"Could anybody give an advice about what kind of conversion I should do to the videos my HD DVB-T MPEG4 Receiver produces in order to make them smaller (in terms of disk space, not screen size)? "
I don't know that there's an easy, one-size-fits-all answer, but it does bring up some stuff that might be of interest or use to you & others...
Now the first step if you want to save your video in some other frame size, make the file smaller, &/or do some sort of editing, is to determine what it is you've got -- the free Media Info is a great place to start. Something to be aware of, lots of broadcast video, whether it's carried by cable or satellite or whatever, actually uses reduced size frames that are reported as full size by the video stream or file. If/when this is the case, it will never look quite as good as video using a full sized frame, either as-is or after conversion. If you know what sort of video you have, you can both research conversions & methods, and predict the success of any conversion before you go to the time & trouble.
HD broadcast, whether OTA [Over The Air], cable, satellite etc., normally has an internal DRM flag that may or may not be set. Many [most?] recorders, software & hardware, will not record HD as-is, but convert it to SD beforehand. HD video hardware like TVs & players etc. use HDMI or component cables. HDMI is capable of 1080p, but nothing that accepts HDMI as an input will record that HD video. Using component cables [basically 3 audio cables with RCA connectors] you skip HDMI's HDCP DRM, but max out at 1080i [interlaced] & may have color issues [there is no standard for component input/output]. That said, I think most OTA, cable, & satellite HD video is interlaced 1080i rather than 1080p anyway.
I'd suggest learning what your original HD video signal is to start with, and how your box saves or stores those files. If it stores the video as SD, &/or if you want more control over the format of the saved video, one solution might be one of those boxes or PC cards that records from the component out, assuming your set top box has component out jacks. While that usually works, there *may be* a better alternative, but it doesn't work nearly as often, and that's to get the original programming stream via a firewire connection on the box. As far as your set-top box itself goes, you'll need to write down the brand, model, and maybe production date &/or serial number and use Google. I'd expect you also might need firmware version number, but there's a great chance you'd have to use Google anyway to find out how to get that version number. And another possibility is to check out what you can do with a TIVO.
At any rate, once you've got the best original video you can get on a hard drive, then you can set about re-encoding it for smaller files if you want. The odds of you being happy with the results however go down when/if the video's already been degraded either before broadcast or before it's saved by your DVR.
* * *
".wmv file it said “failed to import file”... It plays fine also plays in VLC media player so the file is fine."
Video standards all relate to the player -- they have a list of requirements that the player has to handle, but how you get there encoding the file is up to you. I'm Not saying the .wmv file was good or bad -- I've no way of knowing -- but rather want to say that if the file will play, that's absolutely no guarantee that you'll be able to open it in any software other than a player. In this case since it failed in iSkysoft Video Converter Ultimate, trying another converter would be the right move. But if it failed there too I wouldn't start trying every converter I could get my hands on, but would take a good hard look at the video file.
* * *
"So far I’m impressed. I can actually just drop a WTV file and, bingo, it converts. If you’ve ever went through the process of first converting to DVR-MS and then finding programs that can deal with this format, you’ll know what I mean."
If you record/save a video file in win7's Media Center it uses the .wtv format to add DRM if the original video had the DRM flag set. You can't do much with .wtv files with DRM other than play them on the PC where they were recorded -- .wtv files without DRM can easily be copied to files in the traditional Windows Media Center format of DVR-MS via the win7 context menu. When win7 was young it was a bit of a hassle, but today there are loads & loads of apps that'll accept .wtv without DRM along with DVR-MS files.
* * *
"Xmedia recode is still better and free. It allows you to covert video and simply copy the audio so as not to lose say DTS-HD master audio [5.1] or True HD [also 5.1]. Most of these converters change the audio to AAC or MP3, AC3 if you’re lucky. I convert all my blu rays and dvd’s to mkv using it saving lots of space and discs. II can put 3-4 movies on a dvd as a data burn, the br player plays back perfectly. Or put them onto a flash drive. MKV is the format of choice as it allows for uncompressed audio in any format."
I don't mean to sound contradictory -- if this is what works for the poster, cool. And there's nothing wrong with Xmedia Recode. But it does reflect some misconceptions, so it's not something to learn from. First off, HD audio requires special hard & software to play, while offering little or no advantages if you're not running a home theater setup, complete with a nice, somewhat expensive audio system. And providing you don't have neighbors within earshot BTW. DTS HD is also huge, taking up over 1 GB for the average movie. That's not to say AC3 5.1 is tiny at Blu-Ray bit rates, at 1/2 a GB or better -- the AC3 5.1 on a DVD is about 200 MB smaller. AAC OTOH is the successor to mp3, so sounds better, & is relatively tiny. It can also contain 5.1 audio, though not everything can handle it, especially 5.1 AAC. The advantage for a cell/tablet is they're almost certain to work with AAC -- AC3, not so much. The only uncompressed audio that most every player will handle is .wav, which is really huge.
Now about .mkv... It's a container format, in that respect no different than .avi, .mov, .mp4 & so on. It's designed to be flexible, holding all sorts of content, but it takes a special player to handle anything beyond 1 video + 1 audio track. Mkv files are popular among enthusiasts & folks who put video on-line or download it, but when it comes to players [software or hardware] & video apps, .mkv support is iffy.
How much video can you fit where?... *IF* you use a high quality encoder set for high or near highest quality [& therefor SLOW], you can fit most Blu-Ray movies on a DVD & have them look extremely close to the original, though only one movie per DVD, & you'll want to use dual layer discs. Remember that the max DVD bit rate is ~9, so you might want to downsize the video to 1440 x 1080, or if you don't mind the quality hit, 720p. If you reduce the frame size to 720 pixels wide, with letterboxing cropped off so height varies, a high quality AVC video will take up around 1.5 - 2 GB, with audio stored as AAC stereo. You don't want to go smaller than that 720 width unless quality is unimportant -- you don't want to go down to that 720 width unless you'll be watching the video on a 7" tablet [or smaller] or on a standard, CRT TV.
* * *
" it did hammer all my CPU’s four cores even though I have a graphics card and at medium quality it estimated it would take around 50 mins to convert."
Using all the CPU cores is what you want -- in fact the closer you can get to 100% CPU use the better. Encoding video the same amount of calculations or work is going to be done whether you use 10% or 50% or 100%, so you want the higher usage to go ahead & get it over with. That said, you'll very rarely hit 100% unless you've got a single core CPU -- that was one of the problems when they 1st came out with dual core processors, much lower CPU use, & the percentage number became a measure of how good the software was, the higher the better.
On GPU assist...
Very briefly there are 3 types of assist possible -- DXVA [Direct X Video Acceleration], proprietary methods of accessing the GPU [Quick Sync, CUDA, & Avivo/Stream], & OpenCL. The max amount of assist in all 3 categories depends on the hardware -- with add-on graphics cards the more expensive the card the more powerful it is, while with built-in graphics it depends on both the ranking of the CPU & how new it is, as Intel & AMD have both been increasing capabilities. Working with video most software will use DXVA, normally at the least for reading, decoding, & playback, & while you may or may not see your video playing, all 3 steps are used to get the data that will be re-encoded. OpenCL like the name says is an open standard, a means of accessing the GPU's number crunching capabilities. AMD, Intel, & Nvidia all have OpenCL support, though Nvidia backed off OpenCL a bit in favor of CUDA.
Think of Nvidia's CUDA, Intel's Quick Sync, & AMD's Avivo or Stream as coding shortcuts -- in that respect it's a bit like Direct X -- where software might send a relative few calls to the GPU, getting more-or-less the same results as if they wrote much more code targeting the CPU. One catch is it's proprietary rather than open, e.g. CUDA only works on Nvidia. Another is that like DX, what you get is not always what you want, since you can't do everything that you can with regular programming, and so may have to write a bunch of code to transform the results you do get. Using this proprietary tech usually means slightly lower quality -- that applies to DX & OpenCL too, but not as much. How much quality is lost is hard to say, since 1) there have been very few comparisons published, & 2) software that concentrates on something like CUDA very often focuses on speed more than quality anyway.
Nvidia's CUDA is the most widely used, but then Nvidia did a Lot of developer outreach & support. Intel's Quick Sync is probably the fastest of the 3, but they've hardly supported it, so you'll find much less software using it. Another problem using Quick Sync has been that Intel's built-in GPU was rather anemic a few CPU generations back, & even with the latest, greatest from Intel, a lot of people still run add-on cards from AMD or Nvidia -- when you're running a graphics card you can't access the Intel GPU the same way as when it's powering your display. This is changing, 1st with a technology called Virtu, later with win8.1. Using either you *may* be able to use an Intel GPU's Quick Sync alongside Nvidia or AMD cards -- ffdshow & the LAVFilters may be your best shot, & then probably for the video decoding phase, which is after all a pretty important but often user-neglected aspect of transcoding video. That said if you're interested you'll need to do a bit of research & jump through a hoop or two setting things up to access Intel's GPU.
Now, the Achilles heel of GPU assist is that it can't do everything, plus you only want to use those functions that provide some benefit, so you still have to do a lot, often most video processing the old fashioned way, using your CPU. Getting the GPU & CPU working in tandem is hard, while getting them to work together in some sort of sync is much harder, especially when CPU & GPU horsepower vary so widely. Very often you'll see the CPU sitting, waiting to receive data back from the GPU -- often you'll see the reverse, with the CPU somewhat rarely sending data the GPU's way. When you're re-encoding or converting video it's pretty rare to see both high CPU & GPU usage -- usually the CPU % will drop, & you might only see a small GPU % monitoring with something like MSI's Afterburner or GPU-Z. [One way Afterburner is handy is that it lets you use a fan speed curve for your graphics card. A graphics card's built-in, more-or-less hard-wired curve often focuses more on keeping fan noise as low as possible, which can mean a lot of extra heat released inside the case, raising CPU & chipset temps. Some graphics cards can almost double as space heaters.]
At any rate, check total (re)encode times & quality with GPU assist on & off. You might or might not see CPU % drop, you might or might not see GPU % rise, but quality & time to (re)encode are what matters. Something like Process Lasso might well come in handy. Watching, checking the processes for your security software can be a good thing -- I always put Bitdefender in Game mode & it helps. If it's been a while since you fired up or re-booted your PC, try a re-boot to clear the graphics processes & memory, especially if you've been running a game or playing video or anything like that that uses your graphics hardware. If you have video software that you regularly use & that gives you decent GPU assist, be wary of driver updates -- back up 1st in case driver changes effect that... I've seen both increases & decreases in GPU assist after updating AMD graphics drivers. Get & save a short video clip in the format & frame size you usually convert, then use that for all your testing -- that way you have a clear baseline so you compare Apples to Apples.
* * *
"I tested this and another one of the many video converters, which showed up on GAOAT.
- A 1.2 GB .wmv video (res 1280×720) to a 619 MB .mp4 (res 720×480) file in min 11:54. The Aiseesoft Total Video Converter Platinum (GOAT Sep. 2013) finished the same task in 7:35."
If/when your output file is AVC/H.264, you can check the results with the free MediaInfo, & that *may* show some of the x264 command line arguments that were used. Often a ffmpeg-based converter that's a lot faster uses faster encoder settings, though that may or may not show up as a difference in the MediaInfo window. The x264 encoder built into ffmpeg has loads & loads of possible settings -- one of those is speed vs quality, which is an either/or proposition, where as quality increases speed goes down. As long as you're happy with the results at whatever speed setting that's cool. What you need to look for IMHO are all the other settings as well as the specs MediaInfo shows for the file. If you've been using converter "X" & the results play fine, look good where you want/need to play them, & then along comes converter "Y" which is a lot faster, if the stuff MediaInfo shows you is different, that file out of converter "Y" may not play, or play as well as what you've been getting out of converter "X".
If your output is Xvid there are different encoder settings too, but since Xvid uses so much less pre & post processing than AVC, they're not nearly as critical for playback. You can dig up some comparison tools on-line to put some sort of numbers to before & after accuracy, but for general, at home use, personally I think quality is very subjective & you being happy with the results is what matters.
* * *
"I need a video converter which can deal with subtitles in foreign languages, especially with the accent characters. This one can’t do it, so it is not good for me."
Subs are either graphics or text based -- graphics-based subs are used on DVD & Blu-Ray movie discs -- text-based subs are used during creation & editing, & as long as the file name [not the extension] is the same as the video file, they can be played/used by a lot of players when they're in the same folder next to the video file. Most ffmpeg-based converters only *burn* subs permanently on the video, whereas embedded or separate subs can be switched on/off. Text captioning can be embedded in mpg2 files, graphics based subs can be embedded in .mp4 or .mkv, though the number of players for either are limited. Graphics based subs can also be included in .m2ts [used for Blu-Ray] & of course the VOB files used on DVDs. When the video might be re-sized by the player, e.g. used on your PC, tablet, & cell, graphics based subs don't look as good as text based subs that the player renders. Use BDSup2Sub to re-size graphics based subs before embedding them in a video file or muxing into an mkv container etc. Use SubtitleEdit to OCR graphics based subs to text.
Now, special characters... Graphics based subs can be anything -- they don't even have to be text. Text based subs can use &/or contain special characters as long as the text file is unicode -- saving/creating a unicode text file is a separate step in something like Notepad. Software for dealing with subs &/or captioning *may* include restrictions on what formatting is legal, & since text-based subs are simply text files with special formatting provided by certain characters [it's similar in that respect to HTML], you can run into issues converting one format of text sub to another, &/or if you want to display a character normally used for formatting purposes. Since loads of video software comes out of Europe stuff like accented characters aren't often a big problem, though right -> left or top -> down reading may take a bit of hunting & trial & error to find software that works.
* * *
To Fran...
" ... Copy of old family video I obtained over Christmas – consisting of 4 equal in size (0.99GB) Files. All Files are in VOB Format, so before burning them to dvd disks to hand out further – I decided to convert these Files to more familiar MP4 formats to make it easier to be watched on more popular PC Players."
IMHO you might be better off leaving the video you've got alone.
OK, to start with the video's in mpg2, & if it's old video, unless it was film professionally scanned & encoded, it likely originated at best as Super VHS or Hi-8, which by today's standards are poor quality -- if the original was regular VHS, which was far more common, the quality levels are much lower. You can get a quick idea by comparing the quality to a current retail DVD, but look at it frame by frame rather than playing. The quality matters in 2 ways -- 1 is just a matter of personal preference, if you want it to look as nice as possible, & 2 is whether or not the effects of re-encoding, regardless the software, will reduce the quality to where no one wants to watch it. Encoding video is a lossy process -- you'll *Always* lose a bit of quality, & it's called generational loss. *IF* it is wanted/needed, you *might* [& I really want to stress *Might*] improve the quality a bit running the video through filters available for VirtualDub &/or Avisynth -- you can post a few frame shots or grabs as stills in the appropriate forums at videohelp.com &/or doom9.org for advice on the filters & settings. You'll also find loads of info on how to use VirtualDub &/or Avisynth with mpg2 video. Bear in mind that the player you use in Windows as well as your graphics hardware settings can alter the video's appearance, so as possible turn that sort of thing off if/when you're analyzing video.
That said, the only advantage to re-encoding the video would be to use AVC/H.264, which would both reduce file sizes & make it more compatible with current cell phones & tablets etc. PCs/laptops natively handle mpg2 much better than AVC. I can't think of much of any overall advantage to re-encoding the video to Xvid mp4. If it's in DVD format already, with .IFO & .VOB files in a VIDEO_TS folder most anyone can play it in Windows, & Windows Media Center actually does quite a nice job with DVDs on your hard drive. I Use Nero Express but ImgBurn can do quite nicely to burn that VIDEO_TS folder [if that's what you have] to a video DVD. If you want to extract the files to a single mpg2 file, that's fairly easy -- there are several apps/tools available. Note: mpeg2 video can exist in a .mpg container, that contains timing info & *usually* audio, or as a .m2v file without audio or the same timing data. It's pretty easy to mux mpeg2 video into .mpg files, or demux them into .m2v -- you do that to work with the audio separately, & some programs require .m2v or .mpg.
Finally you said that one player gave you problems seeking [fwd - rev]... If the nature of the video is something that you want to be able to repeat certain scenes & such easily, you might consider re-using the video as-is but redoing the DVD layout to include more chapters where you want them. That way the player could very easily skip to the exact points you want, & repeat those scenes as you want. Personally I'd use DVD Shrink to re-write the DVD layout to include just one, larger VOB file to make things easier, or rewrite it all as a mpg2 file, then likely import them into Sony DVDA or similar where you could set your chapters wherever, & add a menu or 3 if you like. I've seen versions of DVDA as cheap as $0 after MIR, or Sony offers trials -- it's been years since I tried one of their trials, but they used to be fully functional for 30 days, which would allow you to complete your project.
"my File was ready to be played – checked the quality of it, initially happy that it shrunk from 0.99GB to 781MB – but NO, the quality was nowhere near to original VOB Format"
Things that effect video quality & file size -- the efficiency of the format used, the bit rate set in the encoder, the quality of the encoder itself, the encoder quality settings you make, and the quality & format of original source that you're (re)encoding. AVC is an efficient format -- a DVD movie will take up roughly 1.5 or 2 GB encoded to AVC with a variable bit rate set to about 1.5, and look as good as the original on my Galaxy Tab 2 7. At the DVD frame width of 720 pixels, using square pixels in Nero Recode 10, with height varying after cutting off any letterboxing, I usually like to stick with a bit rate around 1.5 or so. [I actually convert video to create DVDs so I can use Recode, embedding DVD subs in the video file]. You can research [Google] the x264 AVC encoder that's included in the ffmpeg WinX & similar converters use -- it's a very good encoder, but you can't make many of the settings in these converters so you've little control over things that effect quality. A front end GUI for x264 like Simple x264 Launcher will accept Avisynth input [if you decide to use some of that filtering I spoke of], & let you set everything about how x264 encodes your video, but the process is not user friendly in that there's a lot to read & learn about those settings. Mpg2 itself is a decent format quality wise for video -- it's just not the most efficient thing around -- so the only remaining variable is the source quality. Judging from your file sizes you might have used Xvid [I don't recommend it] -- from your comment on the result I suspect that the quality of your source might be questionable.
"... choosing only ‘Use High Quality Engine’ (previously combined with ‘Deinterlace’) ... quality as appalling as first result...
... opting only for ‘Deinterlace’ this time – resulted in a File size 777 MB, ‘smoother’ outlines of everything on video, but horizontal lines and ‘jumpiness’ from scene to scene made it worst than everything I achieved so far!!"
Don't use deinterlace. There's not enough room here to explain it properly, so I'll just say don't.
"Briefly considered Giovanni’s suggestion of AVC, but the free Version only converts a little ‘teaser’ of large video File, that’s all."
You should see no big difference in the results because ffmpeg is still powering the process & you've not mentioned any frame resize [that I saw anyway], which can very from app to app depending on the code used. If you wanted to access more of the Xvid, mpg2, or AVC encoding settings there are apps or tools that let you do that, but they're harder because you have to learn what settings to make & how they effect the outcome.
