Constant Bit Rate Mode

[K Smith]

Hi friends

I'd like to know using this app is it possible to get Constant Bit Rate (CBR)? I downloaded a HD video but when i view the mediainfo i dont see anything like that.

Cheers

 

[0x0x0x0x0]

https://streaminglearningcenter.com/articles/technical-brief-switch-from-cbr-to-vbr-to-improve-overall-quality-and-avoid-transient-quality-issues.html

Simply put: there is no mathematical way for CBR quality to match VBR unless bitrate of CBR is on par with peak bitrate of VBR

 

[K Smith]

https://streaminglearningcenter.com/articles/technical-brief-switch-from-cbr-to-vbr-to-improve-overall-quality-and-avoid-transient-quality-issues.html

Simply put: there is no mathematical way for CBR quality to match VBR unless bitrate of CBR is on par with peak bitrate of VBR

Hm let me explain a bit. I download a movie from amzn using any stream and the quality was ok. But one of my friend using another method managed to get a better bitrate than mine. I compared both our media info and realised he had bitrate mode : constant. And his bitrate/ file size was better than mine.

that’s why I’m bit confused How someone managed to get a better stream than this app..

 

[0x0x0x0x0]

Hm let me explain a bit. I download a movie from amzn using any stream and the quality was ok. But one of my friend using another method managed to get a better bitrate than mine. I compared both our media info and realised he had bitrate mode : constant. And his bitrate/ file size was better than mine.

that’s why I’m bit confused How someone managed to get a better stream than this app..

You say your quality was OK but then you speak of "better bitrate;" that's comparing apples and... orange pickers? Maybe they used a screen recorder that had no choice but to throw bitrate at the problem because it had to encode in real time...

 

[K Smith]

You say your quality was OK but then you speak of "better bitrate;" that's comparing apples and... orange pickers? Maybe they used a screen recorder that had no choice but to throw bitrate at the problem because it had to encode in real time...

I was hoping to get the best copy out of any stream. I said the quality was ok because it was good quality. But then again when someone show me a better rip I’m just wondering why I couldn’t get that output using any stream. Trust me it’s not a screen recording app. I’ve used screen recording apps before. It won’t even come closer to AS. My stream bitrate was around 3Mbps and 5gb file size his one was around 6Mbps and 8gb file size.
It’s not comparing apples vs orange because it’s not 2 different things.


His audio stream and mine was the same. Just his video stream had a better bitrate and he had this additional media info parameter called “Bit rate mode : Constant”

 

[Vzxy]

Hm let me explain a bit. I download a movie from amzn using any stream and the quality was ok. But one of my friend using another method managed to get a better bitrate than mine. I compared both our media info and realised he had bitrate mode : constant. And his bitrate/ file size was better than mine.

that’s why I’m bit confused How someone managed to get a better stream than this app..

He was likely using an app that records vs just downloading. He is not getting better quality since both ways are using the same source file for a particular title from Amazon. His recording app is just recording at whatever bitrate he has selected. That doesnt make it better, and in many ways its worse since the quality will not get any better doing so and it will only take more disk space in the process.

 

[0x0x0x0x0]

I was hoping to get the best copy out of any stream. I said the quality was ok because it was good quality. But then again when someone show me a better rip I’m just wondering why I couldn’t get that output using any stream. Trust me it’s not a screen recording app. I’ve used screen recording apps before. It won’t even come closer to AS. My stream bitrate was around 3Mbps and 5gb file size his one was around 6Mbps and 8gb file size.
It’s not comparing apples vs orange coz it’s not 2 different things.


His audio stream and mine was the same. Just his video stream had a better bitrate and he had this additional media info parameter called “Bit rate mode : Constant”

What is "better bitrate"? Why do you say it's better? If it's what I think you mean, using MPEG-2 will always result in better bitrate? Bitrate is not everything: it's quality that matters! (hint: even at Main L4 CBR will need more bits to store similar quality as VBR at High L4)

PS: I didn't say you were comparing apples to oranges, I said apples to orange pickers, because they are even more different: you don't even tell us what level or profile the CBR file is...

 

[K Smith]

What is "better bitrate"? Why do you say it's better? If it's what I think you mean, using MPEG-2 will always result in better bitrate? Bitrate is not everything: it's quality that matters! (hint: even at Main L4 CBR will need more bits to store similar quality as VBR at High L4)

Ok I don’t know about cbr setting and all. It was a competition kind of a thing and mine was defeated because of the bitrate.

like I mentioned earlier. Mine was Around 3Mbps and his one was around 6Mbps so if it’s like that in the media info means you really think mine is better than his one?

if that’s the case any idea how to really compare the quality? Do I need to download the video and do some tests or can I do some test with screenshots etc?

 

[0x0x0x0x0]

Ok I don’t know about cbr setting and all. It was a competition kind of a thing and mine was defeated because of the bitrate.

competition in what: who can waste more bandwidth to store a medium?

like I mentioned earlier. Mine was Around 3Mbps and his one was around 6Mbps so if it’s like that in the media info means you really think mine is better than his one?

Bitrate literally means jack: I can take the file that you downloaded and turn it into a 100Mbps bloat, what would that achieve, other than a waste of bits?!

if that’s the case any idea how to really compare the quality? Do I need to download the video and do some tests or can I do some test with screenshots etc?

Comparing screenshots is another "great" idea that plagues amateur video encoders: why does anyone care what a picture looks like if it stays up on the screen for under 42 milliseconds at the longest (assuming 24000/1001 frame rate). Doom9 and other forums are littered with people's onanistic obsession with artefacts or blurriness of a specific frame: they spend more time posting the damn thing than it literally stays up for on their screen!

Chances are, your mate's encode is something like Main@L3 CBR, well congratulate them on wasting bandwidth...

 

[zanetti]

What is "better bitrate"? Why do you say it's better? If it's what I think you mean, using MPEG-2 will always result in better bitrate? Bitrate is not everything: it's quality that matters! (hint: even at Main L4 CBR will need more bits to store similar quality as VBR at High L4)

interesting. but, ive downloaded arrival (2019) from prime and it was 1.9gb with bitrate of around 1900kbps @ 1080p. surely, if the bitrate was 5000kbps the overall quality would be better, right? mind you, this is an exception and the only movie from prime that was so ridiculously compressed and small in filesize.

 

[0x0x0x0x0]

interesting. but, ive downloaded arrival (2019) from prime and it was 1.9gb with bitrate of around 1900kbps @ 1080p. surely, if the bitrate was 5000kbps the overall quality would be better, right? mind you, this is an exception and the only movie from prime that was so ridiculously compressed and small in filesize.

Not always the case, as one of the x264 devs put it:

You can improve quality and reduce bitrate at the same time.

https://forum.doom9.org/showpost.php?p=1116743&postcount=16

just depends on whoever is doing the encoding actually knowing what they're doing


Edit: as an example, there are literally tons of recommendations out there in the wild to bump qcomp for both x264 and x265 to something like .75-.85, and if you know what you're doing, that's not what you want to be doing (hint: flix's qcomp is usually at .5)

 

[K Smith]

competition in what: who can waste more bandwidth to store a medium?



Bitrate literally means jack: I can take the file that you downloaded and turn it into a 100Mbps bloat, what would that achieve, other than a waste of bits?!



Comparing screenshots is another "great" idea that plagues amateur video encoders: why does anyone care what a picture looks like if it stays up on the screen for under 42 milliseconds at the longest (assuming 24000/1001 frame rate). Doom9 and other forums are littered with people's onanistic obsession with artefacts or blurriness of a specific frame: they spend more time posting the damn thing than it literally stays up for on their screen!

Chances are, your mate's encode is something like Main@L3 CBR, well congratulate them on wasting bandwidth...

thanks for explaining. So I checked his media info agai. And it says :

Format profile : Main@L4

but mine was : High@L4

can you help me with this one thing? I just want to compare my copy (quality) with his one. What is the best way to do that? Opening both videos not gonna help because it looks almost the same

 

[Vzxy]

Ok I don’t know about cbr setting and all. It was a competition kind of a thing and mine was defeated because of the bitrate.

like I mentioned earlier. Mine was Around 3Mbps and his one was around 6Mbps so if it’s like that in the media info means you really think mine is better than his one?

if that’s the case any idea how to really compare the quality? Do I need to download the video and do some tests or can I do some test with screenshots etc?

Thats hard to say. But I do know I have never seen a CBR download yet from AP using AS. I don't check every file so I can't say they're not out there, but I just haven't ever seen one and I've downloaded tons. Usually, NF files are far smaller than the same title @ same resolution from AP, but they use high profile on some of those and get very good quality at a much smaller file size. Bitrate doesn't really tell the entire story. I do video production for a licensing company and I use ffmpeg all day long to convert files and sometimes I get very good clips with very low bitrates. Like crystal clear 1080p at 1.2Mbps because they used better options during the encoding to achieve better compression, finely tuned to the type of source video - ie: animation vs live action vs broadcast news which resulted in the best quality and smallest files.

I've seen all sorts and AS is simply downloading the source files, as encoded by Amazon, directly to your machine, no transcoding at all. It's the best and purest way to handle it since Amazon has way more processing power to crank out these files at the best settings than we do. Capturing that same source at a higher bitrate will only increase the file size and CBR will, by definition, be higher bitrate than VBR in almost every case. But also be aware that AP and NF are always updating how they encode files and this will keep changing.

 

[cartman0208]

interesting. but, ive downloaded arrival (2019) from prime and it was 1.9gb with bitrate of around 1900kbps @ 1080p. surely, if the bitrate was 5000kbps the overall quality would be better, right? mind you, this is an exception and the only movie from prime that was so ridiculously compressed and small in filesize.

Well ... "Arrival" contains a lot of dark scenes, slow camera moves and many flat surfaces ... thats where the encoders try to save bitrate.
The version I backed up from Bluray actually has a bitrate of 900kbps and still looks pretty good, x265 though

 

[Vzxy]

thanks for explaining. So I checked his media info agai. And it says :

Format profile : Main@L4

but mine was : High@L4

can you help me with this one thing? I just want to compare my copy (quality) with his one. What is the best way to do that? Opening both videos not gonna help because it looks almost the same

high profile uses more reference frames for a better compression at lower bitrates or from a lower quality source file, such as black and white or old movies. Generally high profile will look better, but requires more processing power to watch it, so older devices may not be able to handle it. Main profile is the most widely accepted settings that most players can handle easily.

The basic rule of thumb is if its 720 or less or B&W it will probably be high profile with options specifically designed to get the best quality it can. If its 1080 or higher, or a modern digitally filmed movie then you can get away with a main profile encoding and still get good quality. Thats not for every case, but thats the nutshell version.

 

[0x0x0x0x0]

thanks for explaining. So I checked his media info agai. And it says :

Format profile : Main@L4

but mine was : High@L4

can you help me with this one thing? I just want to compare my copy (quality) with his one. What is the best way to do that? Opening both videos not gonna help because it looks almost the same

Correct me if I'm wrong, but prime contend delivery has no specification for Main@L4, so it sounds like he was using a screen recorder to compress an already well compressed stream...

If you can't tell the difference between the two visually, then why on Earth do you still care?

 

[K Smith]

Thats hard to say. But I do know I have never seen a CBR download yet from AP using AS. I don't check every file so I can't say they're not out there, but I just haven't ever seen one and I've downloaded tons. Usually, NF files are far smaller than the same title @ same resolution from AP, but they use high profile on some of those and get very good quality at a much smaller file size. Bitrate doesn't really tell the entire story. I do video production for a licensing company and I use ffmpeg all day long to convert files and sometimes I get very good clips with very low bitrates. Like crystal clear 1080p at 1.2Mbps because they used better options during the encoding to achieve better compression, finely tuned to the type of source video - ie: animation vs live action vs broadcast news which resulted in the best quality and smallest files.

I've seen all sorts and AS is simply downloading the source files, as encoded by Amazon, directly to your machine, no transcoding at all. It's the best and purest way to handle it since Amazon has way more processing power to crank out these files at the best settings than we do. Capturing that same source at a higher bitrate will only increase the file size and CBR will, by definition, be higher bitrate than VBR in almost every case. But also be aware that AP and NF are always updating how they encode files and this will keep changing.

thanks for the reply. I completely agree with you. I’ve tried some “old” manual method to capture the stream and it was exactly the same as AS output. So that means AS is doing the same thing downloading the original stream and decrypting it.

I just want to know how to differentiate 2 videos with different bitrates

 

[Vzxy]

Well ... "Arrival" contains a lot of dark scenes, slow camera moves and many flat surfaces ... thats where the encoders try to save bitrate.
The version I backed up from Bluray actually has a bitrate of 900kbps and still looks pretty good, x265 though

Yep, anywhere there is not a lot of perceptible motion on the screen, such as dark scenes or still graphics, etc. This of course can be fine tuned in the encoder, but that's what VBR is all about. HEVC/x265 is really good anyways. I reencode all my downloads to HEVC to save tons of space and avoid having to always buy more drives. Plus I don't store anything on single drives, always RAID, so that means I have to buy two every time, plus there's a limit to how many drives you can physically attach to a PC at once. I don't want to have to swap USB drives to watch a movie. All this means HEVC is a life, and data saver for me.

 

[K Smith]

high profile uses more reference frames for a better compression at lower bitrates or from a lower quality source file, such as black and white or old movies. Generally high profile will look better, but requires more processing power to watch it, so older devices may not be able to handle it. Main profile is the most widely accepted settings that most players can handle easily.

The basic rule of thumb is if its 720 or less or B&W it will probably be high profile with options specifically designed to get the best quality it can. If its 1080 or higher, or a modern digitally filmed movie then you can get away with a main profile encoding and still get good quality. Thats not for every case, but thats the nutshell version.

it was a latest movie and 1080p. So I guess AS won’t be using main@ profile by any chance?

 

[0x0x0x0x0]

it was a latest movie and 1080p. So I guess AS won’t be using main@ profile by any chance?

Even Apple with their crappy devices back in the early days of 1080p (think around 2010) was using High@L4