It's interesting that people slate this codec when it's around the same if not a little higher bitrate than the long GOP compression used by most high end DSLR's AND samples at 10bit!!! If storage is an issue then proxy is fantastic. Beats H.264 any day of the week so let's all smile and be friends
Long GOP compression used by most DSLR's (H.264) is not the same as ProRes Proxy even if they're the same bit rate. H.264 uses IPB frames (most of the time, unless selected like on the GH4 camera) and ProRes uses all I-frame compression. So Long GOP at 50mbps equates to probably something like 200mbps ProRes, I don't know exactly but it's a lot more. Which is more like ProRes HQ. That why if you convert a .mov file off your camera to ProRes Proxy, you can see a lot of artefacts.
I am not sure what point you are trying to make if i am honest. The point of this video was a simple test showing that prores proxy was a very usable codec. This video was made at a time where people were reluctant to use it seeing as it was the "Lowest quality" version of prores to come out of the bmpcc, I even heard some people use the word "unusable" to describe it. Given that this was at a time were the 5dmk3 was in full swing as the go to camera for low budget filmmakers, i decided to make this video proving that not only is proxy out of the bmpcc usable, its produces better quality, colour information and robustness than the common .mov produced by canon cameras at the time.
Agreed. The point of high bit-rate codecs is not what JUST what the image looks like in different modes - It's how the footage holds up under intense color correction. The DIFFERENCE between the codecs mostly lies in the areas you cannot see when until you go and try to apply some extreme grading, then the higher bit-rate will shine as it is recording more information than the eye can see - where as the lower rate codecs rely much more on perceptual encoding which basically throws away the information that the eye can't see in order to save on space... But the moment you go and try to jack up the shadows you'll see the difference quite quickly!
I agree. I was switching between LT and 422, and then at one point one of the settings on my camera got changed without my knowledge. Some of the footage ended up as proxy. It wasn’t until I got the project finished that I noticed. Everything still looked great. As long as it doesn’t require heavy grading, or significant post, I think proxy is an excellent way to maximize footage time while preserving quality.
Daniel, thanks for the test but I think we may be missing one important point here. The only difference between the ProRes formats is the bitrate. In other words, the only time we must expect to see any difference is where bitrate issues compete with each other and the content in your video is not one of them. A great part of the image content here is flat and almost 80% of the content is single color. The said bitrates are the maximum amounts, i.e. the limits of information that can be recorded within one second of movie and the content here does not even come close to it. To fully test the difference between these formats, I think the image must contain a deal of information and then the limiting bitrate issues become visible. With a content full of moving details (such as waving braches of a tree full of leaves etc), then, I assume, the proxy will suffer while higher formats can easiliy deal with the bitrate demand. Thank you for your tests.
I really wish people would read the description, this was a simple not scientific test. The whole point I was trying to prove is that there is barely any noticeable difference to the the image quality, at least not enough to warrant the amount of contempt people give to the thought of using proxy. I am well aware that under proper test conditions that the differences will become more apparent, but since when do we film with the intent of pushing the camera and its output to destruction? There is a reason proxy exists and in comparison with the format we have been used to using from DSLRs, its immensely better. Thanks for the comment.
***** Please don't get me wrong. I did read the description and was not expecting for a scientific test either. What I was trying to say was that "the difference" would be more pronounced if a different content were used. It is only natural that you do not see any difference because there is none in the content. Under the circumstances even the h.264 clip survived relatively well. I do agree that Proxy does the job for casual shooting. Thanks again and best wishes...
3:15 I can see the difference pretty clearly... Slight difference in color (Uncontrol daylight sun maybe). But look at the LT light green, you can see the printing pattern between dark green and white, which those small dot can't be seen on the Proxy. Also the font is softer
These differences are SO minute that you would NEVER notice them in actual video. Look, we're zoomed in at ridiculous levels on a still, fine-text image. You're telling me you're gonna see the color cast on a wrapper as the actor walks past a trash can and be horrified? LOL. I myself am guilty of having GAS just like all of you but jesus, to nitpick this level of difference is straight up stupid. Take a breath, people.
on the LT proxy close up you can see the breen dots on the LT on the Proxy it a blur....its still good, what would happen if you put in camera sharpen on at all three levels
I don't know if anybody has said it before but your test is not accurately. You have used a highly contrasted image with almost no information at all. If you want to know the difference between HQ, LT and PROXY, I suggest that you shoot a video which contain a lot of information and details. Try shooting (with the focus in infinitum) a landscape, a city scene, etc. Try both something static and something with movement. Also you have to consider that HQ format support much more iterations than PROXY. And that's one of the main differences. Alvaro.
Yeah people don't realize it comes into play with movement and how much is going on in the scene. Also when it comes to manipulating in post you will see the difference.
You'll mainly see differences when you introduce motion and heavy colour grades between the different flavours of Prores, I still image like this isn't really a very demonstrative comparrison.
im using the onboard intel graphics from i7 3770. i converted to prores a .mov from my dslr canon. after this, by comparison, both has the same results. the difference is that the original .mov from camera has a little more contrast. the prores version has a flat appearence, but any improvent about sharpness, dertails, etc. need i a best video card?
Converting a .mov to prores wont improve the videos detail and sharpness, once a file has be compressed to .mo or h.264, all the information that the camera threw away to make that compressed file is gone.
Well, i wouldn't say the dynamic range is higher, its more to do with its ability to handle grading, the amount you can push the brightness, contrast or saturation before the image starts to deteriorate. But yeah, my point with this video wasn't to see if there was a difference, it was to show that there wasn't, as far as image quality goes. I had been hearing a lot of chatter from people talking about proxy like it was the plague of codecs, using words like unusable, atrocious, and horrible. My point was to show that these people are ripping into a codecs that performs better than the codecs from the best dslr 10 times over. Thanks for the comment. :)
Yeah, Your able to push those things because of the dynamic range information that the different codecs record. The different codecs dont have much really to do with initial image quality so the comparisons are a little bit irrelevant but it was an interesting comparison. The real comparison to me would be the trade off of size of files to grading ability. With proxy, the sizes are nice and small allowing for longer recording and it allows you to use 45mb cards. The downfall, just dont expect alot of leverage during the grade processs. Anyways thanks for sharing.
I guess on a mini-doc setup, run n' gun... Pro Res Proxy would work ? Not trying to be dumping footage all day.. Seems like Proxy is still solid and gives a nice amount of record time ?
I came here as I'm sick of dealing the AVCHD which is very taxing on my not so strong machine. I was looking for a file-size-friendly prores codec that would make editing. Point is, as long as prores proxy isn't lower than AVCHD in quality I'm going for it!
what i understand is that the proxy codec is meant for offline/online use for the coding speed ? so not much for a quality difference... www.4kshooters.net/2015/01/26/choose-the-version-of-prores-best-suited-to-your-project/ The Prores Proxy stores less data so it will run better on slow hardware in post after grading & editing quality will be less than the other formats, so it's a choice you make, if you have good hardware and fast memory, you don't use the proxy codec...
This video was made around a time when people were arguing over whether or not proxy should be used for filming projects and in the groups i would hear words like "Unusable" and "worthless" regularly to describe the Proxy shooting format on the bmpcc. I made this just to see if what they were saying was true. Needless to say, if i need to shoot long continuous shots, say like a conference or stage performance i wont hesitate to use proxy in camera if it means 3-4 hours of uninterrupted shooting.
Daniel I think the main (better say only) difference between these formats is when you try to colorgrade them. Thats where you gonna need more information.
***** I think you have your answer. I you can't see a difference then, you'll be happy with Prores Proxy. If, however your client or the end result requires color grading, you will find that the results will not be so close. Proxy has less information to grade. You will notice a difference in how much you can "bring out" in post. End of the day... if you or your non paying clients are happy, then it's a win/win!
Myron Davis you say "If your client requires color grading" .. well, EVERY footage from bmpcc has to be graded. Just saying. This test says "there is almost no differece between HQ and Proxy without grading" .. I say okey, but every footage has to be graded, so it is not so usable information (with no offence)
This comparison is worthless unless you are shooting something with no motion or movement whatsoever. You didn't actually test the codecs, you tested encoding still images. lol
