Welcome to our community

Be a part of something great, join today!

  • 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 vs. DVX andromeda vs. HVX200 vs. XTi

Matthew Bennett

Well-known member
Joined
Sep 3, 2007
Messages
486
Reaction score
0
Points
0
Website
www.truenorthmedia.ca
Special thanks to Denis Betsi and Sasha Moric on this test:

Special note:
I outputted all the jpgs directly from Quicktime, and I think that people using browsers other than safari may not be seeing them. try Safari as a browser if you have it..



Last weekend we shot the RED, HVX, Andro, Canon XTI, M2 and Brevis side by side.

First, A wide shot of a room: Stills and Clips: HVX, Andro, RED, in that order.

HVX_Room.jpg


Andro_Room.jpg


RED_Room.jpg


CLIPS: Andro and RED are Prores, HVX is DVCPRO HD (native)

www.qmediasolutions.com/guest/RED_Camera/RED_Cam_Comparison/HVX_ROOM_clip.mov

www.qmediasolutions.com/guest/RED_Camera/RED_Cam_Comparison/Andro_ROOM_clip.mov

www.qmediasolutions.com/guest/RED_Camera/RED_Cam_Comparison/RED_ROOM_clip.mov

RESULTS:

We exposed the HVX and Andro DVX the same shutter, f stop, etc, RED was 320 ISO.

RED seems to have a bit of dynamic roof over the Andro, plus all its other lovely qualities. You can really see the effects of quantizing the 14bit head of the RED vs. the 12bit Andro, giving such an large range of tonal gradation into the lights, mids and darks. Even when a RED image is a bit dim, it can still appear luminous.
 
You can really see the effects of quantizing the 14bit head of the RED vs. the 12bit Andro, giving such an large range of tonal gradation into the lights, mids and darks. Even when a RED image is a bit dim, it can still appear luminous.

Thanks for doing this test, very interesting! Unfortunately I'm not on a mac so I can't view the clips- is there any way to see Prores 10-bit in windows?

If you have a 14-bit Red it must be a special edition... the regular Red is 12 bits, I believe.
 
Next - Macbeth, Andro, RED and XTi

Please excuse the 'hard' grade of the Andromeda, I did a mid sacrificing, high levelling curve on it during the LUT. RED is graded with a very light curve, I'm really liking its linear response.


Andro_Chart.jpg



RED_Chart.jpg



Canon_Chart.jpg


Pro-Res CLIPS:

www.qmediasolutions.com/guest/RED_Camera/RED_Cam_Comparison/Andro_Chart_Clip.mov

www.qmediasolutions.com/guest/RED_Camera/RED_Cam_Comparison/RED_CHART_clip.mov

Not really a super fair comparison here, due to the harder LUT curve on the Andro, but obviously the RED footage is quite stunning here due to all the lovely tonal gradations.

Also check out the funky green bit on the Andro clip, the light on top frame where the sensors have clipped. RED handles business like this in such a gentle way...
 
Never realized how much noise Andromeda has until now.
 
You can actually get Andromeda footage to be noiseless, but you have to shoot the DVX at what the chip is actually rated at: 32-50 ISO. Which is ridiculous in most shooting situations.

If you use a linear LUT (no gain) as a capture with Andromeda, you can see what the chip actually exposes.

Chris Nuzzaco posted a long thread on this over at the reel-stream forum:

http://forum.reel-stream.com/viewtopic.php?t=778
 
It is always great that the good folks on this form take the time to do and share tests... Many Thanks!

The difference is clear to my eyes but I think that controlled lighting is important in side by side camera/stock tests.

The HVX clip is overexposed to my eye and the RED clip seems to be in range. So it makes it a bit difficult to evaluate the differences.

But Again thanks, I know you are giving up your free time for this
 
Ok I might have missed a beat here. R U exposing for the room and letting the sky/highlights do what they want? is that the test?

But the interior room seems brighter in the HVX clip as well. I am confused :)

the DVX seems to match the RED in exposure, hoever of course not in quality.
 
Fair enough, as I lack the scientist-mind that it takes to do these tests absolutely proper.

For the room, for example, it gets a bit apples and oranges.

We shot the HVX at ND-1 f6, 1/60 shutter, and the DVX at the same. But the HVX is rated at 320 ISO and the DVX is rated at 640 ISO (from what I've heard..). On top of that, the DVX andromeda mod actually rates the DVX at 32-50 ISO at a linear chip-level response, which I then rendered from RAW andromeda files using a LUT curve which rates that 32-50 ISO signal at about 320 ISO (i would estimate).
Then we shot the RED at 320 ISO (proper rating for the chip), i forget which f-stop, 1/60 shutter, a couple of hours later, it could have gotten darker or lighter outside.

We've all heard that the HVX is about 7-8 stops, DVX andro around 9.5 stops, and RED in the 10-11 range. Yes? (Sorry If I haven't kept up with the latest estimates of RED's range..)

Again, you'll have to excuse me calling these 'tests' as they are more 'lets just shoot all the cameras at sort of around the same, same scene and see what we get.."

Not 'tests' more rough comparison frames of roughly the same view.
 
Here's the Linear response from the DVX andromeda:

Andro_DVX_Linear_ROOM.jpg


As you can see the green CCD has completely lost detail in the upper highs, giving us the dastardly magenta highlights.
Andro user Chris Nuzzaco uses a 'short linear' LUT to record with, throwing away the top part of the scale, and then rendering the RAW with another LUT to bring the gamma back up to a simulated 250-300 ISO.. smart idea, fairly noiseless response too, and also gets around the magenta problem.
 
too bad the 2k version of Andro didn't make it to the market. it will blow away RED.
 
I'm not trying to be confrontational, im just wondering how a camera hack on an HVX200 could hold up to, well, any single aspect of the RED?

Apart from how a traditional DVX100 becomes when converting it into Andromeda? There's specs on their website I think, but they aren't going to make it.
 
Back
Top