PDA

View Full Version : FS100 and Ninja: Yes, No, Maybe?



matteo1
07-14-2011, 01:12 PM
I'm just about to buy a FS100 (with much help from this forum) and am wondering about adding a Ninja to my order. The project I"m working on is in the bush in Africa. The FS100 is the "B" camera – Sony 800 is the "A" camera – and the FS100 will be used for run-and-gun. I know FS100 is not the perfect beast for running around baobobs, but for the first shoot I was using a 5D and at least with the FS100, I won't spend so much time syncing. The producer plans to do an Academy run next year, meaning that the documentary will be projected at some point.

Right now I'm planning on purchasing the Sony HXR_FMU128_Flash_Memory_Unit, but I'm wondering if it might make more sense to buy a Ninja instead. I know they do completely different things. We're shooting 1080p 24 and I plan to do some 1080p 60. I realize that the Ninja doesn't really do 24P natively (it takes in the 24P at 1080i 60 and then you have to run it through the Ninja "Stripper" app). And, of course, like all the other digital recorders, it doesn't do 1080p 60 at all.

So . . . . Knowing that I'm shooting 24P and will definitely shoot some 1080p 60 for slow motion, does running everything into a Ninja gain me all that much in image quality to buy the Ninja? Will it help out appreciably down the road when we're color grading the footage?

Also, if you would recommend getting the Ninja, which SSD would you recommend (I don't think it makes sense to use HD for this kind of thing)

Thanks.

legrevedotcom
07-14-2011, 01:22 PM
I'm sticking around for a listen on this as well... The various threads around here go from one extreme to the other in regards to what is possible / sensible wether to shoot on card or recorder.

Have I understood correctly that the FS100 more or less records the same quality to card as one would get from a 5D?? And it's not until you go recorder you get the good stuff? I'm asking since one of my mates have been able to create decent stuff in post with the mediocre quality of the 5D.

millerandmiller
07-14-2011, 02:00 PM
Join the club, really hoping a solution will come out soon. Would be great to record out in the best quality.
However being a 5D user since launch I have taken very fondly to the FS100 the resolution increase over the 5D is apparent when editing and that follows all the way though.

I'm still mostly testing but here are my results so far.
http://vimeo.com/26419694
http://vimeo.com/25638703
http://vimeo.com/26319752
http://vimeo.com/26032986


James

Barry_Green
07-14-2011, 02:08 PM
Have I understood correctly that the FS100 more or less records the same quality to card as one would get from a 5D??
No, the FS100's codec is substantially advanced over that which is in the 5D.


And it's not until you go recorder you get the good stuff?
Well, that depends on the recorder. If you're using the Sony CF recorder, it's no different -- it's exactly the same as what's recorded on the internal cards. If you're using a Ninja or Nano, then the codec can be substantially higher bitrate and with higher color sampling than what gets recorded to the onboard recorder.

zeke
07-14-2011, 02:32 PM
If you don't mind recording 8bit 4:2:2, the nano has proven itself in demanding environments. Recording at 100mb mpeg2 will enable you to maintain quality in post.

mico
07-14-2011, 03:18 PM
Proof positive that the Ninja works very well. Perfect test lots of motion.

http://vimeo.com/25060362

addax
07-14-2011, 03:37 PM
I heard the "regular" battery of the FS100 doesn't feed the Ninja in order to get smooth recording, can anybody clear that?

matteo1
07-14-2011, 03:43 PM
Mico,

What drive were you using?

mico
07-14-2011, 03:57 PM
It's not my video but Atomos has repeatedly said that Intel drives are the most reliable. I believe they were talking SSD's. Ask the vimeo poster or go to Atomos site to see their recommended drives because not all drives work well for the ninja. Theres also a thread in the Af-100 forum about the ninja where I first saw this video.

P. Harrill
07-14-2011, 07:54 PM
As far as I can tell, the OP has not shot with the FS100. True?

And as far as I can tell, no one on this thread has both a Ninja and a FS100, so no one posting here has direct evidence of the Ninja producing appreciably better results than the camera itself. True?

The video linked to above is for a Panasonic AF-100... yes, that camera shoots to AVCHD like the FS100, but it's a different sensor. Hell, it's a different manufacturer. Is anyone seriously willing to take a test like this as proof -- actual proof -- of anything with regards to the FS100?

As Barry suggests, the implementation of AVCHD in this camera is quite good.

If you've not shot with the camera, the OP should do so before buying anything like an external recorder. That's my advice.

Here's more advice:

1) Get your FS100 and rent or borrow a Ninja.
2) Shoot your own tests approximating the conditions you're going to be filming under.
3) Consider how the equipment will make things easier or harder.
4) After you've done YOUR tests, ask yourself: Will my audience notice? And if they do notice, will they care?
5) Finally, consider the budget, especially what you have to sacrifice if you get that item. (All budgets are limited, so all choices involve sacrifices in one direction or another.)

I'm not knocking the usefulness of these forums, but to base purchasing/rental decisions on some vimeo test of a recorder with a different manufacturer's cameras… sorry, that wouldn't be enough information for me.

David Shapton
07-16-2011, 07:36 AM
@P.Harrill

Improved quality is not the only advantage of using an external recorder, but I won't go into that here.

Just want to mention a couple of things.

First, the Ninja has it's own batteries, made by us, but in the Sony NP format. The supplied batteries will give you a minimum of 5 hours recording time and probably more. It does not take power from the FS100.

Second, we have a lot of satisfied users with the FS100.

Third, as Barry says, AVCHD is an efficient codec that is capable of looking good, and most of the time it does. So casual side-by-side comparisons with ProRes won't show drastic improvements - especially through the fog of YouTube. BUT... when the going gets tough; when you reach the limits of AVCHD, that's when ProRes comes into it's own. You can think of it as a safety net in that sense: it's always going to get the shot, even when AVCHD crashes out with artifacts and other unpleasant stuff. Just try capturing rippling water or rapid motion with high detail in AVCHD. If you're just filming a sleeping dog, then you're not going to see much difference.

Think of it like this. ProRes is visually lossless. It looks the same, even to experts, as uncompressed. Does AVCHD look the same as uncompressed? No.

Dave Shapton
Atomos

Mark Crabtree
07-16-2011, 10:23 AM
Dave,Does CineForm converted from prores look as lossless as CineForm converted from uncompressed?

David Shapton
07-16-2011, 10:42 AM
Dave,Does CineForm converted from prores look as lossless as CineForm converted from uncompressed?

Mark, I don't know because I've never tried that - especially under laboratory conditions. ProRes isn't mathematically lossless; only visually: so I'd imagine that Cineform converted from ProRes would not theoretically be as good as directly from uncompressed, but I would still expect the result to be very good and quite possibly still visually lossless, depending on the exact definition on the term, since these codecs are built to withstand multiple generations of encoding (with generally better results if you stay within the same codec). But this is something you should really be asking Apple and Cineform.

Dave Shapton
Atomos

Mark Crabtree
07-17-2011, 12:42 PM
Dave,Have the playback features been enabled yet?

David Shapton
07-17-2011, 02:37 PM
Dave,Have the playback features been enabled yet?

I think I've answered this recently elsewhere. We're very close to releasing a new version with playback, and a whole bunch of other features. We've implemented playback in a really good way which we think all Ninja users will appreciate.

Dave Shapton
Atomos

Ian-T
07-17-2011, 02:49 PM
As far as I can tell, the OP has not shot with the FS100. True?

And as far as I can tell, no one on this thread has both a Ninja and a FS100, so no one posting here has direct evidence of the Ninja producing appreciably better results than the camera itself. True?

The video linked to above is for a Panasonic AF-100... yes, that camera shoots to AVCHD like the FS100, but it's a different sensor. Hell, it's a different manufacturer. Is anyone seriously willing to take a test like this as proof -- actual proof -- of anything with regards to the FS100?

As Barry suggests, the implementation of AVCHD in this camera is quite good.

If you've not shot with the camera, the OP should do so before buying anything like an external recorder. That's my advice.

Here's more advice:

1) Get your FS100 and rent or borrow a Ninja.
2) Shoot your own tests approximating the conditions you're going to be filming under.
3) Consider how the equipment will make things easier or harder.
4) After you've done YOUR tests, ask yourself: Will my audience notice? And if they do notice, will they care?
5) Finally, consider the budget, especially what you have to sacrifice if you get that item. (All budgets are limited, so all choices involve sacrifices in one direction or another.)

I'm not knocking the usefulness of these forums, but to base purchasing/rental decisions on some vimeo test of a recorder with a different manufacturer's cameras… sorry, that wouldn't be enough information for me.


But P.Harrill, you can apply the result of that AF100/Ninja video to most cameras that use AVCHD. Sure they are not the same sensor... But in a prosumer camera like the FS100's price range I would expect the HDMI out to be of better quality than a "compressed" AVCHD. This should give someone at least a good idea on what to expect. IMO the only difference among all of these cameras would be the actual RAW image being fed from the sensor.

Mark Crabtree
07-17-2011, 08:39 PM
I think I've answered this recently elsewhere. We're very close to releasing a new version with playback, and a whole bunch of other features. We've implemented playback in a really good way which we think all Ninja users will appreciate.Dave ShaptonAtomosSo, you are going to release a firmware update that will give all Ninja's playback and other features? Are we talking days or weeks?

P. Harrill
07-17-2011, 11:58 PM
I was making two basic points:


1) You're the best person that understands what you need for a project, so if you want to know whether a piece of equipment is going to work well for you, the best way to know is to test it for yourself.


2) If you're going to forego doing your own tests, use tests that can give you an actual depiction of the items you plan to use and take advice from authorities with direct, first-hand experience (e.g., don't watch tests of an external recorder with a different manufacturer's camera, even if the codec is the same).


If these are controversial opinions on DVXUser, geez, I might as well check out now.


Yes, Ian, the FS100 and the AF100 are kinda the same. But the question isn't whether the Ninja (or any other external recorder) does a better job recording than the camera's codec. (I should hope so, considering you're paying extra for it.) The question -- that is, the question that the OP originally asked -- is:


Is something like this WORTH IT?


And, still, no one has answered that question.


Why? Simple. Because no one on this board (other than David) has any experience with the Ninja (see point #2 above) and because it's difficult to answer questions like that for anyone other than yourself (see point #1 above).

DVC
07-18-2011, 03:55 PM
I think I've answered this recently elsewhere. Dave Shapton, Atomos

David, sorry to intrude on FS100 thread, but have you implemented a FORCE PROGRESSIVE (1080p25 or 1080p30 for ProRes recorded files) menu item for use with 1080i50/25PsF and 1080i60/PsF60 such as that from the VG10?

Dermot
07-22-2011, 08:47 PM
i have been told there was a firmware update yesterday, has anyone had a chance to test with the revised firmware?

d

millerandmiller
07-23-2011, 02:49 AM
I was told the new firmware would be in a couple if weeks.

David Shapton
07-23-2011, 04:27 PM
The new firmware is in the final stages of testing. As soon as this stage is completed it will be released. Keep an eye on our website.

Dave Shapton
Atomos

rejdmast
07-23-2011, 04:54 PM
Dave,

What is the status of the Ninja being able to be remote triggered by the FS100 HDMI output with TC? I heard you guys are looking into it.

Ryan Lightbourn
07-23-2011, 06:15 PM
So . . . . Knowing that I'm shooting 24P and will definitely shoot some 1080p 60 for slow motion, does running everything into a Ninja gain me all that much in image quality to buy the Ninja?

*edit* apparently I misunderstood what the Ninja was supposed to do *

I ordered one for my FS-100 thinking it would hold more detail in the darks (and possibly highlights), and maybe a bit more latitude...but this is not the case. I'm actually sending my Ninja back to B&H.

Barry_Green
07-23-2011, 06:16 PM
Well, no, it shouldn't show any more info in the darks or the highlights. That's not what it's meant to do. What it should do is give you some higher color sampling, and more resistance to codec mush in highly detailed scenes.

Ryan Lightbourn
07-23-2011, 06:19 PM
Ok well that makes sense...I edited my post!

morgan_moore
07-24-2011, 12:27 AM
..What it should do is give you some higher color sampling, and more resistance to codec mush ..

...Which should enable you to shoot a with a flatter camera profile gaining you wider DR and due to the increased color sampling the footage will not break when graded up to a decent contrast resulting in..


hold (ing) more detail in the darks (and possibly highlights), and maybe a bit more latitude

??

S

David Shapton
07-24-2011, 12:35 AM
Ryan,

Working in a 10 bit space will also help with grading. It's not going to improve the quality of your original 8-bit footage, but any changes you make will have 1024 levels per colour channel to map onto rather than 256. This will give you some of the benefits of increased latitude as the darks will not degrade as quickly when you push them.

Dave Shapton
Atomos

morgan_moore
07-24-2011, 12:43 AM
Dave you might clarify my 'understanding' of this

The 8 bit codec breaks in a grade really easily due to those lack of colours (256 vs 1024)

So if you shoot with a flat profile in the camera to gain DR the (8bit) footage will often artifact (particularly colour banding) when graded from flat to viewable

Super flat profiles are therefoer not advisable in an 8bit space

If you use a 10 bit recorder the footage can take more of a grade due to more colour information

You can therefore gain DR by shooting a flat camera profile and grading

ie

The recorder does not give more DR unless used hand in hand with a new camera profile

Shooting a 'regular' profile will not see any advantage beyond retention of detail (due to bit rate)

To get best results (from a recorder) one should be shooting with a profile aimed at the grading process

S

David Shapton
07-24-2011, 12:56 AM
Morgan,

Please understand I'm not an expert when it comes to grading!

Recorders can only record what they're given, so, yes, if there's an 8 bit input to the recorder, that's what gets recorded.

The sole advantage with grading (apart from the increased bitrate/better codec and all of those pluses) is that there is more precision available when a colour is changed from the original. If you stay in an 8 bit space, then any change you make will have to be "forced" onto another of those 8 bit levels, whereas in a 10 bit space, you'll have four levels for each original one.

I wouldn't say that you'll get more DR; the 8 > 10 bit conversion doesn't put new information into your video, but you'll be able to modify it in more subtle ways.

Dave Shapton
Atomos

morgan_moore
07-24-2011, 01:28 AM
OK I was talking as if the camera outputs 10bit to the recorder

S

kinoeye
07-25-2011, 07:15 PM
What software can be used right now to do a reverse telecine on the Atomos footage?

kinoeye
07-25-2011, 07:39 PM
Seems that I can't get rid of the aliasing from the Ninja.
Here's a screenshot from FCP. the image on the left is from the Ninja, and the one on the right is from the SD card in the FS100.37364

NeedCreative
07-25-2011, 08:55 PM
Seems that I can't get rid of the aliasing from the Ninja.Here's a screenshot from FCP. the image on the left is from the Ninja, and the one on the right is from the SD card in the FS100.37364 I have one coming tomorrow. Curious- how are you converting the footy back to 24p?

Alpern
07-25-2011, 09:10 PM
After Effects, Compressor, MB grinder. I have used compressor and after effects and they work well. I have not tried grinder but Atomos told me it works.

millerandmiller
07-26-2011, 09:19 AM
Guys this is what I used. Old mac Quicktime dev app called 'Dumpster'.

No transcoding or even saving of the entire file, its instant.


Open the file from the Atomos and change "Fields = 2 Details = 1" to "Fields = 1 Details = 0" click save (instant) import in to FCP, FCP will see the clip as progressive and drops into a progressive sequence.


http://www.mmfilm.co.uk/twitterpics/dumpster_grab.jpg


P.S. 'Dumpster' is an old Apple Dev program. You won't be able to run it in Lion. Just up to Snow.


I found someone who put a link on an Apple support page. (This is not my link) http://idisk.me.com/jrwalker4-Public/Dumpster.dmg

MattDavis
07-26-2011, 09:33 AM
37364

I don't think that's aliasing. It looks like the kind of trouble you get putting a Progressive clip on an Interlaced timeline. Try setting the timeline interlace settings to 'None'.

millerandmiller
07-26-2011, 11:05 AM
I don't think that's aliasing. It looks like the kind of trouble you get putting a Progressive clip on an Interlaced timeline. Try setting the timeline interlace settings to 'None'.

Unfortunately that does not work with these clips. They look soft at 100%. Zoom in and you will see the soft interlaced artefacts. The only quick way is by changing the flags on the file to show as Progressive. As my post a couple above http://www.mmfilm.co.uk/twitterpics/dumpster_grab.jpg . Then they will show perfect in FCP.

NeedCreative
07-27-2011, 06:52 AM
After Effects, Compressor, MB grinder. I have used compressor and after effects and they work well. I have not tried grinder but Atomos told me it works.

Thanks. I couldn't get grinder to work at all (latest version); it didn't seem to reverse telecine correctly, always leaving artifacts no matter what I did. Compressor worked perfectly after I got my frame rate correct. (oops!)

I hope Atomos' Stripper does the change in the metadata (?) of the file similar to James' method so that a recompress isn't necessary. You're not losing much if anything when recompressing ProRes but there is a time hit for sure.

Dermot
07-27-2011, 08:32 AM
Or just use an Avid?

::::::FLEES::::::

d

JardaBar
10-09-2011, 03:56 AM
Is 1080/50p (1080/60p) video format supported on Ninja? Thx.

nomad-3
10-09-2011, 04:55 AM
Nope. And it'll never be – hardware limitations.

JardaBar
10-10-2011, 01:00 AM
Thank you. So we must wait for some new toy. :-(

FelixGER
10-15-2011, 06:18 AM
Im wondering if there is any possibility to mount the Ninja on the FS100 and still use the camera handheld.
What do you guys think about the extra weight (1 Kilogramm) und where to place it?

nomad-3
10-15-2011, 10:01 AM
Sure, why not? But I'd place it on the back of a shoulder support as counterweight. The screen is not any better than the Sony's.

Only Problem: no remote start/stop (yet?). I hope Atomos could bring us TC start in an update, since the FS100 is delivering TC via HDMI.

JMartin
10-15-2011, 06:46 PM
TC support over HDMI would certainly be good, but I would have thought that LANC would have been easier.
Not sure why exactly but the implementation on the FS100 is problematic.
http://atomos.activehosted.com/kb/article/cameras/fs100/fs100-and-lanc