

[Björn Benckert]

RC-X sits on top of the SDK, so this kind of functionality does indeed make its way to the SDK.

Graeme

Aha, So it will be there in the next version of Resolve? Flame will take ages as always I guess