AF-100 vs AF-100A in JAPAN as for December 2012

wow. I took the footage right off the card for the avchd and right out of the Samurai, un-treated. I guess I should abandon this post as it is a bit techie for me. The Samurai simply produces a better picture IMHO and stands up very nicely in post, way better than the avchd. I am happy with the camera and with the Samurai + DNxHD. I think its time I got to work!

DNxHD works just fine in Cs5.5. You guys!!!!
 
Think you are better of quality wise with prores but if you like dnxhd use it. I find dnxhd worksbest for avid nle. Also look at some colour grading tutorials like this one for premiere http://vimeo.com/29256397

DNxHD and ProRes are on the same level as far as quality with a slight edge actually for DNxHD technically. Your remark about DNxHD working best with Avid is only true for the MXF files. But the Hyperdeck also records DNxHD in QuickTime, which is what people not editing on Avid should be using. This is why Avid created DNxHD in QT and they say so.

Back to the subject of better or not than the internal AVCHD, I can pull more info from blown highlights in DNxHD than in the native AVCHD. I have done this test several times.
 
i only meant that premiere reads dnxhd files wrong and to put a filter on and change it back to 0-255 to get right black and white levels might introduce poorer quality. Thats why i changed to prores. But i agree that external recorder brings more quality especialy in fast motion scenes
 
i only meant that premiere reads dnxhd files wrong and to put a filter on and change it back to 0-255 to get right black and white levels might introduce poorer quality.
Have you seen evidence or documentation of this behavior in Premiere Pro CS5.5-6? My understanding is that Premiere takes an agnostic approach to video levels and allows the installed and bundled decoders to render each video as it sees fit. In the case of the DNxHD decoder, which is supplied by Avid, it always uses studio swing levels of 16-235. ProRes in Final Cut Pro works similarly. Some background on studio swing issues can be found in this article from 2009:

http://digitalfilms.wordpress.com/tag/video/
 
In premiere 5.5 and 6 using dnxhd from shuttle2 i get same results as preston. Avchd and dnxhd doesnt match on timeline. dnxhd has wrong black and white levels.
 
In premiere 5.5 and 6 using dnxhd from shuttle2 i get same results as preston. Avchd and dnxhd doesnt match on timeline. dnxhd has wrong black and white levels.
Yes, this is a built-in property of DNxHD encoding, it always uses studio swing 16-235 color levels. If you open an After Effects project and select "HDTV (Rec. 709) 16-235" Working Space under Color Settings, AE will identify the color space in each file you import. With that Working Space selected, AE will import DNxHD files without remapping its color space. But when you import AF100 AVCHD files, AE will recognize them as using a different color space, and offer to convert them to studio swing. Premiere Pro does not do this, because unlike AE, Premiere is not a color managed application.
 
Working with my AF101 and BM hyperdeck Shuttle2, i've seen that several times.. You just have to know that internal Pany (and Sony too on FS100) AVCHD and external BM (or Atomos) PRORES HQ codecs record the signal into 0-110 IRE density space. That mean there are details in the hyper white area (100-110) you cannot watch in the Premiere pro CS5.5 (orCS6) timeline monitor if you don't use a filter (Amplification for example) in order to "compress" the hight levels under 100 IRE. And then you'll see exactly the same details in the hight lights (in your example the window) comparing with DNxHD record. DNxHD compress all the SDI signal (0-110 IRE) density space into (10-100) - these figures are mine coming from many footages with the AF101 and BM hyperdeck Shuttle. So there is no difference. The matter point is 4.2.2 8bits and now 10 bits SDI signal which provide more fine resolution détails in 4.2.2 color space, and a lot more in 10 bits space where you'll be able to graduate footages without banding... Nothing more.
 
Last edited:
Working with my AF101 and BM hyperdeck Shuttle2, i've seen that several times.. You just have to know that internal Pany (and Sony too on FS100) AVCHD and external BM (or Atomos) PRORES HQ codecs record the signal into 0-110 IRE density space. That mean there are details in the hyper white area (100-110) you cannot watch in the Premiere pro CS5.5 (orCS6) timeline monitor if you don't use a filter (Amplification for example) in order to "compress" the hight levels under 100 IRE. And then you'll see exactly the same details in the hight lights (in your example the window) comparing with DNxHD record. DNxHD compress all the SDI signal (0-110 IRE) density space into (10-100) - these figures are mine coming from many footages with the AF101 and BM hyperdeck Shuttle. So there is no difference. The matter point is 4.2.2 8bits and now 10 bits SDI signal which provide more fine resolution détails in 4.2.2 color space, and a lot more in 10 bits space where you'll be able to graduate footages without banding... Nothing more.

Not sure about Premiere as it's not a program I like or use. But in DaVinci Resolve 9, I can't bring back the same amount of detail in the highlights I get from DNxHD when grading the same shot in AVCHD.
 
You should try Premiere, because with Premiere you can !!
I have hundred examples where you can watch the things. no problem. However it's not fair to compare AVCHD and DNxHD, because of the different color space - 4.2.0 and 4.2.2. Even if you just want to compare the luma, it's not right. But comparing the PRORES HQ and DNxHD records would be right, and you would see the difference between two ways recording... where DNxHD file looks flat and seems to be a log compression beside the PRORES HQ file which look very similar from the AVCHD, but different...
 
Not for nothing, but I got the same positive results from the Samurai with DNxHD camparison with AVCHD in After effects that I got in Premiere...
 
When you say "colour space" are you thinking information? As in bit depth, more shades of grey etc? 420 vs 422 is a 50% jump in colour resolution. Finer noise, better keys, possibly less banding in high coring scene files but essentially its the same as AVCHD minus the ugly macro blocking and is encoded in I frames.
 
I don't know what happened. I also look for footages online but can't find any footages. Anybody is buying this camera?
 
seems like no one buys it at all, cant find anything online

I start to believe nobody buys this camera too. It is so strange that panasonic release a new camera and nobody cares. Nobody cares to review it not even bloggers. You can't find anything online about it. You can't find anybody here who buy buys it. It is like the camera doesn't exist.
 
Well it's not a different camera as far as IQ goes. It has a few new features but the image is going to be the same.
 
But it has a new processor no? Maybe this will help.

That was needed for 1:1 punch in focus assist and I think they managed to rewrite the SDI out part of the firmware, I read the "new" camera as an AF100 with the paid firmware included and a new processor for higher quality sound, 1:1 punch in focus assist and in camera dithering via the SDI output as you're not getting a true 10 bit out. Thats pretty much what Jan said, not a quote but how I read it.

Its not a new camera at the end of the day, they just found a way of adding a feature that there other cameras have and repackaged it with the firmware update and discontinued production on the old AF100 design, they are still available of course but I don't think they are being made, only the A series is being made now.

Great for new owners but not an update for us. If the economy wasn't strained this could of been a send in hardware upgrade for us existing users via Panasonic but I guess it's not economically viable now.
 
Back
Top