How Important is 10-Bit Really?

[anonim]

2 hours ago, Yurolov said:

For me you have to do your homework on one particular camera and then compare it with others and see for yourself if 10 bit actually makes a difference.

This, I think, summarize topic very nicely - with other lovely compose thread of argumentation.

But than "And the GH5 is not good because it has 10 bit" - I don't understand? From my experience, its strong side is that it has 10bit, not just 8bit - not because its 10bit is better than 8bit from some other cameras (especially those with better DR - and it is difference between EVA1 and GH5), but simple because GH5's 10bit is better than GH5's 8bit.

I think that just there's no any reason to think of quality of 10bit or 8bit in general - of course, we can, but than it is theoretic question.

[jonpais]

No, @anonim, @Yurolov is right - no one in their right mind would use the GH5 for paid work. But why does he keep bringing up the GH5? I thought we were talking about bit depth, not Panasonic. And since he gets his authority from Dave Dugdale and Max Yuryev, I’m inclined to believe he doesn’t know what he’s talking about.

[Deadcode]

Those people who mostly complain about 8 bit limitations are usually backyard flower shooters.
10 bit is awesome, but those who cannot get decent results from 8 bit wont get satisfied with 10 bit either.
Next step is blame the bitrate, then blame the camera and lens, and the NLE.

And these people never think they have a lack of knowledge...

In conclusion: 10 bit for most of the users in EosHD forums wont give any benefit over 8 bit. 

GH5 always get in picture because that's the only affordable consumer camera (+GH4) which is able record in 10 bit.

In the future there will be several choices with 10 bith recording in affordable range of course, but when it happens 
everyone will cry for raw recording because it's soooooo much better (not) then compressed 10bit..

[jonpais]

@Deadcode Right. Log is for losers. All EOSHD forum members shoot flowers. We all blame our equipment for our failures. Thank you for adding to the discussion.

[anonim]

2 minutes ago, jonpais said:

...no one in their right mind would use the GH5 for paid work.

I understand, it is irony - or, as Matthew Hartman learned me, at idiom-English it is little bit "salty". I hope other side will not be too angry.

I think that It is very hard for many talented reviewers not to project some inclinations (or also something else) in comparing task once when climb to the level that could easy acquire many different cameras, lenses etc. That's why rather choose to follow reviewers who are openly fond to one system and achieve exceptional results with it - Brandon Lee with Sony, or White in Revery with Panasonic.

@Deadcode

"In conclusion: 10 bit for most of the users in EosHD forums wont give any benefit over 8 bit."

Let's help each other to get more benefit - I like @Yurolov sugestion's path: let's compare (who can and have resources) concrete cameras in respective fields and relevant variables and provide results. It will certainly help (at least me) in future decision where is the best for me to invest.

 

[Deadcode]

10 minutes ago, jonpais said:

@Deadcode Right. Log is for losers. All EOSHD forum members shoot flowers. We all blame our equipment for our failures. Thank you for adding to the discussion.

Actually you are the who shoot HLG because you cannot get decent result from VLOG. 

LOG is not for those who "grade" with LUT's. 

But you are far more professional, you grade HLG with LUT's :P

@anonim you are right, this should be the reason of this topic, but troll on duty just arrived without constructive comments.

[jonpais]

21 minutes ago, anonim said:

I understand, it is irony - or, as Matthew Hartman learned me, at idiom-English it is little bit "salty". I hope other side will not be too angry.

I think that It is very hard for many talented reviewers not to project some inclinations (or also something else) in comparing task once when climb to the level that could easy acquire many different cameras, lenses etc. That's why rather choose to follow reviewers who are openly fond to one system and achieve exceptional results with it - Brandon Lee with Sony, or White in Revery with Panasonic.

@Deadcode

"In conclusion: 10 bit for most of the users in EosHD forums wont give any benefit over 8 bit."

Let's help each other to get more benefit - I like @Yurolov sugestion's path: let's compare (who can and have resources) concrete cameras in respective fields and relevant variables and provide results. It will certainly help (at least me) in future decision where is the best for me to invest.

 

C’mon now. Backyard flower shooters? Most filmmakers who complain about 8-bit are idiots, in other words? Or nature photographers are somehow inferior? @Deadcode‘s comments about ‘most EOSHD shooters’ is nothing if not condescending.

14 minutes ago, Deadcode said:

Actually you are the who shoot HLG because you cannot get decent result from VLOG. 

LOG is not for those who "grade" with LUT's. 

But you are far more professional, you grade HLG with LUT's :P

@anonim you are right, this should be the reason of this topic, but troll on duty just arrived without constructive comments.

What in the hell are you talking about? no name, no face, no portfolio, no show reel, no website, no Vimeo channel? Zero credibility.

[Yurolov]

1 hour ago, anonim said:

This, I think, summarize topic very nicely - with other lovely compose thread of argumentation.

But than "And the GH5 is not good because it has 10 bit" - I don't understand? From my experience, its strong side is that it has 10bit, not just 8bit - not because its 10bit is better than 8bit from some other cameras (especially those with better DR - and it is difference between EVA1 and GH5), but simple because GH5's 10bit is better than GH5's 8bit.

I think that just there's no any reason to think of quality of 10bit or 8bit in general - of course, we can, but than it is theoretic question.

I think you misunderstood friend. I said that the 10 bit of the GH5 is not as good as the EVA1. I linked that video above that compares the 10 bit of both cameras. Go check it out; it is done by a professional colorist. I never said the GH5 is a bad camera. I like it, in fact. It produces a nice image.  

1 hour ago, jonpais said:

No, @anonim, @Yurolov is right - no one in their right mind would use the GH5 for paid work. But why does he keep bringing up the GH5? I thought we were talking about bit depth, not Panasonic. And since he gets his authority from Dave Dugdale and Max Yuryev, I’m inclined to believe he doesn’t know what he’s talking about.

Nice strawman argument. GH5 is a good camera. (see above) That wasn't my argument. To dismiss people without even viewing the videos makes me think you are just trolling and don't want to engage me in the topic. I put them as an example, not as an authority. Nor does Dave Dugdale do the comparison, which you would know if you cared to watch. So if you can explain to me why they are wrong then please let me know. I am happy to be proven wrong. We are all learning, even the most senior of us.

But it is clear your curiosity is limited so there is no point for me to further engage with you.  

[anonim]

1 hour ago, jonpais said:

C’mon now. Backyard flower shooters? Most filmmakers who complain about 8-bit are idiots, in other words? Or nature photographers are somehow inferior? @Deadcode‘s comments about ‘most EOSHD shooters’ is nothing if not condescending.

Actually, I think that Yurolov's (actual target of quoted saltiness not @Deadcode) thread of argumentation was very nice, as your also - and that both could be simply united. When now conclusion-claim "And the GH5 is not good because it has 10 bit" is precised, I, for myself, am very glad to read his, as also your, before these-saltiness-intrusion argumentation.

Contrary, Deadcode's comment is obviously very "salty" and include direct personal attack from the lines - he wished it to be so. Sadly, you, obviously, personally irritate him here or irritate in general him - as I irritate someone else or many here. And as every single person here and everywhere whose first act toward others is (explicitly or hypocritic hide) from the position of I-know-better-and-more-than-you - irritate me, equally when someone use words as Bullshit or other Shit-x variants as powerful sound argument. Comment "Those people who mostly complain about 8 bit limitations are usually backyard flower shooters" could be seen as opinion - I don't like formulation, and I'm think that "usually" is a bit of exaggerating.

Do your argumentation in this thread deserved to be seen as irritating in this thread? - I think not at all, and I strongly agree with its conclusion theory side wise - as with @Yurolov's practical side wise.

edit

Appendix - I'm noticing that in the last time I'm becoming irritated with usage of unexplained word "cinematic" from EOSHD self-proclaimed authorities

[Yurolov]

11 minutes ago, anonim said:

Actually, I think that Yurolov's (actual target of quoted saltiness not @Deadcode) thread of argumentation was very nice, as your also - and that both could be simply united. When now conclusion-claim "And the GH5 is not good because it has 10 bit" is precised, I, for myself, am very glad to read his, as also your, before these-saltiness-intrusion argumentation.

Contrary, Deadcode's comment is obviously very "salty" and include direct personal attack from the lines - he wished it to be so. Sadly, you, obviously, personally irritate him here or irritate in general him - as I irritate someone else or many here. And as every single person here and everywhere whose first act toward others is (explicitly or hypocritic hide) from the position of I-know-better-and-more-than-you - irritate me, equally when someone use words as Bullshit or other Shit-x variants as powerful sound argument. Comment "Those people who mostly complain about 8 bit limitations are usually backyard flower shooters" could be seen as opinion - I don't like formulation, and I'm think that "usually" is a bit of exaggerating.

Do your argumentation in this thread deserved to be seen as irritating in this thread? - I think not at all, and I strongly agree with its conclusion theory side wise - as with @Yurolov's practical side wise.

Thank you. I am always willing to talk with people who treat each other and the argument with respect, with a view to reaching the truth. Test it and we will know! 

[anonim]

1 hour ago, Deadcode said:

@anonim you are right, this should be the reason of this topic, but troll on duty just arrived without constructive comments.

I hope that sign :P targeting @jonpais means - Look, I'm little bit salty!

Because "Troll on duty" is the last that I could say about character of @jonpais's involving in hundreds of threads here. I never saw that he esteemed his opinions or capabilities above others. Quite contrary - he was always ready, and inspirativelly sober-mindly, to admit his non-interested lack of grading intervention. But he has his without discussion always relevant opinion and - which is for me the most important - always first in detail, and without any grain of lifting up his nose, explain it. Than, if attacked, or as result of misunderstaning, may be salting-add (besides, very witty, I think). Personally, resisting in Indian's subname of @jonpais as "Troll on duty" - to me more speaks about caller.

[jonpais]

@Yurolov‘s comment about people being ‘obsessed’ with 10-bit is in itself an indication that he does not understand the topic. I do not see anyone here obsessed with anything. Obsession implies being fixated on something, not being able to rid oneself of a particular idea - and I would refrain from characterizing anyone in this thread that way. The OP asked how important 10 bit is, and the answer is, in some cases very, in others, not.

[anonim]

22 minutes ago, jonpais said:

@Yurolov‘s comment about people being ‘obsessed’ with 10-bit is in itself an indication that he does not understand the topic. I do not see anyone here obsessed with anything. Obsession implies being fixated on something, not being able to rid oneself of a particular idea - and I would refrain from characterizing anyone in this thread that way. The OP asked how important 10 bit is, and the answer is, in some cases very, in others, not.

Agree, if @Yurolov indeed found that you are obsessed with 10-bit - it is for me, I have to say, complete nonsense. (But I'm pretty sure it is not such - his arguments and manner of argumenting are too noble for such BS level - if he thought so, I'm sure he would told you that openly.) As about "people" being obsessed - I don't know, as part of his global thinking discourse about topic in form of contrary-view strategy, it may be accepted But hay! - both of you really thought and explained opinions at really-really at high end level not just for local BS-type-accurate often knocking down standard.

[Deadcode]

13 minutes ago, jonpais said:

 The OP asked how important 10 bit is, and the answer is, in some cases very, in others, not.

This is the correct statement.

Advanced colorgrading, VFX and working with power windows requires 10 bit source. Curves/contrast/saturation adjustments not really.

My eariler statement about the community refers to this, most of the users grading capabilites and needs are strickted to Curves/contrast/saturation adjustments. This kind of push on the image requires more bitrate then the need of higher bit depth. This is why i think most of the users dont really NEED 10 bit. If 10 bit is available it's good. But nowday it's expensive. In the future if more and more brand brings 10 bit to the table, it will be awesome. But today it's not worth it for the average user.

[cantsin]

Okay people, I produced an 8bit and a 10bit reference video file for you to test & settle your dispute. They were produced with a high quality imaging pipeline (2.5K BM Cinema Camera Raw -> Resolve -> Adobe Media Encoder) but encoded in such a way that they represent the difference between 8bit and 10bit video recording in a camera like the GH5 under ideal conditions (i.e. supposing that the video signal processing and internal codecs of the GH5 are as good as the pipeline CinemaDNG -> Resolve -> AME).

What I did:

1. Pointed a fresnel spotlight to a white wall so that it created a high-contrast gradient;
2. Shot it with the Blackmagic Cinema Camera 2.5K in CinemaDNG raw, exposed to the right (i.e. with no clipping of the highlights, using the sensor's maximum dynamic range)
3. Imported it into a 1920x1080 project in Davinci Resolve, used Resolve's color space transformation tool to transform the color space to Panasonic VLog.
4. Rendered it out as 16bit uncompressed TIFFs for maximum quality and no loss of color depth.
5. Imported the TIFF sequence into Adobe Media Encoder, and rendered it out as 
   1. 10bit h264 at 200 Mbit/s.
      (Note: Unfortunately, I could only use 4:2:0 color sampling because h264's High 10 profile does not support 4:2:2. I also tried ffmpeg/x264, but it doesn't support 10bit 4:2:2 either. Panasonic's own implementation of 10bit 4:2:2 h264 recording in the GH5 is non-standard, causing the well-known NLE compatiblity problems when the camera came out.)
   2. 8bit h264 4:2:0 at 200 Mbit/s.

Perhaps surprisingly, the 10bit file is much smaller than the 8bit file; I "blame" more efficient compression in the newer h264 high10 profile for this difference.

How you can play with the footage:

1. Import it into your NLE
2. Apply a Vlog-to-Rec709 LUT or color space transformation
3. Compare banding of the gradient

I did, btw., not do the latter yet because I want to be impartial and let myself surprise by the outcome.

gradient_test-8bit_vs_10bit.zip

[Deadcode]

21 minutes ago, cantsin said:

Okay people, I produced an 8bit and a 10bit reference video file for you to test & settle your dispute. They were produced with a high quality imaging pipeline (2.5K BM Cinema Camera Raw -> Resolve -> Adobe Media Encoder) but encoded in such a way that they represent the difference between 8bit and 10bit video recording in a camera like the GH5 under ideal conditions (i.e. supposing that the video signal processing and internal codecs of the GH5 are as good as the pipeline CinemaDNG -> Resolve -> AME).

What I did:

1. Pointed a fresnel spotlight to a white wall so that it created a high-contrast gradient;
2. Shot it with the Blackmagic Cinema Camera 2.5K in CinemaDNG raw, exposed to the right (i.e. with no clipping of the highlights, using the sensor's maximum dynamic range)
3. Imported it into a 1920x1080 project in Davinci Resolve, used Resolve's color space transformation tool to transform the color space to Panasonic VLog.
4. Rendered it out as 16bit uncompressed TIFFs for maximum quality and no loss of color depth.
5. Imported the TIFF sequence into Adobe Media Encoder, and rendered it out as 
   1. 10bit h264 at 200 Mbit/s.
      (Note: Unfortunately, I could only use 4:2:0 color sampling because h264's High 10 profile does not support 4:2:2. I also tried ffmpeg/x264, but it doesn't support 10bit 4:2:2 either. Panasonic's own implementation of 10bit 4:2:2 h264 recording in the GH5 is non-standard, causing the well-known NLE compatiblity problems when the camera came out.)
   2. 8bit h264 4:2:0 at 200 Mbit/s.

Perhaps surprisingly, the 10bit file is much smaller than the 8bit file; I "blame" more efficient compression in the newer h264 high10 profile for this difference.

How you can play with the footage:

1. Import it into your NLE
2. Apply a Vlog-to-Rec709 LUT or color space transformation
3. Compare banding of the gradient

I did, btw., not do the latter yet because I want to be impartial and let myself surprise by the outcome.

gradient_test-8bit_vs_10bit.zip

Thank you very much for this test. Could you please upload the original dng sequence to somewhere?

[Emanuel]

3 hours ago, jonpais said:

Couldn’t disagree more. Why the antipathy toward 10-bit? 

I don't think there's antipathy, but not to understate 8-bit acquisition.

[cantsin]

19 minutes ago, Deadcode said:

Thank you very much for this test. Could you please upload the original dng sequence to somewhere?

Here's one sample DNG frame - the full 168 MB sequence doesn't make a real difference since it's a static image.

Blackmagic Cinema Camera_1_2017-06-21_2318_C0000_000021.dng

[Deadcode]

9 minutes ago, cantsin said:

Here's one sample DNG frame - the full 168 MB sequence doesn't make a real difference since it's a static image.

Blackmagic Cinema Camera_1_2017-06-21_2318_C0000_000021.dng

I did the exact same test with a6300 SLOG2 and it was smooth before/after REC709 correction.

On your sample i can see banding on the 8 bit footage, but the bitrate is very low (16Mbps). So i would like to recreate your sequence but with Resolve > ProRes HQ, and 8 bit forced high bitrate reencode from that ProRes to h264. 

Or maybe i post my footage later

[cantsin]

In the end, I was too curious and just had to test the outcomes myself.

Workflow:

1. Imported the two test clips (8bit Vlog and 10bit Vlog 200Mbit h264) into a new 1920x1080 Resolve project;
2. Applied Resolve's "Color Space Transform" tool to convert both clips from VLog to Rec709; no further color correction used.
3. Used the "Grab still" function in Resolve's Color menu to create one still from each of the two clips, exported them as PNGs.

The results are attached here. "8bit" and "10bit" only refer to the color depth of the original video file; both stills are 8bit PNGs.

 

3 minutes ago, Deadcode said:

On your sample i can see banding on the 8 bit footage, but the bitrate is very low (16Mbps). So i would like to recreate your sequence but with Resolve > ProRes HQ, and 8 bit forced high bitrate reencode from that ProRes to h264. 

The bitrate is low because of the static frames/no motion; actually, the bit rate is even much lower for the 10bit file. 

But I will do another experiment and export high-quality 8bit 4:2:2 DNxHR and transform that.