dubzee

Hi folks, I just bought a set of Zeiss Contax lenses and am looking to unload my old EF glass. I'm located in Toronto, ON, local pickup or ship. I baby my glass and everything is in what eBay sellers would call 9+++ or 10 shape. HMU for more info. These are good deals I believe especially with UV filters :-) ALL OPTICS PERFECT. Anything wrong is only extremely minimal wear on the barrels. Canon EF 50mm f/1.4 w/UV filter - 95% like new - $275 CAD - wonderful normal FOV lens and shines with Dual ISO, it's so crisp and pleasing! -- SOLD Rokinon 85mm f/1.4 w/UV filter - Mint - $300 CAD - barely used it but it's a killer lens around f/2.8 for portraits and razor thin DOF! Aka Bower or Samsung. Tokina 16-28 f/2.8 Ultrawide angle zoom - 95% like new with 3.5 year warranty left - $650 CAD - needs no introduction. It takes the Canon and Nikon UWA zooms out back to put them out of their misery. BEST IN CLASS FOR THE COST :-) barely used too. -- SOLD Happy shooting and let me know if you're interested. -Ben

Great work!

Test #2 with two lenses and a different post workflow.

That's it exactly, playing devil's advocate. Keep on keepin' on, good discussions are hard to come by online.

I have to disagree that there's a resolution advantage to downscaling from 16:9 to 2.35:1. If it was an analog optical process - sure - but there are fixed quanta and downscaling could introduce aliasing because of it through dithering. Also remember the getting closer thing is abrogated by the fact that the 0.7x wide angle adapter gets me 14% closer. Really what we're doing is comparing apples and oranges. I never said this was a replacement, but just an experiment I was doing to see how it turns out, with some thought-provoking discourse in the process.

Ya that is a great vid! It'll be interesting to try my camera once they release the new debayer algorithms as part of the MLV extractors. As it stands it's kind of pixellated. And hey... 2k GBP vs $300 CDN ain't bad for "better than expected".

Not sure I understand... It's approximately 25% less pixels, not 14%.

Do you mean the vignetting specifically?

I just looked at those, looks awesome! Cheap too. My BUY finger is getting itchy. Thanks for watching, once again.

Remember, just in terms of the resolution is what I was saying. And based on the discussions, turning 16:9 into 2.35:1 even with a real anamorphic isn't a resolution advantage. 4:3 to 2.35:1 perhaps :) Cheers!

Avez-vous du vidéo comme example?

Not sure, best bet is to create an account - http://www.magiclantern.fm/

I'm just post-processing some of the footage I took. Unfortunately I had "Vignetting" control turned on in Magic Lantern so the colours are a bit weird, but it looks decent. Will post soon :)

They're working on ML V1.2.3 for 5D Mark 3 where you'll be able to preview both.

Yes it works on my Lilliput.

The only way you can select those higher resolutions is by going into Crop Mode. To do so, make sure Magic Zoom is off and then press the Zoom button on the left of the camera. That takes a 1:1 crop from the centre of the sensor albeit at 3x zoom.

YouTube disabled my sound cuz I used a copywritten song so I took it down. I'm off to High Park to see how this works out in the field. There's no belittling going on, just sound technical discussion. Perhaps you have a passion for arguing on the internet but I don't share it. I don't have to justify but I will, changing my user name to the user name I use on all bulletin boards and release dance music as. Shocking as it may be, I'm not doing any of this to mess with you. Just hate how my default user name was my Facebook name which is hidden from the world :-) Cheers

I guess we just have to wait and see... going to run a bunch of tests tomorrow and see how close I can get. At the end of the day it's all about experimentation and having fun.

We agree.

BTW, David Mullen ASC agrees with what I've posited: anamorphic lenses on a 16x9 sensor give no resolution advantage. http://www.cinematography.com/index.php?showtopic=62529

And I disagree with you. If you're recording in HD and then doing post at 2560x1080, you're stretching the image since it was captured at 1920x1080. Read another way: you're blowing up and the computer is interpolating what's in between. Like I said early in the post: yes there is more data, but it's still limited by quantization. I stuck a 0.7x wide-angle adapter on the front. It's got a 93mm front filter AKA it's huge. So I get the advantage of both horizontal AND vertical compression so that there's no need to further manipulate the image and dither/cause aliasing. Added bonus is my lenses take advantage of the shallow depth of field with a 1.42857x wider FOV. The footage I've captured on my 5D Mark 3 in MLV, with the Cinemorph, the WA adapter, and the Helios 58mm f/2 looks absolutely incredible. Once I've got it rendered I'll provide proof.

Other than to not have to crop hence my original question.

In this case I'm only speaking about the resolution. I have a Cinemorph anamorphic bokeh/flare filter so I'm sorted in terms of the optical aberrations. In front of the filter I have a huge 0.7x wide - angle adapter which provides h+v compression (as opposed to just horizontal with the anamorphic lens). Side benefit is I get a 40mm f/2 Helios and a 60mm f/1.4 Rokinon due to the compression. As you can tell I know my stuff. I'm just trying to figure out why I'd spend the money on a real anamorphic instead of my poor man's hack job.

So here's a thought... and sorry for the long post but this has been on my mind for a while. We keep talking about how an anamorphic lens uses the entire imaging sensor and is then compressed in post. We always follow up by saying that just cropping isn't the same. But isn't it? Let me explain. With my camera, the 5D Mark 3, there's a fixed number of horizontal and vertical pixels which are available to be exposed. There's no in-between or random arrangement like there'd be with silver halide grains in analog film. When I record MLV at 1920 x 1080 I'm getting 2,073,600 pixels. When I record 1920 x 804 (scope) I get 1,543,680. So yes, 25.56% less pixels. But! An imaging sensor isn't a perfect pickup mechanism. It doesn't have pixels all over it. That's why it's digital - it's quantized. It's a representation, as best as possible, of an analogic wave, which turns into a stair-stepping wave when it goes through the ADC due to there being no infinite quanta (even the current standard 64 bit is "only" 18,446,744,073,709,551,616 steps). What happens when you squeeze an image from 1920x1080 to 1920 x 804? You don't gain any resolution. In fact you may introduce aliasing due to lines being squeezed which weren't squeezed before, and the stair-stepping we're all familiar with (jaggies) comes out. Yes, we started with more pixels, and I understand the value in starting with more instead of starting with less. Analog mediums - sure - squeezing leads to resolution advantages. But digital is a fixed number of lines. But if in the end it all has the same aspect ratio, just one didn't have any squeezing done to it (and thus no pixels deformed and/or aliased), wouldn't just cropping make sense?