SCCE (large single sensor camera shoot out) part 3 goes live on Oct 5....

Duke M.

Duke M.

Veteran
Only three days away. A trailer is up now. (FYI: SCCE = singe chip camera evaluation, but I think "large single sensor camera shoot out" is more descriptive.)

http://www.zacuto.com/the-great-camera-shootout-2011/episode-three

Another real life comparison chapter is said to be in the works.

Usually each clip is not viewable at first in Firefox due to domain referral blocking. Internet Explorer works.

For those that haven't seen them, the overview is here:
http://www.zacuto.com/the-great-camera-shootout-2011

part 1 (Tipping Point/ dynamic range) is here:
http://www.zacuto.com/the-great-camera-shootout-2011/episode-one

And part 2 (Sensors and Sensitivity) is here:
http://www.zacuto.com/the-great-camera-shootout-2011/episode-two
 
S-Log was not available, but they did use an external recorder.

Epic was not available, but they did have a Red MX.

GH2 was not tested, but they did test the Canon 7D and 5D.
 
When describing the test results there were several comments that put the F3 in the highest category with the Alexa and Red MX. With the extra stop with S-Log the F3 has a bit more dynamic range than the Red MX, but S-Log will not be demonstrated or tested in any of these results. The actual tests were done about 9 months ago. I'm sure they had an 'expert' Red MX operator, but who was an expert on the F3 back then?
 
Certainly not me, I can't even get my PL mount to come off.. :D

Each manufacturer was invited and allowed to have a camera expert there but not all sent one.
 
The MX holds about 12. The standard F3 holds 12, and the log F3 holds 13.5

This is the benefit of 13micron pixels. Giant light gulping surfaces.
 
LeavingTheCandy said:
What's the Epic do?
Click to expand...

Someone please correct me if I'm wrong, but I thought they advertized that the Red MX chip was supposed to be the same basic type of chip as the Epic. That may have changed later. IDK. If that is so the Epic will be about the same DR in normal mode as the Red MX.

Now the Red fans are going to say; "But with xHDR it will blow them all away!" However, my understanding is that xHDR (high dynamic range) works the same way HDR works for photography. That is two pictures are taken one after the other at different exposures and then merged. A great technological feat to be sure, but they aren't taken at exactly the same time.

I think of it this way. Two images captured sequentially with 180 degree shutters would leave the shutter open all of the time with ghosting. Or two images with 90 degree shutters will give a jerky play back and less ghosting.

Or its just going to process the signal twice, once dark and once light, then combine them anyway for faked dynamic range. No ghosting that way, but not true dynamic range either.

The question then becomes how much movement is allowable before xHDR causes ghosting? And, will this really high definition capture the ghosting in even greater detail? Or are they going to fake the DR which any of us can do in post with Photoshop and AE just like photos? Inquiring minds want to know.
 
Last edited:
The sensor is just one factor in DR. The real improvements in DR have come from improvements in signal to noise ratios. Across all types of chips improvements in materials and fab processes have added more than 10dB of signal to noise ratio in the last 3 to 4 years. Nearly two stops worth. Epic is certainly benefiting from this compared to Red 1 MX since the original FPGA processors in Red 1 did not get upgraded with the sensor package. I would expect some improvement here, but comparison tests haven't been done yet.
 
Last edited:
Since there are several ways of measuring DR, you really can't compare from test to test. The cameras need to be tested in exactly the same way, preferably at the same time and facility.

Its been a minor contention with me that Red has often posted higher numbers than other people doing tests. Its not that they aren't the correct numbers for Red's tests, but that you can't compare the separate tests and differing methods. Minor differences in background light, back lit illumination, etc can make a difference.

My expectation is that the Epic will be around 12 without xHDR because even with the advances in noise there are more (therefore smaller) photo sites.

Its really a shame that Red didn't allow the Epic to be tested in the SCCE.
 
Last edited:
That is the issue measuring Dr at this point is very subjective as to where the cut off point is. It ends up that most of these tests people weigh in more closely to their own personal bias whether they mean to or not, so in the end it's best to test for yourself rather than rely on what others are telling you.

From what I can tell (uneducated) epic has more than 12 stops or else all other cameras have one stop less as well, from comparisons I have seen the Dr is not that far off from alexa in some cases (could be user error) it looks to have even more... but this is all subject to how it was shot, when, by who, so..... yeah I'll wait until I have the money and test the camera's in my range accordingly.

P.S. I can't imagine red would blatantly lie like that(or deceive)... though they could, if I remember correctly there were outside witnesses to the test, but still it is best to compare yourself in the end, because unfortunately just about everyone in a position of authority these days has their own agenda and biases... it's very hard perhaps impossible not to, now a universal standard testing mount like a focus collimator that tests the DR of a camera would be very interesting, but mighty difficult to produce given the differences in sensors and even post processing as all the steps in the chain matter not just the pixel size.
 
Last edited:
I'm not saying that Red tried to deceive anyone, but simple things make a difference. For instance, many tests use a back lit chart. There can be slight variations in how bright the bulb is behind the chart, and the older the bulb the dimmer its likely to be. Even if its the same manufacturer of chart and same manufacturer of bulb, the bulb can be from a different batch. So different back lit charts, or even the same one at different ages, can produce different results.

My recollection is that SCCE tested the Red MX and F3 at 12.3-12.5 stops, which is very good for both of them. But the Epic is 5k in the same space, so smaller photo sites. My expectation is that Epic will be 12ish too without xHDR. When AbelCine tested S-log on the scope at 13 stops the chart was at full stop increments. 13.5 may be about right for S-log if SCCE did it.
 
Last edited:
S. Matthews said:
Epic DR tests...
http://provideocoalition.com/index.php/freshdv/story/epic_vs_alexa_dynamic_range/
http://www.reduser.net/forum/showthread.php?47529-EPIC-and-Alexa...

plus a very good detailed comparison between Red One MX and Alexa
http://www.gunleik.com/tests/alexmx/RED1_MX_and_Arri_Alexa/Introduction.html
Click to expand...

I'm Gonn have to call Matthew Jepsen about this one. The PROvideo Coalition test just doesn't look right.

the chip charts on the Alexa were rendered accurately, where as the chip chart on the epic were weird.

By using his counting method, the epic has 14 stops, but then the alexa would have 16. just doesnt make sense. Its like, the bottom grey chip counts for the epic, but not the Alexa? and the top chip... the one thats rendered accurately... isnt counted for alexa either.

I expect the epic to have about 13 stops when all is said and done. a bit more than the MX ( as its the same chip) but with better signal processing, to bring out a bit more.
 
the look on the chart for the epic (in the red thread) was described by Jim and an observer as being lines on the chart itself, and them showing up being a result of resolution of the epic over Alexa. Now once again i was not there have not seen the chart so I can not say, but the whole how many stops either have, is not how many you can see but how many can be useable and that falls under what each user/client is willing to accept.

If you want to examine it further or read others who did I suggest reading the thread... i know it's a lot but there are some good questions and answers in there.

Use the red team button at the top Jim's and others posts to fly through the thread easier, though you will miss some other useful information.
 
Last edited:
You must log in or register to reply here.
Back
Top