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 :)

Comparative tests

Status
Not open for further replies.
Gunleik,

thanks for a nice test.

Also must add that I was totally right few months ago talking about both cams.
 
Thank you Gunliek!

Jarred Land said:
3. Speaking of the blooming... we know the Alexa uses a Contrast Filter to roll off the highlights.. LightIron has done a test with Contrast Filters on RED to match the one on the ALEXA if you prefer that look.. Maybe Michael can post those results.
Click to expand...

Yes please!

Jarred Land said:
I keep hearing a 1/2 lowcon seemed to be almost a perfect match.
Click to expand...

Is that what Michael thought as well?
 
Gunleik,

That was one highly informative analysis.
I have to find a day where I can re read this while checking out the several sample files you provide.

I remember reading posts where you mentioned this tests going back several months ago. I can now say that it was well worth the wait. Thanks for your meticulous hard work.

One slightly off topic question.
On the first page where you feature both cameras side by side, I notice what looks to be a modified RED Cradle holding the camera battery on the RED. It looks like the articulating piece of the Cradle combined with a battery plate without the drive section of the cradle. Wondering what pieces you used to put this together or is this a custom machined piece?

I always liked the versatility of the RED cradle but never liked it's bulkiness plus I use the ET drive holders. The design as shown in that photo is pretty ideal.

Again, great work!

Brian Timmons
britim-media
 
Thanks for doing the tests, Gunleik.

Are there any original files from both cameras available ?

I tried links on the page and they don't seem to work.

The only ones available for download were from the "Matchtest" page where there seem to be missing .r3d grabs for primary and ND6 for a proper comparison.




I'm also interested in the decision to compare Alexa ISO1600 to MX ISO3200.
There's something which seems like sharpening/resolution issue with the latter image also.


 
Big thanks for the write-up!

Bottom line:

Alexa has a slight but measurable edge over R1-MX in DR, R1-MX has a large and measurable edge over Alexa in resolution... no real surprises there. Basically a 2X difference in cost for comparable packages (Rocket shouldn't be in that comparison IMO), with Red the clear price winner any way you slice it. Pro-Res versus Raw is another factor to consider. I see how these results leave room for debate over which camera is the best choice, at least for those who don't care about price.

If I were going to imagine a fantasy camera to end the debate, I'd come up with something that exceeds the dynamic range AND resolution of both cameras, pack it all in an even smaller/lighter form-factor, and keep the price well below Alexa... then I'd add framerate options, faster read-reset, an awesome touch-screen menu and remote, still capability with autofocus, modules with various transcode and wireless streaming options, SSD storage, and world-class color science... I sure wish there were such a camera someday -- that would be epic! :)

- Tim
 
Hi.

I will be going through all the downloadlinks.

apparently some point at mxalex, some at alexmx

My bad...

I will also go through all the packages once more.

Now I will have to run for a plane, but I will go through this and answer other questions on Friday.

Cheers and thanks for the gentle words!

Gunleik
 
Best camera comparison test regarding Red up to now.

How did you develop the MX footage? At 800 ASA, at 320 ASA? I find that MX footage is best developed at 320 ASA, at least when exposed normally (800 ASA RedGamma).

Your test shows very well the individual strengths and weaknesses of both cameras.

Thanks!

Hans
 
thank you so much for doing this test. an excellent comparison!
 
Very useful and educational test Gunleik. Thanks for making the effort, I'm sure it took a lot of work.
 
Hrvoje Simic said:
Thanks for doing the tests, Gunleik.

Are there any original files from both cameras available ?

I tried links on the page and they don't seem to work.

The only ones available for download were from the "Matchtest" page where there seem to be missing .r3d grabs for primary and ND6 for a proper comparison.

http://img864.imageshack.us/i/filesad.jpg/


I'm also interested in the decision to compare Alexa ISO1600 to MX ISO3200.
There's something which seems like sharpening/resolution issue with the latter image also.


http://img43.imageshack.us/i/isoz.jpg/
Click to expand...

Looked like compression artifacting to me.
 
Thanks again

I would gladly accept second opinions to the DR and other issues in my test, but then Oi would need original R3Ds and tungsten lightening going with them.

If I get that, I will add them to the test.

Cheers

G
 
Processed RAW leaves variables.
.r3d grabs remove variables. Same ISO, same NDs.

IMO
 
It is such well documented, well thought test. It better be so, if you followed the reactions towards Geoff Boyle's test on this forum.
I have to admit that I always had an rather critical eye for Red, from were I'm standing it seems at times that you're even a bit cautious when commenting on Red performances.
I have in no way such a deep and analytical knowledge regarding the Red camera's workflow.
So from point of view of a young cinematographer(I haven't touched Arri's camera yet), Alexa seems to be forgiving were it matters when you are in a tight corner, when you are not able to control the contrast of the scene, and it comes with a smooth, gradual roll of the densities in both ends.
But what I found reassuring in your tests is that is not much room to fiddle with the settings in Alexa.
Regarding monitoring it goes closer to the old by now HD saying: What you see is what you get.
Maybe finally we can cool down with reinventing this craft and get people interested again in the art side of the business,lighting, composition, grammar and storytelling.
Thank you for demystifying endeavor.
 
Last edited:
Mr. Roger Deakins says on his forum that someone told him Alexa and the RED MX (hence EPIC as well) use the same exact sensor - just optimized differently.

I've never heard it before... could this be true?
 
These tests are great reading materials for anyone who is thinking of working with either RED ONE MX or ALEXA. They give a great look into the thinking and practical workings of both cameras, and I am glad to see such a well documented and well thought out set of tests coming to light for EITHER camera, let alone one which does a great job of comparing the two!

You have done a great service for anyone thinking of shooting a project on either of these cameras, and I would be shocked if anyone got up in arms with any of the views shared.

One question I have - which I would be keen for you to answer - is that you didn't seem overly concerned about chroma moire in the Alexa image when you were doing the 150-200% blow up.

My suspicion is this either
A ) because you have a workflow for fixing this that is neither time consuming nor difficult to achieve.

or

B ) scenarios which would generate such chroma moire (like test charts!) are relatively unlikely to come up in the real world.

or, I suppose, both? Be very keen to have a response to that.
 
Medavoym said:
Mr. Roger Deakins says on his forum that someone told him Alexa and the RED MX (hence EPIC as well) use the same exact sensor - just optimized differently.

I've never heard it before... could this be true?
Click to expand...

Probably meant that the Red One and the Epic use the same M-X sensor -- obviously the Alexa uses a different sensor, it's 3.5K (2.9K image area) instead of 5.1K for one thing.
 
I guess it is time to respond.

The test seems like a good one but it was done with a few holes. The Alexa had more DR than MX. We have acknowledged that from the beginning. But it is more like 1/2- 3/4 stop. Not 1.8 stops. Using old color science is not the way to test. We had warned about this several times. LightIron has done side by side testing with current color. Michael has said that the difference is what we have said... 1/2 to 3/4 stop. And he has processed a lot of footage from both. He has also said that using a 1/2 contrast filter matches Alexa footage "almost exactly". EPIC is a different story altogether.

It is very easy to match color with Alexa to MX. It is impossible to match resolution.

I love side by side tests... but every detail matters.

Thanks to Gunleik for these. But I will also encourage Michael to post his comprehensive tests. Michael has posted a ton of footage from both cameras and apparently has come to a bit different conclusion than Gunleik.

Jim
 
With my concept of HDR(y) that use two different ISO setting of the same shot in R3D with exposure fusion to merge two images I can extract exactly about 14 or even 15 stops of DR with RED1-MX

also without any motion and electric light artifacts that supposed to be found at HDRx that involves shutter speed on EPIC.

That means I can get out with that kind of HDR concept from RED1-MX footage the same amount of DR as an Alexa using Log-C or could be even more...

Also I saw that EPIC is much cleaner at the higher ISO (much+much+much better S/N ratio) that means perfect for that...
 
Status
Not open for further replies.
Back
Top