Cinema 5D says A5100 has 13 stops of dynamic range

cinema5d on aug 22 backed off from their 13 stop claim and settled on 12 stops for the 5100.

(updated info in same link as above)

ymmv

be well.

rob
 
BlackLegSanji said:
And their test method is just as valid as yours. Different methods, different interpretation, different results.
Click to expand...

But without giving a breakdown of methods to arrive at that number its all useless (I feel like a grade school math teacher)

Cinema 5D by their method rated the 5D3 in h.264 as having a higher dynamic range than the C300 in Clog. Having used the two cameras extensively I know that is not true at all. 5D3 Raw? Yes of course that makes sense but h.264? I call shenanigans on that.
 
I thought they did say how they got the numbers, no? They shot a test chart and put it into a software to help count the number of steps that pass a certain signal to noise ratio. I'm baffled by the 5D3 vs C300 difference, too, but one questionable result doesn't mean we should discount the site altogether. Over at DxO they gave the A99 a low light score of 1500 which left a lot of people scratching their heads, but on overall their results are pretty accurate.
 
Let me be the first to say that software is not required to understand the dynamic range of any camera if you understand

1) How to use a chart
2) How to read a chart

There is an absolute lack of understanding for both 1 & 2 by many people who consider themselves (perhaps mistakenly) a professional.


I don't care how expensive your chart may be, if you don't understand at a base level #1 & #2, you will not get accurate (that is factual) results.


When you read the claims of "measured DR" you must first.... Consider the SOURCE of the claim.

My first group of criteria for asking, "Is what I'm reading trustworthy?"

Does this (typically) blogger, writer, "filmmaker" or website administrator....

Have legitimate professional credits to their name? (not, spec/uncommissioned shorts only to be found on-line).
Has this person ever been employed by a camera manufacturer, or reputable top-tier camera rental company?
Is this person widely respected by other who are particularly technical and member of one or more technical organizations who understand this sort of thing?
Is this person known to be an legitimate and recognized by their peers, expert in optics, imaging or color science?

I can think of probably 4-5 people in the entirety of the internet that blog with any sort of regularly that you can answer yes/true to the above questions.


More often, I think the source of such claims on-line comes from someone with one or more of the following second group of criteria for asking "Is what I'm reading trustworthy?"

Is this person's primary job it is to administer a website, get hits, clicks, likes or tweets on a blog or piece of "news?"
Does this person have no formal technical training? (not the same as artistic training, film, art or liberal arts school)
Does this person have no legitimate credits?
Does this person admit to not reading a manual?
Is this person an "affiliate" marketer and posses a website with one or more of the big box shops?


If you are generally answering yes (or true) to the second group, keep in mind that your source is unreliable at best, and at worst laughably incompetent.

Now each of these second categories are not mutually exclusive of the first group of criteria I listed, but for some reason they generally tend to be.
 
BlackLegSanji said:
And their test method is just as valid as yours. Different methods, different interpretation, different results.
Click to expand...

Not really. They just posted a nice graph, but no mention and examples on how the measurement was performed. I'm geniunly very curious about this.
 
Yes, the 13-stop results of the A5100 gives me a lot of doubt, too, but then I also think that measuring DR by over and under expose can't really tell what the camera is actually capable of. This DR measurement thing really got me confused.
 
Well, they are successful in driving traffic to their site with all the hyperbole. I have read some of the comments of people challenging their findings and/ or explaining their 'scientific' process but I have yet to see their detailed response. I think Illya said it best in the post above.
 
macgregor said:
Not really. They just posted a nice graph, but no mention and examples on how the measurement was performed. I'm geniunly very curious about this.
Click to expand...

There was a claim of software being used to calculate the results. Which could work, if done competently- although entirely unnecessary for measured DR (rather than theoretical DR.)
 
wgtwo said:
Out of interest, who would you say these 4-5 people are?
Click to expand...

That's something that I will only share in private rather than in public, as by saying who's actually worth listening too, in a sea of misinformation, I'm basically calling 99.9% of folks out there doing this sort of thing half-worthy, unworthy, untrustworthy or incompetent.

There is fairly large group of people who know this stuff really well. However, they turn off their brain when they go home from the office (or set), work for large companies which forbid them from contributing to on-line discussions or are just far, far too busy to also do this stuff in their "spare time".

The funny thing is every April and September I get together with a fairly large group of friends and inevitably we start talking about the lies and ineptitude that we read about on-line or in a supposedly technical publication, or white paper.

Always consider the source before you believe anything. Aggregator sites are some of the worst offenders, because they perpetuate misinformation faster than anything else out there. It's crazy how much B.S. are in some of them, but it's essentially blind leading the blind- even if you care enough to try to correct something, you end up arguing with morons and trolls in a terrible comment system, which I'm convinced is almost never read anyway.

Archives get an air of authenticity on these aggregator sites which is undeserved. Just because the information wasn't challenged by someone doesn't mean it's true, and usually it underwent no vetting.
 
Thank you for pointing all that out Illya. Agreed 110%, and I used to spend hours on the internet arguing with those those kinds of people until I realized my free time was far too precious after 14 hours a day on set to waste that way.

Here's what I tell our interns and PAs when they ask where to learn about cinematography: Everything on the internet is wrong - read a book instead. Now that's obviously not strictly true, but it's close enough to be a good practical rule to live by. The basics don't change and the old-fashioned printed information is still valid - and everything that is newer or fast-changing enough to not be practical to get on paper isn't really all that important anyway (like exact numbers for dynamic range for a consumer camera - knowing it won't make your cinematography any better at all).

I used to have a list of trusted websites/blogs from which to gather information - but many of them have recently become suspect as well, as advertising sponsorship is an sneaky insidious beast that is the enemy of truth. Luckily I've been doing this dozens of years and my bullshit detector is finely tuned, but less experienced readers are likely to be less discerning and crap spreads fast on the net.
 
Yeah, I get caught up in all these numbers and "what's best" but then I realize that the people who are paralyzed by only working with the "best" equipment or are scrutinizing charts all day end up doing nothing at all. I've happily been working with "unusable quality" stock h.264 video canon dslr's since 2009 and it funny, the work keeps coming. Sure, I really like the image when I get to hire a c300 or an F3 or F5, but the client doesn't seem to be any happier when I do. I just cashed a check for $61k for 1 month's work with a 5D mkIII, all shot stock h.264, and the client loved the final product. I'm going to gleefully exchange my 5D for a A7s which seems like it will be an upgrade in image, but more important, is lighter and more versatile in several different situations. I really don't care if it has 12 stops or 11 - it looks as good or better than my current setup, is lighter, and has a few new tricks up it's sleeve (low light, remote control, timelapse app, versatile lens mount, autofocus during video with native lenses, ability to shoot log, 4k path). I suspect the people who are declaring it "unfit" really don't work much, or would have a problem with any camera.
 
After receiving my a7s and running some preliminary tests, I'm very intrigued by the a5100. I thought slog2 would be the killer feature of the a7s, but it's not: it's xavc-s, which helps usable DR a lot more than slog2 does.

I just posted it here, but let me say it in this thread again...

Usable DR: it really depends on how you measure it, so you need to devise a test that fits whatever you're going to shoot. I have done three preliminary tests, with three different results:
- total DR (if it shows up in the waveform, it counts; this would be the closest test to what cinema5d has done): slog2 and cine4 have around the same DR... and they are not much different from what I have in my 5n. Did you read that? Same DR as a 5N.
- usable DR, caring a lot small detail (I found a building made of black bricks, joined with dark-grey mortar; standing 20m away from it, with a 50mm lens, the shadows only count if I can clearly see individual bricks): slog2 has one stop better DR than cine4... and my 5N has 3 stops less DR than slog2. Did you read that? A 3-stop improvement!!
- usable DR, caring just about medium detail (same black building, same lens, but now standing just 5m away, so the bricks are much bigger in the screen; again, shadow detail only counts if I can clearly see individual bricks): slog2 has one stop better DR than cine4... which has the same DR as my 5N. So, just a 1-stop improvement this time, and only if I stick to slog2.

The reason I'm intrigued about the a5100 is that the usable DR advantage of the a7s over the 5n is biggest when you care about small detail, which the codec in the 5N is killing completely (the black bricks wall looks like it's just painted black; you can see windows and corners, but it's completely smooth and shows no detail at all). So maybe the a5100 can indeed come within one stop of the a7s in all situations.
 
Samuel H said:
The reason I'm intrigued about the a5100 is that the usable DR advantage of the a7s over the 5n is biggest when you care about small detail, which the codec in the 5N is killing completely (the black bricks wall looks like it's just painted black; you can see windows and corners, but it's completely smooth and shows no detail at all). So maybe the a5100 can indeed come within one stop of the a7s in all situations.
Click to expand...

Honestly, i wouldn't count on that. To my eye, the 100% C5D crop chart pic of the a5100 looks like crap in terms of fine detail and codec effectiveness. Do they share same XAVC-S codec? Both capturs are internal 1080p... Perhaps uncompressed 8-bit 4:2:2 via hdmi ...? Very intrigued to see more a5100 samples though.
 
Yes, it's the same XAVC-S codec, but it seems the a5100 is doing some line skipping (as usual with a high-mpix sensor), that's what's killing the detail and creating some artifacts. It should have much faster readout, though, so you're trading RS for a cleaner image. I know which one you prefer... In any case, it's not a direct replacement, it's just different trade-offs.

An a6100 with the same image of the a5100 plus an EVF, for $650, would be a very nice replacement for my 5N.
 
You must log in or register to reply here.
Back
Top