We discussed that recently re: the Mainconcept encoder.

I tried a couple x264vfw encoders inside Vegas, but never warmed up to either of them.

2-pass rendering with those codecs gives you precise the average bitrate you gave in.
1 pass rendering does not and two pass rendering has the advantage to exactly match a predicted filesize for your goal if that is important.

 

 

The two pass render is more efficient - - that means it can produce the same quality at a smaller file size.

I use x264 encoder for rendering

If that means you are using the X264vfw encoder inside Vegas, the rest makes no difference. It's inefficient as hell, because it's a crippled format.

Question, if a vídeo is an hour long, 720p, youtube bitrate (4-10 average? I don't have it memorized)

Is it worth it? In your opinion? Because it takes longer, and let's say you have one of those computers who can only record to 720p30fps meaning they are not super computers nor anything, they are not powerful so it will take long, more than an hour

What do you think guys?

IMHO, it’s not worth it in the situation you described.

IMHO, that video is going to be out there a long time (or at least until you delete it), do you want it right, or do you want it fast?

It takes too long :/ but it has better quality

I think there are more factors to decide on whether to use 2 pass rendering, I use x264 encoder for rendering (can't use hardware since it sucks)
 

Somebody correct me if i'm wrong, but a potentially good use of GPU hardware is for the first pass of a 2 pass rendering. That should really fly and as all it needs to do is give the bitrate peak values for each slice/frame quality should not be an issue. Then on 2nd pass where it actually renders it uses a quality AVC software render with the values obtained from the gpu hardware read..

2pass rendering gives no better quality than singlepass rendering, it's like Musicvid already explained, a more efficient use of the selected bitrate which results in a smaller filesize. If Quality is Pre, just use singlepass rendering with CBR at max Bitrate.

I used to always use 2Pass FLV encoding for YouTube back when you could upload your own video files and YT would no re-encode with their low quality renders if uploaded in flash and under a certain bit rate. So the best option for best quality at predetermined bitrate was 2-pass.

I used to always use 2Pass FLV encoding for YouTube back when you could upload your own video files and YT would no re-encode with their low quality renders if uploaded in flash and under a certain bit rate. So the best option for best quality at predetermined bitrate was 2-pass.

So YT can distinguish your upload if it's two-pass rendered or single-pass? A predetermined bitrate is that a predetermined average bitrate?

YT didn't care about single or dual pass, it was the average bitrate which could be accurately determined via dual pass. If it was too high they would re-encode with their inferior encoding. They re-encode everything now though. It's part of the process of checking for copyright music/video.

IMHO, that video is going to be out there a long time (or at least until you delete it), do you want it right, or do you want it fast?

Your opinion? I already have mine, it depends

You?

It takes too long :/ but it has better quality

I think there are more factors to decide on whether to use 2 pass rendering, I use x264 encoder for rendering (can't use hardware since it sucks)
 

Somebody correct me if i'm wrong, but a potentially good use of GPU hardware is for the first pass of a 2 pass rendering. That should really fly and as all it needs to do is give the bitrate peak values for each slice/frame quality should not be an issue. Then on 2nd pass where it actually renders it uses a quality AVC software render with the values obtained from the gpu hardware read..

I guess I got you....? A first pass rendering doesn't take long is only the second pass

I don't know what's your point but yeah that's kind of how it works I guess

2pass rendering gives no better quality than singlepass rendering, it's like Musicvid already explained, a more efficient use of the selected bitrate which results in a smaller filesize. If Quality is Pre, just use singlepass rendering with CBR at max Bitrate.

You sure :/ do you have any proof or anything? Explanation?

Well, my opinion is I want it right.

It takes too long :/ but it has better quality

I think there are more factors to decide on whether to use 2 pass rendering, I use x264 encoder for rendering (can't use hardware since it sucks)
 

Somebody correct me if i'm wrong, but a potentially good use of GPU hardware is for the first pass of a 2 pass rendering. That should really fly and as all it needs to do is give the bitrate peak values for each slice/frame quality should not be an issue. Then on 2nd pass where it actually renders it uses a quality AVC software render with the values obtained from the gpu hardware read..

I guess I got you....? A first pass rendering doesn't take long is only the second pass

 

I don't know what's your point but yeah that's kind of how it works I guess

 

2pass rendering gives no better quality than singlepass rendering, it's like Musicvid already explained, a more efficient use of the selected bitrate which results in a smaller filesize. If Quality is Pre, just use singlepass rendering with CBR at max Bitrate.

You sure :/ do you have any proof or anything? Explanation?

Do you have any proof? For me 2pass rendering is just a waste of time, I don't see any difference. 2pass makes only sense if you have limited storage space, has nothing to do with quality.

File size affects upload time. OP is apparently in an underserved area.

OP needs to balance the render : upload time to his satisfaction. If it were me, I would use 2 pass and High Profile to optimize upload time on relatively slow connections.

My observations based on actual tests:

I rendered from a common source in Handbrake to the youtube reccomended upload bit rate, 8000 kbps for 1080-24p.

For a 20 second source:

1) The resulting output file size was 23.815 Mbps for the 1 pass file  and 20.768 Mbps for the 2 pass file. This is about a 13 % reduction in file size.

2) The peak bit rate of the 1 pass file was 27.871 Mbps while the peak bit rate of the 2 pass file was 12.085 Mbps. [opinion] This variation is significant. [/opinion]

Placing both files back on a Vegas timeline and comparing to the source, I observed no visual difference between either of the rendered files and the source in the Preview Screen. The Videoscopes showed something going on.

Placing both files back on a Vegas timeline and comparing each to the other, I observed no visual difference between the rendered files in the Preview Screen. The Videoscopes showed something going on.

My Conclusions:

2 Pass renders more closely match the bit rate target with lower variation from that target.

I see no visual evidence that one render methodology is better than the other.

The upload time improvement and its benefit to mankind from a 13% reduction in file size is a value-judgment that the user has to make for him or her self.

My tests are recorded for peer review.

Yes, 2) is significant. For any streaming media, lower peak bitrates mean less stuttering on delivery. The amount of difference you found though, is surprising.

"The amount of difference you found though, is surprising."

[SWAG] My source was outside on the ocean. It had lots of motion which would raise the average, but no flashing strobes, which might raise the peaks.[/SWAG]

Somebody correct me if i'm wrong, but a potentially good use of GPU hardware is for the first pass of a 2 pass rendering. That should really fly and as all it needs to do is give the bitrate peak values for each slice/frame quality should not be an issue. Then on 2nd pass where it actually renders it uses a quality AVC software render with the values obtained from the gpu hardware read..

I guess I got you....? A first pass rendering doesn't take long is only the second pass

I don't know what's your point but yeah that's kind of how it works I guess

I don't think you've ever used 2pass rendering in Vegas. Due to my former bad upload speeds I used it daily.

A 2pass software render takes about twice as long as the single VBR render. My 2m30s VBR takes 5m15s in 2pass VBR. My idea is a codec that uses HARDWARE NVENC or QSV to do first pass lightening fast read & collect VBR bitrate values, then on 2nd pass, software encoding is used for the superior quality.

I notice though there is no 2-pass option with MAGIX AVC hardware encoding using QSV or NVENC. Strange.

Actually, water in motion can murder peak bitrates if there are lots of sparkly (specular) reflections. Like thousands of little strobes.

​​​​What does it take so long? Damn, waste of time

So, let me get this

​​​​1 pass rendering: -rendering time, +quality, +file size, +upload time

2 pass rendering: +rendering time, -quality, -file size, -upload time

My connection is good so I don't care about upload time and my memory is alright

So well, I may do my own test since I am not sure about what you guys are saying but I guess you think something was wrong with his test. My computer is weird and Sony Vegas has been crashing a lot lately (probably because of windows update)

If you don't care about upload time, render single pass.

I also suggest you download Handbrake and play with CRF encoding, which is the best of both worlds.

Also suggest you download Handbrake and play with CRF encoding, which is the best of both worlds.

What's handbrake? And I may stay with VRB ^^ll I mean, that's what I record at and that's what youtube recommends anyways (if I am not wrong)

Why CRF?

Note:How many of you think I should go with single pass?