Jump to content

Nx1: 16-235 vs 0-255


Marco Tecno
 Share

Recommended Posts

I think this has already been discussed in other threads, but never as a stand alone topic.

 

I don't really get the differences about these two encoding modes. The 0-255 seems to have more discrete steps in representing the (same) dr that the sensor can output and the codec can encode.

 

Is this generally better to preserve colors and dr? Then why most ppl suggest to use 16-235?

 

 

Link to comment
Share on other sites

EOSHD Pro Color 5 for Sony cameras EOSHD Z LOG for Nikon CamerasEOSHD C-LOG and Film Profiles for All Canon DSLRs

With 16-235 I see more detail in the shadows.

Premiere seems to take the 0-255 footage and clip it to 16-235, so you lose a ton of dynamic range.

Since there is no way to directly edit H265 files in premiere and a transcoding step is required, did you try transcoding to H264 with the proper 0-255 range? Premiere should be able to handle 0-255 footage if the codec supports it. My understanding is that Prores is a broadcast coded that uses the limited range (16-235 for 8 bit). So the clipping either occurs when you transcode the H265 into Prores or when the prores actually uses the illegal values that are clipped then by Premiere. 

The problem is that using the limited range 16-235 you are throwing away 1/3 of the space. 

 

Link to comment
Share on other sites

my understanding is both settings record the exact same information. what you are changing is the flag that tells your NLE how to display the image data -- across the full range (0-255) or across the broadcast legal range (16-235). but I don't think it inherently changes the data captured by your camera, only how that data is mapped out when you're looking at it. and as andrew noted, when you have a file flagged to mapped the data across the full range and an NLE that overrides that flag to adhere to broadcast values of 16 to 235, it's going to clip 0-15 and 236-255.

I shoot 16-235 because everything I produce has to adhere to broadcast standards. if I didn't things would look pretty effed up on blu-ray to television. one of the reasons I love media composer is the full control you have over the luminance range (data/full vs video/broadcast) when importing clips.

Link to comment
Share on other sites

Actually, Premiere does not clip footage shot in 0-255. You have a broader range to work with, so it might look like you lost info. The only reason I see for shooting in 16-235 is if you need to turn in your footage directly without post work because feeding a 0-255 image directly to TV monitors for example is going to lose info at the top and bottom. If you do post work though, you will always have the opportunity to make it broadcast safe after all the processing, no need to shoot in 16-235.

As an extreme example, here's a clip from footage I shot of some mangoes we just bought. I transcoded to ProRes 422 with FFMpeg and imported to Premiere Pro (the waveforms you see are from SpeedGrade). As you can see, by default it looks like the image is hopelessly clipped at the top and crushed at the bottom. But adjust the lift and gain a bit, and you see that the info is all there.

 

image.PNG

before.PNG

after.PNG

Link to comment
Share on other sites

Ok, so for a 'video newby' like me, to check if I understood: both the settings keep the same data and map them in two different ranges (0-255 or 16-235). Hence the former should have more discrete steps then the latter, thus giving less banding? Or this is not related at all?

Besides, are luts designed to work better with one mode then with another? Does this depend on the sw used?

 

Uh...difficult matter indeed ;-) 

Link to comment
Share on other sites

Ok, so for a 'video newby' like me, to check if I understood: both the settings keep the same data and map them in two different ranges (0-255 or 16-235). 

the camera tries to fit the same data into two different ranges. But as soon as it puts them in the smaller range you loose some information. as you say there. are less discrete steps that might result in banding when grading. 

as far as the lut I doubt there is a difference since the data are mapped in the 0-255 range and rgb space when you are in the nle program.

Link to comment
Share on other sites

  • Administrators

Forgot to mention, transcoded to ProRes first with EditReady of course.

Only reporting what I see.

16-235 works for me and doesn't give me any 'side effects'

the camera tries to fit the same data into two different ranges. But as soon as it puts them in the smaller range you loose some information. as you say there. are less discrete steps that might result in banding when grading. 

as far as the lut I doubt there is a difference since the data are mapped in the 0-255 range and rgb space when you are in the nle program.

You will not see more banding with 16-235 compared to 0-255.

We are talking luminosity here with the 0-255 'steps' not tonal precision which is what causes the banding. A blue sky for example could have such subtle variation in tone that it gets compressed away and you end up with 4 coloured bands. 0-255 allows for 256 so that isn't to blame, the compression is. Banding on the NX1 is mainly caused by the scaling from 6.5K to 4K internally and compression, plus the fact that at low ISOs the image is extremely clean so there's no noise to dither the 8bit bands smoothly together. You can try applying some dithering noise in post though.

Link to comment
Share on other sites

  • Administrators

LUT will work with 0-255 as well but you might have to do a workaround in post to see more of the shadows. LUT or no LUT it doesn't matter, shoot 16-235 so that when EditReady converts to 16-235 ProRes you see into the shadows.

Link to comment
Share on other sites

Thanks, confirms what we've been advocating: shoot in 0-255 unless you plan to skip post.

To throw another option out, if you'd like to capture fine gradients on the low end, you might want to also try the 16-255 option (not 16-235). It spreads the low end info out a bit more at the cost of the mids/highs, but might help you with banding on the lows.

Link to comment
Share on other sites

Ok now I'm really confused. Andrew says:

"... shoot 16-235 so that when EditReady converts to 16-235 ProRes you see into the shadows."

While the link from editready seems to state the exact contrary of this!!!

And....what to do if using rockymountain + adobe premiere in windows??

Help!!

Link to comment
Share on other sites

  • Administrators

"In order for the EditReady file to look “right,” you’d need to apply a LUT or color correction to adjust the signal."

When I developed my LOG LUT for the NX1 I had better results in Premiere when the camera was set to 16-235.

Trust me 0-255 is asking for trouble, it is a whole host of complexity on top for virtually no gain!

Whether 0-255 or 16-235 we are splitting hairs because the NX1 has banding in both modes.

As the EditReady blog post said, NLEs expect 16-235 and the ProRes files won't look right in them if you shoot 0-255.

They don't explain HOW the LUT should correct for this.

Even if you use a LUT in EditReady like my LOG converter, Premiere will still crap it up in 0-255.

Ok now I'm really confused. Andrew says:

"... shoot 16-235 so that when EditReady converts to 16-235 ProRes you see into the shadows."

While the link from editready seems to state the exact contrary of this!!!

And....what to do if using rockymountain + adobe premiere in windows??

Help!!

Is it worth getting confused about? Just shoot 16-235 and be happy!

Rocky Mountain probably remaps 0-255 to 16-235 anyway.

Link to comment
Share on other sites

Interesting article about this topic but for GH4. Basicaly the guy says that there is nothing to gain with 0-255 vs 16-235.

http://blog.josephmoore.name/2014/10/29/the-three-most-misunderstood-gh4-settings/

Yes, we know both capture the same info (nothing gets dropped at the top or bottom), the difference is in how they get ENCODED. I.e., which part (lows, mids, highs) gets preference in terms of space on the 8 bits. See comments on the page for the difference and effect on banding.

Here's my summary and recommendation:

  1. 0-255. Works for most situations, especially recommended when mids are more important than lows and highs. Uses the entire 8 bit range, so it's the best you're going to get for mids and overall image. Not good if you aren't planning post work.
  2. 16-235. Necessary when delivering out of the camera without post work. Also good on high contrast scenes where you want to preserve details in lows and highs at the cost of mids.
  3. 16-255. Good when you have critical info in the lows and can't lift the overall exposure because it will blow the highs.
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

  • EOSHD Pro Color 5 for All Sony cameras
    EOSHD C-LOG and Film Profiles for All Canon DSLRs
    EOSHD Dynamic Range Enhancer for H.264/H.265
×
×
  • Create New...