Barry_Green
Moderator
Please show an example of this.
The Emperors-new-hack
- Thread starter Letzring
- Start date
Just look at the motion from some stuff shot with the stock Inter-frame codec... and then look at some stuff with low-GOP or intra-frame recording. It explains itself. The intra-frame looks more "sticcato" where the inter-frame has weird motion that I can't really put my finger on. It kind of looks like "chunks" of the images are moving independently from the background, or vise versa. Sometimes, when using a 180 shutter, it looks like only the subject has the 180 shutter look, while other parts of the image look like they were shot with an open shutter.
Like I said, some people either don't see it, or don't care. I personally, much prefer the intra look. If you look at Red or any other digital cinema footage... aside from the resolution... it has a very pronounced staccato motion to it that looks like a rapid succession of still images. Inter-frame captured motion looks like the frames are sliding into each other or something... I don't know any other way to explain it.
They aren't composed exactly the same. Can't tell how the banding is with the AF100 (although probably less pronounced)
And he was shooting lower ISos. That's just a given.
dustylense
Veteran
@Gh2user...
My camera is bigger than yours.
My camera is bigger than yours.
This description is technically incorrect. What the AVCHD encoder does in predicted macroblock encoding is to use the motion vector only as an initial approximation of the image. It then calculates a residual image which contains the detailed differences between the image and motion vector. This residual image is encoded at same quality level as an intra-frame but requires far fewer bits to encode. The image is not "interpolated", it is reconstructed.
A subtle point that many intra fans are not aware of is that an AVCHD encoder uses motion vectors not only in P and B-frames, but in I-frames as well. The difference is that P and B-frames may use motion vectors found in previous and/or subsequent frames, while macroblocks in I-frames can only use motion vectors found in that same I-frame. When an AVCHD encoder runs out of bitrate, you can sometimes spot crudely quantized macroblocks that are encoded with low-quality motion vectors. These macroblock artifacts are often more noticeable in I-frames simply because they require more bits to encode than P or B-frames.
Brian@202020
Veteran
My experience is the GH2 hack is better than stock, but doesn't surpass the AF100. I applaud Vitaliys efforts and will continue to follow the progress on Personal View. The GH2 is a great B camera now because of the hack, and the fact it uses the same lenses as my AF100 makes the GH2 a perfect companion to the AF100. I've been meaning to do some test between the two cameras, but I've been working too much, mainly with my AF100. Things will get a lot more interesting on the GH2 front when/if they can get a clean HDMI out and a slog type color profile. Until then the AF100 will be my A cam.
In case anyone wants a link directly to the latest patches for the GH2, here is a link to a thread I started on Personal View. All the testers and Vitaliy are supporting the thread and posting the latest patches.
http://www.personal-view.com./talks...ult-most-popular-patches-in-one-place#Item_24
In case anyone wants a link directly to the latest patches for the GH2, here is a link to a thread I started on Personal View. All the testers and Vitaliy are supporting the thread and posting the latest patches.
http://www.personal-view.com./talks...ult-most-popular-patches-in-one-place#Item_24
Last edited:
Jester2138
Well-known member
Having done just such a test myself, yes.
no hard evidence? you should show 2 videos A:intra B:inter-frame same footage
PaPa
Major Contributor
I am unbiased as i own both and have shot with both. The hacked GH2 has an iamge that knocks the socks off of the AF100. HAnds down. Form factor aside.
Brian@202020
Veteran
Well each to their own PaPa. I prefer the skin tones, colors, and the curves of the AF100. My experience is the AF100 has about 1/4 to 1/2 a stop more latitude. Also the AF100's ISO is more accurate than the GH2. The hacked GH2 running an I-Frame codec does have a little better motion for internal recordings, so I'll give you that, but for me that just isn't enough. Also the ability to record a clean uncompressed image out of the SDI and/or HDMI from the AF100 is just the icing on the cake. Like I said earlier, once Vitaliy makes the curves able to be tweaked as well as a clean HDMI out, I might reconsider.
Last edited:
PaPa
Major Contributor
You hit the nail on the head with that. Skin tone is really a preference, however, and i found something lacking with the skin tones i got from the AF100 that i get slightly better with the GH2. But that might not actually be skin tone, and just general levels of detail. This is an unfair comparison because im going hacked GH2 vs Stock AF100. A better comparison would be hacked GH2 with something like a samurai image from the AF100, But the big advantage here is price, and less focus on form factor because the AF100 is still, all in all, a better cinematic tool for indy filmmakers.
Check this out for comparison. It's in a thread i made in the GH2 section a while ago. And it also taught me to never shoot under 320 ISO on the GH2. At 400-800, posterizing is a non issue.
AF100
Clean:
link: http://img716.imageshack.us/img716/7433/af100clean.jpg
Curved:
link: http://img13.imageshack.us/img13/5527/af100curve.jpg
Clean Crop:
link: http://img815.imageshack.us/img815/5364/af100cleancrop.jpg
Curve Crop:
link: http://img546.imageshack.us/img546/7316/af100curvecrop.jpg
GH2
Smooth Clean:
link: http://img855.imageshack.us/img855/4170/gh2smoothclean.jpg
Smooth Curve:
link: http://img694.imageshack.us/img694/3055/gh2smoothcurve.jpg
Smooth Clean Crop:
link: http://img221.imageshack.us/img221/2553/gh2smoothcleancrop.jpg
Smooth Curve Crop:
link: http://img267.imageshack.us/img267/6555/gh2smoothcurvecrop.jpg
This test was really helpful in comparing the overall feel of compression between the two, study the detail in the shadows, and how detail was resolved. From this, i concluded that their dynamic range was close enough that any differences i could ignore. The AF100 was more sensitive to light, however, which is nice.
The other thing, about dynamic range. I'm tired of numbers. What's more useful to know is how many stops of useable dynamic range you have in a camera, and with the hacked GH2, there is a lot you can recover from the darks. The af100 and the chroma clipping in the cine gamma modes is a widely known issue, but he GH2 suffers from this as well in cinema film mode. And here is my take on it:
The af100 in the Cine gamma modes is trying its best best best to keep as much information as it possibly can even if it means having channels clip at different spots. This is great, because in terms of numbers, they can advertise that it has X amount of dynamic range, but it's simply not useable. If you used the camera to it's strengths, that yellowy bluey looking highlight would have simply rolled off nicely as the GH2 and most other DSLR's do because they would rather have a pleasing rolloff that try to hold on to that detail. Clearly users here and other's i've spoken to don't like the way it looks but are torn to use anything else because the cine gamma's have the best dynamic range. If the AF100 just kept to it's strengths and rolled off smoothly, it would approach soemthing much closer to what we already have in the 7D, GH2, etc... Just don't let yourself be fooled by numbers, cause they are not important.
Last edited:
Brian@202020
Veteran
When I was speaking of latitude, I was referring to usable latitude. There is easily a full stop of latitude difference when you include the unusable yellow highlights.
PaPa
Major Contributor
Sorry, would have to disagree with you there. Definitely not a full stop. Perhaps a quarter to a third.
PaPa
Major Contributor
Now before i post this and start a flame war, let me just be clear. I'm not saying the GH2 is a better camera, or worse than the AF100. I am simply examining the image it produces. Clearly the GH2 doesn't have so many of the things that the AF100 has which makes it much more difficult to shoot with. I'm just comparing the image here. You can conclude if you find this meaningful or not. I do and that's why i would like to share it.
Hopefully this serves as enough evidence to clearly show side by side comparisons between the two. Lenses were shot at same F stop, only thing that changed was shutter.
My reasoning for how i exposed the images. I made sure that the portrait of my child self was exposed exactly the same across all three. But at the end of the day, it doesn't really matter because there is so much contrast in this shot that you can easily see where each camera blows out, and at the same time, how much information it keeps in the darkest parts of the image.
I.E. Even if the AF100 blew out sooner than the GH2, that means i either over exposed it slightly, or under exposed the GH2. But then it wouldn't make sense why the GH2 could still see the same detail/more in the darks.. tu comprends? Not trying to do the fanboy thing. Just want information to be clear so that we can avoid confusion and statements that may or may not be true. The images tell us whats going on. But the AF100 with the external recorder i know would be a whole different ballgame. No question there. So why the test? Quality versus price. How much can i achieve with something so inexpensive.
open images in separate tabs for the best comparison.
AF100 - 24mm lens F2.8 ( focused on the white fluffy thing on the desk )
CineD
Norm 1
detail at -5
chroma at -5
AF100
Link: http://img717.imageshack.us/img717/1665/fullpl.jpg
400% Zooms
Link: http://img694.imageshack.us/img694/1204/af100close01.jpg
Link: http://img10.imageshack.us/img10/1276/af100close02.jpg
GH2 in Cinema Mode with Quantam V2 and 5DtoRGB ( -2 -2 -1 -2 )
Link: http://img843.imageshack.us/img843/6276/cinemal.jpg
400% Zooms
Link: http://img804.imageshack.us/img804/2421/cinemaclose1.jpg
Link: http://img834.imageshack.us/img834/4065/cinemaclose2.jpg
Smooth mode ( -2 -2 -1 -2 )
Link: http://img836.imageshack.us/img836/5775/smoothl.jpg
400% zooms
Link: http://img600.imageshack.us/img600/3812/smoothclose1.jpg
Link: http://img804.imageshack.us/img804/1419/smoothclose2.jpg
Hopefully this serves as enough evidence to clearly show side by side comparisons between the two. Lenses were shot at same F stop, only thing that changed was shutter.
My reasoning for how i exposed the images. I made sure that the portrait of my child self was exposed exactly the same across all three. But at the end of the day, it doesn't really matter because there is so much contrast in this shot that you can easily see where each camera blows out, and at the same time, how much information it keeps in the darkest parts of the image.
I.E. Even if the AF100 blew out sooner than the GH2, that means i either over exposed it slightly, or under exposed the GH2. But then it wouldn't make sense why the GH2 could still see the same detail/more in the darks.. tu comprends? Not trying to do the fanboy thing. Just want information to be clear so that we can avoid confusion and statements that may or may not be true. The images tell us whats going on. But the AF100 with the external recorder i know would be a whole different ballgame. No question there. So why the test? Quality versus price. How much can i achieve with something so inexpensive.
open images in separate tabs for the best comparison.
AF100 - 24mm lens F2.8 ( focused on the white fluffy thing on the desk )
CineD
Norm 1
detail at -5
chroma at -5
AF100
Link: http://img717.imageshack.us/img717/1665/fullpl.jpg
400% Zooms
Link: http://img694.imageshack.us/img694/1204/af100close01.jpg
Link: http://img10.imageshack.us/img10/1276/af100close02.jpg
GH2 in Cinema Mode with Quantam V2 and 5DtoRGB ( -2 -2 -1 -2 )
Link: http://img843.imageshack.us/img843/6276/cinemal.jpg
400% Zooms
Link: http://img804.imageshack.us/img804/2421/cinemaclose1.jpg
Link: http://img834.imageshack.us/img834/4065/cinemaclose2.jpg
Smooth mode ( -2 -2 -1 -2 )
Link: http://img836.imageshack.us/img836/5775/smoothl.jpg
400% zooms
Link: http://img600.imageshack.us/img600/3812/smoothclose1.jpg
Link: http://img804.imageshack.us/img804/1419/smoothclose2.jpg
dustylense
Veteran
I think both cameras do a fine job. The Gh2 with the right hack is no doubt sharper than the af100. I have also as Papa has shown seen comparable or better highlight handling with the gh2. But I won't give up the af100 to the gh2 because of it's features on the AF100 (waveform, ND's, sound, ....). But the 2 of them together work wonders. For the price, the Gh2 ROCKS!
Also, I think editman mentioned using 5dtorgb to transcode footage from both cameras. I agree as you can get a nice wide trans from the program.
Also, I think editman mentioned using 5dtorgb to transcode footage from both cameras. I agree as you can get a nice wide trans from the program.
Brian@202020
Veteran
I love both cameras and and there is a time and place for each one.
PaPa
Major Contributor
Just put the AF100 footage through 5DtoRGB and the results were quite impressive. The only thing that's off now for comparisons sake is all of their black levels. Making sure they hit 0 IRE would be a good way to measure the actual shadow detail. But this conversion really really helped the lighlights. Big time. Smoothed it out and brought back detail. Amazing
AF100 5DtoRGB
Link: http://img535.imageshack.us/img535/561/af1005dtorgb.jpg
400% Zooms
Link: http://img841.imageshack.us/img841/4826/af1005dtorgbclose1.jpg
Link: http://img195.imageshack.us/img195/7200/af1005dtorgbclose2.jpg
AF100 5DtoRGB
Link: http://img535.imageshack.us/img535/561/af1005dtorgb.jpg
400% Zooms
Link: http://img841.imageshack.us/img841/4826/af1005dtorgbclose1.jpg
Link: http://img195.imageshack.us/img195/7200/af1005dtorgbclose2.jpg
Thanks for doing that PaPa.
You're right ,the AF100 stuff looks pretty good now.
Was that the same exact settings that you used before?
On the blow ups I'm not seeing any crazy difference in the resolution like I saw before in other tests.
What are your thoughts?
Edit: I take that back, I can still see the difference when A and Bing them, and the I like the noise going on in the GH2 much more. But, it looks like the AF100's got more going on in the shadows.
Probably helped by the extended options for the picture.
Is there any way you can try the same test with the Nostalgic profile on GH2?
You're right ,the AF100 stuff looks pretty good now.
Was that the same exact settings that you used before?
On the blow ups I'm not seeing any crazy difference in the resolution like I saw before in other tests.
What are your thoughts?
Edit: I take that back, I can still see the difference when A and Bing them, and the I like the noise going on in the GH2 much more. But, it looks like the AF100's got more going on in the shadows.
Probably helped by the extended options for the picture.
Is there any way you can try the same test with the Nostalgic profile on GH2?