It not 1080p no but is 1080 just interlaced as the signal the signal output from the AF is 1080i or 720p (depending on your scene file), your small HD then scales the image to its resolution own resolution.
AF100 external recorder - Ninja 2 or Samurai?
- Thread starter Kindredmachine
- Start date
It not 1080p no but is 1080 just interlaced as the signal the signal output from the AF is 1080i or 720p (depending on your scene file), your small HD then scales the image to its resolution own resolution.
Hey guys just got the samurai yesterday and looking forward to playing with it tomorrow! Read the manual but just a quick few questions, firstly what apple pro res codec do you guys usually opt for, the HQ (as it's the best quality right?). Is there a noticeable difference in quality between that and the 4:2:2 one? And niave I know, but the HQ, it still records in a 4:2:2 colour space right? As that's the one of the main points of using it over the AF100s 4:2:0. Also, saw a little thing in the manual about monitoring options, a blue screen assist function that is suppose to help detect noise etc when monitoring, how exactly do you use this for that purpose though? :/
Good tip Shooter, I will be sure to play back via the Samurai. And John, just saw the idea about recording the slow mo off the camera, that's genuis! So simple, and it really works effectively!? Can't wait! How do you attach the unit to your cam?
Good tip Shooter, I will be sure to play back via the Samurai. And John, just saw the idea about recording the slow mo off the camera, that's genuis! So simple, and it really works effectively!? Can't wait! How do you attach the unit to your cam?
In my opinion, I always try to record the best quality I can. While uncompressed can be overkill for most applications, when it comes to compressed formats, why skip on the quality to save a few GBs when SSDs are so cheap these days? Going to the trouble to record on a external recorder just cheap out on the quality settings doesn't make much sense IMO.
Can't help you with the specific Samurai questions as I don't use one.
I agree, use 422HQ, 422 is the colour resolution which as you noted is the same for all the formats, 422LT, 422, 422HQ are the compression levels its HQ beng the least. As Adam says, why cheap out on quality now that you are using the recorder?
cant help with the samurai controls I'm afraid. Surely it has a manual or you can get a PDF online?
Have you guys done any A/B tests to back up these claims?
If so, would you mind posting some clips showing the same shot in the native AVCHD codec and in the externally recorded codec - be that a 100 Meg codec or a 250 Meg codec or whatever you prefer? Perhaps demonstrating how you've needed to subsequently manipulate it and have somehow benefitted from the higher data rate?
Sorry, not having a dig at anyone personally - but where's the evidence?
Happy to be convinced - please point me to the right place.
Ben.
As a general rule of thumb, for most uses ProRes 422 will work fine for you. If you plan to do some post processing, then HQ is a good idea. I pretty much just keep mine set on HQ, however.
As far as the "blue gun" setting is concerned (whatever they call it, I don't have my Samurai at hand at the moment), this is a standard feature on most any high end monitor and is used for calibrating the monitor via SMPTE color bars. Unfortunately, the Samurai doesn't have many calibration controls, but fortunately, I find the blue screen image to be right on when checking the color bars from the AF100.
Turn on the "blue" setting and feed color bars and see if all the bars (top and bottom parts) now look solid and all bars across look the same. The PLUGE line should be barely visible.
Search online for full information on setting up a professional, or broadcast, monitor.
As far as the "blue gun" setting is concerned (whatever they call it, I don't have my Samurai at hand at the moment), this is a standard feature on most any high end monitor and is used for calibrating the monitor via SMPTE color bars. Unfortunately, the Samurai doesn't have many calibration controls, but fortunately, I find the blue screen image to be right on when checking the color bars from the AF100.
Turn on the "blue" setting and feed color bars and see if all the bars (top and bottom parts) now look solid and all bars across look the same. The PLUGE line should be barely visible.
Search online for full information on setting up a professional, or broadcast, monitor.
kyriakides
Member
Please someone here can help me to find the kelvin settings in the Af101E? I only find the 3.2 k and the 5.6k and the black balance.
Thanking you in advance Costas Kyriakides
Thanking you in advance Costas Kyriakides
filmguy123
Veteran
Was just browsing this post, I do not have the Samurai but I am familiar with the codecs. I just wanted to point out though, you are asking people to do an awful lot of work to prove something to you lol. You want them to find or make comparison clips for you, manipulate them, and document their findings?? I think a better bet would be to:
1.) Research what Apple says about the various ProRes codecs
2.) Google it
3.) Search the forum
Tests have been done. For many uses you will not see a different. More colorspace and higher bitrate comes into play for the following uses:
1.) Intense color grading
2.) Green Screen work
3.) Other visual effects work
Basically, every pixel and the color space is more distinct and more accurate.
Frankly I agree with their advice. If you aren't using it for these purposes and figure "meh, good enough" then why did you buy an external recorder in the first place?!? They give that advice because anyone buying an external recorded is looking to maximize quality... if you were just gonna shoot with the lowest, why add the extra cost and bulk? The internal SDXC recording does a "good enough" job if you're not after more color space and higher bitrate.
I'm sure you can find comparisons online.
Last edited:
Thanks filmguy. A bit of clarification from my end. I'm very familiar with the Apple ProRes codec. I worked quite closely with the Apple ProApps team, while I was the post production consultant on the BBC DMI/Fabric - their £80 Million project to create a single digital archive and set of production workflows across the entire corporation of 24,000 people. I also was on the small group of people that decided upon the AVC Intra 100 Meg codec for BBC HD Acquisition and Post. We couldn't use ProRes as a codec as it wasn't SMPTE ratified and it would mean getting in bed with one supplier, which is against their procurement rules.
I haven't personally bought one of these recorders. I've played with them and, in the case of my AF101, I couldn't find any discernible difference between a ProRes 422 clip from a Ninja and the identically recorded clip on the SD Card. All done on a BBC R&D broadcast monitor, with the anecdotal help from a couple of highly respected BBC HD engineers. I've also been a supervising film editor and post production manager for BBC, Discovery, Nat Geo etc for 25 years, so my eyes are fairly well trained.
It wasn't a definitive test and was done relatively quickly just to inform a potential purchase, so I'm interested to see if anyone out there has something else that would convince me otherwise. The link Gary posted would confirm to me the lack of discernible difference, if it wasn't for the fact that it's a highly compressed Vimeo file. The obvious places to look for compression artefacts - sky, building blocks, water - anything that might challenge a Long GOP format - look identical when played back on my 27" iMac in Vimeo HD. A horizontal A/B comparison, rather than a distracting vertical red line would have been arguably more useful too.
My suspicion is that these recorders as improvers of quality are the Emperor's New Clothes - I totally get the workflow advantage of running off straight from the shoot with Apple-ready ProRes movs - it's worth it just for that, if you need a quick turnaround. It was an attempt to speed up my ingest workflow that made me look at the Ninja in the first place. But increased perceptible, tangible, qualitative improvements in visual quality? I'm not seeing it.
Go on, somebody, put me out of my misery and please post something that you have done to convince yourself of the differences between the native 24 Meg codec and a parallel HDMI/SDI recorded ProRes codec - beyond buying it and being convinced of the difference anecdotally with subsequent material.
Perhaps you've done a test and you'd be willing to upload the originals? If so, I have an FTP uplink on my company website - www.matobo.co.uk
I'd be delighted to be shown and convinced that I've got this one totally wrong! I just can't get this sneaky feeling out of my head that there's too much group-think going on here.
Ben.
You can download the file from Vimeo to see a better comparison.
The difference will be less noise in heavy grading (i. e. power windows, secondaries, white
balance fixes, ect.) For most whose "color grading" consists of brightness/contrast/saturation tweaks (maybe even some Fast Color Correction action!) there will be little difference.
Of course, the higher bitrate will also help with smaller details retaining quality in complex scenes, stuff in motion, water, grass, ect.
The difference will be less noise in heavy grading (i. e. power windows, secondaries, white
balance fixes, ect.) For most whose "color grading" consists of brightness/contrast/saturation tweaks (maybe even some Fast Color Correction action!) there will be little difference.
Of course, the higher bitrate will also help with smaller details retaining quality in complex scenes, stuff in motion, water, grass, ect.
As someone else says on the Vimeo forum: "The file to download has only 17 mbps 4-2-0 and 8 bits. This is why there is no differences."
I'm afraid the compression down to MPEG4 renders any comparison meaningless. It would be like comparing an MP3 Audio file with an uncompressed AIFF and then uploading the test as a subsequent MP3 - it takes everything down to the lowest common denominator.
Gary, did you shoot the Vimeo test? If so, do you have any of the untouched A/B originals, that you'd be happy to upload to me?
Thanks in advance,
Ben.
I'm afraid the compression down to MPEG4 renders any comparison meaningless. It would be like comparing an MP3 Audio file with an uncompressed AIFF and then uploading the test as a subsequent MP3 - it takes everything down to the lowest common denominator.
Gary, did you shoot the Vimeo test? If so, do you have any of the untouched A/B originals, that you'd be happy to upload to me?
Thanks in advance,
Ben.
filmguy123
Veteran
bengiles,
Again the difference is in post. It probably won't be discernible to your eye. It will be on green screen work, where very fine differences in color and pixel separation are important, or when cranking saturation/midtones/etc. The footage breaks down when pushed hard for color grading, stretching range, etc. Most people who use these recorders do so as they will be abuse it after the fact. It's not just to get something better to the eye. Even if it was, most clients can't even tell the difference between 25mbps+ Blu Ray and a compressed HD web clip.
The benefits are primarily for the 3 tasks I mentioned in the second half of my post.
If you are just storing this for archive to be viewed only, you may not even need something as robust as ProRes LT. If you are shooting footage for visual FX or are planning substantial grading, you will most certainly want 422 HQ for maximum flexibility.
Again I don't own one of the recorders so I'm not defending a purchase, but I have experienced the AVC codec break down in post... making me wish I had one of the recorders, from time to time. I am amused though at how much effort you seem to be pushing people to go through to try and prove/defend to you the codec difference, as if the forum is full of atomos sales reps, instead of asking the rather large favor of their time in uploading sample material for you. You know this is something you could rent and test yourself, right?
Again the difference is in post. It probably won't be discernible to your eye. It will be on green screen work, where very fine differences in color and pixel separation are important, or when cranking saturation/midtones/etc. The footage breaks down when pushed hard for color grading, stretching range, etc. Most people who use these recorders do so as they will be abuse it after the fact. It's not just to get something better to the eye. Even if it was, most clients can't even tell the difference between 25mbps+ Blu Ray and a compressed HD web clip.
The benefits are primarily for the 3 tasks I mentioned in the second half of my post.
If you are just storing this for archive to be viewed only, you may not even need something as robust as ProRes LT. If you are shooting footage for visual FX or are planning substantial grading, you will most certainly want 422 HQ for maximum flexibility.
Again I don't own one of the recorders so I'm not defending a purchase, but I have experienced the AVC codec break down in post... making me wish I had one of the recorders, from time to time. I am amused though at how much effort you seem to be pushing people to go through to try and prove/defend to you the codec difference, as if the forum is full of atomos sales reps, instead of asking the rather large favor of their time in uploading sample material for you. You know this is something you could rent and test yourself, right?
Last edited:
Hi filmguy (sorry, I don't know your name.)
The test I did deliberately involved significant manipulation, as it was shot to be flat/technical. Everything I've worked on over the past 25 years has been graded well beyond a first light.
I cut music promos for U2, Oasis, Blur, Paul Weller, Jamiroquai plus various high quality network commercials back in the early 90s as well as numerous "landmark" BBC documentaries and dramas since then. I know my way around a Pandora Pogle and I'm no stranger to 3 or 4 pass grading.
I guess the thrust of where I'm going with this is to say that I've read many posts with opinions about how a parallel recorded ProRes stands up so much better when it's manipulated in post - I'd just love to see some concrete examples. And to have those must surely be a good thing for the community on this board.
Ben.
The test I did deliberately involved significant manipulation, as it was shot to be flat/technical. Everything I've worked on over the past 25 years has been graded well beyond a first light.
I cut music promos for U2, Oasis, Blur, Paul Weller, Jamiroquai plus various high quality network commercials back in the early 90s as well as numerous "landmark" BBC documentaries and dramas since then. I know my way around a Pandora Pogle and I'm no stranger to 3 or 4 pass grading.
I guess the thrust of where I'm going with this is to say that I've read many posts with opinions about how a parallel recorded ProRes stands up so much better when it's manipulated in post - I'd just love to see some concrete examples. And to have those must surely be a good thing for the community on this board.
Ben.
Last edited:
Yes - I'm sorry, as I said earlier, I've done a test and couldn't find any discernible difference.
I'm just hoping there may be others out there who do actually own one of these and have satisfied their intellectual curiosity by doing their own A/B test.
That's why I think forums like this are great for sharing knowledge (as well as opinions.)
Cheers,
Ben.
You have to go back a year or two and dig up some old threads. There were a bunch of tests when the AF100 1st came out, here and various other websites and forums
eg
http://www.dvxuser.com/V6/showthrea...ot-with-Nano&p=2239992&viewfull=1#post2239992
Shooter
Veteran
Re Debating and Testing various codecs.
IMHO and experience...one can not see any significant of quantifiable differenceunder normal lpreviewing and reviewing situations.
A real valid test is to hand off an identical shot recorded on dual codecs (eg AVCHD and ProresHQ) to a colorist and let him loose on them.
The colorist will be able to make a judgement based on the "user" experience and the comments will be the value of the exercise.
The colorist may be able to show you side by sides as the testing is done and you may be able to get a "visual " result at the conclussion iof the session.
IMHO and experience...one can not see any significant of quantifiable differenceunder normal lpreviewing and reviewing situations.
A real valid test is to hand off an identical shot recorded on dual codecs (eg AVCHD and ProresHQ) to a colorist and let him loose on them.
The colorist will be able to make a judgement based on the "user" experience and the comments will be the value of the exercise.
The colorist may be able to show you side by sides as the testing is done and you may be able to get a "visual " result at the conclussion iof the session.
Shooter
Veteran
Well..I have screeds of dual recorded footage. The issue is uploading them to you (file size wise).
One of your post already mentions that you have some Prores HQ and SD footage. Is that not suitable for this exercise?
My experience with colorists and my own work in Da Vinci has always produced a simple result or determination.
"4 2 2 footage is more flexible and pliant to work with than 4 2 0"
The controls have more "control and subtlety" over the image with 4 2 2. (There is the answer right there)
Pushing cc and grade till something breaks is pointless. Who does that or why?
CC and grade is about extracting the best from the footage so if the footage can be tweaked easier with more finesse...an obvious advantage.
One comment colorists always make.
"Well shot footage on any codec ...is better than badly shot footage on the best codec around".
FWIW: I have got to the point with codecs that the limiting factor for me is now HD rez. (1920 x 1080)
One of your post already mentions that you have some Prores HQ and SD footage. Is that not suitable for this exercise?
My experience with colorists and my own work in Da Vinci has always produced a simple result or determination.
"4 2 2 footage is more flexible and pliant to work with than 4 2 0"
The controls have more "control and subtlety" over the image with 4 2 2. (There is the answer right there)
Pushing cc and grade till something breaks is pointless. Who does that or why?
CC and grade is about extracting the best from the footage so if the footage can be tweaked easier with more finesse...an obvious advantage.
One comment colorists always make.
"Well shot footage on any codec ...is better than badly shot footage on the best codec around".
FWIW: I have got to the point with codecs that the limiting factor for me is now HD rez. (1920 x 1080)