TMPGEnc 3.0 XPress Released

[pepegot1]

Let's hope this one is faster than previous versions.

This new TMPGEnc software provides a completely
reworked interface and a new optimized
MPEG-1/2 encoder engine!
All the main functions of the now famous
"TMPGEnc Plus 2.5" has been improved in both
performance and usability.

TMPGEnc 3.0 XPress
Upgrade Download version price: US$ 58.00
Upgrade Price for registered user: US$ 38.00 35%OFF

[Baldrick]

no ac3 audio encoder included....

[Forum Troll]

I got the same email, cause I bought the DVD Author from them. Do I need to buy this new encoder from them, or can I get the upgrade for free, since I am a customer of their other product?

[Baldrick]

But you can use the tmpgenc dvd author ac3 plugin.

[Baldrick]

I can't say that it is that much faster..., 6 minutes to encode an 1 minute dv clip using 2 pass vbr and standard quality settings. Took 7 minutes with tmgpgenc plus.

[thor300]

Great news, Baldrick saved 1 minute, tmpgenc 3.0 rules. I doubt it will ever be good, by the time they reached a normal encoding speed and compliance, imagine where the other encoders will be? Dont misunderstand me, i HOPE pegasys will make it good some day, i wouldnt mind having a proggy that can do advanced filtering and generally being all-in-one, tmpgenc is just still not an option. Anyway, looking forward to read more about experiences with the new version.

[Baldrick]

Normal encoding speed for me is 24 hours for a 2 hours video...

No, they should spend more time improving the encoder engine speed....it is still very slow compared to cce and mainconcept. Even if I use the fastest encoding options I can't get any faster encoding than 8 fps, 0.3 x realtime.

[pepegot1]

With their paranoid WIN XP type activation, they have gone too far. Imagine working on a project and then the encoder does not work because of their stupid activation system. Sayonara Baby!! Besides-there is no great increase of speed, so I will stay with 2.5 for the present.

[tag]

....Imagine working on a project and then the encoder does not work because of their stupid activation system....

Do all AC-3 encoders exhibit similar behavior or just the one in TMPGEnc? What about the one in Vegas, for example?

[KBeee]

And being Windows XP only makes it useless for anyone trying to keep as much personal info out of MS hands as possible by sticking with older OS's.

[NightWing]

Do my Vulcan peepers deceive me? Its likes a 1280 x 1024 screen size? Wow!

[victor010170]

Any one have any comment on Tag? CCE with Vegas 5?

[SatStorm]

This week I'm really very busy with my real life job, I don't even visit the forum as much.
But I did a short test yesterday

First of all, it works with win2K.
You need to connect to the internet to make the demo work. But after this, you can disconnect. Not practical indeed, expecially if you, like me, still use PTSN dial ups.

Anyway:
It seems that it is simply TMPGenc plus 2.5 with a different GUI. It seems somehow faster, but it must be speed optimatisations or the better support of the SSE instructions. On my AMD 2600 the difference is not that much.

Quality wise, I don't see ANY reason to switch to TMPGenc Express 3. The picture is the same. Also, I don't use winXP and I don't connect to the internet all the time. Don't mention that the PC I use for encoding and capturing is dedicated, I even manually unistall internet explorer from it! (there are ways, even for win2K).

So, my first impression is that it is not what I need now to encode. Version 2.5 still holds great for my needs, Mainconcept all the time improves and CCE Basic is always a great alternative, unbeatable compared with avisynth.

But it is really strange... When TMPGenc started, some years ago, it was a freeware application. Now, it is a hardcore protected application. Some things really do change over the years. And as my grandmother says: All the good things sometime ends.

[thor300]

Very informative post Satstorm. I guess you did DVD encoding? What i wonder, and hope anyone tried or will try soon, is the part where 2.5 sucks so badly; SVCD at "low" bitrates. Anyone seen any improvement with 3.0 there? Not much hope for a reply, who makes SVCDs nowadays... When i did SVCD the difference between CCE and Tmpgenc was like night and day. The best thing about tmpgenc IMO is that its existence probably help keep the price of CCE Basic down

[accountwiz]

I used 3.0 to convert Lost in Translation to DVD format and it looked fantastic. Going to convert the same flick to both VCD & SVCD this weekend and see how they look. I think this is a great tool for the money.

[thor300]

Good accountwiz, post here so we can know what you think of the SVCD/VCD encodings. Btw, i seen your avatar before, are you the mod/admin of cdfreaks? The Norwegian guy?

[SatStorm]

I encode 1/2 D1 DVDs, so in a way I encode something close to xCVD.

The last one and a half year I use minimum bitrate 2000kb/s with TMPGenc. That way I eliminated the well known problems with the static scenes . It is the only way I know with this encoder for good results.

If you source is DVB transmissions or avi captures, you don't need to encode to full or cropped CCIR 601. A minnor filtering and a framesize like 352 x 576 is an excellent choice. This of course is my opinion, others may have another.

TMPGenc mention something about new mpeg 2 engine. I didn't notice something, but I'll test further to see if there is indeed a difference and I don't see. Many tests are needed for a new program, also time. And time is something I don't have this week...

[accountwiz]

Well, my testing of 3.0 goes on as time permits. As I mentioned earlier, I did an encode of Lost In Translation to DVD-R size that looked perfect on my 27-inch WEGA Sony TV. The dark scenes in this movie really challenge an encoder and 3.0 did fine shrinking the movie just enough to fit on a disc.

I thought I'd try some real compression next to see at what level I could make it break down and, somehow, I created a 1.3gb mpg of the movie using the 'XDVD' setting that 3.0 now has. This seemed ridiculous, but I went ahead and burned it to an RW and played it on both my Sony and Toshiba Flats. There's a graininess/fuzziness that you really notice at first, but it's better than the VCD/SVCD and even Divx/Xvid stuff I've seen done before.

At 1.3gb, that means 3 movies fit comfortably on a DVD with resolution less than DVD-quality but above the other standards I've seen in action. It looks exceptionally promising for capturing TV shows to DVD, but I'll also use it for movies where I don't want to constantly cart out the original disc...and of course, the convenience of having 3 films on one DVD-R that look pretty good is an incentive, too.

[SatStorm]

Most of the "issues" TMPGenc has, are invisible to mainstream / typical TVs.
Those things some few notice, shows on LCD and Plasma screens mostly.

Myself, when I compare a PAL SxVCD (Sefy's xSVCD variation) with a CVD, I don't notice a difference in my Nokia 29" 100 herz TV. But I detect a huge difference on my philips 36" pixel plus and even more difference on the plasma and LCD flat TVs, I'm able to test in my job (I'm a lucky bastard about this...)

[Sefy]

@SatStorm, when you say you detect a huge difference, is my SxVCD variant looking bad on the large screen TV's ?

[SatStorm]

No, not bad at all...
Those interlace SxVCDs I use to made with CBR at ~ 1500kb/s are looking excellent. Like a good VHS. Those I use to made with 2 pass VBR, with an average of 1200kb/s, look less good. But that was TMPGenc's fault. Bad resizing + bad bitrate handling (we couldn't know those things back then!) results a picture with unoticable noise on a mainstream TV. But on plasma, it is very noticable noise. Like ghosting and moskito noise at the same time! With the LCD screens the noise is less noticable, but you have a motion blurness. LCD don't like 288 interlace frames...
Those projects I use to filter and resize before converting them to SxVCD, look excellent even at 1150 vbr (400 min 1150 average 2520 maximum). They trully look like a good VHS tape!

SeVCD looks sometimes better SxVCD. It has a more crisp picture, but with terrible colours.

But most of my SxVCDs, use to be "the most possible per disc". Those, look terrible unfortunatelly. But not those with enough bitrate!

[Sefy]

That's Interesting! I wonder if there's a way to do a "template" like that for CCE or MainConcept and to see how they would handle it

[andie41]

In my brief experience with Tmpgenc 3.0, I noticed the encode time seemed to be slower than 2.5, but the quality seems improved - smaller file size and, according to Bitrate Viewer, more efficient. However, it iseems to compress the luma - 16-235 becomes 30-218. Anyone else notice this?

[Sefy]

I've only noticed that it is indeed slower from 2.5 since it took me faster then realtime to encode VCD and now it's much slower

[VideoTechMan]

I read about the 3.0 Xpress and I thought it was a good product. I downloaded the demos of Xpress and DVD Author and they both work pretty well for me, so I went ahead and bought them, along with the AC3 plugin. I also dont like the fact that the plugin requires periodic validation, but since my system is connected to my DSL line at home its not too much a big deal.

For me I think the big winner of the plugin is that it works with both Xpress and DVD Author, and that you can encode to MPEG2 and the sound to AC3 at the same time, which is a BIG plus so that you can encode at a higher bitrate right from the start, rather than having to convert the hefty WAV file to AC3 after encoding.

Since I am just getting into DVD production, this suite of progs would be great to start out with, and once I hone in my skills I can move on to higher products.

VTM

[seventhseal]

My time doing 2-pass VBR encodes has been cut by more than HALF with 3.0, with the same outstanding results. Woo hoo!

[Sefy]

My CBR encoding to VCD only has been increased by nearly that much
I used to encode to VCD in faster then realtime, now it's slower, which is why i'm sticking to 2.5 for my encoding

[paulw]

Any one care to comment on their XDVD option . ie what is is and how it works??

TIA.

[seventhseal]

Had a look at my 3.0 version 2-pass VBR encode DVD on a 32-inch TV. It disappoints me that I have to report inferior results.

I did a 2-pass VBR, noise reduction, max 8000 avg 4400 min 1750 1I 7P 1B encode of a 100 minute video on 3.0.

I did the same video with a CBR bitrate 4400, noise reduction, 1I 7P 1B on 2.5.

Would you believe the CBR version looks better? The VBR has pixelization throughout and the noise reduction is terrible.

Also, 3.0 seem to add brightness to the picture for some reason. Not sure if it's the noise reduction or not. But the blacks aren't as deep as they used to be, as if someone pumped the brightness up on the TV or something.

[thor300]

Just give up, tmpgenc will never be good. Tmpgenc fans will sooner or later have to admit that AviSynth with CCE Basic is superior by far and cost the same. Start learning some basic avisynth, its not difficult at all for basic conversions, all you really need to get going is AviSource, lanczos/bicubicresize and a ConvertToYUY2(). 3 lines, that cant be so hard?