Erik Naso
Veteran
Why on earth would he do it that way?
C100: Is there any C100 Atomos example footage available to download somwhere?
- Thread starter Burners
- Start date
groveChuck
U-matic Member
:undecided
Curnsie
Well-known member
Thanks for the tip. I will give it a try.
Nate Haustein
Well-known member
Something with that test is screwy. The difference simply isn't that apparent. In my footage, the untouched clips looks absolutely identical. Color is better and motion artifacting is definitely better with the ProRes, but there's no way I'd say the footage is sharper right off the bat.
The other day I had the camera out for a shoot and I was curious about ProRes vs AVCHD. Here are two still shots with some curves applied and the saturation cranked way up. Personally I like working with ProRes footage, but in terms of "night and day," you guys be the judge...
https://dl.dropboxusercontent.com/u/8338487/C130_AVCHD_SS.jpg
https://dl.dropboxusercontent.com/u/8338487/C130_PRORES-HQ_SS.jpg
The other day I had the camera out for a shoot and I was curious about ProRes vs AVCHD. Here are two still shots with some curves applied and the saturation cranked way up. Personally I like working with ProRes footage, but in terms of "night and day," you guys be the judge...
https://dl.dropboxusercontent.com/u/8338487/C130_AVCHD_SS.jpg
https://dl.dropboxusercontent.com/u/8338487/C130_PRORES-HQ_SS.jpg
Erik Naso
Veteran
I regret posting that video now. I think the test procedure wasn't sound.
I like working with any codec better than AVCHD but as for image quality its pretty good. High detail shots will look better in ProRes and contrasty ones as well with saturated colors. Your samples are pretty even without dark shadows where the noise shows up more.
Hi Erik,
I, too, have had problems with my Ninja2/c100 combination. I'm logging clips from an interview I did a bit ago, and finding about 1/4 of the clips are corrupt (this is DNxHD with AVID workflow). I have the SD card as a backup, but this is very disturbing and makes me nervous with this combination. And the whole point of this setup was to get a broadcast spec, and good image (a-la-c300) at a reasonable cost. But unreliable is not working for me.
I'm curious if you've heard anything back from Atomos about this, and I would consider contacting them as well.
Thanks, Dan
Erik Naso
Veteran
I haven't contacted Atomos yet. it's on the list of things to do this week. Please contact Atomos as well. I think issues need to be addressed.
Curnsie
Well-known member
what issues are you guys having exactly? My Ninja 2 is "glitching" during recording. It stops recording for a split second but starts again immediately thereby creating 2 clips. I only noticed this because the timecode goes back to the start and there is a slight glitch in the picture on the screen. Mine does this relatively infrequently and I've noticed it reduced significantly when I changed the HDMI cable. Thankfully it hasn't happened during interviews. My setup is using a Western Digital caviar Black 500 Gb hard drive. My hdmi cable is a Belkin 3ft right angle (though I've just bought a Gefen Mono-LOC 3ft cable which I haven't tried in the field yet) The Ninja OS is the latest version. I've lived with it a i thought it low quality hdmi cables were the problem - the one supplied with my C100 Ninja 2 bundle was a £1.99 special! It also happened a lot with a slightly more expensive Philips cable. It's happening a lot less with the Belkin cable.
Erik Naso
Veteran
I had two incidents. Once with my AF-100 the Ninja 2 recorded with interlacing so I had lines in the footage but after a few recorded clips it went away and the rest of the footage looked great. This was for the C-Shooter review. I had my C100 in the rig and had to use the AF-100 to shoot the review.
The second time it was with the C300 and the Ninja 2 recorded 30 fps when the camera was set to 24p. Did the pulled down set up and recorder locked it in.
With the C100 I haven't had a issue yet, but I only used it a few times and extensively on the Canon CN-E primes review. I shot for several hours running around Balboa Park turning the unit on and off all day long without issue.
I have seen others get strange noise when recording ProRes HQ. Granted the issues I had are with other cameras than the C100.
I highly recomend using SSD drives.
The second time it was with the C300 and the Ninja 2 recorded 30 fps when the camera was set to 24p. Did the pulled down set up and recorder locked it in.
With the C100 I haven't had a issue yet, but I only used it a few times and extensively on the Canon CN-E primes review. I shot for several hours running around Balboa Park turning the unit on and off all day long without issue.
I have seen others get strange noise when recording ProRes HQ. Granted the issues I had are with other cameras than the C100.
I highly recomend using SSD drives.
My problems are similar. I'm recording to DNxHD and using a Sandisk 240 SSD. This is very worrisome to me, and to be honest, I would not have purchased the c100 if I had known this to be a problem, since I need to be able to record broadcast spec and reliably. I hope this does get fixed! Erik, do you have suggestions of who to talk with at Atomos?
Erik Naso
Veteran
Sorry no. I haven't yet reached out to Atomos. Make sure you remember your settings and if you have samples send them. I will contact Atomos this week and hopefully get one point person to talk to.
I have been messing around with the Ninja 2 tonight and noticed that with my copy I cant disable the HDMI Timecode Trigger Record function. It's stuck on off. This is supposed to be set to disabled. I sent in a ticket with Atomos. I think my recorder is broken. We shall see.
In the meantime check your firmware and make sure it's current. They have a update issued July 30th. I dont think I have this one, so that's the first thing I'm going to check.
Here is the link for the update.
Last edited:
Curnsie
Well-known member
Danny and Erik, it is very worrisome for those who record long form interviews or events!! Thankfully 90% of my shooting is corporate clips with each being a few seconds long so when the Ninja has "glitched" as I call it, I can just do a retake. Thankfully it has never happened during an interview!
I use the term "glitch" as I have 2 young sons who watch Wreck It Ralph where one of the characters "Glitches" constantly lol
Last edited:
MattDavis
Well-known member
I've held off jumping in, but feel strangely compelled to. Mainly because the Vimeo test was at odds with all my findings.
I don't have a Ninja or Samurai, I have a Pix 220 - mainly because I needed better mic audio than my FS100 and EX1R could provide - but hey, it's a 4:2:2 10 bit prores recorder. I have done a couple of jobs with the Ninja 1, though.
My experience with ProRes recorders is that most of the time, AVCHD is 'good enough' and that comparing the same shot from the internal and the external recorder, they look visually identical - though there are obvious differences on an NLE's waveform monitor and vectorscope, which show why there may be tears before bedtime further down the road in a broadcast (multigenerational) environment.
http://www.mdma.tv/c100/ - this shows people's usual experience with corporate shooting: no visible difference that would warrant the expense and/or inconvenience of an external recorder - but hey, this is a really easy test. We're not stressing the AVCHD in any way. No gain, no motion, no extreme detail.
Folks may remember this test with the FS100 - http://www.mdma.tv/420v422/ (note the lower definition, slightly more ragged chroma and dull skin tones - that's not so in the C100). These are blow-ups of a portion of the image, BTW. Point being, AVCHD is perfectly at home in these circumstances.
There ARE times when the internal recording is better than the external recorder - usually when you're using gain (okay, 'high isos') and choosing a lower bitrate codec. This can be awful with ProRes LT, as it really struggles with encoding the grainy image in its limited bitrate (http://www.mdma.tv/c100/noiseandprores-crop.jpg).
So although ProResHQ may seem to be overkill, when the going gets tough (noisy pictures, lots of motion, lots of detail, threats of banding) it may be a better choice. From that 'Noise and ProRes' test, I'm sticking with HQ from now on.
I don't have a Ninja or Samurai, I have a Pix 220 - mainly because I needed better mic audio than my FS100 and EX1R could provide - but hey, it's a 4:2:2 10 bit prores recorder. I have done a couple of jobs with the Ninja 1, though.
My experience with ProRes recorders is that most of the time, AVCHD is 'good enough' and that comparing the same shot from the internal and the external recorder, they look visually identical - though there are obvious differences on an NLE's waveform monitor and vectorscope, which show why there may be tears before bedtime further down the road in a broadcast (multigenerational) environment.
http://www.mdma.tv/c100/ - this shows people's usual experience with corporate shooting: no visible difference that would warrant the expense and/or inconvenience of an external recorder - but hey, this is a really easy test. We're not stressing the AVCHD in any way. No gain, no motion, no extreme detail.
Folks may remember this test with the FS100 - http://www.mdma.tv/420v422/ (note the lower definition, slightly more ragged chroma and dull skin tones - that's not so in the C100). These are blow-ups of a portion of the image, BTW. Point being, AVCHD is perfectly at home in these circumstances.
There ARE times when the internal recording is better than the external recorder - usually when you're using gain (okay, 'high isos') and choosing a lower bitrate codec. This can be awful with ProRes LT, as it really struggles with encoding the grainy image in its limited bitrate (http://www.mdma.tv/c100/noiseandprores-crop.jpg).
So although ProResHQ may seem to be overkill, when the going gets tough (noisy pictures, lots of motion, lots of detail, threats of banding) it may be a better choice. From that 'Noise and ProRes' test, I'm sticking with HQ from now on.
To my eye, without the use of photoshop, this looks identical.
thanks for the pics
Curnsie
Well-known member
+1. I'm flicking between Safari tabs and see no change in the images at all.
morgan_moore
Major Contributor
A little question. Lets say you had a super flat look on the camera increasing your DR, now you would have to give this a hard grade to turn it into a pretty picture. Do you think a recorder would show at that point. If so it would be 'useful' as you would gain wiggle room in post..?
MattDavis
Well-known member
Well, the first issue here is that we're dealing with 8 bit cameras. Not 10 bit. So you really have to maximise your use of dynamic range from the get-go. If you were shooting 10-bit S-Log, yeah - fill your boots with your hard grade. But with an 8 bit camera like C100, FS100/700, C300, even F3 internal recording, you're really only managing the layout of tonality rather than going for a quasi-raw workflow. I'd not want to hard-grade any of these cameras. To paraphrase the President of the Galazy himself: 'it goes to pieces so fast, we get hit by the shrapnel'.
You can either live with WDR for everything and accept a little bit of a compromise, or you can flip between WDR, C-Log and other Custom/Picture Profiles - but what one must never do is shoot, say, C-Log on a really flat overcast day with a brightness range of 'three stops' and then complain about the picture quality when you stretch that little grubby band of info in the middle of your waveform monitor out to fit to actual blacks and whites. Of course it's going to look like you processed your video with a cheese grater. It was only feeble, delicate little 8 bit in the first place, so you need to give it all the help you can give it, including walking the image to a reasonable position in the first place.
There's a time and a place (and a light) for super-flat, and if it needs a hard grade afterwards, maybe raw or 10 bit would be a happier choice.
However (and FINALLY he gets a bit nearer to the topic), shoving a 10 bit 422 recorder onto an 8 bit camera still gets you an 8 bit image, albeit carefully stencilled onto a 10 bit signal, and even that won't really help in the grade.
Oops - into tl:dr territory. Again.
You can either live with WDR for everything and accept a little bit of a compromise, or you can flip between WDR, C-Log and other Custom/Picture Profiles - but what one must never do is shoot, say, C-Log on a really flat overcast day with a brightness range of 'three stops' and then complain about the picture quality when you stretch that little grubby band of info in the middle of your waveform monitor out to fit to actual blacks and whites. Of course it's going to look like you processed your video with a cheese grater. It was only feeble, delicate little 8 bit in the first place, so you need to give it all the help you can give it, including walking the image to a reasonable position in the first place.
There's a time and a place (and a light) for super-flat, and if it needs a hard grade afterwards, maybe raw or 10 bit would be a happier choice.
However (and FINALLY he gets a bit nearer to the topic), shoving a 10 bit 422 recorder onto an 8 bit camera still gets you an 8 bit image, albeit carefully stencilled onto a 10 bit signal, and even that won't really help in the grade.
Oops - into tl:dr territory. Again.
Last edited:
morgan_moore
Major Contributor
probably missed the basic - its out putting 8bit in a 10bit wrapper - not 10 bit?
in which case I get you 100%
S
in which case I get you 100%
S
MattDavis
Well-known member
Yes, 'friad so. Even the FS700 outputs 8 bit in a 10 bit wrapper until you go with the Alphabet Soup converter and its deck. But then it's 12 bit raw. If that, and your Scarlet, is what you've been playing with recently, you lucky chap! Welcome back to 8-bit-land! LOL