abrupt highlights clipping

[Alexander Ibrahim]

Originally Posted by Michael H. Stevens

When Alexander said: "Acquiring in a higher quality codec will help. Chroma subsampling is not the key issue in this particular case, though it certainly wouldn't hurt. Mostly you want a 10 (or higher) bit codec. Such codecs have more ability to handle "superwhite" and "superblack." " does he mean acquiring like with Cineform NEO-HD?

Yes that is exactly what I mean.

[Alexander Ibrahim]

Originally Posted by Bob Grant

I think ideally you want a 10bit recording medium and the only way to get to the 10bit output from the camera is from HD-SDI port and all the options that could be used in the field are rather expensive.

In the field the only option I know RIGHT NOW is the AJA ioHD coupled with a laptop and an external drive- possibly a drive array.

So, yeah that's expensive as it will cost about as much as the camera itself.

In the near future, we have the Convergent Designs Flash XDR.

Also Cineform has begun developing its own solid state recorder, which I have nicknamed "SOLID."

Once encoded to 8 bit on the SxS cards most NLEs can handle the super whites and blacks. Vegas certainly doesn't have any issues with them although most delivery formats might. Only a matter of wrangling them and there's a number of ways to do that.

This statement is perfectly true and completely misses my point.

The key in my statement is that a 10 bit codec has more ability to handle superwhite and superblack.

8 bit codecs can certainly represent these levels.

8 bit codecs like XDCAM, HDV, DV, DVCPRO HD etc can only represent 256 levels of luma and color intensity.

10 bit codecs can represent 1024 levels. There is just a lot more luminance data, all of which will be available to represent your image.

7.5 IRE to 100 IRE represents your average NTSC image. 0 to 109 IRE represents the range of the EX1. If you use the "super" ranges for image composition you are asking your codec to handle 17% more data. Well strictly speaking you are asking the codec to handle 17% greater range in the data- which compromises its ability to compress efficiently. If you ask this from the codec designed to handle four times the levels, you get better results.

In other words- less abrupt clipping.

[Piotr Wozniacki]

Originally Posted by Alexander Ibrahim

8 bit codecs like XDCAM, HDV, DV, DVCPRO HD etc can only represent 256 levels of luma and color intensity.

10 bit codecs can represent 1024 levels. There is just a lot more luminance data, all of which will be available to represent your image.

7.5 IRE to 100 IRE represents your average NTSC image. 0 to 109 IRE represents the range of the EX1. If you use the "super" ranges for image composition you are asking your codec to handle 17% more data. Well strictly speaking you are asking the codec to handle 17% greater range in the data- which compromises its ability to compress efficiently. If you ask this from the codec designed to handle four times the levels, you get better results.

In other words- less abrupt clipping.

This is an excellent point, Alexander - I mentioned earlier in this thread that it's the lack of mid-tones in the partially clipped area that looks bad; as if banding occured due to the chroma/luma resolution not being enough.

I'd like to see the same areas using 10-bit output from HD-SDI!

[Michael H. Stevens]

Nice post Alexander. Hope we will hear more from you.

I'm asking, if the mpg4 files on the SxS card are 8-bit how does Cinform NEO-HD turn them into 10-bit avi files?

[Bob Grant]

Originally Posted by Alexander Ibrahim

Snip...

7.5 IRE to 100 IRE represents your average NTSC image. 0 to 109 IRE represents the range of the EX1. If you use the "super" ranges for image composition you are asking your codec to handle 17% more data. Well strictly speaking you are asking the codec to handle 17% greater range in the data- which compromises its ability to compress efficiently. If you ask this from the codec designed to handle four times the levels, you get better results.

In other words- less abrupt clipping.

There's no 7.5 IRE setup in the digital realm.
The impact of using the extra range from superblack to superwhite on an mepg-2 encoder would be negligable, certainly nothing like 17%. The biggest issue for the XDCAM or any mpeg-2 encoder is noise and motion.

Recording 10bit or 10bit log certainly can gives you more wiggle room. With the SI-2K I'm effectively under exposing compared to the Z1 or EX1 and then digging into the shadows. It's nice to delay those decisions to post for sure and with care you can get a much more pleasing image. I can just as easily if not more easily mangle it too.

In the end you're pretty much forced to deliver in 8 bits. You can make the hard decisions in a camera like the EX1 or you can spend a lot of money on more expensive cameras or recording systems and shift the decisions to post. Unless you start with a sensor with high dynamic range you might be spending a lot of money to record higher quality noise.

[Piotr Wozniacki]

As an update to my research of best PP for different lighting conditions. Today we have the first snow this winter (February used to be all freezing temperatures and plenty of snow, but due probably to global warming I didn't have an opportunity to test my EX1 with snow, until today).

Below is a grab with Cine4 (my PP4); as you can see it's very safe - no abrupt clipping (in fact, hardly ANY clipping). With Cine1 (my PP3), it is very difficult to keep the snow on the safe side, while avoiding the "abrupt clipping" behind the trees silhuettes and still getting the dark parts bright enough... Also, I'll be experimenting in the remaining two cine gammas (Cine2 and Cine3), as my PP4 seems to be very conservative in terms of using this camera's latitude potential.

What's interesting, when I switch to 32bit in Vegas, not only does the picture look better, but I'm getting the scopes range from well below 0 up to full 110 (see pictures to the right side).

Sorry I shot my old, good barn again, and trough the window, too - but it's really freezing and that was just a PP experiment :)

[Piotr Wozniacki]

For comparison's sake, similar grab with my PP3 (Cine1), also from 32bit Vegas timeline. It's juicier, but more difficult to handle (also, the sun has gone down by some 2 hours ...).

[Bob Grant]

Originally Posted by Piotr Wozniacki

What's interesting, when I switch to 32bit in Vegas, not only does the picture look better, but I'm getting the scopes range from well below 0 up to full 110 (see pictures to the right side).

In 32bit Vegas converts from StudioRGB to ComputerRGB, on the output side you need to / should add the inverse conversion. You can achieve much the same image improvement without going to 32bit and suffering the extended render time.

[Piotr Wozniacki]

Bob, I'm getting similar results when I convert from Studio to Computer RGB while still in 8bit (time-saving) mode - is this what you mean?

[Bill Ravens]

Originally Posted by Alexander Ibrahim

The key in my statement is that a 10 bit codec has more ability to handle superwhite and superblack.

8 bit codecs can certainly represent these levels.

8 bit codecs like XDCAM, HDV, DV, DVCPRO HD etc can only represent 256 levels of luma and color intensity.

10 bit codecs can represent 1024 levels. There is just a lot more luminance data, all of which will be available to represent your image.

You guys absolutely amaze me with your compulsive analyzing of things. For example, the statement that one should move from 256 levels to 1024 levels. Keep in mind that each channel carries 256 values. Therefore for all three channels(and ignoring alpha channel), we're talking 256x256x256=1.68 MILLION colors. OTOH, 1024 levels would be over 1 BILLION colors! There isn't a display screen in the world that can display that.

[Steven Thomas]

True, the real benefit for 10 bit is in post.

[Michael H. Stevens]

Piort:

Did you read my post that addressed this issue? It looks to me that the Vegas scopes can not cope with the super blacks or whites as they show both ends as clipped. So what use are the Vegas scopes now??? I wish someone could explain in detail just how NLE cope with these extreams.

[Piotr Wozniacki]

Originally Posted by Michael H. Stevens

Piort:

Did you read my post that addressed this issue? It looks to me that the Vegas scopes can not cope with the super blacks or whites as they show both ends as clipped. So what use are the Vegas scopes now??? I wish someone could explain in detail just how NLE cope with these extreams.

Michael,

No, I am myself a bit lost here. In fact, while I can see the scopes reaching from below zero on the left, to well over 108 on the right - the picture itself doesn't show any traces of clipping.

I think it takes Glenn Chan or somebody equally knowledgeable to finaly explain all this to us, regular Vegas users :)

[Michael H. Stevens]

Piort:

I started a post on this on the Vegas forum so hopefully Glenn will give us his wisdom.

[Piotr Wozniacki]

Originally Posted by Michael H. Stevens

Piort:

I started a post on this on the Vegas forum so hopefully Glenn will give us his wisdom.

Michael,

I started one there, too - since yesterday Glenn hasn't responded yet; still hoping he will :)

Piotr (like Peter or Pietro; not Piort :))