What is the best encoder out there?

[TyMac]

Cleaner, TPMGEnc?...etc What is you opinion on the best encoder out on the market today?

[Xesdeeni]

I've never heard anything good about Cleaner, either in speed or quality. TMGPEnc is good, but CCE Basic is the fastest and IMHO does a better job.

Xesdeeni

[Kika]

CCE ist the fastest and gives you the best over all quality. In some special cases, TMPGEnc can be much better, but it is much slower, and you have to know a lot about MPEG and TMPGEnc to get the best out of this encoder.
For interlaced Movies, the best Encoder ist ProCoder.

[Arky]

It depends on numerous factors, including your intended average bitrate, your source material (for example, DV, or ripped .m2v), the time you have available for multiple passes, the codec you are using, whether or not you are using Interlaced or Progressive source &/or output, whether or not you have extremely fast action sequences.... the list is endless.

The four main contenders at the top of the market for image quality are (in no order of preference):

#)Cinemacraft
#)Canopus ProCoder
#)MainConcept
#)TMPGENc

For a fairly thorough discussion on this topic see this very recent thread.


Arky ;o)

[Eric Tetz]

I found a clear quality difference between CCE and ProCoder: the ProCoder output was much better. I tried everything to get better results out of CCE, because I wanted it's speed, but the CCE footage always had much more artifacts, especially mosquito noise. Quality is my primary concern, so I ruled CCE out.

Now I see people here claiming the CCE quality is actually better the ProCoder, which is shocking, given the results I've seen with my own eyes. So this is only with non-interlaced footage (mine is NTSC DV)? Would I be better off deinterlacing my footage and using CCE? Is this what you guys do?

[TyMac]

Originally posted by Eric Tetz
I found a clear quality difference between CCE and ProCoder: the ProCoder output was much better. I tried everything to get better results out of CCE, because I wanted it's speed, but the CCE footage always had much more artifacts, especially mosquito noise. Quality is my primary concern, so I ruled CCE out.

Now I see people here claiming the CCE quality is actually better the ProCoder, which is shocking, given the results I've seen with my own eyes. So this is only with non-interlaced footage (mine is NTSC DV)? Would I be better off deinterlacing my footage and using CCE? Is this what you guys do?

How much is Procoder? Do you have a url?

[Eric Tetz]

Sorry to be flippant, but I can't help myself...

Google is your friend.

[Happygolucky]

Originally posted by Eric Tetz
I found a clear quality difference between CCE and ProCoder: the ProCoder output was much better. I tried everything to get better results out of CCE, because I wanted it's speed, but the CCE footage always had much more artifacts, especially mosquito noise. Quality is my primary concern, so I ruled CCE out.

Now I see people here claiming the CCE quality is actually better the ProCoder, which is shocking, given the results I've seen with my own eyes. So this is only with non-interlaced footage (mine is NTSC DV)? Would I be better off deinterlacing my footage and using CCE? Is this what you guys do?

I have to agree, as I've written in another thread, my tests convertining a very high quality Divx AVI movie to SVCD using CCE (2.57.00.09), Procoder (1.01.35), and TMPGEnc Plus (2.513) showed CCE to be horribly inferior to the other two. The picture quality of the CCE encode was terrible, it looked as if viewing through a piece of gauze. Procoder was excellent though it took nearly twice as long to encode as TMPGEnc Plus. The TMPGEnc Plus encode produced the best results. The only difference between it and the Procoder encodes were Procoder had slightly more blockiness in dark areas. I was using DVD2SVCD beta7 and each encoders highest quality settings. For speed and quality, TMPGEnc wins hands down. CCE doesn't even come close.

[Arky]

Originally posted by Happygolucky
I have to agree, as I've written in another thread, my tests convertining a very high quality Divx AVI movie to SVCD using CCE (2.57.00.09), Procoder (1.01.35), and TMPGEnc Plus (2.513) showed CCE to be horribly inferior to the other two. The picture quality of the CCE encode was terrible, it looked as if viewing through a piece of gauze. Procoder was excellent though it took nearly twice as long to encode as TMPGEnc Plus. The TMPGEnc Plus encode produced the best results. The only difference between it and the Procoder encodes were Procoder had slightly more blockiness in dark areas. I was using DVD2SVCD beta7 and each encoders highest quality settings. For speed and quality, TMPGEnc wins hands down. CCE doesn't even come close.

Well, IMHO, the reason CCE didn't come close in your tests is because Div-X is invariably a very noisy source, regardless of how (relatively) well it has been encoded (this comment applies to the squeezing of 90minutes+ onto a CDR, not with shorter duration, higher bitrate Div-X, using a legit codec which doesn't 'max-out').

My point is that ProCoder, as I, and others, have discussed before, looks suspiciously like it uses a significant amount of pre-processing to boost it's subjective output-image quality, even though it does not admit to the user that it is doing so. TMPGEnc Plus also offers considerable pre-processing options to smooth out noise and make the MPEG encoding more effective when using a relatively noisy source. This means that we are not talking about a level playing field here. Had you used AVIsynth to perform similar pre-processing on you Div-X source, prior to feeding it to CCE, I strongly suspect the results would have been much closer.


Arky ;o)

[Happygolucky]

Originally posted by Arky
Well, IMHO, the reason CCE didn't come close in your tests is because Div-X is invariably a very noisy source, regardless of how (relatively) well it has been encoded (this comment applies to the squeezing of 90minutes+ onto a CDR, not with shorter duration, higher bitrate Div-X, using a legit codec which doesn't 'max-out').

My point is that ProCoder, as I, and others, have discussed before, looks suspiciously like it uses a significant amount of pre-processing to boost it's subjective output-image quality, even though it does not admit to the user that it is doing so. TMPGEnc Plus also offers considerable pre-processing options to smooth out noise and make the MPEG encoding more effective when using a relatively noisy source. This means that we are not talking about a level playing field here. Had you used AVIsynth to perform similar pre-processing on you Div-X source, prior to feeding it to CCE, I strongly suspect the results would have been much closer.


Arky ;o)

DVD2SVCD does use AVISYNTH to feed all of the encoders, I believe. Regardless of whether the source was "noisy", if Procoder and TMPGEnc produce output which is crisp and clean but CCE produces output which looks as if it is out-of-focus or viewed through gauze over the lens, the problem would be with CCE given that everything else was the same except for the encoder. And again, perhaps CCE can be tweaked and massaged to produce exceptional results, but the point is that Procoder and TMPGEnc produced excellent results without having to tweak and massage.

[Arky]

Originally posted by Happygolucky
DVD2SVCD does use AVISYNTH to feed all of the encoders, I believe. Regardless of whether the source was "noisy", if Procoder and TMPGEnc produce output which is crisp and clean but CCE produces output which looks as if it is out-of-focus or viewed through gauze over the lens, the problem would be with CCE given that everything else was the same except for the encoder. And again, perhaps CCE can be tweaked and massaged to produce exceptional results, but the point is that Procoder and TMPGEnc produced excellent results without having to tweak and massage.

Ok, well I'm happy for us to agree to disagree , but I will just say 2 things: simply using AVIsynth does not actually mean that you are necessarily using AVIsynth *filters*, and secondly, I find your remarks that: "CCE produces output which looks as if it is out-of-focus or viewed through gauze over the lens" quite remarkable. I've never come across this phenomenon and I do not believe the encoder is at fault, even though you assure me that only the encoders differ between jobs, and not the procedure itself. In my experience, CCE produces exceptionally sharp encodes (if anything, ProCoder has proven to yield more blurred results in my past tests, as I mentioned before). I suspect this may be a resizing issue, in which case, resizing should ideally take place before feeding CCE. You are, after all, encoding for SVCD, and I tend to use CCE for full D-1 DVD, but I have used CCE with GoCCE for brief SVCD edit checking on T.V. and must say I was most impressed with the results.

As I said, though, my remarks aren't meant argumentatively, and I'm content to peacefully agree to disagree with you.


Arky ;o)

[Happygolucky]

Originally posted by Arky
Ok, well I'm happy for us to agree to disagree , but I will just say 2 things: simply using AVIsynth does not actually mean that you are necessarily using AVIsynth *filters*, and secondly, I find your remarks that: "CCE produces output which looks as if it is out-of-focus or viewed through gauze over the lens" quite remarkable. I've never come across this phenomenon and I do not believe the encoder is at fault, even though you assure me that only the encoders differ between jobs, and not the procedure itself. In my experience, CCE produces exceptionally sharp encodes. I suspect this may be a resizing issue, in which case, resizing should ideally take place before feeding CCE.

As I said, though, my remarks aren't meant argumentatively, and I'm happy to peacefully agree to disagree with you.


Arky ;o)

I used AutofitCD, which uses AVISYNTH to do the resizing. Anything else?

You miss the point entirely. While you say CCE can be made to produce excellent results, it obviously doesn't do so by itself or without a lot of tweaking. The other encoders, namely Procoder and TMPGEnc Plus, produce excellent results without any added helpers or tweaking. The only thing CCE excelled in was speed, but frankly I'd rather have a quality result that took a bit more time.

I've read all the posts and reviews and it does seem CCE can be made to produce very high quality output, but I have a life and don't have all the time in the world to play with this filter and that utility just to get CCE to do something others do out-of-the-box.

As to the out-of-focus results from CCE, I'm not the only one to have noted that fault, and again, perhaps tweaking and using helper apps and filters would solve that problem. But the other encoders don't need filters and helpers to produce great quality.

If you want to play all day with finding the right settings and filters, that's great, I'm glad you have the time to do so. I have other things to do.

[Eric Tetz]

"CCE didn't come close in your tests is because Div-X is invariably a very noisy source..."

I did my tests (ProCoder,TMPGenc,CCE) with relatively clean DV footage (NTSC TRV900, outdoors on a sunny day), and came to the same conclusions. ProCoder's output was significantly better, in terms of visible artifacts. CCE was the worst, with the most mosquito noise and the greatest distortion of smooth areas.

"ProCoder, as I, and others, have discussed before, looks suspiciously like it uses a significant amount of pre-processing to boost it's subjective output-image quality..."

Has no one verified this? Have you guys tried sharing the source files, encoders settings, and resulting MPEGs so that more than one pair of eyes can evaluate the results?

For me the differences were extremely obvious. I fed all three encoders the same, difficult footage -- my son's face moving around a very busy, rapidly moving background. I turned all (optional) filtering off in all apps, and cranked every setting up in favor of quality (over file size or encode time). With CCE, my son's face was very distorted, unrecognizable. With TMPGenc, his faced was smoothed out, blurred -- more pleasing to the eye, but still distorted. With ProCoder, his face was clear as day. I had to look hard to even find artifacts.

"Had you used AVIsynth to perform similar pre-processing on you Div-X source, prior to feeding it to CCE, I strongly suspect the results would have been much closer."

How can one perform "similar pre-processing" when no one is even sure that pre-processing is taking place? I would love to get ProCoder results out of CCE -- I want that speed.

[Arky]

Well in terms of sharing files, there are MAJOR issues of storage and bandwidth, sadly! (although very few of us have not wished this was viable. In time, I have no doubt that it will happen).

I don't claim to have all the answers on this topic - as you are no doubt aware, there was an extremely extensive discussion here

I'm not saying either of you are wrong, I'm simply trying to point out that, like it or not, there are mitigating factors, many of which you will find mentioned in the above thread. There are indeed differences between the encoders, and some are better at one type of encoding and vice versa - I do not dispute this. It's just that I do not think CCE should be written of so broadly or severely - it is an extremely capable encoder.


Arky ;o)

[DaveQB]

what bout avi2mpg2 or called bbmpg
it seems to use both those names

its a freeware encoder,

i havent used it yet, but i do have it in my video editing directory and intend on giving it a shot next time i need to encode mpeg1/2 (i dont do it that much)

[Happygolucky]

Originally posted by Arky
It's just that I do not think CCE should be written of so broadly or severely - it is an extremely capable encoder.

Arky ;o)

I don't doubt that CCE has uses to certain professionals and given its extensive array of options and plugins I'm sure that professionals prefer that sort of action in an application. But for the casual home user, there are better and easier choices out there that produce exceptional results without having to take a college course in it to get it to work.

[Eric Tetz]

"Well in terms of sharing files, there are MAJOR issues of storage and bandwidth, sadly!"

I have 80 megs of Yahoo! Briefcase space, which is more than enough to compare like 10 seconds of footage, right? If anyone has suggestions on what good test footage should include, or if they have some appropriate footage already, let me know.

If 80 megs is not enough, I will gladly pay for more storage space. I would love to post the source file and say "have at it, give me your best DVD-compatible encode of this" and see what the CCE gurus can produce.

"I don't claim to have all the answers on this topic - as you are no doubt aware, there was an extremely extensive discussion here"

I read that, and it was pretty much the same. No general agreement.

"there are mitigating factors, many of which you will find mentioned in the above thread. There are indeed differences between the encoders, and some are better at one type of encoding and vice versa - I do not dispute this. It's just that I do not think CCE should be written of so broadly or severely - it is an extremely capable encoder."

CCE hasn't been written off! Notice that this message board has a TMPGenc and CCE forum, but no ProCoder forum. People jump through flaming hoops to get good results out of CCE, yet write-off ProCoder because it is "probably cheating". It just doesn't make any sense to me.

"I don't doubt that CCE has uses to certain professionals and given its extensive array of options and plugins I'm sure that professionals prefer that sort of action in an application. But for the casual home user, there are better and easier choices out there that produce exceptional results without having to take a college course in it to get it to work."

For the record, that's not my position at all. I'll gladly take a college course in it, if it will deliver the goods: super high quality encodes. In reality, it's not that complex, certainly less complex than TMPGenc. It's just that the results I've seen with NTSC DV have been pretty poor.

[mrlipring]

if you guys want to share a big file, check out bittorrent, it was made for that, and i'm sure you'll be impressed.

[auenf]

Originally posted by Eric Tetz
[i]I have 80 megs of Yahoo! Briefcase space, which is more than enough to compare like 10 seconds of footage, right? If anyone has suggestions on what good test footage should include, or if they have some appropriate footage already, let me know.

If 80 megs is not enough, I will gladly pay for more storage space. I would love to post the source file and say "have at it, give me your best DVD-compatible encode of this" and see what the CCE gurus can produce.

yahoo might have 80mb's, but whats the daily transfer limit?

Enf...

[echooff]

if you take the time to learn how to use cce, you will be impressed with the results. there can be no doubt it is a great and fast encoder. i use tmpgenc for most of my encoding, but it has truble taking some files without additional tweaking. i have even had it reject some file completely. procoder will take anything and produce an pretty good output, and thats right out of the box. it also has tweaks. procoders big problem is; it is sooooo sloooow. thats why i use tmpgenc most. good result, fair speed. i am slowly learning to use cce, but the avisynth scripts test my patience. i admit i'm not good at it yet. the few i have gotten all the way through were very good. btw all i'm talking about is home movies from vhs i recorded over a 5 year span in italy. some have extremely fast motion in them. not from what is being flimed but from the cameraman. he he people i have managed to trap and bore for hours have gotten seasick from watching my early videos