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 :)

RED ONE MX and SONY F35

I think if you made the blacks a touch darker in the F35 version, the noise difference would be less clear. And we are talking about a pretty high ASA rating -- the differences at more "normal" ratings like around 500 ASA are less pronounced. In those tests, mainly you just see the resolution difference.

The fact that so many great-looking TV shows like "Smallville" are shot on the Genesis & F35 is testament that these cameras do fine work.

It seems to me that where this new MX sensor really shines is in terms of low-light photography (where the F35 was the previous king) and for large screen images where resolution becomes more important.

Anyway, now that Red has something as good as the new MX sensor, it seems they are more than happy to be compared to the competition...

It all starts to get a bit silly though, this constant adolescent "who's winning, who's losing" game-playing going on. I know it's hard for some people here to believe, but the competing products from Sony, ARRI, Canon, etc. are all made by decent and talented folks as well, and they all deserve a bit of respect for providing us with nice tools to choose from. There is a level of machismo around here that I find a bit distasteful. I feel like I'm in a college dorm as a bunch of male students watch the Superbowl. If you are not careful, you guys are going to turn Sony and ARRI into the "underdogs" that people will start rooting for to beat the "big guy", i.e. Red.
 
Jannard said:
Zak... if you can't figure out how to read the results, I can certainly understand why you can't see the winner/difference. :-)

I'm not sure that "brings the camera right up on par with F35" is exactly the way I would describe it.

Jim
Click to expand...

Jim your certainly right.. I didnt see the photoshop files.. I just saw the quick shots that were on the first page.. AFTER i saw the photoshop files.. I clearly see the winner... OBVIOUSLY being RED.. please dont look at my origional statement.
 
David Mullen ASC said:
I think if you made the blacks a touch darker in the F35 version, the noise difference would be less clear...
Click to expand...

If you crush the blacks, you end up with less DR. The test shows that MX has at least a stop or two more DR (lower noise level).

David Mullen ASC said:
And we are talking about a pretty high ASA rating -- the differences at more "normal" ratings like around 500 ASA are less pronounced. In those tests, mainly you just see the resolution difference....
Click to expand...

No. ISO ratig is for film, sensors are compared at high ISO because then you can see sensor noise more clearly. The noise level at "500 ASA" is the same except that you can not see it as well as when you bump the sensitivity up.

David Mullen ASC said:
...In those tests, mainly you just see the resolution difference....
Click to expand...
There is much more there than resolution because this is a crop of 4k enlarged 200%. I can't see half-size grain on the MX, can you? It is just not there.

However, I do agree that the F35 is a superb camera and produces gorgeous images. This is exactly why I'm stoked and speachless. Kudos to the RED team for an amazing achievement!
 
I think, as David points out, it could be very good for this whole community if people did not take things so personally.

It is very exiting seeing how fast things are moving right now. RED has really started an "arms race". Right now there are 3 digital systems, imho, that are interesting as an alternative to film. (plus one for high frame rates).

These systems (and future ones) are of course going to replace a lot of film work in the coming years. But the really cool thing is that "normal" people (audiences) won't notice a thing. They might even see an improvement when it comes to large venues, digitally projected and less grain/"noise" on their LCD tv's.

For me, in post on a pilot, shot on RED, I just directed. I can't see why I should use film again...
 
I have no doubt that in an ideal world, 4K is better than 1080 for acquisition, even when delivering 1080. I have no doubt that Mysterium-X is an excellent improvement in terms of low light and color rendition. I also have no doubt that when properly processed to take advantage of 12-bit linear (instead of 10-bit log) image structure, Red footage has excellent latitude in post.

BUT, some people's responses to this test are ridiculous. You are looking at UNPROCESSED S-Log footage from the F35 and comparing it to Red Color and Red Gamma footage. OF COURSE the Red footage is going to look better. It has to. S-Log in not for release viewing. It's going to look washed out and b/c of that, soft. It has to be graded. For example, S-Log maps the darkest black to 90, not 0, not 16, not 64. So the blacks are going to look murky if no viewing LUT is applied.

Not to say Red doesn't win this test anyway, but it's not some crazy vanquishment that it's being made out to be. And it's ironic that Red just recently complained about unfair comparisons due to not appreciating file format differences. Yet that's EXACTLY what's being done here.
 
Beauty of REDgamma is that because our recording bit depth is higher than HDCAM SR, and our decode bit depth much higher, we don't have to go as loggy in an image and still keep all the DR with enough code values to be totally gradable. REDgamma is as nicely gradable as REDlog, but looks nice for viewing too.

Now, if you remap the F35 slog levels to match the appropriate RED images, then measure the std deviation in the shadows the F35 is still giving me higher numbers. But do that yourself and don't take my word for it.

Graeme
 
Graeme,

I tried what you said, and actually the F35 had a lower SD than the Red. For the lower left-hand gradient bar of the 320 1080 test, I got mean value of 32.32 and an SD of 1.91 for the Red. Before adjusting the level, I got a mean value of 44.92 and an SD of 1.83 for the F35. After adjusting F35's values, I got an SD 1.68. Now I'm sure that I did not perfectly match the Red's values, even though the mean for the F35 was now 32. To truly accurately do it, I need to apply a LUT. I can apply the inverse of the S-Log formula to get to linear, but doesn't RedGamma have a gamma of something like 2.2? Actually if it's like REC 709, I'd bet the acquisition gamma (gamma used for encoding) is something closer to 1.95 but still meant to be viewed at 2.2. So I'd have to apply another curve to come close to matching RedGamma. I'm sure you're a lot better at this than I am.

I do think the Red images look very good, and resolution wise they win. The Red footage also looks less noisy, and it might be, even though my test of one swatch showed that not to be the case.

But my main point remains, most of the viewers here are comparing by eye what are essentially un-comparable images. And after the F35's images are adjusted for viewing, the differences will be much less dramatic. And adjusting for viewing is not trivial.

There are other issues with the test as well. How was ISO normalized? Was camera gain used? Zero gain on the F35 is not truly 0. You can lower the gain even further, so depending on the composition, you can lower the gain--and noise--but lose some dynamic range and sensitivity in the process. But maybe that's the right way to do it.

But this test illustrates EXACTLY the problem w/ comparing camera images. There are so many variables, from composition, to camera settings, to in-camera processing, to post processing that it's very difficult to evaluate who the winner is.

My issue with this test is that ungraded S-Log looks like crap, b/c it's not supposed to be viewed for release. But people are comparing it to RedGamma and RedColor images and saying Red waxed the F35; that's just not a fair way to evaluate both cameras.

And being that Red has recently been on the other end of just such a comparison, you guys know all about the limits of these types of test. (I realize you didn't do the test and and that you DID mention that S-Log is not for viewing.) But people are still going to walk away thinking "Red looked great and F35 looked like sh!t" b/c the side-by-side is fundamentally flawed.

So why not post your re-leveled F35 images? Not saying what you did will be the same as using the appropriate LUT in Pablo or Lustre, but it should be a little fairer nonetheless.


P.S. I don't mean to bash the work that LIGHTIRON Digitial did. It's great that they went through all the effort and it's definitely helpful. But it's not conclusive and there are aspects of the presentation that people should keep in mind when judging the images.
 
I don't have the source files for either the F35 or the RED - this was Michaels test. I'm just doing the same as everyone else - looking at the images as is. Same as you, to do it accurately, I need a LUT and I don't have an slog one.

Graeme
 
Graeme,

BTW, if you do do a comparison between the two some time in the future, realize that the viewing LUT for S-Log is not the same as the LUT used for post. The viewing LUT blows out everything above about 90%.

A decision was made to sacrifice the highligts to show more shadows. So the image will look more blownout than it actually is. The LUT used for post and grading of course retains the full range of values.
 
What I'd want to do is just take it back to linear light. Got a lut for that?

Graeme
 
I don't have a copy of 10SLogAto12FL.lut or 12SLogAto12FL.lut which convert 10 or 12 bit S-Log to 12-bit linear light, but I have to believe Michael has them. Really, he could just load the F35 footage into whatever grading system he uses, apply the LUT and export TIFF's. Don't know if he'd be up for it.

I'm trying to figure out what I'd guess you'd call the inverse of the S-Log formula, plug in an S-Log value and get a linear light value, but I'm getting some crazy results. If I figure it out, I'll post it.
 
Thanks Peter.

Graeme
 
You're very welcome.
 
Michael Bravin said:
For now I am interested and concentrating on the test of the Red One with MX against a camera that I know VERY VERY well, the F35. Since I know what the F35 can do, for me it is especially instructive to see shots comparing to it.
Click to expand...

The problem for Jim's competitors in the digital cinema market is that by the time they are ready to take on Epic X (for most, that will be years from now), he will already be moving on to yet another generation of his modular DSMC system - namely, the full-frame 35mm "Vista Vision" Epic. That camera is going to be very significant.
 
Would someone at this point actually be willing to say what they think
the ISO/ASA of the new Red MX sensor is, as well as what they believe the
new Red MX's dynamic range is?

Also, what is the true dynamic range of the F35?
And yes I know this is all subjective. Just want to hear opinions. Also, at what ISO
does the new Red MX sensor provide the most dynamic range???
 
M-X DR doesn't vary with ISO. All you change is how many stops above or below, but the total remains the same.

Graeme
 
Graeme Nattress said:
M-X DR doesn't vary with ISO. All you change is how many stops above or below, but the total remains the same.

Graeme
Click to expand...

OK....so then at what ISO would we get 6.5 stops over and 6.5 stops under?
 
I've not actually figured that out....

Graeme
 
edit.....

different serial numbers
 
You must log in or register to reply here.
Back
Top