While not everyone agrees on the details, most will agree that the RED ONE revolutionized digital cinema, either directly or indirectly. Nearly two and half years after it’s introduction RED turned the industry on it’s head again by offering a cost effective major  upgrade to their cameras. This upgrade to the Mysterium X sensor has the potential to radically improve the camera already in the hands of it’s owners — some owners, like us, had their RED cameras since 2007. This is a big deal, something never done by Sony, Panasonic, Ikegami, Arri in the history of the industry. Sure you could repair a camera or replace a part – put I have never been able to essentially upgrade a camera I already own without buying an entire new model!

Well? Did RED  do it? Does the Mysterium X outshine the Mysterium, and is it worth the investment?

We decided to put it to the test and results are posted below. Here are some stills that represent the tests:

Low Light Compare - Tungsten

Mysterium on left, MX on right, both cameras at ISO 2000. You can immediately see the noise floor and additional image information.

Exterior Night Comparison

This vehicle headlights and dusk shot is dramatic. Notice the low noise in the MX image, the better sensitivity and contrast range (ability to handle brights and darks). ISO 2000.

We exposed this scene at proper exposure and then opened up 3 stops to make the recording.  Using REDcine FLUT controls in POST we wanted to see if scene could be recovered. As you can easily see, the MX was easily recovered (look at the blown out leaves in the sky if the before and after) and would have been completely useable with some additional color correction, but the Mysterium  was missing so much information that it would have been impossible to correct. Both cameras were at ISO 800.

The video clips speak for themselves. Enjoy. Both clips are the same, We are testing out the both YouTube and Exposure Room – let me know which player works best for you.

On Exposure Room (download link available – file is 250mb)

On YouTube. Make sure you choose the 720P version and zoom to full screen for best view.

We believe the MX is worth the investment and have upgraded both of our cameras to the new MX sensor. Looking towards the future, our upgraded RED ONES will be a nice compliment to the Scarlet and the EPIC cameras.

Details for the Camera Test and the processing:

We upgraded our camera #15 to Mysterium X and held back our second camera, #16 to Mysterium. The test was conducted with gathering of some  industry friends at our OC facilities on Feb. 26, 2010. The tests were informal, yet provide valuable insight into the major differences between the Mysterium and Mysterium X sensor upgrade. The firmware is a beta release, as is the RED Cine program used to convert the R3D files to PRORES and to export snapshot TIFF images.

The process for the test was simple – the same set-up was used for each cameras series of shots. We used the same lighting, tripod, subjects and lenses. Between each test, as few minutes of time lapse and the subject movements and poses were similar, though not exact.

Our goal was to demonstrate any differences between the two sensors. We were expecting a difference in the noise floor and the contrast range. Both were evident and significant to us, see if you agree by reviewing the representative samples we provided. A log of the complete test is available here.

Both cameras set to factory defaults and black shaded before test. Shooting 4K 2:1, RED Code 42 Camera B is Serial No. 16 “Connor”, Mysterium Build 21.4.1 Camera A is Serial No. 15 “RIpley”, Mysterium X upgrade, Build 30.2.4 White balance was matched in Red Cine and all clips were set to REDcolor and REDgamma. No other corrections were made.

The R3D files were exported from REDcine as Apple Pro Res HQ, then brought into FCP. The video files were edited in Final Cut Pro using Apple Pro Res HQ and then exported as a reference file. The reference file was exported to H264 1080P using Telestream Episode. The H264 file was uploaded to web video server where they did their own conversions.

For the split screen comparisons, I exported TIFF still frames from REDcine and, using Photoshop created the split screen and text overlays. The files were exported from Photoshop as TIFF files and then converted to 80% quality JPEG files. As a reminder to all, TIFF single frame exports from the Red 4K 2:1 RAW file are 48MB each frame. The JPEG exports seen her are only 2.5MB so some details are missing as a result. If you are interested in seeing the TIFF files I can make links available on request, however, I feel the additional artifacts don’t distract from the comparative results, but always always happy to hear your thoughts on this.

Thanks to all of our colleagues that participated in the test at our Orange County facilities and share our enthusiasm for RED technology.

PS: We are happy to say that our both of our cameras are now equipped with the MX sensors and are available for rent.

  6 Responses to “Mysterium X vs. Mysterium RED shoot out – the results”

  1. Hi Brad – It was fun shooting the test with you and and everyone that showed up to help. It’s great to see the corrected images – both stills and video. They’re very instructive and really show how much better the MX sensor is vs the M. It makes the camera very much more DP user friendly with more sensitivity, latitude, contrast, and maybe even sharpness. Here are my observations on the stills and video…

    1 – Lighter Test – This is one where the grain or lack there of is most evident, even in the JPEG. I know you have a lot of range to change color in post, but I also think the color rendition of the MX vs the M “out of the box” is also demonstrated here. The M has a definite greenish cast which is missing in the MX.

    2 – 2000 iso Ext Night – I think the lack of noise in the MX vs M is only part of the story, though the difference is very dramatic between the two shots. I think the contrast in the MX is better in the M. The apparent sharpness also seems better in the MX vs the M. This is probably because there is less noise/grain on the MX vs the M. Per contrast, look closely at the windows on the right of the 1813 building. You can actually make out subtle differences in reflections in the MX shot which are lost on the M. I also think the color rendition looks better on the MX vs the M. That is just my personal opinion. BTW, this is a shot that, to me, is more dramatic in the moving example…especially for sharpness, contrast, and color rendition.

    3 – EXT/DAY – 3 stops overexposed. It is a dramatic difference in over exposure latitude and the detail in the shoulder of the exposure curve. With the MX there is actually detail in the very over exposed areas that can be brought down in post to give you an exposure that is useable. The M sensor lacks the detail in those areas so when you bring it down you can really see the clipping of the sensor. The corrected footage gives the M sensor a very video feel where as the MX sensor has a much more filmic feel. Also, the MX is able to distinguish between the light toned building and the sky which the M is not able to do.

  2. i am in red camera asst my name is sivakrishna i interisted plz tell me plz fored me up date in red camera news plz fored me

  3. Thank you so very much for the time and the detail taken for these tests. Very useful indeed
    Can you please give me a link to download the TIFFS ?
    Thank you again.
    Daniel

  4. Thanks for the compliments. I could probably pull the TIFF files from the archives, but by now, most industry folks agree that the MX chip offers a vast improvement to the earlier version. The movie Social Network confirms this. If an upgrade of your RED or choosing an MX for rental is possible, always go that route no matter the situation is my advice.

  5. Thank you so very much for the time and the detail taken for these tests. Very useful indeed
    Can you please give me a link to download the Manual?

  6. To be fair, the blown out MX sample is focused on the leaves & not on Dave so it’s not a completely apples & apples comparison.

 Leave a Reply

(required)

(required)

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

   
© 2010 shootwithred.com Note that "shootwithred.com and Video Resources, Inc. are not affiliated with or endorsed by Red.com, Inc./Red Digital Cinema Camera Co." Suffusion theme by Sayontan Sinha