Quality loss after video editing / lossless conversation

[DunnoNo]

Hi

As far as I understand every time you convert a file you lose quality.

I have a couple of questions

1) I know you can extract video and audio from one container and put it into another one without quality loss. But Do I still lose quality if a converted video has the same setting or as an original video? What if it has higher settings than the original? Still loss quality?

2) Related to the first one. If edit some video in , let's say, Premiere and then export it. Will the original video track loss quality no matter how high my settings are?

[redwudz]

If you re-encode a video file with a lossy codec like H.264 or Xvid, you will have quality loss.

1) Same settings, codec, for a re-encode as the original, still quality loss. Higher bitrates for a re-encode, less quality loss but larger file.
If you use a loss-less codec like Lagarith or similar, no quality loss but much larger file. Sometimes you may convert to a loss-less codec, do your edits, then convert back to the original lossy codec.
I do that with .mov files from my camera. I think very little loss, but I'm not an expert there.

2) Higher bitrate re-encode = less loss = larger filesize. Very high bitrate re-encodes may not play on most players either.
Depends what you plan for the final product.

But others here can probably give you better information.

And welcome to our forums.

[DunnoNo]

1) ohh ok If I re-encode into exactly the same thing with the same codec, settings and etc. I still lose quality. Strange hmm higher settings than original still loss. Would appreciate if anyone could explain why or give a link. I'm just curious

2) Players are not a concern. Just was wondering if there's quality loss.

[jagabo]

The process of re-compressing a video involves first decompressing it, then compressing it again. To the lossy codec used in the second compression the artifacts from the first compression are just details. It attempts to preserve them, creating more artifacts and losing more detail.

[poisondeathray]

It also depends on the type of video and editor you are using , and type of "editing" you are referring to. Certain types of video can be passed through or "smart rendered" with cuts only editing. For example, many NLE's like premiere can smart render video types like DV, XDCAM, DVCPro, AVCIntra, DNxHD. This means original quality is kept, exports are fast. Only sections where you applied filters or overlays are re-encoded and degraded, the other segments are passed through untouched

[Bernix]

Add 1) You can try it on jpeg. If you set 95% quality, try to recompress it several times and then compare with original. I dont mean 100% quality, it is like loss less codecs.

Bernix

[DunnoNo]

It also depends on the type of video and editor you are using , and type of "editing" you are referring to. Certain types of video can be passed through or "smart rendered" with cuts only editing. For example, many NLE's like premiere can smart render video types like DV, XDCAM, DVCPro, AVCIntra, DNxHD. This means original quality is kept, exports are fast. Only sections where you applied filters or overlays are re-encoded and degraded, the other segments are passed through untouched

So if i just use h264 videos they're not smart rendered if i let's do voive over?

[poisondeathray]

It also depends on the type of video and editor you are using , and type of "editing" you are referring to. Certain types of video can be passed through or "smart rendered" with cuts only editing. For example, many NLE's like premiere can smart render video types like DV, XDCAM, DVCPro, AVCIntra, DNxHD. This means original quality is kept, exports are fast. Only sections where you applied filters or overlays are re-encoded and degraded, the other segments are passed through untouched

So if i just use h264 videos they're not smart rendered if i let's do voive over?

Depends on the type of h264 video. Probably not for what you are using; the common types you typically encounter cannot be smart rendered. But AVCIntra is subset h264; professional Panasonic camcorders use it.

But if you're not actually editing the video (ie. not cutting or adding frames, video effects, color manipulations, transitions, or overlays) , and only doing a voice over to add audio - then you can export the audio separately and mux it with the original video stream. The original video will be stream copied, thus no visual quality loss. But the audio would need to be re-encoded (audio track will incur loss if you use lossy audio format)

[SameSelf]

It also depends on the type of video and editor you are using , and type of "editing" you are referring to. Certain types of video can be passed through or "smart rendered" with cuts only editing. For example, many NLE's like premiere can smart render video types like DV, XDCAM, DVCPro, AVCIntra, DNxHD. This means original quality is kept, exports are fast. Only sections where you applied filters or overlays are re-encoded and degraded, the other segments are passed through untouched

pdr, do you know when Adobe added "smart rendered"? IOW, is this a functionality added in CC?

[poisondeathray]

pdr, do you know when Adobe added "smart rendered"? IOW, is this a functionality added in CC?

For things like DV, DVCPro, many years ago maybe CS3 ? or for DV probably even earlier

CS6 added DNxHD, ProRes (Mac version) , AVCIntra, XDCam

http://blogs.adobe.com/kevinmonahan/2012/10/11/smart-rendering-in-premiere-pro-cs6-6-0-1-and-later/

[SameSelf]

Awesome thanks. The reason I asked is because of all the issues I had on my lossless thread which made me think that the only way to pass through PP without PP transcoding and introducing generational losses was with UYVY. I need to do some more testing for sure.

EDIT: On closer reading of your link, it looks like they only added Long-GOP MPEG2 OP1a and XDCAM EX support in CS6. Everything else, DNxHD, ProRes, AVCIntra was added in CC.

[DunnoNo]

Thanks guys


poisondeathray

Thanks for the muxing audio tip.