Video Test in Low Light: Panasonic GH3 versus the BMPCC Shooting RAW

That is interesting - I expected the BMPCC to do better than the EOS, but not by that much.

What really surprised me though was the shadow noise in the BMPCC images when comparing it to the Epic. Of course, the Epic starts at $17,500 for a box that can't really do anything without another $3k+ in "accessories", and they shot 5k for the test - the Red *better* produce a cleaner image :)
 
Before talking about monitors, I would advise practicing more in focusing. Especially if the intention is so-called testing. As we all know, Mark is a great proponent of auto focusing. But both images are focused differently (and I am not even talking about DOF - it's another issue) - the difference is subtle but it's there and this one thing along makes OP's talks about RAW ridiculous.
I'm not sure I follow - what does focus have to do with raw? Care to elaborate?
 
Before talking about monitors, I would advise practicing more in focusing. Especially if the intention is so-called testing. As we all know, Mark is a great proponent of auto focusing. But both images are focused differently (and I am not even talking about DOF - it's another issue) - the difference is subtle but it's there and this one thing along makes OP's talks about RAW ridiculous.
I'm not sure I follow - what does focus have to do with raw? Care to elaborate?
Focus by itself has nothing to do with RAW. But it has to do a lot with the "test" where two frames are offered that we are supposed to look at and compare.
 
Came across this one today as I was googling for some info about Resolve. In my opinion, this shows where the BMPCC strength lies -shadow and highlight details. I suspect that most of the older Canons - EOS M included, since it is basically a 7D sensor - will have similar dynamic range performance compared to the 50D. Raw from the 5D Mk III would likely fair better than any of the other older Canon DSLRs. In fact, that is a comparison I would really like to see.

 
The BMCC RAW grabs appear to "win," much as an HQ still shot will beat most video frame grabs. I merely wonder what this means in practice. For instance, in the following side-by-side comparison (internal AVCHD versus external Atomos), the grabs of the latter may look slightly better to the connoisseur, but I confess not to see any difference in the moving image, or could easily be fooled.

http://vimeo.com/55678698

Furthermore, to edit RAW video must be on par with turning still frames into video: something that will choke my PC, take forever, and possibly yield less benefit in the final result than other factors requisite to snare and keep viewer attention.

Is the Pocket BMCC, even if you had enough batteries and cards, something you'd really want to use to shoot video of a trip, casual social gathering, or a sports event?

How many GH3 owners shoot in its ProRes mode, and how many do much with that that they'd not do with the plebian AVCHD?

Do BMCC-ers who shoot RAW have the time or need to grade, overlay, and render the stuff? Will their viewers notice (or pay for) the difference?

Frame grabs aside, what is left in the "pudding" or the compressed stuff people finally see?

Too much of what people actually get (or manage) to shoot involves bad light whose WB is no easier to correct in post than to fly through a storm with manual navigation. If AVCHD is lossy and not the pro's choice for primary capture, at least the smart-auto WB does the (at times hopeless) job at lest 75% right, and a single pass edit to do the necessary selection, titles, effects, voice-over or whatnot saves time and entails IQ loss one will notice barely in Blu-ray and an eagle eye.

Frame grabs may be the fine pallet test, as with vintage wines. But, if they start to taste the same as the $7 / bottle brand after a glass or so, and you have to host 100 guests, out of your own pocket, does the $37 / bottle brand make any sense?

I am a sucker for gear, too, but more likely to be lured to a 4k GH4 or an FDR AX1, but only after I learn what the workflow and hardware requirements are. Very likely, I'll decide that my video ails mainly for want of a better stabilizer or, better yet, more comic content. Not a slider, but a banana peel.
 
Before talking about monitors, I would advise practicing more in focusing. Especially if the intention is so-called testing. As we all know, Mark is a great proponent of auto focusing. But both images are focused differently (and I am not even talking about DOF - it's another issue) - the difference is subtle but it's there and this one thing along makes OP's talks about RAW ridiculous.
I'm not sure I follow - what does focus have to do with raw? Care to elaborate?
They are just trying their hardest to discredit me and the comparison - faulting the focus (auto), my capability to focus (really?) and even the logic (!), anything to avoid coming to grips with the advantages of RAW over the cameras they own. It is expected (when you see words like "ridiculous", you know that people are behaving emotionally).

C'mon guys, let's all learn, ok? I will post my experiences with the BMPC in the field and with editing RAW and post some video from RAW, and then we can all discuss RAW and look critically at the video to learn from it. I am not trying to sell you anything or dis your cameras (I have as much invested in them as you). Do watch your spelling though :).
 
Last edited:
Markr041 wrote:
Of course you could in an editor adjust the colors of the GH3 clip, but because the GH3 video is highly compressed, 8-bit 4:2:0 color sampling you would end up with an even worse-looking clip. And if you have to use an editor, then using uncompressed RAW is the way to go: 14-bit, 4:2:2.
The trade-off: The GH3 offers a very nice video ooc; the BMPCC requires editing in post and with that work one can come up with a better-looking video, with more creative possibilities.
I think that the trade-off is more extensive than that. How much time and effort did it take to edit this RAW output from the BMPCC? What tools did you use? What steps were involved?

We know that you spend zero time editing the GH3 clip. It is straight from the camera. But how large was this clip, and what did it take to get it to look that good?
Good question.

It's simple - import RAWdng from the camera into Lightroom, edit as you like, and then export as jpeg (in this case). Once one has the settings you like (WB temperature, saturation, etc.) you can save the settings and then for other scenes like the one you just edited it is one click to apply the exact same settings. To apply the same settings to all the frames in the clip, it is just one click (synch).

This is much less work than the EOS M RAW, which requires first removing pink dots, then converting the RAW from the camera to RAWdng before any editing. And the Canon files are 1.5X bigger, because the BMPCC uses losslessly compressed RAW.
 
The BMCC RAW grabs appear to "win," much as an HQ still shot will beat most video frame grabs. I merely wonder what this means in practice. For instance, in the following side-by-side comparison (internal AVCHD versus external Atomos), the grabs of the latter may look slightly better to the connoisseur, but I confess not to see any difference in the moving image, or could easily be fooled.

http://vimeo.com/55678698
I see quite a bit of difference without even looking at individual frames. There is no denying that a high bitrate codec will reduce the macro blocking effect, which is what I see a lot of in that video.

That said, this type of scene is brutal for video - lots of quirky details, and minimal opportunity for the camera to compress efficiently, which forces visible macro blocking. Thankfully, most people will rarely punish their cameras like this (because it is BORING :) ). That alone justifies cameras with AVCHD...
Furthermore, to edit RAW video must be on par with turning still frames into video: something that will choke my PC, take forever, and possibly yield less benefit in the final result than other factors requisite to snare and keep viewer attention.
You are absolutely correct. I have 30 minutes of raw footage of a band I shot Friday night. Just loading that footage into the media pool in Resolve took about 5 minutes - it was an exercise in patience to say the least. I'm nowhere near done with it, and at this point, I just want it in an intermediate format that I can dump into Vegas so I can get the video back to the band. I will get it done, but ProRes is much easier to deal with when you have longer clips.
Is the Pocket BMCC, even if you had enough batteries and cards, something you'd really want to use to shoot video of a trip, casual social gathering, or a sports event?
I don't even bother shooting those videos any more, at least not on high end gear. In my family they rarely get watched, making photos more effective. Your mileage may vary. Still, I would opt for a $300 camcorder or my cell phone for personal video before dropping $1,000 on the BMPCC. Same with photography though - I always laugh at the soccer mom's lugging around their Canon 5D Mk IIIs to the events at my daughter's elementary school. Chances are very likely that the photos from my wife's Olympus ZX-1 are just as good, and I did not have to look like a tool :)
How many GH3 owners shoot in its ProRes mode, and how many do much with that that they'd not do with the plebian AVCHD?

Do BMCC-ers who shoot RAW have the time or need to grade, overlay, and render the stuff? Will their viewers notice (or pay for) the difference?
If you shoot raw, you have post work to do. Otherwise, nobody will be able to view it.

As far as the viewers noticing, that depends. Look at the opposite end of the spectrum. A lot of money was put into the look of the Marvel movies (Iron Man, Avengers, etc). Very little was put into Blair Witch Project, Tiny Furniture, and El Mariachi. All of those movies were wildly successful though, because as it turns out, a compelling story is more important than really good looking footage (unless you are producing something like the "Planet Earth" series - that had to look good). The lo-fi look of the latter films was either embraced (Blair Witch) or simply the best they could do with what they had. In the end, the audience was willing to overlook the lack of hundreds of thousands of dollars of camera gear. I doubt they would give The Avengers the same pass though.
Too much of what people actually get (or manage) to shoot involves bad light whose WB is no easier to correct in post than to fly through a storm with manual navigation. If AVCHD is lossy and not the pro's choice for primary capture, at least the smart-auto WB does the (at times hopeless) job at lest 75% right, and a single pass edit to do the necessary selection, titles, effects, voice-over or whatnot saves time and entails IQ loss one will notice barely in Blu-ray and an eagle eye.
Pros don't use AVCHD because they have a budget. There is nothing wrong with AVCHD - it's just that there are better options out there, especially if you want or need something visually compelling.
I am a sucker for gear, too, but more likely to be lured to a 4k GH4 or an FDR AX1, but only after I learn what the workflow and hardware requirements are. Very likely, I'll decide that my video ails mainly for want of a better stabilizer or, better yet, more comic content. Not a slider, but a banana peel.
A very wise conclusion - the world needs fewer Philip Blooms and more JJ Abrams. I love Mr. Bloom's videos, but I am not compelled to watch them a second time.
 
Before talking about monitors, I would advise practicing more in focusing. Especially if the intention is so-called testing. As we all know, Mark is a great proponent of auto focusing. But both images are focused differently (and I am not even talking about DOF - it's another issue) - the difference is subtle but it's there and this one thing along makes OP's talks about RAW ridiculous.
I'm not sure I follow - what does focus have to do with raw? Care to elaborate?
They are just trying their hardest to discredit me and the comparison
Actually, that takes very little effort.

The point that is being made is that you proudly posted what you apparently thought was a frame grab from a RAW video crushingly better than a frame grab from a GH3 video.

You asserted that only you knew which one was accurate, color-wise, implying of course that it could have been either. I personally found the GH3 more pleasant to look at, and as we all know accuracy is rarely the most important aspect of color reproduction -- case in point being the popularity of the heavily color shifted JPEGs from Fuji and Olympus.

You went on to assert that the GH3 was too smooth and that the shadow performance of the RAW was far better. My own eyes on a calibrated monitor found instead that the GH3 was smooth while the RAW grab was grainy. So there is little to celebrate in my opinion.

Of course, a better job of processing the RAW would return equally smooth video. But then it would dull some of the sharpness that was allegedly there (yet not really evident from what I could see) so until you actually perform the test correctly, everyone is simply speculating.

Including you.
- faulting the focus (auto),
That was a valid comment, since you commented at some point on detail and focus matters a great deal.
my capability to focus (really?)
No. Your love of auto focus was what he commented on. Which as we all nkow will be different between two entirely different cameras.
and even the logic (!),
Which was entirely absent. Largely because you never stated the premise of the test and you chose to not provide your analysis, leaving it up to the rest of us.

Badly constructed test + no real premise + no real analysis == black hole of information
anything to avoid coming to grips with the advantages of RAW over the cameras they own.
Not at all. Everyone knows that a RAW image has more latitude than a JPEG image. Ergo, a video made up of a string of RAW images can be processed into sharper and cleaner video than one made up of JPEG images. By definition.

But you did nothing of the sort.

So whining that we all didn't make the same leaps as you did based on a terrible test is not going to change minds.

I have never seen playing the victim card win an argument. Just sayin' ...
It is expected (when you see words like "ridiculous", you know that people are behaving emotionally).
I think you should reread the post to which I am responding to see what emotion looks like on paper ...
C'mon guys, let's all learn, ok?
If you want people to learn from you, then you ought to try to actually teach.
I will post my experiences with the BMPC in the field and with editing RAW and post some video from RAW, and then we can all discuss RAW and look critically at the video to learn from it. I am not trying to sell you anything or dis your cameras (I have as much invested in them as you).
Suggest that you start telling people what you were testing for, how you tested it, and why your results are in fact valid and worthy of the effort to examine them.

I don't think you are in any danger of selling anything to anyone.
Do watch your spelling though :).
Watch it do what?
 
Pros don't use AVCHD because they have a budget. There is nothing wrong with AVCHD - it's just that there are better options out there, especially if you want or need something visually compelling.
It's more than having a budget for good equipment, it's the fact that you're also paying a cast and crew. If you miss a shot or if a shot needs to be pushed or pulled beyond what's possible with AVCHD, then it's very, very expensive to re-shoot that scene. You don't want to be the cause of that kind of a problem, so it's in your best interests to shoot in a format that allows you the greatest degree of post-production flexibility and latitude. The cost of the equipment to capture and edit in high resolution lossless RAW is pretty trivial compared to the rest of the production cost.
 
Pros don't use AVCHD because they have a budget. There is nothing wrong with AVCHD - it's just that there are better options out there, especially if you want or need something visually compelling.
It's more than having a budget for good equipment, it's the fact that you're also paying a cast and crew. If you miss a shot or if a shot needs to be pushed or pulled beyond what's possible with AVCHD, then it's very, very expensive to re-shoot that scene. You don't want to be the cause of that kind of a problem, so it's in your best interests to shoot in a format that allows you the greatest degree of post-production flexibility and latitude. The cost of the equipment to capture and edit in high resolution lossless RAW is pretty trivial compared to the rest of the production cost.
Yes. And no. It depends on the project. I worked on a web series a few years ago that was shot entirely on a Panasonic HPX 170. It was the DP's personal camera, but I suspect they would have used whatever she had - they wanted her, not her gear. She toyed around with my GH2 for about 10 minutes once, called it "cute" and handed it back :)

Upstream Color was shot entirely on a GH2. An unusual movie, but it did well at Sundance and SXSW. While it's not exactly a blockbuster, I understand that it grossed over $300,000 in it's first month or so, and personally, I thought it looked brilliant.

Jack Daniel Stanley shot his short "Depth of Phil" on a GH1. I shot a concert DVD a couple of years ago on a Canon XH-A1 and a Canon 7D (it was included as part of a special edition on the bands next CD release). Crank 2 was shot entirely on several low end Canon camcorders (they had a $20,000 budget). "Tiny Furniture" was shot on a 5D Mk II. Modern productions are all over the place, depending on budget, skill, and requirements. I still know several indie producers who will not relinquish their 5D Mk II, because it is "good enough".

I think cameras like the BMPCC are a step up from the GH2 and GH3, but as has been beat to death on numerous forums, it is not a camera for everyone. All I am saying is that this should not discourage anyone - you can make a film on any camera you want. The first thing you need is a compelling story, and the rest will follow...
 
Before talking about monitors, I would advise practicing more in focusing. Especially if the intention is so-called testing. As we all know, Mark is a great proponent of auto focusing. But both images are focused differently (and I am not even talking about DOF - it's another issue) - the difference is subtle but it's there and this one thing along makes OP's talks about RAW ridiculous.
I'm not sure I follow - what does focus have to do with raw? Care to elaborate?
They are just trying their hardest to discredit me and the comparison
Actually, that takes very little effort.

The point that is being made is that you proudly posted what you apparently thought was a frame grab from a RAW video crushingly better than a frame grab from a GH3 video.

You asserted that only you knew which one was accurate, color-wise, implying of course that it could have been either. I personally found the GH3 more pleasant to look at, and as we all know accuracy is rarely the most important aspect of color reproduction -- case in point being the popularity of the heavily color shifted JPEGs from Fuji and Olympus.

You went on to assert that the GH3 was too smooth and that the shadow performance of the RAW was far better. My own eyes on a calibrated monitor found instead that the GH3 was smooth while the RAW grab was grainy. So there is little to celebrate in my opinion.

Of course, a better job of processing the RAW would return equally smooth video. But then it would dull some of the sharpness that was allegedly there (yet not really evident from what I could see) so until you actually perform the test correctly, everyone is simply speculating.

Including you.
- faulting the focus (auto),
That was a valid comment, since you commented at some point on detail and focus matters a great deal.
my capability to focus (really?)
No. Your love of auto focus was what he commented on. Which as we all nkow will be different between two entirely different cameras.
and even the logic (!),
Which was entirely absent. Largely because you never stated the premise of the test and you chose to not provide your analysis, leaving it up to the rest of us.

Badly constructed test + no real premise + no real analysis == black hole of information
anything to avoid coming to grips with the advantages of RAW over the cameras they own.
Not at all. Everyone knows that a RAW image has more latitude than a JPEG image. Ergo, a video made up of a string of RAW images can be processed into sharper and cleaner video than one made up of JPEG images. By definition.

But you did nothing of the sort.

So whining that we all didn't make the same leaps as you did based on a terrible test is not going to change minds.

I have never seen playing the victim card win an argument. Just sayin' ...
It is expected (when you see words like "ridiculous", you know that people are behaving emotionally).
I think you should reread the post to which I am responding to see what emotion looks like on paper ...
C'mon guys, let's all learn, ok?
If you want people to learn from you, then you ought to try to actually teach.
I will post my experiences with the BMPC in the field and with editing RAW and post some video from RAW, and then we can all discuss RAW and look critically at the video to learn from it. I am not trying to sell you anything or dis your cameras (I have as much invested in them as you).
Suggest that you start telling people what you were testing for, how you tested it, and why your results are in fact valid and worthy of the effort to examine them.

I don't think you are in any danger of selling anything to anyone.
Do watch your spelling though :).
Watch it do what?

--
http://kimletkeman.blogspot.com
Is there something wrong with you? I am sure we could have endless fun jousting with words and poking, as almost everything you say above is empty rhetoric and uninformative. But that is not what this forum is about.

I notice you do not have a GH3 or a BMPCC, or any apparent experience with RAW.

Anyway, if you read my experience in the post about the BMPCC in use for shooting RAW, you might find it informative. If you want more info on that I am happy to respond.
 
Last edited:
Before talking about monitors, I would advise practicing more in focusing. Especially if the intention is so-called testing. As we all know, Mark is a great proponent of auto focusing. But both images are focused differently (and I am not even talking about DOF - it's another issue) - the difference is subtle but it's there and this one thing along makes OP's talks about RAW ridiculous.
I'm not sure I follow - what does focus have to do with raw? Care to elaborate?
They are just trying their hardest to discredit me and the comparison
Actually, that takes very little effort.

The point that is being made is that you proudly posted what you apparently thought was a frame grab from a RAW video crushingly better than a frame grab from a GH3 video.

You asserted that only you knew which one was accurate, color-wise, implying of course that it could have been either. I personally found the GH3 more pleasant to look at, and as we all know accuracy is rarely the most important aspect of color reproduction -- case in point being the popularity of the heavily color shifted JPEGs from Fuji and Olympus.

You went on to assert that the GH3 was too smooth and that the shadow performance of the RAW was far better. My own eyes on a calibrated monitor found instead that the GH3 was smooth while the RAW grab was grainy. So there is little to celebrate in my opinion.

Of course, a better job of processing the RAW would return equally smooth video. But then it would dull some of the sharpness that was allegedly there (yet not really evident from what I could see) so until you actually perform the test correctly, everyone is simply speculating.

Including you.
- faulting the focus (auto),
That was a valid comment, since you commented at some point on detail and focus matters a great deal.
my capability to focus (really?)
No. Your love of auto focus was what he commented on. Which as we all nkow will be different between two entirely different cameras.
and even the logic (!),
Which was entirely absent. Largely because you never stated the premise of the test and you chose to not provide your analysis, leaving it up to the rest of us.

Badly constructed test + no real premise + no real analysis == black hole of information
anything to avoid coming to grips with the advantages of RAW over the cameras they own.
Not at all. Everyone knows that a RAW image has more latitude than a JPEG image. Ergo, a video made up of a string of RAW images can be processed into sharper and cleaner video than one made up of JPEG images. By definition.

But you did nothing of the sort.

So whining that we all didn't make the same leaps as you did based on a terrible test is not going to change minds.

I have never seen playing the victim card win an argument. Just sayin' ...
It is expected (when you see words like "ridiculous", you know that people are behaving emotionally).
I think you should reread the post to which I am responding to see what emotion looks like on paper ...
C'mon guys, let's all learn, ok?
If you want people to learn from you, then you ought to try to actually teach.
I will post my experiences with the BMPC in the field and with editing RAW and post some video from RAW, and then we can all discuss RAW and look critically at the video to learn from it. I am not trying to sell you anything or dis your cameras (I have as much invested in them as you).
Suggest that you start telling people what you were testing for, how you tested it, and why your results are in fact valid and worthy of the effort to examine them.

I don't think you are in any danger of selling anything to anyone.
Do watch your spelling though :).
Watch it do what?
 
Is there something wrong with you? I am sure we could have endless fun jousting with words and poking, as almost everything you say above is empty rhetoric and uninformative. But that is not what this forum is about.

I notice you do not have a GH3 or a BMPCC, or any apparent experience with RAW.

Anyway, if you read my experience in the post about the BMPCC in use for shooting RAW, you might find it informative. If you want more info on that I am happy to respond.
I wouldn't like to generalize, but in this particular discussion you are the one who behaved in an arrogant way and exhibited many other qualities that do not characterize you as a good, kind and wise person.

Your position has been very simple: you disagree with my test - then, you are against RAW. Here is a quote from you "...and obviously have no clue about raw." You made this rude statement with very vague idea what equipment I use and what's my knowledge regarding RAW and, anyway, what I am doing as a cinematographer.

In your subsequent statements you tried to rationalize your approach by saying that people who disagree with you simply try to justify their use of their own equipment. You obviously have no clue that some people simply rent equipment when it fits their needs - there is nothing to defend. You live in your small world with very limited experience (oops, I forgot that James Cameron is your best friend) of shooting flowers and poor editing skills.

In your multiple posts here you accused people of many things. What you still do not see that you are doing yourself exactly what you were accusing other people of. Ironically, that even includes your attempts to ridicule me for misspelling. But when you misspelled something (in not less embarrassing way) you immediately asked for mercy that other people showed to you.

You've got your new toy, good for you. Enjoy and tell us about your experience.
 

Keyboard shortcuts

Back
Top