2K Soft Focus on all Red Ones

[Jeff Kilgroe]

It seems obvious to my naked eye that the 2K shots are soft. Are we looking at the same samples? I haven't done any Photoshop test but just by looking at them it seems obvious to me that the wine bottle graphics are very soft in the 2k version compared to the 4k version in Ryan's test.

Can't say I'm finding any meaningful results in Ryan's chart tiffs he posted (sorry, Ryan). The exposure and lighting are not conducive of a proper test here. The framing between 4K and 2K shots changes for no apparent reason (super-flimsy tripod?), the lens is wide open at T3 so that will soften up the image and focus isn't consistent between shots. The 16foot 18mm 4K shot cropped to 2K is noticeably softer than the 16 foot 18mm 2K shot. Several of the 4K shots are noticeably darker by 1/2~2/3 stop so what changed and why?

Lighting and camera position aside, I see more optical flaws than anything else. In other words, not only should a proper testing environment and procedure be in place to give meaningful results, but repeating tests with an assortment of lenses. No lens is perfect, not even the coveted Master Primes. We often see people here experiencing image problems and it's quick and easy to point a finger at the camera, but many times it is the optics or other part of the system that may be a more direct cause.

His older wine bottle shots suffer from some of the same issues. The exposure and amount of noise changes between the 2K shot and 4K shot cropped to 2K.

[David Mullen ASC]

Though it sounds like people are saying that there is no difference between 2K mode and 4K cropped to 2K, the way to do the test would be to line up a line resolution chart (at a decent stop) in 2K cropped mode on the camera, shoot it, and then switch to 4K mode so that the chart now only occupies half the field, and then shoot that and crop it to 2K in post, and compare the resolution.

But this is a separate issue from whether 2K on the RED, no matter how you get there, looks "too soft" for whatever uses you had planned for the footage. But one would expect it to be softer than shooting in 4K and not cropping.

[Shawn Booth]

And yes, Sanjin, your test is a bit soft.

Really? Would you mind pointing out how his image is soft?

[Alex Carr]

I would try using other sets, Cookes are a little soft and Ultra 16 Lenses were not necessarily 'Made' for the characteristics of the Red. Red Zooms are notoriously awful Wide open, and with Focus, I shot with a 14mm Ultra Prime and a 28-76 Optimo. Never once noticed any issues worth bringing up. Because we know 2k is softer than 4k in ALL situations. Its the nature of the camera system. Use Sharper Glass to achieve better results.. I suppose...

[keicol]

the way to do the test would be to line up a line resolution chart (at a decent stop) in 2K cropped mode on the camera, shoot it, and then switch to 4K mode

Please do this with the subject at least fifteen feet away so you can see what we're saying.

Also try and use a lens that's at least a 21mm or wider.

In the real world you might be shooting 4k at a 4/5.6 so try going to 2k 100fps from there without changing the lighting or the iso which means you'll be darker and wide open (expected) and also very soft.


.

[Ryan E. Walters]

No worries David- I stated that it was just quick and dirty. However, these are some stills from another test that I did previously with some text, and there seems to be something with the 15 - 20 feet measurement:

http://www.ryanewalters.com/downloads/ResTest.zip

I hope to do some more methodical testing soon, but I have been busy on other projects ...

[Nook Kim]

I thought I'd chime in here and give my experiences a little bit. I run a very small post house in Korea, but I've dealt with a bit more than seventy shot-on-Red projects, mainly TV commercials and music videos. From time to time, there were projects bought in for either transcoding or DI that looked soft. Some were softer than others. I thought, at that time, it could be the back-focus issue, the lens, or the focus puller. However, as time went on and more soft shots appeared, I realized that they were all recorded in 2K high-speed. So I started asking the camera crew what lenses they were shot with, and most of them said they were shot on wide lenses. The most popular wide lens seems to be the Ultra 16mm.

Recently, there was a project that was to be shot all in 100 fps or higher, and when I processed the footage, the camera crew and I couldn't believe how soft the shots were. I could sharpen them up a bit during the DI process, but some shots just don't sharpen up nicely due to the condition of the shoot.

I think this is not a small problem for the camera, because one of the bigger reasons that people here rent Red is it's ability to shoot high speeds. There are more people bring it up this fact (yes, I said it) to me week by week. There are more than a few crews that tested more than a dozen Reds with more than a couple dozens of wide lenses based on my advice, and they all found that it is very risky to use Red camera for high speed shooting.

I'm very glad this issue has been talked about on Reduser, but I'd really love to see what RDC has to say on this topic. As keicol said, I'm not trying to bash Red by any means. I started my business based on it. I hope this issue is addressed by RDC and get resolved if it's possible.

[Robert Horwell]

I have certainly come across this problem in exactly the same circumstances as mentioned here. 14mm Ultra Prime, at 12ft 100fps and soft. Anyone from RED care to chime in with ideas?

[Jacek Zakowicz]

Though it sounds like people are saying that there is no difference between 2K mode and 4K cropped to 2K, the way to do the test would be to line up a line resolution chart (at a decent stop) in 2K cropped mode on the camera, shoot it, and then switch to 4K mode so that the chart now only occupies half the field, and then shoot that and crop it to 2K in post, and compare the resolution.

But this is a separate issue from whether 2K on the RED, no matter how you get there, looks "too soft" for whatever uses you had planned for the footage. But one would expect it to be softer than shooting in 4K and not cropping.

David, Keith,
The optimator always generates the same image. how about using to level the playing field for 2k and 4k cropped. It should be very easy to see the difference. The resolution target goes up to 2000lph. It should be pretty obvious if the 2k and 4 k frames differ...
Jacek Zakowicz, optitek-dot-org