Blackmagic Shuttle DNxHD issue

[James Hughes]

sorry… that last post… directed to Liam was supposed to be directed to Kristian Lam….. Sorry about the spelling mistake.
cheers.

[James Hughes]

Kristian,
Just wondering if you’re planning on addressing this levels issue in the next firmware release?
Cheers

[Evan Burns]

Looks like I’m encountering the same problem in CS5.5. Just to be clear, can I fix this with color correction or is the information being clipped in some way?

DNxHD (Hyperdeck Shuttle 2)

AVCHD (straight from camera)

[James Hughes]

Everyone tells me that there is no information lost. Technically this seems to be the case, but the picture is crap compared to something like ProRes422.
I’m told DNxHD is supposed to be superior to ProRes, so I can only assume that having to process the DNxHD file is causing the artefacts I can see.
If BM were to implement the codec in a more user-friendly manner, or give us the option to select the 0-255 range, then I’m lead to believe this would solve our issues.
I know I’m repeating things already said, but nothing has been done about it yet.
BM……..???????

[Bill Ravens]

I’ve been engaged in a similar discussion over on the COW Davinci Resolve Forum. It seems Resolve has already dealt with this issue by providing the option to select full range or remapped range.When I import the Hyperdeck DNxHD files into Resolve, and select ‘full range’, I get a ‘properly’ colored image.

It looks like the Hyperdeck is exporting the recorded data with the wrong flag set in the header. The flag is telling your NLE to remap all input values to a reduced range. My workaround is to import into resolve, first, and export to OP1a. Alternatively, you can reset the levels after you import into your NLE.

At this point, however, I’m not convinced that the superblacks aren’t being clipped. I sure would like BMD to give us some feedback about this.

[James Hughes]

Yes, I agree. The lack of feedback is what bothers me the most.

[Evan Burns]

Tomorrow I will compare the Quicktime uncompressed with the corrected DNxHD and perhaps highlight some of the artifacts or potential clipping. Unless someone has already done that. I’d like to know the severity of the issue before resulting to buying additional SSD dives.

[Bill Ravens]

FWIW, here’s my comparison of the FS100’s native AVCHD and the DNxHD on the hyperdeck…
https://forums.creativecow.net/readpost/124/879660

[Jay Bloomfield]

I tried to upload a complete metadata listing for a BMD Shuttle 2 DNxHD MXF file, but I’m not sure where it ended up. Anyway, here’s the snippet that describes the video essence:

CDCIEssenceDescriptor
InstanceUID = {daa94d91-6dd8-fae0-e4a2-739abd196d03}
ContainerDuration = 1377
SampledWidth = 1920
SampledHeight = 1080
StoredWidth = 1920
StoredHeight = 1080
DisplayWidth = 1920
DisplayHeight = 1080
SampleRate = 30000/1001
ComponentDepth = 10
PictureEssenceCoding = [060e2b34.0401.010a.04010202.71070000]
ColorRange = 897
WhiteReflevel = 940
BlackRefLevel = 64
ColorSiting = 4
VerticalSubsampling = 1
HorizontalSubsampling = 2
ImageAlignmentOffset = 8192
AspectRatio = 16/9
VideoLineMap
VideoLineMapEntry = 26
VideoLineMapEntry =
FrameLayout = 0
DisplayYOffset = 0
DisplayXOffset = 0
SampledYOffset = 0
SampledXOffset = 0
EssenceContainer = [060e2b34.0101.01ff.4b464141.000d4d4f]

As you can see, the range is from 64-940, not 0-1023.

[Evan Burns]

Thanks Jay,

That’s hard evidence and very clearly shows the issue.