These two images are not night and day to my eyes. They're pretty much exactly the same, aside from the gamma shift, and both look bad because the source is pretty messy to begin with. 5DtoRGB's just exposing the limitations of your camera more than Cineform.
My hacked Gh2 is killing my Af100
- Thread starter dustylense
- Start date
These two images are not night and day to my eyes. They're pretty much exactly the same, aside from the gamma shift, and both look bad because the source is pretty messy to begin with. 5DtoRGB's just exposing the limitations of your camera more than Cineform.
Mark Smith
Veteran
The Cineform frame does look better on my set up here. I'll try to dig out some other examples that seem to highlight the differences.
As for this frame being messy- yup it sure is: times sq in the "dark" every kind of light source imaginable bouncing around, hand held camera shooting at a 5.6 or so. Sub Optimal.
As for this frame being messy- yup it sure is: times sq in the "dark" every kind of light source imaginable bouncing around, hand held camera shooting at a 5.6 or so. Sub Optimal.
Thomas Worth
Active member
I really hate to have to do this, but this isn't a fair test.
For starters, you're using QuickTime Player to compare the images which is known to perform post-processing. Unfortunately this disqualifies the test right away because we can't be sure that the output from QuickTime Player is accurate. Also, I noticed that the QuickTime controller in the CineForm image is blurry, which means the image isn't 100%. And the images are 1862x1047 (5DtoRGB) and 1888x1032 (CineForm). They should both be 1920x1080 to make this fair.
A better approach would be to use After Effects or some software that does not apply post-processing. You could load each clip into AE, then export a 1 frame PNG image sequence of each. That would be a much better way to judge the differences between the two.
I wouldn't mind getting a hold of some of your footage, if you wouldn't mind sending me a clip or two. I still want to see why you are having audio problems. That should be fixed.
dustylense
Veteran
For what it's worth, mr. worth...
I just tested my af100 converted through 5d2rgb and the sound is just fine. I bought the program through the app store for mac. Must be operator error on the other end.
I just tested my af100 converted through 5d2rgb and the sound is just fine. I bought the program through the app store for mac. Must be operator error on the other end.
Thomas Worth
Active member
Thanks for the heads up.
Mark, you're welcome to contact me if you'd like help working out the audio problem.
Mark Smith
Veteran
Don't think I've gotten a clean audio result from any clip yet with 5d2rgb. So yeah there must be something wrong some place. The audio is fine on the clips- I can play them back from the camera, also in hdvu. 5d software isn't that complex that I have over looked a setting so it must be something in the configuration of my system where I have the software installed or something like that. If I send you an mts file is that sufficient or do you need something in the complete AVCHD file structure.
Thomas Worth
Active member
No, just the unmodified MTS file will be fine. If you've got a small file (a few seconds of recording) that shows this problem, that would be ideal.
You can use YouSendIt, etc. to send the file to the Rarevision support address.
Thomas Worth
Active member
Final Cut is one way of doing it of course, but there are better ways that yield higher quality results. 5DtoRGB is one of them. You can read about it here:
http://rarevision.com/5dtorgb/
And there are plenty of tests/blogs that show the difference between Final Cut/QuickTime transcoding and 5DtoRGB.
Cheers, I read that and still didn't understand what it was doing in the AF100 sub forum. That's why I posted initially. It says it's for DSLRs. I'm assuming it also does
wonders for AF100 footage. I also looked into Cineform Neo. Hmmmmn. Dunno. Is this stuff all standard now or is it just a matter of preference for pixel peepers?
I'm in the dark here. Enlighten me.
Mark Smith
Veteran
I ran a couple of random clips through 5d2rgb again and have a good audio track now. I didn't change a thing in settings, just reprocessed the clip.
I've run massive, long files through 5DtoRGB and spit out 24GB chunks, non hacked or lightly hacked footage (1080/60i) and never had an issue with the audio. Hacked footage is giving me 1GB a minute, and even then it's still stout on the conversion.
If you're on an older system, and this kinda goes for heavy computing in general, doing too much or taxing the system while it's converting can cause trouble.
If you're on an older system, and this kinda goes for heavy computing in general, doing too much or taxing the system while it's converting can cause trouble.
Mark Smith
Veteran
New Mac mini running lion , plenty of memory, I don't think the system is weighed down by the processing.. And as I said a moment ago I just did get a couple of files with good audio without doing anything except reprocessing. It's just a random occurrence I suppose.
Mark Smith
Veteran
I am going to try to get both cine form and 5d2rgb processed files into after effects later this week and try to get an apples to apples comparison as far as exported stills go. I can probably even add a fcp transcoded file to the mix and then we can see atl three exported as png's from after effects.
Bern Caughey
Veteran
As beige is the new (next) black, the Mac Mini is the new tower.
Not quite as glam, but I see them in editing bays all over Hollywood.
Thomas Worth
Active member
That would be great. Thanks, Mark!
Mark Smith
Veteran
files.me.com/markso7/jrk9qz
here is a zip with 3 stills exported from AE as PNG's . THis took me a little while as I am definitely not an after effects person, but i fumbled through it.
Some other notes
the FCP generated pro res lt version of the clip the still was pulled from is 3.32 GB
the 5D2RGB generated pro res clip is 4.51 GB
the Cineform pro res clip is 2.74 gb
I find the file sizes interesting because there is no real visual difference between them playing back on my edit monitor.
Besides the gamma shift in the 5D2RGB file they all seem pretty close though the FCP and Cineform file are truer to the original image in terms of contrast.
here is a zip with 3 stills exported from AE as PNG's . THis took me a little while as I am definitely not an after effects person, but i fumbled through it.
Some other notes
the FCP generated pro res lt version of the clip the still was pulled from is 3.32 GB
the 5D2RGB generated pro res clip is 4.51 GB
the Cineform pro res clip is 2.74 gb
I find the file sizes interesting because there is no real visual difference between them playing back on my edit monitor.
Besides the gamma shift in the 5D2RGB file they all seem pretty close though the FCP and Cineform file are truer to the original image in terms of contrast.
Emanuel
Cross-Examiner
For some reason, Canon never dealt very good with 720p for some reason... :-D(...)
Perfect example, the C300 footage at 720p to me was LESS than stellar. The colored moire was horrible and something that would keep me from using it. Yet, at 1080p I thought the image was FANTASTIC. This means of course there are good and bad things to deal with when shooting with the C300, as with the GH2, as with the AF100.
That's why I just want one of all of them
Bern Caughey
Veteran
In my brief testing 720 H has not proven fail safe.
http://www.dvxuser.com/V6/showthrea...-for-30MB-s)&p=2516927&viewfull=1#post2516927
http://www.dvxuser.com/V6/showthrea...-for-30MB-s)&p=2516927&viewfull=1#post2516927
Hi guys i just got one of the hackable gh1 cameras and live in the UK. Can anyone recommend what bit rate/settings are a decent place to start? I own the af100 as well and plan to use the gh1 as a b cam and for docu use. Thanks any advice would be much recommended, there's so much info on this cool hack I'm not sure where to begin! But mucho excited!