Graeme Nattress

Phil, you told me it was 1k last year at IBC, so you've improved the camera 250% in the last year. Thanks for your help! Graeme

Gavin, my guess is the level of optical low pass filtering the Dalsa is using. Best point both cameras at a zone plate and see what levels of aliassing you see. Graeme

Sure it's interesting, but only if you're you're working on the design and electronics of the camera. And it's not so much as guarding unexceptional things, but guarding "everything" in a certain area incase anything accidentally slips out that you or I might not think important, but someone else does.

You seem to delight in asking questions you know you won't get answered, so to delight you further, this one won't get answered either.

Yes it's CMOS. We've always said it's CMOS. It's also RED Proprietary.

Sigma cameras avoid using the necessary optical low pass filter. This makes their images "sharper" by allowing the sensor to alias and produce fake detail in the form of aliassing artifacts. This is probably because of the "low" resolution of the sensor and if they'd properly filtered it, it would be lower resolution still. Aliasses are bad enough in a still, but at least you could go in by hand and pixel paint them out if you so desire. Doing that on a motion picture would be terminal. There's no way to automatically detect an alias from real resolution, and even downsampling the image won't remove the aliasses, so you've got to avoid them getting into the the system in the first place. Graeme

If it's shot 1080i, 24p with 3:2 pulldown, then the pulldown should be removed before you edit the footage. Do not de-interlace! If you remove pulldown you get back to the 24p frames and all should look fine. Graeme

Of course, if you go over to CML and their camera comparison images, you can download dpx files from a number of cameras. If you do so, you'll see, and in no particular order, and this certainly isn't a complete list: vertical pattern with a 2 pixel repeat over most of the image, blacks crushed completely out of existance leaving small square blocks of noise compression artifacts, chroma fringing, strange glows, chroma aliassing, image channel offsets and halos on bright objects on dark backgrounds. Some of these are lens related and some camera related and some tape / compression related. Either way, with a few minutes work in photoshop (add a bit of saturation here and there) and you can put together a collection that makes you view the above image in an entirely different light.

And when we've finished tweaking, that's the kind of thing we can do. Graeme

Natasha and Boris measure a little over 11 stops. Graeme

Walter, you hit the nail on the head. The magic word is "gamma". This is a non-linearity in the signal that is corrected by a non-linearity in a CRT display (and faked in LCD) that allows for a greater dynamic range than 8 stops to be represented by 8bit data, among other things. I'd check out Poynton's Gamma FAQ for more info. www.poynton.com Graeme

If you play a DV deck out via component, you get the correct chroma smoothing that say, the apple codec does not, and you get a superb image that would be superior to that of the a BetaCamSP component feed to the same monitor. You will get a visibly sharper image, with lower noise. Graeme

You speak in riddles. 13.5Mhz is the luma bandwidth for all SD digital formats from DV to Digibeta. They all have identical luma resolution. That's your 4, in 4:2:2, the 2 being half of 13.5Mhz, or 6.75Mhz. DV halves the chroma bandwidth again, so it's 4:1:1. Properly smoothed, on a decent monitor, you won't really see that, but it does effect things when you start doing keys and other effects. Luma bandwidth / resolution is one factor. There are others, like signal to noise ratio, and if you don't take that into account then yes, you'll get the nonsense that S-VHS looks better than 1". Graeme

Because luma resolution is more important than chroma resolution. Without doing the extraction of the chroma and making it more visible, you couldn't eyeball a chroma difference between the three formats (when DV was correctly smoothed because of the way the codec works). However, the luma difference was instantly visible as a blurring of fine detail. No matter what you do to it, you're not going to magically get more resolution out of it than 400 or so lines, whereas DV and DigiBeta will always be giving you the full 550 or so, and that difference is quite visible. The full methodology used was given so you, if you want to, or have the time and money to spend doing so, can reproduce it - if you don't think that methodology was right, you have the option to perform and write up the experiment of your choice. Graeme

DVCAM is identical in image quality to DV, and both are superior to BetaSP, not to mention cheaper and easier to get into your computer. Graeme

Of course, there's a good reason why we didn't do component straight into a capture card..... The cards available at the time were all SDI only. The budget for the experiments was already pretty high, so we could only do what we had the gear for or could borrow the gear for. "The BetacamSP was brought in by multiple methods: buy using the Digital Betacam deck as a high quality component to SDI converter to allow the analogue video to be brought in via SDI uncompressed 4:2:2, and the BetacamSP was brought in via a component dub to DVCAM, and from there via SDI and Firewire." It's a long article, lots of tests. I probably should have done 5 articles out of it instead of 1. Graeme

Nope, it was played from a BetaSP deck, over component, to the Digibeta deck which did a passthrough YPbPr to SDI conversion, and out to the Mac capture card. And I didn't do it - a big post house in Toronto did it all for me as they were very interested in the results. I processed all the images and wrote up the results, and presented it to them all at a meeting of the Toronto FCP user group. Yes Walter - loosing battle. Somedays it's not worth getting out of bed. Not to worry though, I've got much better stuff to be working on. Graeme

NO!!!! Read the article again. BetaSP was converter from YPbPr to SDI via DigiBeta and then brought in 10bit uncompressed 4:2:2 and 8bit uncompressed 4:2:2 via SDI. Graeme

Not off topic, really..... Yes, the outputs are smoothed, but they're ananlogue, and hence they degrade the picture a little too. That's why SDI works so well, or smoothing or fixing in software. Graeme

Nope, we used the Digibeta's pass-through (no compression) YPbPr to SDI conversion. Graeme

Here's links to tests I performed, with full details of what was done and why. http://www.lafcpug.org/Tutorials/basic_chroma_sample.html A pristine 35mm source was used, and a high quality Digibeta telecine of it was also used as source material. You can clearly see that the DV copy is closer to the original than the BetaSP copy, especially in the luma, where SP is visibly blurred compared to the DV or Digibeta. Graeme

It amazes me to read such forum postings! BetaSP has 100 more lines than DV! 525 v 480! That's vertical resolution and defined by the format. Sure, NTSC has 525 lines but we know that only 486 are for active picture information. ALL NTSC formats have 486 lines, from VHS through to DigiBeta (although some, like DV have 6 of those lines as black, effectively). Resolution, in video, usually goes off the horizontal resolution, as all SD formats have the same vertical resolution. When tested BetaSP has about 3/4 of horizontal resolution of DV or DigiBeta. Chroma resolution is harder as it's generally sub-sampled. In Digital formats, it's on a number basis like DigiBeta is 4:2:2, and DV is 4:1:1 etc. BetaSP is not digital, so you've got to measure it. It's chroma resolution turns out to be not as good as Digibeta, but a little better than DV - somewhere in the middle. Overall, DV looks a bit better than BetaSP, and is obviously cheaper to get into your system. Graeme

REDCINE is cross platform, Mac (intel) and Windows PC.

" Scaled Bayer images have huge problems with sharp colour edges." Since when? Pro photographers have used bayer pattern DSLRs for years now without these so called issues stopping them superb images. I'm sure a lot of people here use high end DSLR's as part of their production work, again without issue. Phil you're still saying that 1k RGB = 4k Bayer which is utter nonesense. Surely you've got a digital camera with a bayer filter sensor so you can post up some example issues??

David, thanks for trying to be a voice of reason, but Phil came up to me at IBC and told me that the RED was really a 1k camera. On other forums we've had very reasonable discussions about Bayer CFA sensors, pros, cons, benefits, drawbacks, even resolution, but it's not easy to have such informed discussions here. Perhaps the only worse place to discuss such things is the Sigma forum at DPReview :-) As someone who's done some serious R&D into upscaling algorithms, I've got to say that such a comparison of a HD resolution 1920x1080 RGB image from an RGB sensor or 3 chip / prism system to a 4k Bayer pattern image would be most interesting. I'd be pretty confident in saying that the equivalent resolution of an optimal RGB image would have at least 3k as the horizontal dimension, and that as for which looked better, the native 4k or the 3k upscaled would be very dependent on the scaling algorithm chosen, and even then, as scaling always introduces artifacts, the 4k could still very well be preferable. Graeme