Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Moire / Interlacing Issues in 4k R3D files

Almost always moire is caused by a poor scaling choice.

Graeme
 
Okay, got a correct clip uploaded and ready to review. This is a clip from the R3D file and is 216mb. About five seconds long. The interlacing issues have existed on the chains near the woman's legs, chains in the background and on the edge of the bed. As I mentioned a few posts ago, I have now discovered that the lab was only watching this in 1920x1080 and were not, in fact, watching this in true 4k. Thanks!
https://www.dropbox.com/s/lexbrisnr8wg9xt/A103_C007_0123S5_001.R3D
 
I see no obvious issues in the R3D file Jon. Even when I do a rough downsample I'm not seeing anything that could look like interlacing. Are you abel to get a grab what what the QC thinks is wrong, because the issue doesn't appear to be in the source data.

Graeme
 
Jon, that movie appears to have been made via an very very poor downsample. I don't know the full details of your workflow, but at some point in your chain you've got some bad downsampling going on or you're not decoding the R3D at full resolution.

Graeme
 
Hi Graeme. Thanks for the info. My editor used Red Cine X for creating the ProRes files that we edited from. We have done everything from the ProRes files including grading since we never intended to need to go back to the 4k footage. So you are thinking that it has something to do with the conversion from the R3D to the ProRes file? I'm also a bit baffled at why the lab is saying that they are seeing it in the R3D files.
 
Jon Keeyes said:
Hi Graeme. Thanks for the info. My editor used Red Cine X for creating the ProRes files that we edited from. We have done everything from the ProRes files including grading since we never intended to need to go back to the 4k footage. So you are thinking that it has something to do with the conversion from the R3D to the ProRes file? I'm also a bit baffled at why the lab is saying that they are seeing it in the R3D files.
Click to expand...

I guess you need to question them RE their signal chain. What are they viewing it on, through and at what resolution?
 
Yes Jon, you've got to look at how the ProRes was made from the R3D. There's plenty of settings you can decode the files with - some fast and only suitable for a quick proxy file, and others for a proper full quality decode.

To understand what the lab is seeing, I'd have to see what they're seeing.... But when I see what you've shown me in the pro-res movie, that's where your major issue lies.

Graeme
 
Jon... the R3D looks just fine here too.

As mentioned... The prores looks quite awful and someone messed up along the way with the transcode.
 
Thanks guys. We were given some suggested settings to look at in the transcode and they were set to those. Any suggestions on what we should be trying, setting, etc. would be gratefully appreciated.
 
In the export window of Redcine X ensure your "Debayer Settings" under "Output Resolution" are set to "Full".

Everyone in this thread is correct, there are no issues with the R3D. If you see aliasing in the R3D it is likely because you are still in fact viewing a downsampled image unless you truly have a 4k monitor. If you zoom in to 1:1 with a full Debayer you will not see any issues. This is likely the culprit at the post house where they are viewing the R3D. Tell them to view at 1:1 with full debayer, and it should be fine.

They or you can also try bringing the footage through Da Vinci. It's actually really nice to be able to have access to the metadata when colouring, you should try it out, it's pretty neat!

One time I did see a shimmer of moire on set once when our set was made from corrugated cardboard. We checked with DIT and they couldn't see it on their end, and it is not present in the final output: http://www.youtube.com/watch?v=N0sHb3YXygI
 
looks like 1/4 debayer to me
 
Thanks for the info. Anything else that comes to mind in settings, please let me know. In the meantime, my colorist is going to dig into it tonight and try converting again. Thanks!
 
Best setting : change the lab!

If not possible, you could try to online the R3D footage and replace the Prores files keeping so your grades... Dont know if this could even work... What colorgrading software do you use?

Pat
 
Bob Gundu said:
The transcoding is definitely the problem here. Plus, there were also some odd settings in your RMD file that leads me to believe your lab is not as proficient as they say they are.
Click to expand...

Yeah, this "lab" has been quick to assign blame to the footage, and hasn't seemed conscious that the workflow itself may be the issue.

Similarly, a colorist familiar with R3D's would have probably seen the RAW files themselves were in good shape.

On the flip side, I'm always impressed by the generosity of these boards. It's impressive how many talented people have stepped in and offered help.

Comforting to see also that the original shoot files themselves...are looking good!
 
Last edited:
Graeme Nattress said:
Scaling is the operation that when improperly performed will induce aliasing. Marcel, if you want me to take a look at the R3D to check that the data is good, just post it up here or get it to me. If you can post a grab of what you're seeing in premiere that will help too.

Graeme
Click to expand...

Hi Greame, many thanks for the offer, I have upload the r3d file where i see the moire to dropbox, here is the link for anyone interested:

https://www.dropbox.com/sh/r3do8t431a4t19z/1dPJ29V9kH?n=256344153

I see the moire in the actor shirt when I am editing it in Premiere CS6, after reading all previous post I feel safe now since probably it is just the premiere downscaling which is causing the issue, and i will appreciate if anyone with more experience in Premiere could show me the right settings to correctly export the final project to a blue-ray disk.

Regards!
 
Jon Keeyes said:
Hi Bob. Just to clarify for everyone, I'm the director of the movie and not an editor by trade. My editor is out of pocket for a week so I'm having to take the lead on this as we're dealing with QC and distribution deadlines. We have seen some slight moire throughout editing but really didn't think anything of it. Once the final movie was sent for QC, that's when it became an issue because of the "rejected." In backtracking through the problem, we started wondering if it had to do with the color grade, but discovered the moire was in the base ProRes files converted from the 4k footage. We then started thinking it had something to do with the conversion of the 4k footage but the moire kept appearing in the ProRes files (my editor's broadcast monitor is broken so he's only being able to see footage on his hd computer monitor). Finally, the lab requested the 4k footage to review and they determined that the moire is in the 4k footage. I have not personally seen it yet in the 4k footage. I just got RedCine loaded onto my computer last night so I can start seeing what the lab is seeing as they provide examples.
Click to expand...

Jon, from what you described here, it sounds like your colorist is grading using the ProRes transcodes and NOT the original 4K R3D files. That is your problem. One, you're losing detail by using the transcodes. Two, you're completely losing all the benefits of shooting RAW. I'm a colorist myself and I ALWAYS conform back to the R3D for the color grade. That will result in a MUCH better quality master in the end, even if you only master at 1080p. Bottom line, make sure your colorist is grading with R3D's. If he's not, tell him you need him to do that or get a new colorist. That's my opinion.
 
Mike Halper said:
I'm a colorist myself and I ALWAYS conform back to the R3D for the color grade. That will result in a MUCH better quality master in the end, even if you only master at 1080p. Bottom line, make sure your colorist is grading with R3D's. If he's not, tell him you need him to do that or get a new colorist. That's my opinion.
Click to expand...
I know of some fairly major theatrical projects that used DPX files and not R3D's. This is not as big a concern as it used to be, particularly on big-horsepower daVinci Resolve Linux systems, but it's still an issue for older systems.

I personally think you can get 90% of the quality out of 2K files, but I don't dispute I'd rather work with the 4Ks for blow-ups and repos. As long as the time and budget are there, I'm happy to work with R3D's all the way. God help you with a quick turnaround and low budgets.
 
In short:
If you are seeing moire in r3d footage, it is either your preview/export quality setting thing or someone else in post is doing it wrong.

Shortest:
If you are seeing moire in r3d files, someone in post is doing it wrong.
 
You must log in or register to reply here.
Back
Top