Canon C300MkII vs C200

[Django]

34 minutes ago, IronFilm said:

I'd say raw is overrated, especially for one man band stuff (i.e. low budget shoots), you're not needing raw now after all! As the C100 is a long long way from that. 

I shot RAW ages ago on 5D3 and BM Cinema Camera.. which the work from those cameras coincidentally ended up on big screen & major art installations.

Shooting RAW & developing it takes more time but doesn't necessarily require more staff (unlike manual focus which may require an AC).

I've been doing a lot of corporate / digital content stuff with fast turnover lately but I'm definitely hoping to get back into the more indie/artistic projects where RAW will be put to good use again (part of the reason for upgrading my personal equipment). For that to work into my workflow it's better if it's internal and supported by my favourite NLE. That kinda rules out ProRes RAW options for me.

So imho, RAW isn't overrated at all -even for low budget shoots- for certain projects that might involve heavy grading or require ultimate IQ / DR.

[zerocool22]

5 minutes ago, Django said:

 

So imho, RAW isn't overrated at all -even for low budget shoots- for certain projects that might involve heavy grading or require ultimate IQ / DR.

Agree, projects that might end up portfolio material I shoot in RAW (but really now with braw I shoot everything in braw anyway, as its the same bitrates as prores, its a no brainer).

[Django]

Yeah correct RAW has never been so relevant with the various light options not taking more space than prores & with native drag n drop support from NLEs

[Ed_David]

On 4/14/2019 at 5:24 AM, Django said:

Reviving this topic as I'm currently hesitating in between both cams as much needed upgrade to my C100. There is a new C300mkII kit that replaces the cumbersome monitor+XLR clamshell design with the C200's touchscreen, separate XLR module + a shogun inferno (allowing ProRes Raw) for $10K.  :

 https://www.newsshooter.com/2019/01/07/canon-c300-mark-ii-now-available-with-the-c200-touch-screen/

This really makes the C300mkII design more appealing imo.

Doing a lot of green screen work with fast turnovers and the manageable 10-bit 422 internal 4K (and 12-bit 444 2K !) will be welcome over the limited C200 internal choices. 

Thoughts?

I have the new lcd. Makes the c300 ii so so so much more ergonomical .

I shot some tests c300 ii and c200. Email me at ed.david@gmail.com and i can get you the raw dpx files

[Dave Maze]

Lol! I was totally color blind in that comparison video. Did a terrible job matching everything 

[HockeyFan12]

23 hours ago, Ed_David said:

I have the new lcd. Makes the c300 ii so so so much more ergonomical .

I shot some tests c300 ii and c200. Email me at ed.david@gmail.com and i can get you the raw dpx files

Did you notice any difference in skew?

On dvxuser I think someone mentioned the C300 Mk II has 8.3ms of rolling shutter and the C200 has 16.7ms (same as the original C300, I believe). Not that I trust everything I read online, but it's enough to make me curious.

Please share your conclusions!

[Kisaha]

4 minutes ago, HockeyFan12 said:

Did you notice any difference in skew?

On dvxuser I think someone mentioned the C300 Mk II has 8.3ms of rolling shutter and the C200 has 16.7ms (same as the original C300, I believe). Not that I trust everything I read online, but it's enough to make me curious.

Please share your conclusions!

Just some numbers from that thread. Their method seems to be ok and I accept the results presented there. Some are from other sources, also.

a7R III 4k S35 ---- 16.0 ms (15.6-16.4)
XT3 4k 60fps ------ 16.0 ms (15.6-16.0-16.3)
C200 (4k=1080p) --- 16.1 ms (4k: 15.5-16.2-16.9, 1080p: 15.5-16.0-16.6)
BM p4k 4k --------- 16.2 ms (official)
a7III 4k 1.2x 30p - 16.7 ms (dpreview)

[HockeyFan12]

18 minutes ago, Kisaha said:

Just some numbers from that thread. Their method seems to be ok and I accept the results presented there. Some are from other sources, also.

a7R III 4k S35 ---- 16.0 ms (15.6-16.4)
XT3 4k 60fps ------ 16.0 ms (15.6-16.0-16.3)
C200 (4k=1080p) --- 16.1 ms (4k: 15.5-16.2-16.9, 1080p: 15.5-16.0-16.6)
BM p4k 4k --------- 16.2 ms (official)
a7III 4k 1.2x 30p - 16.7 ms (dpreview)

I'm not saying that those results are wrong, just that the C300 Mk II has 8.3ms of rolling shutter so it's interesting that the C200 apparently has twice that. I haven't had the chance to put them side-by-side, so I'm curious given that they apparently use a similar sensor (well,  the C200 apparently has a C700 sensor, which is even weirder).

I figured this was an opportunity to see just how different the rolling shutter is.

Also do you know where the results are coming from? I remember reading the Cinema5D methodology was flawed, but I believe someone else came up with a more conclusive test. I'd believe either number, though the C200 seems to have less skew than the C100 to me... but a LOT more than the Alexa, so perhaps it's the same, but just heavier/steadier, hence less wobbly. In that case, it's a big win for the C300 Mk II.

16.7ms (close enough to 16.1) would make sense. It's a 60p camera and that would indicate a 60hz sensor, I think. Then again, the C300 Mk II is a 30p camera but it has a 120hz sensor (confirmed by Canon).

[Kisaha]

1 minute ago, HockeyFan12 said:

I'm not saying that those results are wrong, just that the C300 Mk II has 8.3ms of rolling shutter so it's interesting that the C200 apparently has twice that. I haven't had the chance to put them side-by-side, so I'm curious given that they apparently use a similar sensor (well,  the C200 apparently has a C700 sensor, which is even weirder).

Also do you know where the results are coming from? I remember reading the Cinema5D methodology was flawed, but I believe someone else came up with a more conclusive test. I'd believe either number, though the C200 seems to have less skew than the C100 to me... but a LOT more than the Alexa, so perhaps it's the same, but just heavier/steadier. In that case, it's a big win for the C300 Mk II.

http://www.dvxuser.com/V6/showthread.php?303559-Measuring-rolling-shutter-put-a-number-on-this-issue!

my opinion is that the C200 is kinda of lost in translation. I use it many times per year because I am renting it very cheap from a friend,  but if it isn t available I prefer the C100mkII for 1080p and the C300mkII for higher end projects.

[HockeyFan12]

2 minutes ago, Kisaha said:

http://www.dvxuser.com/V6/showthread.php?303559-Measuring-rolling-shutter-put-a-number-on-this-issue!

my opinion is that the C200 is kinda of lost in translation. I use it many times per year because I am renting it very cheap from a friend,  but if it isn t available I prefer the C100mkII for 1080p and the C300mkII for higher end projects.

I can understand that. Have you noticed any difference with the skew in practice?

I'd always been confused about this and saw this as a good opportunity to clear things up. I'd believe either number, though neither completely makes sense to me. 

[Kisaha]

Just now, HockeyFan12 said:

I can understand that. Have you noticed any difference with the skew in practice?

I'd always been confused about this and saw this as a good opportunity to clear things up. I'd believe either number, though neither completely makes sense to me. 

These are jobs that I give the footage to the clients, so I have no clue, sorry.

Obviously the lesser numbers are the best, but in real life, I am not sure that cameras at around 15ms have any issue at all.

I would like to see smaller numbers for 1080p at least, though.

Check the numbers and read the intro at the dvxuser thread, they explain the methodology.

[IronFilm]

26 minutes ago, HockeyFan12 said:

I'm not saying that those results are wrong, just that the C300 Mk II has 8.3ms of rolling shutter so it's interesting that the C200 apparently has twice that. I haven't had the chance to put them side-by-side, so I'm curious given that they apparently use a similar sensor (well,  the C200 apparently has a C700 sensor, which is even weirder).

It is not just the sensor, but also the processing power behind it that matters. 

Should be no surprise if the C700 and C200 have different levels of grunt powering it. 

[HockeyFan12]

9 minutes ago, Kisaha said:

These are jobs that I give the footage to the clients, so I have no clue, sorry.

Obviously the lesser numbers are the best, but in real life, I am not sure that cameras at around 15ms have any issue at all.

I would like to see smaller numbers for 1080p at least, though.

Check the numbers and read the intro at the dvxuser thread, they explain the methodology.

I looked it over, it's just that it contradicts the official number for the C300 Mk II, which I thought had the same/similar sensor, and there's no number for it to compare with. So I'm wondering if it's a difference between the two cameras in terms of skew, or if it's a dual read out mode as with the original C300 sensor, which had a 16.7ms and 8.3ms read out mode.

For my purposes 16.7ms of skew can be an issue, though I agree it's visually minor. Rarely at 24fps is it even visible but at 60p conformed to 24p it's quite visible and I do match moving, which can be seriously troubled by rolling shutter, even barely visible rolling shutter. Part of the reason I'm curious about this is to enter the correct figure for match moving purposes, though mostly curiosity.

3 minutes ago, IronFilm said:

It is not just the sensor, but also the processing power behind it that matters. 

Should be no surprise if the C700 and C200 have different levels of grunt powering it. 

I'm not disputing this, just curious what Ed's findings are. The sensor readout is sensor-level, though, I believe, not processing. I got to play around with the C700 at Canon Burbank... it is a beast and on a wholly different level I agree. Really just curious if these two cameras have the same amount of skew or if the C300 Mk II has half as much. Information online indicates half as much, but difficult to get the comparison from a consistent source so I was hoping Ed could chime in. 

[IronFilm]

3 minutes ago, HockeyFan12 said:

The sensor readout is sensor-level, though, I believe, not processing.

By the time you see the image on your screen and are measuring it then it has gone through a LOT of processing. 

And just because a sensor "can be" read out at a billion zillion lines a picosecond, doesn't mean that it will be. The processor handling that data needs to be able to keep up, otherwise that will be a bottleneck which will throttle it. 

[Kisaha]

3 minutes ago, HockeyFan12 said:

I looked it over, it's just that it contradicts the official number for the C300 Mk II, which I thought had the same/similar sensor, and there's no number for it to compare with. So I'm wondering if it's a difference between the two cameras in terms of skew, or if it's a dual read out mode as with the original C300 sensor, which had a 16.7ms and 8.3ms read out mode.

For my purposes 16.7ms of skew can be an issue, though I agree it's visually minor. Rarely at 24fps is it even visible but at 60p conformed to 24p it's quite visible and I do match moving, which can be seriously troubled by rolling shutter, even barely visible rolling shutter. Part of the reason I'm curious about this is to enter the correct figure for match moving purposes, though mostly curiosity.

BM UMP2 2k --------- 3.2 ms (official)
RX100 V 250 fps ---- 3.8 ms (3.8-3.8)
a7R III 1080 S35 --- 6.1 ms (6.6-5.7)
BM 4.6K (2K crop)--- 6.3 ms (official)
BM UMP2 4k --------- 6.3 ms (official)
1DX II 1080p 60&120- 6.7 ms (6.6-6.7)
XT3 1080p 120fps --- 7.4 ms (7.4-7.5)
BM UMP2 4.6k ------- 7.6 ms (official)
a7R III 1080 FF ---- 7.7 ms (8.7-6.8)
p4k HD crop -------- 7.8 ms (official)
NX1 1080p ---------- 7.9 ms (7.7-8.0-8.1-7.8)

Ursa Mini Pro 2??!!

(or a used NX1 ? !!)

[HockeyFan12]

8 minutes ago, IronFilm said:

By the time you see the image on your screen and are measuring it then it has gone through a LOT of processing. 

And just because a sensor "can be" read out at a billion zillion lines a picosecond, doesn't mean that it will be. The processor handling that data needs to be able to keep up, otherwise that will be a bottleneck which will throttle it. 

That's exactly what I'm curious about. The C300 mk II has a 120hz sensor (confirmed by canon) which would equate to 8.3ms of rolling shutter. 

I don't believe either camera has processing to account for rolling shutter. Usually when I enter a rolling shutter figure for skew correction or motion matching purposes, I go with the official number from the camera manufacturer. I haven't yet match moved C200 footage. With Alexa footage I don't even bother to enter anything it just works fine assuming no rolling shutter.

It seems the C200 has 16.7ms of rolling shutter (which would equate to 60hz readout). And I'm wondering if this is consistent between it and the C300 Mk II, since I believe Canon specified 8.3ms, or perhaps they just meant a dual 60hz/120hz readout like the previous generation had. That's all.

6 minutes ago, Kisaha said:

BM UMP2 2k --------- 3.2 ms (official)
RX100 V 250 fps ---- 3.8 ms (3.8-3.8)
a7R III 1080 S35 --- 6.1 ms (6.6-5.7)
BM 4.6K (2K crop)--- 6.3 ms (official)
BM UMP2 4k --------- 6.3 ms (official)
1DX II 1080p 60&120- 6.7 ms (6.6-6.7)
XT3 1080p 120fps --- 7.4 ms (7.4-7.5)
BM UMP2 4.6k ------- 7.6 ms (official)
a7R III 1080 FF ---- 7.7 ms (8.7-6.8)
p4k HD crop -------- 7.8 ms (official)
NX1 1080p ---------- 7.9 ms (7.7-8.0-8.1-7.8)

Ursa Mini Pro 2??!!

(or a used NX1 ? !!)

I can live with 16.7ms of rolling shutter, even if the UMP2 looks pretty awesome, I'm just curious what Ed's findings are and whether this is the correct figure for match moving purposes. (Which I doubt I'll do a lot of with this camera anyway, to be fair, but it could influence my rental choice/recommendations in the future.)

[Django]

I see no reason to doubt the official 8.3ms claim.. cinema5d even rated it at 6ms:

 

..they’re probably a little off but we’re definitely in that ultra low zone it seems which is indeed an extra bonus point for the 300mk2..

[Yehouda]

I never had any problem of rolling shuter with the C300 II even handheld with quite fast movements.

PS: I have never  used C200 so I can't talk about this camera.

[IronFilm]

7 hours ago, HockeyFan12 said:

That's exactly what I'm curious about. The C300 mk II has a 120hz sensor (confirmed by canon) which would equate to 8.3ms of rolling shutter. 

I don't believe either camera has processing to account for rolling shutter.

You totally need to account for the processing power of the camera. 

It doesn't matter if your 16K sensor can run at 12,000KHz if behind the scenes if you've got a Rasberry Pi which is only reading a dozen lines per milisecond off the sensor. 

Think of it as a bit like later on in the pipeline, with your media, it doesn't matter if you camera can record 4K ProRes HQ if you're only putting a 10MB/s card into the camera! As you'll by throttled by the bottleneck which is the card write speed. 

Another point to think about: why do some cameras have different degrees of rolling shutter depending on what resolution you're shooting as? (for example 4K vs 1080 from the Samsung NX1) Because they've got the same sensor in both cases, which can run at the same max read speed surely! Right?? Nope, it is because the camera's processor is being limited by how much bandwidth it can handle, thus at full resolution there is more rolling shutter, yet at reduced resolution there is less bandwidth to handle so in return the processor can boost the rate at which the sensor is read (thus increasing the amount of data read up to the max bandwidth the processor can handle)

[HockeyFan12]

1 hour ago, IronFilm said:

You totally need to account for the processing power of the camera. 

It doesn't matter if your 16K sensor can run at 12,000KHz if behind the scenes if you've got a Rasberry Pi which is only reading a dozen lines per milisecond off the sensor. 

Think of it as a bit like later on in the pipeline, with your media, it doesn't matter if you camera can record 4K ProRes HQ if you're only putting a 10MB/s card into the camera! As you'll by throttled by the bottleneck which is the card write speed. 

Another point to think about: why do some cameras have different degrees of rolling shutter depending on what resolution you're shooting as? (for example 4K vs 1080 from the Samsung NX1) Because they've got the same sensor in both cases, which can run at the same max read speed surely! Right?? Nope, it is because the camera's processor is being limited by how much bandwidth it can handle, thus at full resolution there is more rolling shutter, yet at reduced resolution there is less bandwidth to handle so in return the processor can boost the rate at which the sensor is read (thus increasing the amount of data read up to the max bandwidth the processor can handle)

I'll be the first to admit I don't know much about this, so anything's possible. That does contradict my understanding of how sensor readout works (the debayer process is done separately from the readout and entirely after, rather than being done simultaneously line-by-line, and so the sensor readout is affected by earlier sensor-related issues like binning, line skipping, refresh rate, other on-sensor processes, but not by the speed of the subsequent debayering process–i.e. it doesn't "wait up"). But I know nothing about how it works, so fair enough. My bad for not reading more literature on this. I really just meant to ask the question, not start an argument. It still confuses me since the C300 Mk II has a 120hz sensor but 30hz debayer process, resulting in ≤8.3ms of skew, consistent with the 120hz number, not the 30hz number, hence my understanding that the processes were distinct–or at the very least that the sensor doesn't "wait" up for the debayer process, even if it's technically done line-by-line, too. (Again I don't know.) But I REALLY know nothing about this and you're right that the sensor is interacting with other parts of the camera and that can alter the readout speed, I just believed that was more to do with autofocus, binning, etc. (anything sensor-level) than with debayer, which was done entirely after readout. Regardless, that was just my ignorant misunderstanding. Ultimately, I'm just trying to figure out a number for matchmoving.

Regardless, I contacted Canon directly and got the answer regarding the readout speed on both cameras. Which also contradicts what you wrote, but again, I'm not an expert and I'll ask their sensor tech tomorrow for a clearer explanation of what you mean. I don't want to spread any more misinformation or post any information I shouldn't, either.

My apologies again for posting any misinformation. Was just curious if the two cameras have the same amount of rolling shutter or not.