Emery Wells
Well-known member
- Joined
- Feb 28, 2007
- Messages
- 705
- Reaction score
- 0
- Points
- 0
Last week, Jason Diamond and I visited Michael Schrom & Company at Silvercup Studios in NY to do some side by side testing with Katabatics RED-MX #40 and Schroms RED ONE. This was a no frills technical test to see how these two sensors compare and the results are interesting to say the least.
This test was made possible thanks to the fantastic team over at Schrom:
Dir/DP: Michael Schrom
Exec Producer: Carl Sturges
Lighting Director: John McCabe
Camera Assistant: Woody
First let me say we have been shooting with the RED ONE for two years with phenomenal results but the RED-MX feels like a whole new camera. My high level takeaway is the MX delivers on most if not all of what RED has been claiming. We found that a comfortable rating for the camera was ASA 400. Now Im not DP and I know every DP has there own special magic to get the most out of the camera but with the first gen RED ONE I found ASA 320 to be a generous rating under most circumstances and If I wanted to get the cleanest possible image out of the sensor I tended to rate at 200. So I feel like the MX gives you an extra stop, HOWEVER, you can push this sensor FAR beyond what the Mysterium could handle. ASA 800 - 2000 are very 'comfortable.' You wouldn't use this on a commercial but for features, low light, run and gun, whatever, 800 - 2000 is a very usable range. Even beyond 2000 you can get acceptable results. In some of our testing we found that the LCD was seeing more detail than our EYES could see in the room. No joke, no exaggeration.
I chose to process everything in RedCine-X using Red Color/Red Gamma for both the RED-ONE and RED-MX footage. We felt this would give the best apples to apples comparison but on a side note, Im really loving the new Red Color/Red Gamma. It does breath new life into old footage.
First up - Noise Test. For this I took a 100% crop of a 4K Tiff processed out of RedCine-X and lined up every ASA we shot for RED-MX and RED-ONE.
This image was scaled down to better fit in the forum. Download the FULL RES TIFF to really see whats going on here.
The obvious thing is noise. There is less of it in the MX. Additionally, I've noticed the MX is natively warmer than the Mysterium. Both cameras were reset to all factory settings before we got started and the metadata/processing were identical. There is a noticeable difference in color temperature between the two but I haven't heard RED mention this anywhere. As we went up in ASA on the MX sensor, everything got cooler until it eventually turned bright blue.
The temperature difference is also noticeable in the test images below. The RED ONE image appears 'muckier' than the MX which has a much more natural color reproduction. This is really positive because one if the things I and others haven't been crazy about is how RED handles skin tones. I think there is a markable improvement on the MX. Other interesting note is the RED ONE image looks about half a stop under. Both images had identical lenses (Arri Zeiss 100mm) @ T22. One of the things RED has mentioned about the MX is improved accuracy when used with external exposure monitoring tools like light meters, etc. The MX seemed to more accurately match what our light meter was telling us.
At the low end of ASA's the RED ONE seemed to perform better which makes sense because the middle grey has shifted in the MX. I wouldn't shoot ASA 200 on the MX. 320 seems like the comfortable low end.
We then wanted to really stress test the cameras so we turned the key light off.
And then we turned off ALL lights and lit the scene with a single Kino bulb.
I have a lot more to publish as I do more tests this week. You also really need to see either 2K or 4K stills to really analyze these properly. I will try to get those up soon.
This test was made possible thanks to the fantastic team over at Schrom:
Dir/DP: Michael Schrom
Exec Producer: Carl Sturges
Lighting Director: John McCabe
Camera Assistant: Woody
First let me say we have been shooting with the RED ONE for two years with phenomenal results but the RED-MX feels like a whole new camera. My high level takeaway is the MX delivers on most if not all of what RED has been claiming. We found that a comfortable rating for the camera was ASA 400. Now Im not DP and I know every DP has there own special magic to get the most out of the camera but with the first gen RED ONE I found ASA 320 to be a generous rating under most circumstances and If I wanted to get the cleanest possible image out of the sensor I tended to rate at 200. So I feel like the MX gives you an extra stop, HOWEVER, you can push this sensor FAR beyond what the Mysterium could handle. ASA 800 - 2000 are very 'comfortable.' You wouldn't use this on a commercial but for features, low light, run and gun, whatever, 800 - 2000 is a very usable range. Even beyond 2000 you can get acceptable results. In some of our testing we found that the LCD was seeing more detail than our EYES could see in the room. No joke, no exaggeration.
I chose to process everything in RedCine-X using Red Color/Red Gamma for both the RED-ONE and RED-MX footage. We felt this would give the best apples to apples comparison but on a side note, Im really loving the new Red Color/Red Gamma. It does breath new life into old footage.
First up - Noise Test. For this I took a 100% crop of a 4K Tiff processed out of RedCine-X and lined up every ASA we shot for RED-MX and RED-ONE.
This image was scaled down to better fit in the forum. Download the FULL RES TIFF to really see whats going on here.
The obvious thing is noise. There is less of it in the MX. Additionally, I've noticed the MX is natively warmer than the Mysterium. Both cameras were reset to all factory settings before we got started and the metadata/processing were identical. There is a noticeable difference in color temperature between the two but I haven't heard RED mention this anywhere. As we went up in ASA on the MX sensor, everything got cooler until it eventually turned bright blue.
The temperature difference is also noticeable in the test images below. The RED ONE image appears 'muckier' than the MX which has a much more natural color reproduction. This is really positive because one if the things I and others haven't been crazy about is how RED handles skin tones. I think there is a markable improvement on the MX. Other interesting note is the RED ONE image looks about half a stop under. Both images had identical lenses (Arri Zeiss 100mm) @ T22. One of the things RED has mentioned about the MX is improved accuracy when used with external exposure monitoring tools like light meters, etc. The MX seemed to more accurately match what our light meter was telling us.
At the low end of ASA's the RED ONE seemed to perform better which makes sense because the middle grey has shifted in the MX. I wouldn't shoot ASA 200 on the MX. 320 seems like the comfortable low end.
We then wanted to really stress test the cameras so we turned the key light off.
And then we turned off ALL lights and lit the scene with a single Kino bulb.
I have a lot more to publish as I do more tests this week. You also really need to see either 2K or 4K stills to really analyze these properly. I will try to get those up soon.
