Resolution concerns, 2K 4K?

[Karl H]

It may be too early in the day for this, so dont flame me, it's merely observation.

When viewing the few 4K tiff's that have been posted, I couldnt help noticing that at 100% they look, well, soft compared to a 100% crop from my DSLR (canon350D). The 350D is only an 8Mpixel camera yet it seems to me to have a lot more resolution than the stills coming from Red at a higher resolution.

So I played a bit more. I put up a 100% crop from one of the downloaded 4K tiffs, then I took the same image and scaled it down to 2K (1/4 size) - then upscaled it again back to 4K. In theory this should now be a lot softer. Looking at the attached image I cannot see much difference (if any) in the upscaled 2K compared to the 4K original.

http://darklineonline.co.uk/resolution.jpg

Now Im wondering what the reason for this is. Has Red Alert done some kind of upscale? If not, and this is indeed a typical result from the sensor is there any advantage at all to shooting 4K over 2K? Personally for me, I would shoot 2K and save a ton of bandwidth.

My DSLR is only 8Mpixel, I forget but I know Red is around 11/12 Mpixels. But so far the images coming from my DSLR seem sharper natively, that just seems wrong to me?

Before I get slammed, I am generally curious as to why this is, am I missing something? I've been following the board for years and I have reservation 394 myself. I am really hoping there is something I've overlooked or Red Alert has introduced some kind of downscale on export. I'm not looking to stir up trouble, I think what Red has achieved is fantastic. But if I can save space shooting 2K scaled, and it has very little quality difference to shooting 4K, well that's worth talking about (tho it might be premature)

I'm sure I'll be finishing in 2K anyway, and I'm happy with the image quality at 2K. But at 4K, im not convinced yet. Hopefully more 4K footage will change my mind.

[Miltos Pilalitos]

darkline you are right in your observations but i suggest you to wait until some resolution charts are shot in more controlled environment.

[Greg M]

remember this is a motion camera not a still camera.

[Graeme Nattress]

Um... the 4k would have to be pin sharp focus for you to properly perform that test.

When I look at 4k on proper charts, I see significantly more fine detail than I on either 2k export (from 4k) or 4k downsampled to 2k and brought back up to 4k.

Sure DSLR images are sharper - there's a bit of software in their decode that adds "sharpness" even if you set sharpness to 0 in the raw converter (usually....).

We also have a detail control in the bayer demosaic, so you can extract more or less detail from the 4k. Useful for shooting pretty women is the "leading lady" mode that doesn't try to extract much detail at all. Keeps them looking pretty.

Hope that explains what you're seeing.

Graeme

[Álex Montoya]

I was wondering the same as well.

That didn't happen in the milkgirls pics where the difference is visible enough.
http://img230.imageshack.us/img230/6966/comparisonxy0.jpg

Don't know, maybe it's the lens that's a wee soft in that particular f-stop.

I am almost certain that Jim's pics at the races would show more resolution at 4K.

[Karl H]

oh no doubt the Red Camera is the cream of motion HD digital cameras available, but if (and its a big if), there is not a huge difference in shooting 2K scaled over 4K, then it's worth talking about I think.

That said, should it matter if this is motion picture or digital? thats partially why I'm asking. As far as I know the imaging device is the same technology (CMOS bayer pattern) as used in digital SLRs. So it seemed an obvious thing to comare it to that technology.

True, resolution charts will clear up the whole thing, my post could be premature.....

[Michael Schrengohst]

That frame grab is just that, Evin "grabbed" that shot and I would not
spend time doing a critical analysis.

[Karl H]

thanks for the explaination Graeme. Perhaps this isnt the best image to perform the test on.

[Drew Mylrea]

Also, as Greame alluded to, that frame grab isn't in focus at all. Perhaps try the eyes pic as a better comparison.

[Álex Montoya]

I've tried the eyes pic and it's the same as in the girl's one.

[Karl H]

Hi Concrete,

do you mean the milk girls or the original girl pic I posted?

[Álex Montoya]

The one you posted.

[Graeme Nattress]

The eyes shot is not pin-sharp focus either....

[Jack Wester]

At these resolution, good lenses and good focus is everything. If the "antialiasing pixels" on the edges of a detail are more than a few pixels wide, higher resolution is not going to add much.
As Jim said in a post some month ago, one of the thing to watch out for when moving to Red is going to be keeping stuff in sharp focus. Concretes milk girl upscaling/downscaling compared to sample in original post of this thread is a good example of that.

I'm sure that the milk girls shirt could be even sharper if it was the exact focus target.

[Mike Zinner]

I also was worried there for a minute. But simply take a look at this frame grab, which is sharp as hell:
http://www.reduser.net/forum/showpost.php?p=76457&postcount=1

So my conclusion is that non of the footage posted in this sub-forum yet is actually in 100% focus, nor does it have the lighting required.

Actually I know this from my Nikon D50. If I shoot inside or after sundown with random lighting and manually try to focus with f1,2 to 5,6 - I get the exactly the same pictures as the one above.

You need to proper light every shot and hit the focus 100%, otherwise you'd better scale down to 2K to make it look good.

This is what now separates the pros from the indy shooters. And we have to realize that it is not as easy to just pointing the RED somewhere and get stunning, award winning images.

But don't worry. Just watched Transformers the other day - and there are a few shots that are *absolutely* out of focus as well - so much that I went "WTF is that!?!". And don't get me started about the grain...

Even the pros don't get it right all the time on a multi-million dollar production. So why should we expect that from the first demo shots by our first 25 heroes in the field?

Just my 0.02€

[G.A. Kokes]

At these resolution, good lenses and good focus is everything.

Even at 720 and 1080. We use focus pullers on Varicam through 35mm shoots. There are certain productions, such as those not requiring the camera and subject to be in motion where you can get away without a focus puller.

As for these examples, they look fine to me. In order to get a "lab test" quality report on resolving power though, they need to be tack sharp in focus as noted above.

Also, more needs to be known about the lens and its characteristics not to mention lens and camera setting etc. for each shot examined.

I am sure in due time we'll get all that... Right Gibby???

Cheers,
G

[Álex Montoya]

Mike, the frame you posted is 2K.

Anyway, as I have aleady stated, I believe that Jim's footage at 4K will be astonishingly clear.

[Karl H]

well we've had very few 4K grabs to go on, everything at 2K looks great as you'd expect.... im also wondering how far the shot of the couple indoors was 'pushed' in Red Alert. I know if you try and bring the shadow detail up from a real low part of the histogram then the sharpness of that part of the image will be compromised.

[Graeme Nattress]

Of course, what I forgot to add:

to get a 2k image with enough information in it, so that when upsampled to 4k it still looks good is not easy. Take an image that comes from a 2k sensor (and I don't care if it's a bayer or a 3 chip) and that necessarily has to have an optical low pass filter to stop aliassing. You don't stop all the aliassing, just enough to make a really nice image (which is what we do at RED). Now that 2k image has about 1.5k measured resolution. Scale that up to 4k and it will look scaled up. On the other hand start with a RED 4k image of over 3k measured resolution (75% to 80% of pixel count resolution, as we've always said for a bayer camera with an good optical low pass filter) and scale that down to 2k. In that 2k, you should get close to 2k resolution with a proper downsampling filter, not the close to 1.5k you'd get with a proper optically low pass filtered image from a 2k camera.

You see, it takes a 4k camera to get a properly anti-aliassed image at 2k. To get a 2k camera to measure 2k, you'd have to omit that filter and that can produce nasty results.

Graeme

[Salem Kapic]

Hm, what about lenses?

probably, Jim use Cooke or Zeiss instead RED cine lenses!

[Álex Montoya]

Stu of The Orphanage already talked about this in his blog. You need a 4k camera to get the perfect 2k image.

And no, Salemas: Jim has stated that he used the RED lenses.

[Karl H]

I dont think the lens is a huge factor here as you might think.

My DSLR has a lower resolution CMOS chip and also a $50 lens on it (a 50mm prime). In my opinion so far, it produces clearer/cleaner 4K images than I have seen from the Red camera with any lens on it. As we've said though, this could be more to do with focussing and/or Red Alert settings issue.

edit : exception.. the original milk girls pic was very sharp and clean too!

[Graeme Nattress]

The lower the resolution of the CMOS chip, the larger the photosites and the less sharp a lens they need :-)

And as I pointed out, there are vast differences in how the RAWs are developed.... And that we use a OLPF designed to stop aliassing, which is totally unnaceptable in a motion picture, but you "get away with" in stills a little more as you can go and paint out any baddies, which would be a pain in motion.

Graeme

[Salem Kapic]

Please compare Tonaci image eyes.tiff(4k) and old test frame of girl who blow balloon 4K! (Very similar light condition)

Every image from Tonaci test team is "not in focus" like old Jim's images

[Jack Wester]

You see, it takes a 4k camera to get a properly anti-aliassed image at 2k.
Why don't you make more of your posts sticky? You pass a lot of wisdom on this board and it would be sad if it fell into digital oblivium.

In a great deal of the discussions on reduser in the last month (especially in the 2K workflow "wars") this very important fact was often forgotten. Getting the "optimal pixel values" for your output is vital. Otherwise, your output doesn't look as good as your screen/projector has the capability of doing. Any professional graphics designer knows this only to good, but I would guess that a cinematographer in an analogue world, not working at the pixel level on a daily basis, could easily forget the magic of "pixel perfect" images as opposed to just having high resolution.

To me, a megapixel is not publishable as a megapixel if it hasn't been properly downsampled to give each pixel a really good average. Only in a world when sensors are perfect and noice is non existing does your sensor resolution give proper respect to the output resolution.

[Joel Kaye]

That said, should it matter if this is motion picture or digital?

You posted a handheld shot at 1/48th of a second for one thing. Most DSLR's sharpen in camera for another another thing. Most DSLR's force a shutter speed higher than the lens focal length to help freeze camera motion. RED can't do that. Camera motion will create the illusion of "soft".

Without a controlled environment test your questions can't be answered.

[Graeme Nattress]

A megapixel is purely a measure of the number of photosites that are used to create the image.

As you downsample you push more an more quality and per-pixel resolution into the image - it's like distilling malt - you end up with delicious single malt whisky, but an awful lot less liquid than you started with.

Even if sensors were perfect, you'd still need optical low pass filtering to avoid aliassing. And downsampling cannot remove aliassing that gets into a system at the start because aliassing is indistinguishable from real detail (other than looking "wrong").

Graeme

[Karl H]

I think my Canon lenses will be good enough to resolve the full 4K (and I'll be buying more expensive ones too when my Red arrives).

I have to admit, I dont understand the whole process of producing an image from a bayer sensor, I'm only going on what I'm seeing with my eyes.

I know 4K sensor does not equate to a 4K image as RGB needs to be derived from each pixel. Hence my 350D with an image size of 3456x2304 should in theory give out less than a 3.5K image, maybe more like 2K on a resolution chart.

Im still hoping that with mysterium and the correct settings I should be able to get images that are sharper and clearer than what I get from my stills camera. I'm just not seeing them yet, which as discussed could be down to many reasons.

Graeme, is it possible to play with the sharpness during the demoisaic process in RedCine? I, like you, hate artificial sharpening on cheapo video, but whatever canon use it seems smooth but hightly detailed. nothing about it looks artificial.

[Karl H]

>>>You posted a handheld shot at 1/48th of a second for one thing.


good point, perhaps some wide shots outdoors at 1/250th will give us a better idea

[Matt Uhry]

The lower the resolution of the CMOS chip, the larger the photosites and the less sharp a lens they need :-)

Graeme

What would be a good target of LP/MM as seen on a lens projector for getting 4k out of a Mysterium?


Matt Uhry
www.mattuhry.com

[Jack Wester]

I guess I'm saying the obvious, but an unsharp edge of 20 pixels in a micro perspective is just a gradient between two sets of values. And downsampling and upsampling a gradient would not cause great damage to the image.
Depending on the upsampling method a perfect gradient would actually upscale to a mathematically indentical image.

Simply zoom down to the sharpest and most focused detail on the image and count the number of pixels forming the edge to see of it's sharp enough for the kind of experiment discussed in this thread. The image at hand in this thread is far from a good candidate. The Dickies logotype from the milk girls is a much better subject. I'm sure that with optimal focus and lighting conditions, even bigger differences would be evident.

[Karl H]

The image at hand in this thread is far from a good candidate.

I know, dont keep reminding me:) but it has brought up an interesting discussion!

One thing I never thought of was the amount of sharpening thats acceptable in a still image. Of course when that's moving you might get some sort of 'shimmering' effect, and so Red has to use much less sharpening processes than a still camera.

To me it doesnt look like sharpening in my still images as im processing these also from RAW files; it looks like more resolution. Like everyone here I WANT to see something that will dissolve my concerns. Maybe some sharp 4K images in the coming weeks will shut me up!

In all honestly I have some really bad DSLR pics that I could compare this too aswell, so it isnt fair perhaps to compare this handheld 1/48th second indoor shot (possibly pushed in post too).. to a 1/400th second outdoor crystal sharp image:sarcasm:

[Jack Wester]

http://img230.imageshack.us/img230/6966/comparisonxy0.jpg
The difference between 2K and 4K is actually bigger than what this example shows. The 2K on the left was downsampled from 4K before it was upsampled. So the 2K image shown have better pixel averages and less noise (even if noise under this lighting conditions would be very low) than what would be the case if it was captured at 2K by the sensor.

EDIT - The conclusion is this. We REALLY benefit from 4K even if we're producing 1080p.

[Karl H]

I see what you're saying. You need a 4K sensor to derive a 2K image. If Red had a 2K sensor then the image would never have been close to 2K in real world resolution.....but as it relates to recording 4K, are we saying the resolution difference between the two is not THAT critical. Certainly not a visable 4x resolution difference.

If I understand Graeme corectly, we're probably geting a real world resolution from Red's 4K bayer of around 3K? If thats the case, why dont we have a 3K recording option and save some bandwidth? ;-) or alternatively turn up the bitrate so we get better quality compression at 3K.

[Graeme Nattress]

Darkline, yes you can play with both detail extraction and "sharpness". You'd probably measure around 2.5k on your camera, depending on the strength of the OLPF and how the detail extraction is done on the bayer pattern.

Remember, a lot of these images we're looking at are at a slow shutter, and wide open - hardly a recipe for sharpness.

I've not found any lenses I've used to be limiting resolution yet, Matt.

[Graeme Nattress]

Darkline, if only it were so simple.....

The measured resolution from the 4k is over 3k, but say we downsampled it to 3k, we'd now need to be in RGB, not raw, and therefore the file size just got bigger, not smaller. Downsample to 2k, and you've got a 2k full of nearly 2k resolution, but the file size is 3/4 of that of the 4k and you've lost 1k of resolution for that. That's why 4k is best recorded as 4k.

Graeme

[Jack Wester]

I see what you're saying. You need a 4K sensor to derive a 2K image.
If a pixel recorded by the sensor contains an edge within its boundaries and the sensor would report a pixel average that properly takes and average of all things going on within that pixel boundary, you could argue that 2K would give you 2K, but in the real world, this does not happen. A sensor would never give you such a good pixel.

If you, however have 4 pixels to begin with, or even better 16 pixels, your downsampling would convert those pixels to a single pixel with a very representation of what was going on within that small rectangle of light by calculating the average of those 4 or 16 pixels.

If the optics and focus would not be the bottleneck, a 4K image would look better on an 8K or even a 16K sensor.

If I'm allowed to dream and guess, I'm sure that even if projectors does not go beyond 4K, Red would create for us an 8K sensor in the distant future, To go beyond that would probably mean manual focus combined with automatic finetuning to squeeze out that last bit of focus. That would look better in 4K than 4K looks today.

[Graeme Nattress]

Jack, downsampling reduces noise, but what you're not talking about is aliassing. If you put a > 4k resolution into a 4k sensor, you get aliassing, and that's nasty and not removable by downsampling. That's why we only get >3k out of the 4k as we're low pass filtering the image. I wish you could get "brick wall" optical low pass filters, but you can't, so we have to start filtering early to stop bad aliassing. That's the way of life for the moment for all sensors.

But, when you downsample, you can use digital filters with a very defined bandwidth that can push a lot more resolution into the downsampled image than what you'd get with optical filtering. And you get lower noise (as you have larger effective pixels).

Graeme

[Jim Arthurs]

I'd sure not be too concerned at this point... look to the example image from JJ to get a glimpse of what sharpness can be with this camera... while we'd love to see the full 4K of it, I've no doubt there are great advantages to recording 4K no matter what the actual resolution... it's certainly in the 3K or greater range and that's just fine... gives you all the VFX and re-framing wiggle room you need for a stunning 2K final.

So far, the test team images have just proven how spot on you have to be with critical focus with this camera... I predict a great deal of folks that were jonesing for an eyelash wide sliver of depth of field will be very soon learning to love f11-f16 work simply to protect their hineys on jobs!

BTW, we've covered most of this in a very old thread, where Graeme details (no pun intended) all the benefits of the 4K original solution, then we go rounds on how we want to work for finals... I argue the very point of this thread... that 2K from 4K is the sweet spot of this camera, but it's a manly full-bodied 2K dripping with saturated resolution... not the whimpy 2K from current cameras!

http://www.reduser.net/forum/showthread.php?p=1365#post1365

Besides, I have to remind... if it wasn't for RED we'd be hanging on every detail of what ever incremental feature upgrade Panasonic would be delivering for whatever replaces the HVX or the like from Sony... here we're just arguing the degree of superiority over all other potential solutions at ANY pricepoint.

[Evan Owen]

Like several have already mentioned, focus is critical at 4K. I've noticed that quite a few of the shots so far have been a little out of focus. Once Graeme's focus assist is operational, I'm sure that will be minimized. Until then, we'll probably continue to see soft images for awhile.

I found one of the sharpest images so far on www.fxguide.com (article here) (http://www.fxguide.com/article446.html). Scaling to 2K and back up reduced the detail substantially. If you're careful with the focus, 4K is worth it.

[Álex Montoya]

Yeah, the blond girl's one is the most beatiful pic I've seen from the RED.

You may see the resolution difference better here: http://img67.imageshack.us/img67/7862/comparison2yy5.jpg

Though this picture (http://www.fxguide.com/modules/NewsUpload/files/07Sep/RedShootOne/INT_Window_Cooke14mm_tk1Still.jpg) kinda preoccupies me. It's not the blown-out window but the harshness of the transition that I don't like.

[Justin O'Neill]

I think the issues you are seeing with the footage from the Blair, Tonaci, Evin, Brook crew, are simply due to the footage not being 100% in focus.

This is entirely understandable when you consider their situation. Blair's RED LCD and EVF were not ready yet so they were using Evin's Panasonic monitor for focusing. Brook mentioned they were zooming in at not quite 1:1 and judging focus with their eyes on a 8" screen. Also Gaeme's magic pixies aren't enabled yet.

They were out in the field sharing their learning experience with us in realtime. I would be surprised if the images coming back were 100% technically correct.

[Jack Wester]

Graeme,

I agree, but resampling 4x4 pixels to a single pixel would effectivly and beautifully get rid of any aliasing as well as reduce some noise, right? So when can we have a 16K red camera? :-) Just kidding. IMHO, I don't think we need more resolution at all. My problem would be being out of focus all the time. I would never reach the limits of this camera without some serious help with fine tuning focus.

In your opinion, given a really good dialing in on the focus, how important is the lense to get really sharp details for the resolution of your sensor?

[Michel Hafner]

And as I pointed out, there are vast differences in how the RAWs are developed.... And that we use a OLPF designed to stop aliassing, which is totally unnaceptable in a motion picture,
Graeme
I agree, but digital intermediates of past and present often have aliasing nonetheless. :angry03:

[Graeme Nattress]

No, resampling does not get rid of, and cannot get rid of aliassing. Once aliassing is in the image, it's practically impossible to get rid of and downsampling does not get rid of it. Correct downsampling can only stop aliassing occuring - it cannot remove it.

Getting the lens sharp is not perhaps as important as getting the MTF high in the lower frequencies which govern overall image contrast. This means good glass with good filtering. Maximum resolution, for me, is not the overall lens requirement - "look" is.

Graeme

[Graeme Nattress]

I'm a founder member of:

1) the "Aliassing is bad" club

2) the "linear means a straight line" club

Graeme

[Deanan]

There are a few unknowns for the first images that make it hard to judge
the image for sharpness like...
What is the sweet spot of the lens vs what aperture it was shot at.
How steady was the camera or subject?

It's also important to note that the RED one was designed to produce beautiful images that are practical to acquire rather than hyper sharp images with great looking specs.

A 4k camera designed to produce hyper sharp images will have endless problems with aliasing and will make it a giant headache to pull focus because the perceived depth of field falloff is much steeper.

Deanan

[Graeme Nattress]

Not least to add that edge sharpness contributes towards the perception of motion judder - something you don't want to amplify when shooting at 24p or 25p.

Graeme

[Karl H]

Darkline, if only it were so simple.....

The measured resolution from the 4k is over 3k, but say we downsampled it to 3k, we'd now need to be in RGB, not raw, and therefore the file size just got bigger, not smaller. Downsample to 2k, and you've got a 2k full of nearly 2k resolution, but the file size is 3/4 of that of the 4k and you've lost 1k of resolution for that. That's why 4k is best recorded as 4k.

Graeme

gotcha.... but didnt Jim mention that you had found a way to do scaled 2K raw? So maybe scaled 3K raw too?

[Deanan]

I'm with Graeme.

Aliasing is true evil. Even if you took a 4k image with aliasing
and downsampled it to 2k, the aliasing would still be just as offensive.

DI's I've seen that have aliasing are generally because of shortcuts in
the workflow and excessive sharpening prior to filmout.

Deanan

[Lauri Kettunen]

If you put a > 4k resolution into a 4k sensor, you get aliassing, and that's nasty and not removable by downsampling. That's why we only get >3k out of the 4k as we're low pass filtering the image.

Graeme, would you agree with me that in the end of the day this, let's say, "lost of data" becomes a competetive edge of the RED sensor? I mean, since you anyhow have to add the low pass filter, that's also a reason why the r,g and b sensor pixels need not to be above each other, but instead it's enough that they are side by side.

This also explained why you can save so much of the original information in the deBayer process. It's a sort of optimal compromize making a very good use of the fact that the aliassing filter has to be there. In terms of the discussion we had in the other thread, the filter guarantees the data is "somewhat regular" or should I say "smooth" as it smoothens just a tiny but necessary amount of any kind of "fractal details".

Quite many on this forum also have wondered about the lacking noise. Graeme, to which extent the wavelet compression has an effect to this? Is it another place where the necessity is found as an advantage by the users? In my eyes the aliassing filter, the sensor, and the compression all work together in a saddle balance such that the necessities do not end up as being compromizes of quality. Finaly, it's almost like a jackpot that one wins also with the lenses; We all know how expensive it is to design high quality lenses to HD 3 CCD systems.

Still don't understand, how come the big manufacturers missed all this and "got surprised the trouses on their ankles" as we use to say in Finnish?

[Graeme Nattress]

Actually, to my eyes, aliassing gets more offensive on a downsample!

Graeme

[Graeme Nattress]

Engineering is always a series of compromises. I think we've taken those that lead to a very good, high resolution, low file size image.

Graeme

[Evan Owen]

I'm with Graeme.

Aliasing is true evil. Even if you took a 4k image with aliasing
and downsampled it to 2k, the aliasing would still be just as offensive.

DI's I've seen that have aliasing are generally because of shortcuts in
the workflow and excessive sharpening prior to filmout.

Deanan

Deanan (or Graeme), most RED footage that I've seen so far has been great in this regard–just very, very smooth. But going back to what Concrete mentioned, I too have noticed that areas with blown-out highlights often have very hard edges, even to the point of being badly aliased.

Is this something that will be worked out in the future?

[Graeme Nattress]

I think the output sharpening was turned on for that shot, and perhaps it should not have been.... Without access to the raw file, I cannot guess further. I'm working on "better" "nicer" output sharpening for those who need it though.

Graeme

[Jack Wester]

No, resampling does not get rid of, and cannot get rid of aliassing.I feel the knowledge and experience I had about aliasing is flying out the windows :-) And I'm a 100% sure that you're right, as it would be presumptuous and arrogant for me to believe otherwise. So this simply means that I've got things to learn.

From my days as a graphics programmer drawing vector graphics, the predominant aliasing effect was getting jaggy edges and anti aliasing ment getting rid of those jaggy edges. That kind of aliasing would definitelly go away when outputing to a higher resolution and then downsizing using averages of 4x4 pixels to a resulting in a single pixel. Expensive but effective. But as aliasing is a concept valid in signal processing in general, my view on aliasing does not seem to be valid to the world of sensors and sampling, so I'll just shut up and listen on that matter.
:poster_stupid:

EDIT - I Found this definition on Wikipedia for the kind of simplistic antialiasing I often used in my programming. Not to prove you wrong, because I honestly don't think I really understand the aliasing problem you're talking about and that I'm at fault here

"...anti-aliasing by supersampling usually means that each full frame is rendered at double (2x) or quadruple (4x) the display resolution, and then down-sampled to match the display resolution..."

[Evan Owen]

I think the output sharpening was turned on for that shot, and perhaps it should not have been.... Without access to the raw file, I cannot guess further. I'm working on "better" "nicer" output sharpening for those who need it though.

Graeme

That might be the issue there, but the rest of the image doesn't look sharpened at all, to my eye anyway. It looks silky smooth – except for the edges of the blown-out highlights.

Thanks for your work on the sharpening algorithm though, I'm one who likes careful (or no) sharpening. :biggrin:

[Evin Grant]

The pudding...
http://www.reduser.net/evin/4k2k.jpg
This 100% 4K grab was sharpened using smart sharpened in photoshop, then the image below was down sacaled to 2K and then upscaled back to 4K.
I think the result speaks for it's self. Look especially at the fine detail in the shirt stiching. That being said movement is a big factor, anything moving will not achieve this level of detail, but it really does'nt have to.

Tiff
http://www.reduser.net/evin/4k2k.tif

[Karl H]

Yup I can see that clearly.

what ISO and shutter speed was this taken with Evin? there is much less noise than in the indoor shot

[Graeme Nattress]

Yup, that's all you can do for drawing digital imagery - plot at a much higher resolution and average, or better still calculate the percentage of the line passing through a pixel and plot it that brightness.

However because aliasses are frequency mirrored artifacts, they fold back into the data in such a way that they can't be removed by downsampling techniques. For downsampling, basically you low pass filter the data, and then discard any data not needed to create the new downsampled image. So for 4k to 2k, you'd filter with a filter that removes any frequency above that which 2k can sustain, then discard every other sample. Because aliasses fold back into the data, the contaminate the full frequency spectrum, so no filter bases approach can remove them without removing real information also.

Graeme

[Brook Willard]

Darkline, that was 320 with a 4 stop ND at 2.8 or 4 on the 300mm.

Remember, folks... we focused everything *by eye* using a 720p LCD. We never used a tape measure.

As for the harshness of the transition from inside to outside on the "HDR" shot, outside was unreasonably brighter. I don't want to give you a figure [A MILLION STOPS!] because I don't know. But what you're seeing there is the difference between being in shadow indoors with no lights on to being in direct specular sunlight in cloud-free 95 degree weather in CA.

[Kholi Hicks]

It just looks like soft-focus, to me. There've been countless amounts of times that I've gone to the theaters and caught slightly soft focus on the big screen, or even sitting and viewing on a large LCD at home; it's going to happen. And, it's not really soft focus, considering that you've got to blow that up to like four times it's size to see that. With moving images, you'll see it a lot less.

These aren't controlled environments, it's literally run-and-gun and all-for-fun.

S'all good, peoples. The best time to go hyper-analytical is when someone comes 'round with controlled and planned tests.

[Graeme Nattress]

Think of this the good way - if you shoot at 4k and don't nail it, you still are getting 2k, which is way better than what the Star Wars etc. were shot on :-)

Graeme

[Karl H]

Thanks Brook....I see, sounds like the HDR shot was pushed quite far in post, it must have been a few stops under as well as at a low shutter speed. Then people like me come along and treat it like it's a 'normal' exposed image... tut...

If you can focus that well with a 720p LCD then I think once the focus assist is working we may all find it easier than we imagined.

just to be clear, I'm as excited I think as anyone else on these boards, but like a kid with a new pet, he wants to disect it first. (or was that just me)

[Kholi Hicks]

Or when it goes to HD-DVD/Blu Ray/Standard Def, you'll have to look hard for it. As that's the most realistic distribution method for most projects.

[Brook Willard]

I still find it funny that people are pixel-for-pixel analyzing the stuff we've shot. On a controlled set with lights, direction, actors, marks and some care... any mysterious problems with dynamic range or noise suddenly disappear.

Of course, if I wasn't there... I'd be analyzing it.

[Kholi Hicks]

It'll happen. Everyone's stoked and everyone wants it to be perfect. In the land of reality, it's as perfect as you can get already and in the future it'll only get better.

I mean... look what it can do with available light? It's already cut the hyper-indie no-budget production budget by a few hundred dollars based on lighting costs alone.

[Brook Willard]

Well I wouldn't think that way. Removing lighting from a movie is removing cinematography. Thinking that this camera will let one abandon lighting is the *wrong* way to be thinking.

If anybody should be excited by what they've seen, it's the documentary crowd. The narrative folks already know how good it can look. :)

[Evan Owen]

It'll happen. Everyone's stoked and everyone wants it to be perfect. In the land of reality, it's as perfect as you can get already and in the future it'll only get better.
...

That's so true. I've started noticing my expectations have risen way too high since I started seeing RED footage. Actually, it's starting to bother me... haha.

Realistically, what RED is delivering, even at this early stage, is WAY more than enough for anything I'd ever need. :waaa:

[Jim Arthurs]

I still find it funny that people are pixel-for-pixel analyzing the stuff we've shot.

Think of it this way... it's not like your first few stills were the 4038th stills we'd seen released from PRODUCTION level RED's... they were the absolute first.

Up until that moment we had ONLY seen grabs from prototypes and hand built units and cautioned against speaking out on any issues because they were works in progress.

It was made more difficult because (for reasons I don't fully understand) everyone pressured you guys to start posting light starved sensor images right out of the gate...

[Kholi Hicks]

Well I wouldn't think that way. Removing lighting from a movie is removing cinematography. Thinking that this camera will let one abandon lighting is the *wrong* way to be thinking.

If anybody should be excited by what they've seen, it's the documentary crowd. The narrative folks already know how good it can look. :)


Not abandoning light entirely, it's more like it looks like that it needs less light than my HVX + m2 Adapter. =P edit here: Though probably wrong. Could it possibly eat light like an HVX + M2 setup?

Wouldn't ever ABANDON light, just saying that it could potentially produce a great image if the flexibility is needed on a tighter budget.

[Brook Willard]

Think of it this way... it's not like your first few stills were the 4038th stills we'd seen released from PRODUCTION level RED's... they were the absolute first.

Up until that moment we had ONLY seen grabs from prototypes and hand built units and cautioned against speaking out on any issues because they were works in progress.

It was made more difficult because (for reasons I don't fully understand) everyone pressured you guys to start posting light starved sensor images right out of the gate...

You do make good points.

I wish we had some lighting and somewhere to shoot. No light was allowed to leave the stage and no set was allowed to enter it... so we were fairly limited in what we could shoot. It was mostly an experience for us to learn the camera while posting stuff for you guys.

It looks like the other projects that shot this weekend had a little more budget [even if the budget was still zero], so some nicer imagery should be coming.

I've now seen what the camera can do in a "broken" environment. Now I'm excited to see what it can do on a set with equipment and a crew.

[Jim Arthurs]

I found one of the sharpest images so far on www.fxguide.com (article here) (http://www.fxguide.com/article446.html).

I just took a look at that article, Graeme, for gosh sake PLEASE disable that sharpening setting as default in the next build... it is killing these images when people don't know that the setting is there... look at those hyper-blown out windows and the edges of the sills... I know that's not what you're after... make people have to dig a bit if they want to turn that on... and go through a few alert boxes before they can adjust it... :)

[Graeme Nattress]

I'd mentioned we should have such a warning of the "do not turn this feature on" type thing in camera for things like sharpening.....

It's now defaulting to "off" for the next build.

Graeme

[Jack Wester]

Not abandoning light entirely, it's more like it looks like that it needs less light than my HVX + m2 Adapter. =P edit here: Though probably wrong. Could it possibly eat light like an HVX + M2 setup?

Wouldn't ever ABANDON light, just saying that it could potentially produce a great image if the flexibility is needed on a tighter budget.

Ditto that. Less light, but of course from the right directions, of the right kind and of the right color is less expensive than producing a 10000W for each shoot.

What are your thoughts on matching the RedOne with led lighting equipment. Does anyone have any comments on these (www.litepanels.com) (http://www.litepanels.com/)?

[Jack Wester]

I'd mentioned we should have such a warning of the "do not turn this feature on" type thing in camera for things like sharpening.....

It's now defaulting to "off" for the next build.

Graeme

Good move! :biggrin:

[Damien Molineaux]

Darkline, if only it were so simple.....

The measured resolution from the 4k is over 3k, but say we downsampled it to 3k, we'd now need to be in RGB, not raw, and therefore the file size just got bigger, not smaller. Downsample to 2k, and you've got a 2k full of nearly 2k resolution, but the file size is 3/4 of that of the 4k and you've lost 1k of resolution for that. That's why 4k is best recorded as 4k.

Graeme

Great thread, I've been through all the pages so far. Thank you for all that info, Graeme in particular. I just wanted to add one little thought to the above. Downsampling before recording (in camera), apart from not really saving space also locks in the image. The beauty of Red Raw, is that's it's raw ! Allowing you to tweak the image in post. It's not about making the choice later in the production process, images are nearly always tweaked in post, except usually you're tweaking an image (we're speaking about the digital world here) that's already been tweaked in camera.

To me, Red Raw is the key element of the Red revolution. Making a high quality image is one thing, but being able to handle this easily 24 4k frames per second... genius !

4K on a CF card, it hasn't sunk in yet !

THANK YOU
Damien

[Antoine Baumann]

Yeah great informations.

for the focus, if I am not mistaking Jim did use the panasonic LCD to focus the "engine" picture (http://www.reduser.net/forum/showpost.php?p=76457&postcount=1), so you might even take sharp focus shots with a 720p LCD, but need some practice.

I am very thanksfull to Blair's team for posting the pictures of their test, it is exciting all fo us.

antoine.

[Deanan]

Quite many on this forum also have wondered about the lacking noise. Graeme, to which extent the wavelet compression has an effect to this?

The low noise images are a result of a great starting point (sensor) and maintaining the low noise throughout the system.

Deanan

[Gavin Greenwalt]

Those FXGuide shots were really really painful to look at with the blown out window. Even at 1k . BLACK LINES! AGHHH!! What is this the 1950s? I'm just going to assume it was all sharpening for the sake of my sanity otherwise No white backgrounds ever! ;)

[Deanan]

From my days as a graphics programmer drawing vector graphics, the predominant aliasing effect was getting jaggy edges and anti aliasing ment getting rid of those jaggy edges. That kind of aliasing would definitelly go away when outputing to a higher resolution and then downsizing using averages of 4x4 pixels to a resulting in a single pixel. "

When we talk about aliasing we're talking about the full suite of
aliasing problems you can have. Edge aliasing (jaggies) is one effect
of aliasing. There's also nyquist related aliasing ( moire/sampling patterns caused by fine patterns like certain textiles) and chroma aliasing caused by nyquist issues reacting differently in each color channel (sometimes made worse by inadequate IR filtering).

The last one is the most offensive in motion and consequently downsampling doesn't help because the aliasing has created artificial patterns that spread beyond single pixel variations.


Deanan

[Graeme Nattress]

http://www.nattress.com/4k_Scaling_Test.jpg
Clearly shows the extra resolution at 4k.

Graeme

[Karl H]

oh yeh, clearly there is more resolution, but it doesn't look 4x the amount. My initial dispute was not whether you could tell the difference between 4K/2K as a rule, but whether you could with Red footage going through the same process. I was hoping there was some technical reason why it didn't look as high res as I thought.... and there was (focus, sharpening, exposure, F-stop, 3K real world resolution.... all contributing)...

My concerns have been addressed at least technically speaking :-)

I've yet to see images that match the clairty of the Milk girls footage. I'm hoping it will come in a few days. My guess is the Milk Girls footage was actually lit and therefore nothing was really pushed in post to reveal the noise. Some of the images posted so far are either at high ISO's, shot with low shutter speeds, or have been developed from a few stops under - therefore revealing more noise and compression. Thats not a complaint, it's exactly what I wanted to see. You dont learn a lot from an image exposed with a range of 4 stops due to great lighting.

In the coming weeks I think we'll see the best and worst Red has to offer.

[Graeme Nattress]

4k would never have 4x the resolution of 2k though - it would have at most double the linear resolution, and of course, 2k from a 2k sensor should not have 2k resolution either - now that would be the fair comparison - take a Silicon Image 2k shot and a RED 4k shot and see.

Graeme

[Karl H]

Im sure we would see a big difference from those comparisons.

AFAIK though this is a single chip CMOS problem only right? If you had a 3CCD system of 2K then you would in theory get a full resolution 2K image from that array?

BTW Graeme, thanks for answering all these question, its much appreciated.....

you never answered my 3K Scaled Raw question though :-)

[marco somaini]

we have a lot of people talking about resolution differences. Well, what I have noticed so far, 4K(4520 X 2540) on red is recorded at 72DPI, which is great, but if you compare that to a 11Mpx camera, it records 2731 x 4096 pixels but at 240DPI! that is a lot more information that gets recorded in a still image. So if you really want to compare a canon shot with a red image pulled of the web so for, you need to compare it to an about 1.5Mpx still camera! Just my observation and please red team, let me know and I am wrong on this. I love what yo have done and you guys, keep in mind that 1080 as been the biggest we could record so far on our budget, so cut them some slack, this is not a still camera, it is simply amazing digital video!

[Brook Willard]

...or you can just change the 72dpi file to 240dpi... the pixel count doesn't magically change with ppi resolution.

The 4K images posted so far are all 4096x2048.

[Eirik Tyrihjel]

Dpi has no meaning in video aquisition - it´s a measure of how detailed a printed or scanned image is - Dots per inch.

I can´t explain it better Marco, I am sure someone else can.

For video the only thing you need to look at is the number of pixels, DPI only has a meaning when you define the size of text in an editing application.

[Álex Montoya]

Marco needs an education

[Joel Kaye]

it records 2731 x 4096 pixels but at 240DPI! that

Unless I've lost my mind pixels are pixels. DPI is just a representation of how those pixels are going to be place on a device... usually some type of printer. So basically your whole post is wrong.

[Jack Wester]

it records 2731 x 4096 pixels but at 240DPI!
DPI or dots per inch is synonymous to pixels per inch. If you show Red footage on a one by one inch display its DPI well be several thousands.

DPI is a good measure on resolution for a output display or on a scanner. Given that the surface has the same size, higher DPI always means more pixels. But in this case it has no meaning. A pixel is a pixel. It can have accuracy, channels and range (bit depth or bit length), but other than that, its a pixel.

No worries though. I've posted more errors on this forum than anyone :whistling:

[GlennChan]

AFAIK though this is a single chip CMOS problem only right? If you had a 3CCD system of 2K then you would in theory get a full resolution 2K image from that array?
Not really.

To avoid aliasing, you need to apply optical low pass filtering on your 3 sensor camera. So this means that you won't get full 2K resolution.
Or, you could not apply OLPF. But then you'd have aliasing.


you never answered my 3K Scaled Raw question though :-)
Scaling RAW data would be weird. It probably doesn't work well?

[Jack Wester]

The last one is the most offensive in motion and consequently downsampling doesn't help because the aliasing has created artificial patterns that spread beyond single pixel variations.

Got it. Makes sense. Thanks!

[Floris Liesker]

Not really.

To avoid aliasing, you need to apply optical low pass filtering on your 3 sensor camera. So this means that you won't get full 2K resolution.


And you'd need a prism to feed the 3 CCD's which would introduce chromatic aberration and reduce your effective resolution as well.

[Damien Molineaux]

Im sure we would see a big difference from those comparisons.

AFAIK though this is a single chip CMOS problem only right? If you had a 3CCD system of 2K then you would in theory get a full resolution 2K image from that array?

BTW Graeme, thanks for answering all these question, its much appreciated.....

you never answered my 3K Scaled Raw question though :-)

Actually he did :



Darkline, if only it were so simple.....

The measured resolution from the 4k is over 3k, but say we downsampled it to 3k, we'd now need to be in RGB, not raw, and therefore the file size just got bigger, not smaller. Downsample to 2k, and you've got a 2k full of nearly 2k resolution, but the file size is 3/4 of that of the 4k and you've lost 1k of resolution for that. That's why 4k is best recorded as 4k.

Graeme


You obtain your real resolution once you're in RGB. You can't save a downsampled 4k raw to 3k raw and hope to get a real 3k RGB.

The force of Raw is just that "Raw data", what your sensor sees (ie. records), no downsampling, interpretation, etc.

May the force be with you,
Damien

[Karl H]

Yeh I was being a bit cheeky. Jim mentioned a while ago that Graeme had maybe figure out a way to deliver 2K Scaled Raw. I dont really know how it would be achieved as what you said makes sense.

But Jim mentioned it a few times that it was underdevelopment at 2K scaled, not windoowed, so I was prodding :-) and thought, if you can do 2K you should be able to do 3K

[Jannard]

We still have some amazing things coming (at no extra charge) for the RED ONE. These cameras are not full featured enabled (I'm sounding like a broken record). Some of the features you have heard about, some you haven't. Owning a RED ONE will be like having Christmas on a regular basis. Just about the time you think you have it down pat... poof! There is another surprise that shows up in your email account. :-)

Jim

[Graeme Nattress]

Thanks for your good answers there Glenn. I was flying to Amsterdam so thanks for giving the answers I would have given.

Graeme

[Brook Willard]

Owning a RED ONE will be like having Christmas on a regular basis.

Where's the :laugh: icon when you need it. :)