Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Dragon 6k not eligible for UHD "Tier 1" camera?

Graeme Nattress said:
Foveon is only an answer to an a problem that isn't. To decode data from a Foveon, you're into image processing that's actually slower and more complex than a demosaic
Click to expand...
Everything you said makes sense. But it seems that I have assumed wrongly as to how Foveon works. I always assumed that the formula for a Foveon sensor would be simply R+G+B. I guess it isn't that simple after all.

P.S. Do you have an opinion on quantum dot sensors? What about Sony's per-pixel single exposure HDR method, if you've heard of it?
 
I don't think Foveon went out of their way to market exactly how their sensors work with regards to image processing to extract a nice colour image. The problem is that silicon is a pretty poor colour filter and thus getting proper colour out is tricky. I think if you dig through some of the old Foveon white papers, you'll get a good idea of what they're up to and why it's complex. Similarly if you look at reviews of Foveon based cameras and look to where they see the failings, that will also tie directly in with the complexity of their image extraction process.

Graeme
 
Hello Graeme,

Did someone from the DPP contact you to make a resolution test to evaluate if the Dragon 6k sensors qualifies for UHD Tier1?

I think it could harm Redusers if they do this kind of assessment to the AFC without any empirical testing. This kind of talk is utmost unprofessional from DPP's part.

Pat
 
OK. I freely admit that this thread has now gone way over my head. So if someone would be so kind as to dumb things down I have a question.

I was under the impression that 4K was a larger format than UHD. I was also under the impression from current practices that for people like Netflix etc... Red 5k down to 4K meets or exceeds the spec need - 5.5 or 6K down to 4 K even better.

So when I read that some group has determined that Dragon 6K does not meet their UHD requirements I am lost. I thought it far exceeded that.

Some of what I am reading makes it sound like when Red 6K is debayered it actually falls below the UHD spec which was news to me. So please excuse my scientific ignorance and can someone just spell out what this all means as this does not sound good to me.
 
As far as I understand they are quoting opinion rather than referencing empirical fact. Even if a specific test is referenced, as I note above, it represents only the specific details of that test and thus what is achievable under those specific circumstances. I know from my testing, very small changes can have very different results in resolution. I think a number of the previous posters in this thread have measured resolution of various cameras including their Reds and get a very different result than the opinion in the article.

Graeme
 
I'm lost too Tarquin. The article just doesn't make sense. The Sony F65 has less pixels in a 16:9 aspect ratio image than a 6k Dragon sensor. The Arri pixels don't magically have more resolution than Red pixels.

Graeme
 
Graeme Nattress said:
I'm lost too Tarquin. The article just doesn't make sense. The Sony F65 has less pixels in a 16:9 aspect ratio image than a 6k Dragon sensor. The Arri pixels don't magically have more resolution than Red pixels.

Graeme
Click to expand...


Thanks Graeme as that makes me feel less stupid at least. This all seems like a kind of 16th century 'the world is flat' logic:) When I look at the Weapon 6K footage I am amazed - especially as it is in such a small box. So perhaps this is FIFA grade BS?
 
Does this group have any power since Netflix and Amazon and HULU have already deemed Red Dragon as acceptable for their streaming. And this was determined by their tech heads. So who is listening to them? Not the manufacturers since they are talking directly to the content providers like Netflix.
Hopefully someone can clarify.
 
Does this group have any power since Netflix and Amazon and HULU have already deemed Red Dragon as acceptable for their streaming. And this was determined by their tech heads. So who is listening to them? Not the manufacturers since they are talking directly to the content providers like Netflix.
Hopefully someone can clarify.
 
It's the test that failed

It's the test that failed

Thanks for the link Bérenger. Like some other similar tests, the fail point is measuring an image that has not been fully processed. The "default" rendering of most scenes is relatively flat with low saturation. That's a great starting point for grading and characterizing the image for the intended format - but not to evaluate the full capability of the camera. I firmly believe that any decent colorist could easily "develop" a nice clean RGB image from any reasonably well shot material that would easily eclipse the Tier 1 standard.

My analogy would be testing 40 yard dash times with the athletes in dress shoes and heavy coats.

Cheers - #19
 
Thanks for the link Bérenger,
Interestng to see that the Alexa (in SxS mode) doesn't qualify for an HD tier1 camera with it's 1440x800 resolution limit.

Pat
 
Peter Lyons Collister said:
Does this group have any power since Netflix and Amazon and HULU have already deemed Red Dragon as acceptable for their streaming. And this was determined by their tech heads. So who is listening to them? Not the manufacturers since they are talking directly to the content providers like Netflix.
Hopefully someone can clarify.
Click to expand...

It's unknown exactly where this data goes or who it truly effects, but I have seen EBU influence BBC productions in the past, and below will explain why.

I have no problem with the EBU's Tier definitions, that's all lovely. There's just drastic inconsistencies across the board that don't exactly match up to anybody's tests.

The big gripe I have here is Andy Quested, who is the Head of Technology for BBC HD & UHD and is also a SMPTE Fellow, in this interview references the Alexa 65 (6.5K - 20.3 megapixels) and the Sony F65 (5.7K, interpolated 8K - 17.7 megapixels) as cameras that fit the EBU's definition of Tier 1 and somehow references RED Epic Dragon (6K - 19.4 megapixels) as Tier 2.

The test itself that led to the results that he references also didn't use the Epic Dragon to capture at maximum resolution. Nor were adequate lenses used in the resolution chart test, which seems to have caused the Tier mishap.

The article is from April 3rd of this year referencing Alan's test from January 2015 and honestly this is the sort of testing that has led to misinformation and misleading results. The truth of the mater is especially in 2016 all of the bigger 3 digital cinema camera manufacturers top tier bodies need to be tested the same way, at the same time, in the same space to eliminate the peculiar irregularities that has led to Andy's quote.

The hard thing about various camera tests and papers is once they are published they become the "truth" for somebody somewhere. Unless the proper due diligence is applied often rather misleading concepts float to the top.

I'm nearly on tilt regarding a few camera tests I've seen as of late and I'm thinking of just locking it down with some new standard practices so we can get various entities all on the same page.
 
I think RED needs to go easy on the blur in their OLFPs The URSA Mini tests show that pretty clearly. To me it is the biggest culprit in the sharpness "wars"
I'm pretty sure it (the pixel level sharpness) was pivotal in the test-subject of this thread....IMO
 
AFAIK, most Bayer CFA motion cameras use relatively strong OLPFs to control aliasing. Since the impact of the OLPF is a known value, image processing schemes account for it's effect in constructing the base RGB image. RED takes a conservative approach to this "de-focus/re-focus inversion" piece as part of their "first do no harm" philosophy.

I still contend that with a high resolving lens, and the correct choices in the RAW development step (s-curve, contrast bump, sharper scaling filter, etc) one can construct a UHD RGB image from a 5K Dragon frame that would meet EBU technical standards. Using photo site rasters to make conclusions ignores the proven impact of Bayer decoding algorithms on final output resolution.

Cheers - #19
 
So you don't think the pixel softness we see in RED's cameras is from the OLPF? It does seem odd that the EBU paper on Epic Dragon says there is still aliasing, which is surprising considering how soft the debayered images are at the pixel level, something I always assumed was due to REDs OLPFs being much more aggressive than other manufacturers'. What else would it be?
 
You must log in or register to reply here.
Back
Top