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 :)

Lens Testing Question

A

Andy Post

Member
Joined
Feb 7, 2016
Messages
19
Reaction score
0
Points
0
OK thrill seekers. This is a variation of a newbie question I posted under "Lens...Camera....or operator issue?"

I rented a few lenses to do some testing before I buy (or decide to just keep renting). Tried a Xeen vs Rokinon DS vs the Zeiss DSLR styles I also have.

I put the Raven on a tripod, and initially loaded it with ND so I could set focus wide open. I'm using a SmallHD 701 so I'm zooming in 4X as well. So...I'm pretty sure focus is accurate. Then I shot tests wide open and in stop increments down to 11 to compare.

It's how to evaluate that I'm wondering what I'm doing wrong. Don't have 4K monitor, or probably enough power to play full rez in real time. I was hoping to mimic a 4K view and not resort to "Pixel peeping",so exported frames from RedCine. Then I stacked the frames in a layered Photoshop Document so I could compare at 100%. When do, nothing looks sharp.

I've attached a frame from Xeen 14m at F8. Focus was on the tree on the right. This was exported from RedCine, opened in Photoshop. Then I created a blank 1920 X 1080 file and drag/drop into it to make sure I didn't hose the image by resizing.

I expected a frame grab to look as sharp at 100% view as a DSLR capture would look at 100%. What do I not understand?

Thanks for your always helpful input
 

Attachments

  • Xeen-Test.jpg
    Xeen-Test.jpg
    102.9 KB · Views: 0
What ND are you using? Have you tried without it?
 
Compression ratio? If you post an R3D many questions like that can be answered merely by querying the metadata.

The other answer to your question is that RED RAW has no in-camera sharpening. It is up to the user to select the method, and the amount, of sharpening to apply.
 
If you want to evaluate from a still frame, click the A.D.D button in RedCine to output a TIFF and don't resize it to 1080p as there is no point in doing that (the image you've uploaded is highly compressed so it's hard to judge). Then add one point of sharping in REDCine and export it again, then try 2 points as you should notice a difference. When you were doing your actual shooting, were you using only a SmallHD monitor or the Red monitor as well? Did you use any of RED's tools in camera to judge focus (edge sharpening etc) or were you using SmallHD's only?
 
Michael Tiemann said:
Compression ratio? If you post an R3D many questions like that can be answered merely by querying the metadata.

The other answer to your question is that RED RAW has no in-camera sharpening. It is up to the user to select the method, and the amount, of sharpening to apply.
Click to expand...

I think a lot people are having a hard time conceptionalizing this... Especially those coming from a DSLR background, even more so from the GH4, NX1 and Alpha7s... Who have had some extreme sharpening in-camera even when turned down. Even I who have experimented on post sharpening, probably did it once with bad results. SO... Can someone with a Raven and experience with post-treatment sharpening, please show us an example or two???
Basically show us an example of straight-out-of-camera and then same frame-grab sharpened in post (a before and after pic)

Thanks in advance.
 
Pixel peeping at 100% will appear soft, as others have mentioned for a number of reasons. One is lack of in camera sharpening. Another is the OLPF in front of the sensor. Over sharpening looks worse in motion, so the Red OLPF is a bit stronger than a DSLR. Find a good sharpening algorithm and test it on content in motion.
 
I have a different approach - shoot a resolution target. Sharpness is contrast and resolution- right? Contrast is partially lens design and partially stylistic choice which will vary across manufacturers, but resolution should be directly comparable. And of course OLPF is going to soften a little.
I did that for Scarlet MX, but adjusting the amount of light rather than adding ND to the mix. About to repeat for Raven.
Quick and dirty the other day looked like Raven with Leica 50mm f2 could resolve 1600-1700 line pairs (like 3.5K) - but that was quick test - did not sit there with loupe perfecting focus ;)

I posted this previously but no one had interest - would like to see a 4k version of this vector graphics file - anyone interested in creating? ;)
I would think Weapon 8K owners would want a 4K line pair target to see what their weapon and lens choice can really resolve!!
Trick is to find a printer that can also pass the resolution test - as it will need to in order to print clean sharp version.
http://www.graphics.cornell.edu/~westin/misc/ISO_12233-reschart.pdf

BTW - when you focus 4.5 - 8 k sensor using 1080p HDMI output feed (highest res preview you can get with RED, right?) - how exact can your focus be anyway? you are not looking at every pixel uncompressed, right? when you toggle magnify - you get compressed preview and is it exactly 1:1 pixels? If not your focus will never be perfect regardless.
 
rereading original post - I have 14mm Rokinon and while Xeen is no doubt much better, that lens design exhibits distortion and softness as you get far from lens center. I see you were focusing on the tree that is not centered. So I think you have lens issues in addition to anything else (compression, ND, etc)
I typically use 14 for ultra wide and tell actors not to stay near edge frame for any length of time
I would suggest using 35mm or 50mm Xeen if you have to evaluate sharpness before using 14mm

I always turn off all sharpening with DSLR and Raven looks quite acceptable (sharper than 5DMk3 OLPF I believe) with Rokinon. (IMHO)
 
You must log in or register to reply here.
Back
Top