View Full Version : A small test HD Xvid 1.0.1 against RV10e ( 3500 kbps)
guada
9th June 2004, 23:11
Hello everybody,
Last night, I did a test in 720p on aavi file of 45 mo (surely exempt of sound).
My impressions were very conclusive because, the quality was really to the appointment.
Besides, I noted that the encoding of the codec Xvid 1.01 was very slow in relation to the RV10e; and that even to high bitrate the RV10e was keeping a hot and more definite picture visually than its adversary.
It doesn't change anything of the exploding character of the two codecs, I assure it to you.
My questions:
Does the slowness of the Xvid codec is normal to this bitrate and to the chosen resolution?
The Rv10e codec seems limited in high bitrate how to make to increase its value (more than 4000 kbps)?
pogo stick
10th June 2004, 00:21
Originally posted by guada
Does the slowness of the Xvid codec is normal to this bitrate and to the chosen resolution?
That sounds strange to me.
Because my tests showed me opposite results concerning encoding speed.
iwod
10th June 2004, 04:27
I don't understand........ under default settings, xvid sure always encode faster than RV10. Since you said RV10e so i suppose you have EHQ on. ( And apart from H.264 i can't see any other common codec encode slower than RV10 )
So please mention how you did with Xvid setting.
And i don't understand the point of testing Codec @ 3500Kbps..
Why not use Mpeg 2 and just creat a DVD?
E-Male
10th June 2004, 09:18
maybe becuse DVDs don't support HD content, yet
but i have to agreet hat the speed difference is strange
guada
10th June 2004, 23:29
Hello everyone,
I beg your pardon. I think that myself that i made a mistake about the type of the Xvid codec used.
In fact it was only the Devapi-3 optimized by my care.
But I didn't wish to stop me so early since the opinions were too numerous; I have decided to start again the experience with the last codec Xvid 1.0.1.
And that's, the good and bad surprises:
In time: - Xvid 1.0.1 pass1:14''50 / 2nd pass: 40''20
- RV10E pass1:18''01 / 2nd pass: 19''50
The size: - Xvid 1.0.1 116 mo
- RV10E 121 mo
you will tell me it's strange but it is what says my PC
my work:
an avi file of 45.5 mo converted in AVS sript.
Xvid 1.0.1
- Profiles: unrestricted
- Target size: 123 904 (3396 kbps)
- UH
- 4. wide search
- 250 max I-I
- Quantization: 2/4 2/14 2/14
- treillis
RV10E
- Realproducer plus 10 (erv4 dll OK rv10)
- RME 9 preview gui
- Profiles: NewParams
- Average bitrate: 3396 kbps (3313 kbps after encoding)
- Max startup latency: 25
aaar9800
11th June 2004, 22:44
Don't get me wrong, but why do all people that praise RV9 and RV10 come from France? Again, I have nothing against french, it's just weird.
BTW, why didn't you use startup latency of 60 for RV10. And what about chroma motion and chroma optimization in XviD - did you use those?
Teegedeck
11th June 2004, 23:31
Excuse me, guada, but this test may be fun to do for yourself but it does have little meaning. Please make the effort and learn how to configure these two codecs properly and how to conduct a test and - last but not least - how to announce a test (and this includes discussing the selection of sample-files and avs-script). Then many more people will take interest in it.
Good luck,
Tee
CruNcher
11th June 2004, 23:37
@aaar9800
simple answer :P france->japanese anime freaks culture trade ;)
guada
11th June 2004, 23:55
Hello Aar9800,
I don't have anything against the codec Xvid 1.0.1, even less against that the RV10e codec.
I try to confront some codecs in relation with a present context rightly.
Nowadays, it is the HD that becomes clearer.
The adjustments that I have chosen correspond to my experience concerning encoding.
You must know, that I have a french nationality.....
For more informations on the optimizations of videos codecs, don't hesitate to read a few more the testimonies of the experts, as well as the numerous available guides on quite a lot of video site.
guada
12th June 2004, 00:36
Hello Teegedeck :)
I'm sorry but,I don't know anymore what to think. :confused:
Particularly, when everybody pretends to me the opposite.
I think it's time to stop, sorry :confused: :(
Perhaps, see you soon
tct666
12th June 2004, 00:44
Which "encoding complexity" do you set in your RealProducer Plus 10?
According to your test in 2pass in RealProducer Plus 10 ,it seems that you set RV10-Low.
The higher the EHQ is,the more time RPP10 needs to encode.
EHQ
90 RV10 high
75 RV10 medium
65 RV9 high RV10 low
57 RV9 medium
50 RV9 low
Teegedeck
12th June 2004, 02:46
Guada, it's better to discuss the settings to before actually performing the test... Otherwise the fans of either codec tend to come down on you like a ton of bricks! Don't worry, that's normal. ;)
Better luck next time, try to be as vocal as possible and you'll likely get more positive reactions.
guada
12th June 2004, 08:55
Thank you Teegedeck for your humble character. :) :)
With your experience I can only be inspired.
I think that henceforth your advice will travel with my next steps.
What do you expect " the spirit is willing but the flesh is weak " (the Bible).
Thank you to be at peace with myself.
To soon ;)
aaar9800
13th June 2004, 02:20
guada, i dont think i need to read anymore stuff for configuring rv10 - I have already done about 20 dvdrips using that codec, but i used that only because i was either limited on space or was encoding cartoon stuff.
I dont mean to start a flame, but wtf - dont tell me that i need to go and read something.
Anyway, use whatever suits you best.
SeeMoreDigital
13th June 2004, 12:31
Hi guada,
I'm actually interested in your 720p source. Can you post some details about it please?
Cheers
damrod
13th June 2004, 15:33
EHQ
90 RV10 high
75 RV10 medium
65 RV9 high RV10 low
57 RV9 medium
50 RV9 low
ok oki it explains why rv10 took much time to encode than rv9 ;-)
Sirber
23rd June 2004, 20:00
Originally posted by aaar9800
Don't get me wrong, but why do all people that praise RV9 and RV10 come from France? Again, I have nothing against french, it's just weird.Im from quebec dude :p
damrod
23rd June 2004, 20:26
and back at home no?
SeeMoreDigital
23rd June 2004, 20:47
And I never did get an answer about the source?
damrod
23rd June 2004, 20:50
true!!!
:-))
a 720p video : which resolution?? 3500kbits as bitrate... why so high?
can't we use 2Mbits as bitrate for hd encoding with rv10 or xvid?
SeeMoreDigital
23rd June 2004, 21:30
The source is very important when generating genuine HD encodes!
Many people seem to think upscaling DVD content provides a good indicator but it doesn't!
Of course you can generate HD encodes at around 2Mb/s but only if the source is suitable. Better to aim for 3.5Mb/s because it offers around 150mins (audio permitted) when stored on a DVD and the quality is excellent.
Cheers
unmei
24th June 2004, 11:32
sorry if i come up with something really stupid, but what is 720p ? 1080x720 @ 25/30 or 50/60 Hz? ..because if it were this resolution at only 25/30p, 3.5mbit/s seemed very high for me..
i never encoded at HD, but this would be about 2x the pixel/s i usually do and therefore i would assume 2x the bitrate i'm used to would be enough? Any brainos in this?
SeeMoreDigital
24th June 2004, 12:16
OK,
"720" refers to the vertical quantity of pixels (NTSC DVD and DV cam uses 480 and PAL DVD, DV cam and DVB uses 576)
The "p" means that the video is being shown progressively (just like when you see a film at the cinema) ie: you are seeing one full image frame at a time, and one after the other.
By rights HD images should be broadcast using 'non' anamorphic image pixel frame sizes (NTSC DVD uses 720x480 and PAL DVD uses 720x576 both of which are anamorphic). This means a 16:9 image should use 1280x720 (921,600 total) pixels.
However, when 720p video is broadcast, in order to make the video appear progressive, the frame rate is lifted from the usual 30fps (as used which for the more common interlaced "i" broadcasts) to 60fps.
Technically (from an engineering point of view) there is no reason why 720p has to be run at 60fps. As it can still be progressive at 30fps (or 25fps if it was available in Europe). I guess it's done this way in order to provide maximum flexibility with different receiving / screening devices.
Cheers
Joe Fenton
24th June 2004, 22:22
HD progressive is shown at 60Hz for the same reason that movies are shown at 48 FPS - to minimize flicker. Originally, movies were all shot and shown at 24 FPS. This caused very bad flicker, so projectors were designed to show the same frame twice raising the rate to 48 FPS.
If HDTV were displayed at 30 FPS, it would flicker badly (like a 60Hz interlaced display). So they do the same thing and show each frame twice, yielding 60 FPS.
billou2k
25th June 2004, 14:46
A quick and probably stupid question about real video as I couldnt find the answer elsewhere...
Can RV10 encode interlaced video as interlaced? (with no deinterlacing applied)
thanks.
unmei
29th June 2004, 12:52
mm kay, a lot of info, thanks, but i think i still don't know what i actually wanted to know..
What is the pixel/s rate you are talking about when you want 3500kbit/s ..or even more to the point: how many bit/pixel?
SeeMoreDigital
29th June 2004, 13:23
Originally posted by unmei
mm kay, a lot of info, thanks, but i think i still don't know what i actually wanted to know..
What is the pixel/s rate you are talking about when you want 3500kbit/s ..or even more to the point: how many bit/pixel? Personally I don't know... maybe Karl Lillevold (over in one of the RealMedia threads) can help you with this one!
That said, as all our favourite codecs improve, so has their compressibility. If I remember rightly, even Doom9 was forced to comment recently that the old Mpeg4 bits/pixel ratio needs revising.
In any event compressibility will of course depend on the source content.
Cheers
vBulletin® v3.8.11, Copyright ©2000-2026, vBulletin Solutions Inc.