|
|||||||||
|
Thread Tools | Search this Thread |
February 7th, 2010, 07:19 AM | #1 |
Regular Crew
Join Date: Sep 2007
Location: Bee Cave, Texas
Posts: 151
|
Does Convergent Design Help, Other Than 4:2:2 ?
I have an EX-1 and am seriously considering purchasing the Convergent Designs nano Flash - My primary question is - does using this device for continual I-frames REALLY help much with detail in fairly quick pans shot at 1/120th second 1920 x1080 30p ??? Or maybe it doesn't help at all - personally, I don't care about the improved colorspace from this device instead of using the stand alone typical camera recording.
I AM interested in the opinion of anyone who has actual experience with this device on any of the EX 1R, 3 products as to whether there are ANY advantages otherwise beside the improved colorspace. I am sitting here already with $5,000 worth of SxS cards (purchased at that full price 1 1/2 years ago), and am not interested in spending another small fortune in a device, its cards, and the MainConcept program for use in Adobe Premeire CS-4, unless there are some really substantive advantages! Anyone with actual experience - your reply would be much appreciated! |
February 7th, 2010, 08:57 AM | #2 |
Major Player
Join Date: May 2006
Location: Teaneck, NJ
Posts: 659
|
4:2:2 color sampling is a major improvement over the native 4:2:0 of the EX codec. Gives you significant latitude in color correction and grading and is not to be minimized.
The higher the data rate, the fewer motion artifacts you will see. Shoot MPEG-2 at 50 Mbps with a 1/2" camera and you are at broadcast specs for the toughest of media outlets. Shoot at 100 Mbps and you are at mastering quality. Just compare the frame grabs on the Convergent-Design home page showing sparklers shot with native 35 Mbps 4:2:0 and a nano. If 35 Mbps 4:2:0 is good enough for what you or your clients want, then leave it alone. If you are finding problematic compression artifacting, motion artifacting, images which do not hold up to extensive grading, stair-stepping edges-- then you need the Nano. Otherwise, enjoy your excellent system and continue shooting to SxS. Ned Soltz |
February 7th, 2010, 11:17 AM | #3 |
Regular Crew
Join Date: Sep 2007
Location: Bee Cave, Texas
Posts: 151
|
Dear Ned:
Well, thanks so much for the relavant, prompt reply! I have only one remaining question which you or someone else may answer - and if the response is affirmative, then I'm going for the nano and all the other necessary add ons. Here is something I have been frustrated with from my first shots with the EX-1 since the beginning for me in May, 2008. Where fine detail (very fine horizontal lines or perfectly horizontal very sharp edges/contrast appears), I am getting line twitter - I have, currently, two ways around this. First, if shooting in 1920 x 1080 30p, and editing in Adobe Premiere Pro CS4, before I export the timeline, I add the effect "anti-flicker", which takes FOREVER to render, even on my very powerful BOXX, but it of course reduces the beautiful detail from the picture along with all but the very finest line problems. My second solution is to shoot in 1920 x 1080 60i - beautiful picture, NO twitter, and obviously much smoother pans, but of course, just as obvious it reduces the horizontal resolution. EXCUSE ME< there I go again - finally getting to the question!!!!! Does the use of the nano POSSIBLY reduce the aforementioned horizontal fine line flicker - yes, no, or maybe, for the multiple choice response of those of you kind enough to respond! And thanks again. |
February 7th, 2010, 09:37 PM | #4 |
Major Player
Join Date: May 2006
Location: Teaneck, NJ
Posts: 659
|
I'm going with a maybe here because you're only seeing this in 1080 30p and I have no idea whether there is any relationship to the difference between 30p/60i.
It is definitely a question worth asking in the Convergent Design forum so that you might get a response from Dan Keaton from C-D. One thing that I will say is that the higher bit rate will give you much cleaner images. Wish I had a definitive answer to the flicker question. Ned |
February 8th, 2010, 04:58 AM | #5 |
Trustee
Join Date: Aug 2006
Location: Rotterdam, Netherlands
Posts: 1,832
|
Can you explain this? IMO 1920 horizontal resolution is the same in P and i.
|
February 8th, 2010, 06:33 AM | #6 |
Regular Crew
Join Date: Sep 2007
Location: Bee Cave, Texas
Posts: 151
|
Well, Harm, excuse me, and I apologize for that last, incorrect post before I had my daily allotment of two gallons of coffee!
Vertical resolution, not horizontal, is halved, or am I mistaken on that also. The problem with humility is it is humiliating! But my question still remains, does the nano reduce (or perhaps INCREASE) interline twitter? Or does it result in no discernable effect at all? Anyone out there with experience care to comment? And if not, then I'm about ready to jump over to Cow's (dare I say it here?) Ex and Convergent Design threads! But, if anyone has experience here, and if I haven't "mis-spoken" myself again, any learned response would be much appreciated before I spend another $4,000 on the nano and accessories (including the MainConcept plug-in for Premier!). |
February 8th, 2010, 05:50 PM | #7 |
Trustee
Join Date: Aug 2006
Location: Rotterdam, Netherlands
Posts: 1,832
|
William,
First of, I think this thread is very interesting and I can learn a lot, because I have no experience whatsoever with Nano. However, on the XDCAM, there is no difference in resolution between P and i. Both are 1920 x 1080 with square pixels. How the NLE treats interlaced or progressive can be a complete different world. If you ingest interlaced material into CS4 and output it as progressive, yes, you lose half your vertical resolution, due to Adobe's dumb a** algorithm, which throws away all the odd (or even) fields and just duplicates the remaining ones. IMO with the EX series, you should be of a clear mind what you want to shoot (P or i) and work with that. CS4 is not good at converting from one to the other format. Converting is never good, nor is it easy on your work flow. |
February 9th, 2010, 05:40 AM | #8 |
Trustee
Join Date: Mar 2007
Location: Sherman Oaks, CA
Posts: 1,259
|
Are you sure of the first part of what you said? Interlaced frames are composed of two fields taken at an interval equal to 1/2 * frame rate. Each field contains only even or odd lines of the image. So for any given moment, the vertical resolution for an interlaced image is one half of the vertical pixel count.
I've never heard of XDCAM doing this differently, but I learn new things every day (and forget quite a few too).
__________________
Avid Media Composer 3.1.3. Boris Red and Continuum Complete. Vegas 8.0c. TMPGEnc Xpress Pro 4.0 |
February 9th, 2010, 12:55 PM | #9 |
Trustee
Join Date: Aug 2006
Location: Rotterdam, Netherlands
Posts: 1,832
|
On a field basis you are right, however video is displayed on a frame basis, so the combination of odd and even fields which make up a frame make up a full frame. So per frame there is no resolution difference. The reason for jaggies with interlaced material in caused by time movement, not by resolution differences, same as the stuttering effect with progressive material, which is caused by time movement. Let me see if I can find some more material on the internet to explain the differences between the two forms (I remember I have seen a great explanation once, but like you, I have trouble remembering).
|
February 9th, 2010, 07:19 PM | #10 |
Inner Circle
Join Date: Dec 2002
Location: Augusta Georgia
Posts: 5,421
|
Dear Friends,
I know of no reason why there would be 1/2 the vertical resolution, at least not the way we process full raster Interlace or Progressive footage.
__________________
Dan Keaton Augusta Georgia |
February 9th, 2010, 08:15 PM | #11 |
Major Player
Join Date: Jun 2009
Location: Vientiane (Lao PDR)
Posts: 349
|
We are mixing here two different concepts of resolution: Spatial and temporal resolution.
Interlaced have obviously half of the spatial (vertical) resolution than progressive, but it have double temporal resolution. Rafael |
February 10th, 2010, 09:30 AM | #12 |
Inner Circle
Join Date: Dec 2004
Location: Arlington, TX
Posts: 2,231
|
I think any discussion about progressive and interlaced should also include camera movement.
Interlaced footage of a static image will be a lot closer to progressive from a resolution point of view than if the camera is panning as all of the pixel will line up from frame to frame. So the contect of camera action can be important to determine what you are seeing on the screen. As far as seeing a difference, the only way is to get a hold of one and try it out with the work you do and see. William, I do not know where Bee Cave is, but if it is near Arlington, maybe we can work something out. |
February 10th, 2010, 08:11 PM | #13 |
Major Player
Join Date: Sep 2008
Location: Vancouver, Canada
Posts: 975
|
William, I have a feeling the line twitter you are talking about are caused by the default sharpening settings that are on by default and quite aggressive.
There are more than a few posts here on adjusting this. Unfortunately, I don't have the settings handy. It's in the "picture profiles". "Sharpening" and "Frequency" are some of the biggies that affect this. Turn off sharpening for starters will help in the short term, if this is the problem. You should not get line twitter as you say shooting in progressive. |
February 16th, 2010, 11:46 AM | #14 |
Regular Crew
Join Date: Sep 2007
Location: Bee Cave, Texas
Posts: 151
|
Dear Andrew, and all the rest who were kind enough to reply - sorry for the delay, but I have been away ill, and am now back, after just having purchased the Nano before going away! I just this second returned from over 90 minutes review of "detail", and admit to being brain dead - last year I did at least 5 settings with reduced detail, and the two that really reduced detail also ruined the picture, but that was before I just now reviewed the posts on accompanying settings, which I will likely give a try. Hurridly, as I'm off to finish nano accessories purchase prior to trip out of country (not for two weeks, but delivery sometimes takes a while) I thank you all again.
Bill Urschel |
February 17th, 2010, 08:35 AM | #15 |
Major Player
Join Date: Sep 2008
Location: Vancouver, Canada
Posts: 975
|
Hi Bill,
Just out of interest I did a bit of digging as finding stuff on this topic is difficult due to the "epic" nature of the Picture Profiles thread. One only has so much time in life before it is gone. Here's some info to speed up the process: Search on "frequency" in EX forum (you have do go back a few pages before you hit the meat of the posts on detail settings on the EX3 & 1R's)... http://www.dvinfo.net/forum/search.php?searchid=4137706 Then there is Alister Chapman's post on Sony's explanation on the topic of detail and sharpening (Crispening in Sony's own parlance): http://www.dvinfo.net/forum/sony-xdc...tml#post985655 Alister in the same post gives his detail settings. Spend some time digging through the search listing mentioned above. Hopefully it will shortcut your ability to get settings that suit your needs. |
| ||||||
|
|