Why the downer on the 1D C?

Some very good points in Graemsey post. The one I don't agree with is bitdepth. 8 bit might usually be sufficient for acquisision. However, by converting such footage prior to postproduction we doesn't gain much - there's no way to recreate missing bits, all we can do is pad digital words with binary 0s. The only advantage of this is that there's less chance of artifacts resulting from accumulated processing error. Stretching the footage by applying 'expansion' curve (something done often in grading) will uncover the lack of bits - banding and posterisation might show up.

However, to be fair, there are two additional factors that should be taken into account when speaking about 1DC and bit depth:
1. LOG gamma. Non-uniform sampling, which assigns more bits to where our eyes are more sensitive to level changes. This means much more efficient use of available 8 bits and less risk of banding unless you do something extreme in grading.
2. The fact it's 4K. If destination (delivery) format is 1080 or 2K we can downconvert to such raster size getting REAL ~10bit data (with all 10bits potentially containing useful information). This is because we get 1 pixel from (more or less) 4 pixels (by averaging or, hopefully, more sophisticated, weighted filtering over neighboring area) and while single 8bit pixel can represent 256 levels, 4 pixels can represent 1024 discrete levels, which correspond to 10bits. In other words, 4 pixels (averaged) can represent more 'shades' than single one, assuming they all are 8bit deep.
 
Last edited:
I don't understand this thread. You've asking why the downer on the 1D-C and then asking that we not mention other cameras like the Magic cam. The downer is that the bang for the buck compared with the new Sony FS700 or Magic is really seemingly bad. Two years ago, fine, bring on the 1D-C for this price. But right now it looks like a looser.
 
Zephyrnoid said:
RAW-[/COLOR]...In honesty, I doubt too many of the RAW proponents would be so fond of it if they had to deal with the excruciating nuances of this fine forma
Click to expand...

You've got me curious with this comment - I've never had to deal with raw video, just raw stills in photoshop. The image I capture starts at a place I'm comfortable with, then I get to play with the extra stops and colour depth that the format provides. I don't pretend to have a strong knowledge on the subject though - can you expand on your comment?

Apologies for going ott.
 
hgf4 said:
Dan Chung loves the camera but gives him some grief over the price. The 1DC is the second camera they talk about.
Click to expand...

I really liked this video. Not to get too far off topic, but while I do appreciate all the effort and energy it has taken to provide coverage of NAB by this site and others (how exhausting that must be!), I feel that people have been lobbing too many softballs for too long at these manufacturers. The result is pure marketing without any questioning at all, and whether I see it on this site or others, in the end, if truth be told, we might as well have only one source for this kind of video. It is the same old speech. Until - Dan Chung. I appreciate that someone is being a little tougher, asking the questions we are asking and challenging the order of things. These guys need to be challenged just like in every other industry. A more realistic, harder nose approach also makes for good watching and is more informative. In the past when I have raised this point, people have said this is not journalism, NAB isn't really a place for that, the readership is not really interested or looking for this, etc. I disagree. There is no reason why, with so many image makers here, that a more journalistic, incisive way of conducting these interviews can 't be made.
I would like to suggest that this be kept in mind for next year's work.

Thanks again Luis and Stephan for the hard work they have in done in any case.
 
mcgeedigital said:
Shane Hurlbut lost all credibility on the subject when he declared in front of over 1500 production professional that the 1DC made the Epic/Alexa and F65 "look like sh!t."

That is just ridiculous.

The reason HE likes the camera is because he is paid to pimp it.
Click to expand...
Isnt he the guy that claimed canon lenses have special qaulities that made the 5d mk II fotage look beautiful
 
trez said:
Some very good points in Graemsey post. The one I don't agree with is bitdepth. 8 bit might usually be sufficient for acquisision. However, by converting such footage prior to postproduction we doesn't gain much - there's no way to recreate missing bits, all we can do is pad digital words with binary 0s. The only advantage of this is that there's less chance of artifacts resulting from accumulated processing error. Stretching the footage by applying 'expansion' curve (something done often in grading) will uncover the lack of bits - banding and posterisation might show up.
Click to expand...

That is not exactly true. You can create intermediate bits in and when you apply huge curves to the originally 8-bit material, no banding will show up. Test it up with Neat Video. Make a 32-bit comp in After Effects, take your 8-bit shot in with a good gradient. Apply curves. Notice the banding? Good. Now take Neat Video, apply it to the shot. Just a smidgeon of noise reduction. Neat Video will fill up the 32-bits completely. Now apply the same curves. Notice anything? Way less banding, in my casefiles there is usually no banding added at all.

So yes, it is possible to create a good 32-bit output out of 8-bit.
 
I'm not impressed about the 1D C for the same reason I'm not impressed by the C500 and wasn't impressed by the C300.

Price/performance sucks.
 
trez said:
Some very good points in Graemsey post. The one I don't agree with is bitdepth. 8 bit might usually be sufficient for acquisision. However, by converting such footage prior to postproduction we doesn't gain much - there's no way to recreate missing bits, all we can do is pad digital words with binary 0s. The only advantage of this is that there's less chance of artifacts resulting from accumulated processing error. Stretching the footage by applying 'expansion' curve (something done often in grading) will uncover the lack of bits - banding and posterisation might show up.

However, to be fair, there are two additional factors that should be taken into account when speaking about 1DC and bit depth:
1. LOG gamma. Non-uniform sampling, which assigns more bits to where our eyes are more sensitive to level changes. This means much more efficient use of available 8 bits and less risk of banding unless you do something extreme in grading.
2. The fact it's 4K. If destination (delivery) format is 1080 or 2K we can downconvert to such raster size getting REAL ~10bit data (with all 10bits potentially containing useful information). This is because we get 1 pixel from (more or less) 4 pixels (by averaging or, hopefully, more sophisticated, weighted filtering over neighboring area) and while single 8bit pixel can represent 256 levels, 4 pixels can represent 1024 discrete levels, which correspond to 10bits. In other words, 4 pixels (averaged) can represent more 'shades' than single one, assuming they all are 8bit deep.
Click to expand...

Interesting points.
No matter how sophisticated is your filtering,
0 0
0 1
will be equivalent to
1 0
0 0
and
0 1
0 0
and
0 0
1 0
when downscaled. Because the filter needs to be more or less symmetrical.
So no. Not really 10 bit equivalent, even with smart filters.

1. LOG gamma. Non-uniform sampling, which assigns more bits to where our eyes are more sensitive to level changes. This means much more efficient use of available 8 bits and less risk of banding unless you do something extreme in grading.
Click to expand...
This non-uniform distribution that caters to human eye exists on ALL consumer and pro cameras. It is called (non-log) gamma. ;)
Yes, they may tweak it all they want and try to balance it. And in all fairness, that's all they can do: try to find a decent compromise to stuff it into 8 bits. But it stops being true LOG then and lacks the true exposure latitude options of a true LOG.



And a thing about Shane's short:
The image does look pretty good but we don't know much about it (well, I don't, at least), other than it's mostly night-time shots and flat fluo lighting shots. These are all types of shots where a 8-bit camera (and especially, GOOD low light camera) is at its best. And yeah, detail is nice. This is a 4K camera after all. But what about range in difficult scenes? That's what I would like to see from the 1D-C.
 
Quick question... Does this camera have a unique sensor or is it the same as the 1Dx?

If it is the same... Where is the extra stop or two of DR coming from, the C-log? The codec? 4:2:2?

I guess all of the above... Hopefully, we will see that tech trickle down to the next line of DSLRs.
 
maarek said:
That is not exactly true. You can create intermediate bits in and when you apply huge curves to the originally 8-bit material, no banding will show up. Test it up with Neat Video. Make a 32-bit comp in After Effects, take your 8-bit shot in with a good gradient. Apply curves. Notice the banding? Good. Now take Neat Video, apply it to the shot. Just a smidgeon of noise reduction. Neat Video will fill up the 32-bits completely. Now apply the same curves. Notice anything? Way less banding, in my casefiles there is usually no banding added at all.

So yes, it is possible to create a good 32-bit output out of 8-bit.
Click to expand...

And backing this up further, for years and years, Quantel was an 8-bit system through and through. Commercials would be shot on film, telecined to D1, and then loaded into the Quantel. All of the color correction was done at 8-bit, and they did this dynamic rounding thing to add a bit of noise at each point, so that no banding would show up.

Point is, you can get amazing results out of 8-bit footage if the initial capture is done well.
 
Erich Ocean said:
It says right on Blackmagic's site that the RAW footage is 12-bit uncompressed, in CinemaDNG format.
Click to expand...

Cinema DNG supports lossless compression. (See Adobe's site.) Lossless algorithms mean that you can reduce the size of a file without loosing any quality, as the name suggests, but I believe that they're limited to reducing size by around half.

I'd hope that the Blackmagic Cinema Camera is applying something, because otherwise it's going to be generating around 14 MB for a single 2.5K frame. And I'm not sure how many SSDs can sustain a 430 MB/sec data transfer rate.
 
Brian@202020 said:
Nobody is talking about it because it's kind of a joke. Just because it shoots 4K doesn't mean it's a good quality 4K. Think of it this way, remember when people preferred the HVX200 at 720p over the Sony HDV cameras at 1080? It's because of compression. The HVX had rectangle pixels which was 720x960, but it had a lot less compression leading to a better image. For 4K to have the same compression as 1080 the bitrate would need to be 4-5 times the 1080 bitrate since there is 4-5 times the pixels. Motion JPEG and DVCPROHD are comparable codecs, so a 30p 4K Motion JPEG image would need to be roughly 400-500 Mbps to equal the compression of 30p 1080 DVCPROHD.
Click to expand...
It is 500 Mbps!

I'm sorry, but I am lost if you had clue on it ;) What's your point then?
 
StMad said:
You've got me curious with this comment - I've never had to deal with raw video, just raw stills in photoshop. The image I capture starts at a place I'm comfortable with, then I get to play with the extra stops and colour depth that the format provides. I don't pretend to have a strong knowledge on the subject though - can you expand on your comment?

Apologies for going ott.
Click to expand...

Sure. I was weaned on Photoshop 2 (circa 1992) In photoshop 3 it went to 32Bit and RAW files started appearing that could be clinically nuanced (manipulated) in channel curves using CIELAB (AKA 'LAB') color space . RAW files are the UN-manipulated output of the sensor before ay compression or data manipulation has occured. These files generally look like mud. So we must either deal with them in camera through manipulation of the data compression+optional color and edge engancement OR we manipulate them in post so that the resulting image matches a predetermined standard for display or printing. That's the primary objective of color grading or still image manipulation of RAW files. The controls and monitoring of these activities is complex, requiring a combination of eye-balling the actual image on rather expensive reference monitors as well as metric based control using software tools that bring the RAW files into compliance with a predetermined standard.
This is how professionals deliver consistent and COMPLIANT results from RAW uncompressed data. All professional print as well as broadcast and cinema operators have the ability to metrically do this.
Now. If you're an 'artist' and have a desired result you wish to achieve, the same color grading and manipulation tools that the pros use are available to you . If you are experienced, you can do the color grading without a color monitor even- using vectorscopes and digital numeric readout to control the channels (not RGB... but LAB). To understand what I mean, Open a RAW Photoshop 32bit file from camera into to LAB. Now start making color adjustments. You are no longer adjusting the rather intuitive RGB color wheels, you are now in the netherworld of LAB. It's the same in Video manipulation of RAW video files- to get what you want in a REPEATABLE fashion, you have to know the science behind the color space and how to manipulated to a predetermined result. Or.... spend thousands of dollars on a calibrated reference monitor and try to do it visually.
It;s had work color grading this stuff. That's why a cadre of specialists are busy full time doing nothing but that
So where does that put the budding and cash rich Independent Film Maker in reference to RAW acquisition?
Depends. If they know their software and have been trained or self trained to do it professionally, then great. RAW is of great value to them.
But what if they aren't certified colorists? Then there are three paths.
1) GET THE TRAINING - Self or Professionally instructed
2) Hire a colorist
3) Allow the on camera DSP to spit out a file that uses canned manipulations (White Balance is the most common). In theory, a WB manipulated RAW output has just had it's first manipulation done.

Sorry for the long post to say. "If you know what to do with RAW files, then great, you deserve to have them. If you don't know what to do with RAW, get the training or hire a pro.

The 1DC is really targeting professional outfits that have deep enough pockets to afford a DSLR made especially for them.

UPDATE. I just read the Adobe blurb. It says NOTHING about reducing the RAW file size in fact, it adds less than 1kb of data on top of RAW. It simply reminds us what we've always known, that RAW is much smaller than heavily tagged formats such as TIFF for example
 
Last edited:
You must log in or register to reply here.
Back
Top