Dude's Media Compression Q&A

[Dude]

Hey guys.  I'm not sure if there is any interest in this, but I am in the media compression world professionally and am willing to answer any questions you may have.  This can be anything from "What is the difference between H.264 & AVC" to "What is an MP4", or even "How do I get a similar job when I'm out of college".

If you are really interested in this stuff, the doom9 forums are a great place to hang out.  Otherwise, feel free to and ask me anything 

[jwaz]

I'd like to hear your thoughts on flac vs mp3 (and other high quality audio solutions)

[regack]

What's the next big thing on the horizon?

[TheSoulhunter]

Guess I should join... :3
I'm no more that active in the A/V scene, but I was in the past decade.

[TheSoulhunter]

What's the next big thing on the horizon?

Probably h.265...
Some bitrate savings above h.264 especially for high-res content.

[CPTBadAss]

I'd like to hear your thoughts on flac vs mp3 (and other high quality audio solutions)

I'd like to hear this as well. I'd like to know how you'd compare FLAC to 320kbps MP3.

[Moosecraft]

I'd like to hear your thoughts on flac vs mp3 (and other high quality audio solutions)

I'd like to hear this as well. I'd like to know how you'd compare FLAC to 320kbps MP3.

For me there is a small difference, only able to hear it when im only listening to music and not sitting at my pc.

[Malphas]

How long do you reckon until HEVC/H.265 becomes the de-facto standard? It seems we've only finally finished transitioning from Xvid to x264, and x265 seems to be a long way from completion.

[TheSoulhunter]

I'd like to hear your thoughts on flac vs mp3 (and other high quality audio solutions)

Imo MP3 can give transparent results at higher bitrates, but... As audio files are relatively small compared to other current content (ie. HD video) and storage is kinda affordable (100-200 thousand FLACs on a HDD that costs around 150 bucks!) the actual need for lossy compression is IMO questionable. The only exceptions might be mobile devices with limited storage or compatibility issues (as some devices don't support FLAC, but most do support MP3).

[Dude]

I'd like to hear your thoughts on flac vs mp3 (and other high quality audio solutions)

In my world, it's all about "good enough".  I am not an audiophile, but with that said I do have very nice headphones (Sennheiser HD580's) and a good DAC & headphone amp.  I also have a nice home theater system.  Nothing is crazy extravagant, but it's all very good sounding.

In my opinion, a properly made VBR MP3 sounds great (That's what my music collection is).  I know there are many people who will only have FLAC files in their collection, and I have no problems with that.

A 128Kbps MP3 is not good enough by a long shot for a home audio collection.  For your own collection, I would recommend a minimum of 192Kbps for an MP3 or AAC file.

[TheSoulhunter]

How long do you reckon until HEVC/H.265 becomes the de-facto standard? It seems we've only finally finished transitioning from Xvid to x264, and x265 seems to be a long way from completion.

For personal use or generally? We might see the first really useable implementation(s) in a year, add 1-3 more until it's somewhat matured and tweaked...
But broadcasting, real commercial use, hardware support etc.? The industry is just about to finish the transition from MPEG2 -> h.264, and they had to invest a lot money

[Malphas]

I see almost zero need for lossy audio compression these days. For all archival or buying the material in the first place I'd always go FLAC or WAV. Then if you need a MP3 (Vorbis would be better) to store on your phone or something then you can just transcode as and when you transfer them over (which takes a minute amount of time on modern processors).

[TheSoulhunter]

I see almost zero need for lossy audio compression these days. For all archival or buying the material in the first place I'd always go FLAC or WAV. Then if you need a MP3 (Vorbis would be better) to store on your phone or something then you can just transcode as and when you transfer them over (which takes a minute amount of time on modern processors).

Yep, I do the same...
Also because I sometimes edit the audio or use it for video BGM.

[Dude]

What's the next big thing on the horizon?

For the consumer world, Adaptive Bitrate Encoding (ABR) and H.265 (HEVC).  There are many professional initiatives going on in the background for archiving master copies of films, but you will likely never hear or care about those

ABR has a few flavors right now.  HTTP Live Streaming, Smooth Streaming, HTTP Dynamic Streaming, and MPEG DASH.  HLS, Smooth, and HDS were the first real flavors of ABR encoding.  The problem is that no one has agreed on which one to use.  MPEG DASH is an attempt to create a standard that everyone will agree on.

H.265 is essentially the next version of H.264 encoding.  It's very impressive, however it will take a few years until it becomes mainstream.

[Malphas]

How long do you reckon until HEVC/H.265 becomes the de-facto standard? It seems we've only finally finished transitioning from Xvid to x264, and x265 seems to be a long way from completion.

For personal use or generally? We might see the first really useable implementation(s) in a year, add 1-3 more until it's somewhat matured and tweaked...
But broadcasting, real commercial use, hardware support etc.? The industry is just about to finish the transition from MPEG2 -> h.264, and they had to invest a lot money

Personal use mainly, as in when we can expect release groups, etc. to be using x265. Although I'm also interested in when hardware support will become available (televisions and HTPC devices). What'll probably happen first though is I'll end up buying everything all over again on Blu-ray (if that even still exists in 2-3 years) in 4k and ripping it myself to x265 before it becomes widespread.

[Dude]

What's the next big thing on the horizon?

Probably h.265...
Some bitrate savings above h.264 especially for high-res content.

Yeah, H.265 is a significant bitrate savings over H.264.  I think what they're saying is that for the same bitrate that you typically use for a 1080p encode with H.264, you will be able to spit out the same quality of a 4k stream.  The broadcasters are very interested in this, because they are constantly trying to find ways to lower the bandwidth of their streams while keeping the same quality.  The prospect of using the same bandwidth for their current H.264 1080p streams to create h.265 4k streams is something they really want.

Since HEVC (H.265) was only ratified earlier this year, it will be a few more years until it's really mainstream.  Your computer likely can't handle decoding a h.265 stream right now.  It's extremely difficult, and there aren't many hardware decoding chips.  Probably 2017 you'll really start seeing h.265 being mainstream.

[Dude]

I'd like to hear your thoughts on flac vs mp3 (and other high quality audio solutions)

I'd like to hear this as well. I'd like to know how you'd compare FLAC to 320kbps MP3.

I sort of answered this in another reply, but this is a different way of asking a similar question.

Of course FLAC will be higher quality than 320Kbps MP3.  I honestly doubt you'll ever be able to hear the difference though    The limiting factor to hearing the difference is typically the playback environment.  You will need extremely high quality DAC, amp, and speakers to really get the benefit of FLAC vs MP3.

If you're compressing your own collection though, with storage being so cheap, if you want to go FLAC by all means do it.  There's nothing wrong with lossless compression

[Dude]

How long do you reckon until HEVC/H.265 becomes the de-facto standard? It seems we've only finally finished transitioning from Xvid to x264, and x265 seems to be a long way from completion.

Yep, you're 100% correct.  The compression world is very slow to adapt new standards due to many factors, and it will likely be ~4 years until it's mainstream.

[Dude]

I see almost zero need for lossy audio compression these days. For all archival or buying the material in the first place I'd always go FLAC or WAV. Then if you need a MP3 (Vorbis would be better) to store on your phone or something then you can just transcode as and when you transfer them over (which takes a minute amount of time on modern processors).

Yeah, I am actually a huge advocate of lossless compression in certain cases.  With my personal music collection, VBR MP3 is totally fine for me.  I have no argument against your approach though, and I think it's a very smart one.  The great part about going lossless is that you will always have an equivalent copy to the original CD, and can essentially throw away the CD once you've created your FLAC files since you will never have to re-rip it to a higher quality in the future.

[Dude]

Personal use mainly, as in when we can expect release groups, etc. to be using x265. Although I'm also interested in when hardware support will become available (televisions and HTPC devices). What'll probably happen first though is I'll end up buying everything all over again on Blu-ray (if that even still exists in 2-3 years) in 4k and ripping it myself to x265 before it becomes widespread.

This guy.  He's a smart one.

[Malphas]

I'm just a media hoarder. I wish movie/television studios would join the 21st century and sell DRM-free, 4k content already.

[TheSoulhunter]

How long do you reckon until HEVC/H.265 becomes the de-facto standard? It seems we've only finally finished transitioning from Xvid to x264, and x265 seems to be a long way from completion.

For personal use or generally? We might see the first really useable implementation(s) in a year, add 1-3 more until it's somewhat matured and tweaked...
But broadcasting, real commercial use, hardware support etc.? The industry is just about to finish the transition from MPEG2 -> h.264, and they had to invest a lot money

Personal use mainly, as in when we can expect release groups, etc. to be using x265. Although I'm also interested in when hardware support will become available (televisions and HTPC devices). What'll probably happen first though is I'll end up buying everything all over again on Blu-ray (if that even still exists in 2-3 years) in 4k and ripping it myself to x265 before it becomes widespread.

Releases perhaps in a year or two, mainly depends on how much free time the open source devs pump into the current projects...
For widespread use, we probably look more at something like 10 years, hard to predict, probably also depends on how well 4k and 8k will be accepted and demanded.

[TheSoulhunter]

What's the next big thing on the horizon?

Probably h.265...
Some bitrate savings above h.264 especially for high-res content.

Yeah, H.265 is a significant bitrate savings over H.264.  I think what they're saying is that for the same bitrate that you typically use for a 1080p encode with H.264, you will be able to spit out the same quality of a 4k stream.  The broadcasters are very interested in this, because they are constantly trying to find ways to lower the bandwidth of their streams while keeping the same quality.  The prospect of using the same bandwidth for their current H.264 1080p streams to create h.265 4k streams is something they really want.

Since HEVC (H.265) was only ratified earlier this year, it will be a few more years until it's really mainstream.  Your computer likely can't handle decoding a h.265 stream right now.  It's extremely difficult, and there aren't many hardware decoding chips.  Probably 2017 you'll really start seeing h.265 being mainstream.

First playarounds with the reference ware shows impressive protential, but realistically we are looking at a datarate reduction of perhaps 25-50% imo, not 75%...
Also I feel we are hitting a brickwall here, same as with computing power. For computation the solution could be cloud computing and/or going larger with CPUs (more surface, more cores).
But for video compression there is no real solution... Squeezing it another 50% after h.265 seems almost impossible to me, except if content becomes much more complex (multilayer video? I dunno).

[regack]

This thread is nice and informative, thanks...

[Dude]

First playarounds with the reference ware show impressive protential, but realistically we are looking at a datarate reduction of perhaps 25-50% imo, not 75%...
Also I feel we are hitting a brickwall here, same as with computing power. For computation the solution could be cloud computing and/or going larger with CPUs (more surface, more cores).
But for video compression there is no real solution... Squeezing it another 50% after h.265 seem almost impossible to me, except if content becomes much more complex (multilayer video? I dunno).

You may be correct.  The information I'm providing out about HEVC is not my personal experience with it, only what I've heard anecdotally.  HEVC is still so new that the encoders aren't even close to being matured, so it's difficult to do great comparisons between H.264 & H.265.

Another thing to keep in mind is a dirty little secret about 1080p encoding for playback on your TV.  Often when you view a 1080p stream, it's not true 1920x1080.  The width is sometimes a lower resolution (1440x1080 perhaps) so that they can still say "This is 1080p!" when in fact that's only the vertical resolution.  This means that the difference between 1080p & 4k encode may not be a true 4x overall difference.

This type of compression is one of the oldest compression methods in the book.  Back in the analog NTSC & PAL days, there was regular 4x3 playback, and anamorphic playback.  NTSC, for example, has an actual resolution of 720x480 regardless of if it's anamorphic or not.  Upon playback, if the anamorphic flag is set to "true" like most DVD's are, it will playback at 854x480 (This is simplified, but you get the point).  Also, SD MPEG-2 material is often not encoded at 720x480.  It's usually encoded at 528x480 or 352x480 and resized on playback.

[TheSoulhunter]

Oh yes, good old anamorphic encoding...
Also, to make things even more confusing, alot 1080p content is actually 1920x1088 (to keep it mod16 to fit 16x16 macro blocks).
Well, I'm happy we got rid of the framerate and interlacing complications for the most part, horrible horrible memories still haunting me! XD

[Dude]

Oh yes, good old anamorphic encoding...
Also, to make things even more confusing, alot 1080p content is actually 1920x1088 (to keep it mod16 to fit 16x16 macro blocks).
Well, I'm happy we got rid of the framerate and interlacing complications for the most part, horrible horrible memories still haunting me! XD

Yeah, I'm pretty sure you're correct about that.  Even if you encode at 1920x1080, it will actually throw in those extra lines to make it 1088, but just mask then upon playback.  I know I ran into that a while back and it confused the hell out of me, but the root cause was the way that specific encoder was working.

Unfortunately in my world, interlacing and various framerates have not disappeared!  I've become way too versed out of necessity into manipulating older sources and trying to make them look decent.

[TheSoulhunter]

I see. Do people still use AviSynth sorcery? >.>

[kenmai9]

What major did you major in?

[Dude]

I see. Do people still use AviSynth sorcery? >.>

Yes.  You'd be surprised with how much AviSynth is used in the professional world.

We are trying to migrate off of it though due to the limitations around it.  It's 8-bit only, and really only works well with AVI files (Yes I know you can use other files in there, but they all have various issues).  All of the different professional encoding vendors are coming up with solutions that are removing the remaining need for AviSynth one by one.

[Dude]

What major did you major in?

Recording Engineering.  It was actually an easy transition from digital audio to digital audio/video.

There is no straight path to my job since I'm pretty sure there still aren't college degrees in what I do.  Some of my colleagues don't have college degrees, lots have backgrounds in film, lots have audio backgrounds like myself, some were IT... most of us found this career path by accident and just stayed.

[kenmai9]

What major did you major in?

Recording Engineering.  It was actually an easy transition from digital audio to digital audio/video.

There is no straight path to my job since I'm pretty sure there still aren't college degrees in what I do.  Some of my colleagues don't have college degrees, lots have backgrounds in film, lots have audio backgrounds like myself, some were IT... most of us found this career path by accident and just stayed.

You don't have to answer this, but do you make more than developers? I'm just curious as to what other things I can do in the world besides developing. This media compression sounds really interesting. There's so much out there that I don't know about because I'm still in college.

[Dude]

You don't have to answer this, but do you make more than developers? I'm just curious as to what other things I can do in the world besides developing. This media compression sounds really interesting. There's so much out there that I don't know about because I'm still in college.

I honestly don't know what developers make.  I know that for an entry level position in my field (An operator), they typically make ~$15 an hour.  You don't have to "pay your dues" that long though, and especially in the Los Angeles area (This is the hub for this type of work since there are so many movie studios here), there are so many available jobs (If you are good) that it's easy to bounce around a few times between companies to be at $25-30 after not too long.  Even if you are at a lower pay though, there is often ample amounts of OT waiting for you take advantage of.

If you are motivated and good at your job, it's really rather easy to stand out amongst the other people doing your job and potentially move from an operator position to a lead operator, or even an engineer.

Salaries actually vary pretty drastically, but I think around $80-100k is pretty normal for the engineering guys.  It's not uncommon for them to be higher than that though, since the engineers are in pretty high demand these days 

[Malphas]

Whatever happened to AviSynth 3? Abandoned?

[Dude]

Whatever happened to AviSynth 3? Abandoned?

I honestly don't know.  I think that there is still active work being done, but myself and my colleagues pretty much live in the 2.5.8 world.  I was really active in that world as well for some time, but I haven't kept current for the past 4 or 5 years.

As I mentioned before, many of the encoding vendors are baking the functions that were previously only handled by AviSynth natively into their software.  This means that although it's still used, the reasons it's used are becoming less and less.  This means that I have my bag of tricks that I use it for, and I haven't really had to add to that bag for quite some time.

[TacticalCoder]

What is the OS typically used by digital set-top boxes that people use to watch TV nowadays ?

[Dude]

What is the OS typically used by digital set-top boxes that people use to watch TV nowadays ?

I only use the set-top boxes myself, I've never actually programmed them from scratch.  I'm pretty sure they're almost all just embedded linux systems of some sort though.

[kenmai9]

You don't have to answer this, but do you make more than developers? I'm just curious as to what other things I can do in the world besides developing. This media compression sounds really interesting. There's so much out there that I don't know about because I'm still in college.

I honestly don't know what developers make.  I know that for an entry level position in my field (An operator), they typically make ~$15 an hour.  You don't have to "pay your dues" that long though, and especially in the Los Angeles area (This is the hub for this type of work since there are so many movie studios here), there are so many available jobs (If you are good) that it's easy to bounce around a few times between companies to be at $25-30 after not too long.  Even if you are at a lower pay though, there is often ample amounts of OT waiting for you take advantage of.

If you are motivated and good at your job, it's really rather easy to stand out amongst the other people doing your job and potentially move from an operator position to a lead operator, or even an engineer.

Salaries actually vary pretty drastically, but I think around $80-100k is pretty normal for the engineering guys.  It's not uncommon for them to be higher than that though, since the engineers are in pretty high demand these days 

Awesome thanks for the answer. So what do you actually do? I still don't know from reading this thread haha. I guess you work on getting video and audio to compress better without losing clarity?

My friend is interested in film and is looking for jobs soon.

[HoffmanMyster]

Very cool thread.  I don't know a whole lot about this, so it's neat to read.  And I don't really have a question so I'll just sit here and read the conversation and soak it all up. 

[takadante]

I'm currently a Senior in college and about to graduate. I've been into film production for a while now and was wondering how I would go about getting a job in the media compression world professionally such as yourself?!??!?!? (in LA)

[Dude]

Awesome thanks for the answer. So what do you actually do? I still don't know from reading this thread haha. I guess you work on getting video and audio to compress better without losing clarity?

My friend is interested in film and is looking for jobs soon.

Yeah, I've left my actual position a little ambiguous on purpose.  I will say that I'm fairly high on the technical totem pole

I can describe what people do in my field a little more, just to hopefully help you understand a little better of this world.  Basically we take care of everything between a master tape (Yes, films and tv shows are stored on tape - usually HDCAM SR) and the final compressed deliverable.  There are so many different workflows that we take care of.  Below is a very partial list.

1. The master tape needs to be converted to an equivalent file for downstream deliverables.  This file is used for archival purposes.
2. Rather than ship a tape of tonight's episode of Almost Human, why not securely deliver a master quality file.  Sometimes these files also are edited, subtitled (Or captioned), dubbed in a different language, etc...  The advantage of file delivery is that it's way cheap compared to delivering a tape, and way faster.
3. When you watch a movie on your favorite paid movie/tv subscription service, someone has to deliver them an intermediate file so they can flip it to their favorite flavor.
4. When you watch VOD content from your cable provider (Or on an airplane flight, hotel, etc...), someone has to make those files.

There are many people involved in these workflows.  There are file delivery / metadata people, operators making all of the various files, engineers developing the workflows and validating the in-house tools, storage experts (Digital Media storage is a beast), managers for the different teams, a QA team ensuring the files are correct before hitting your eyes, sales people, client reps, etc...

There are also different types of facilities with their own unique angles at these workflows.  There are the studios, the post production houses who take the overflow work from all the studios, the broadcasters, the online streaming companies, and other places.

If your friend wants to get in the film industry, I wouldn't say my field is the best one to get into.  While I technically work in tv/film, this career path will almost certainly not lead you to a film set or writing room.  There are some simple editing jobs in my field, but that's about as close as you'll come to a "real" film job.

[Dude]

I'm currently a Senior in college and about to graduate. I've been into film production for a while now and was wondering how I would go about getting a job in the media compression world professionally such as yourself?!??!?!? (in LA)

Digital Media operators positions aren't terribly difficult to find.  Monster.com, Dice.com, and Craigslist are good places to start.  It's also a good idea to find the companies you are interested in working for and checking their careers pages directly.  Luckily prior experience usually isn't required for an operator position, and it's a great way to get your foot in the door.

If you want some direction on which companies to look at, just shoot me a PM.

[Dude]

Very cool thread.  I don't know a whole lot about this, so it's neat to read.  And I don't really have a question so I'll just sit here and read the conversation and soak it all up. 

Thanks!  I'm actually surprised this thread is so popular.  I guess it's one of those topics that most people don't think that much about, but it's actually very in depth and interesting (At least I think it's interesting).

[TheSoulhunter]

I see. Do people still use AviSynth sorcery? >.>

Yes.  You'd be surprised with how much AviSynth is used in the professional world.

We are trying to migrate off of it though due to the limitations around it.  It's 8-bit only, and really only works well with AVI files (Yes I know you can use other files in there, but they all have various issues).  All of the different professional encoding vendors are coming up with solutions that are removing the remaining need for AviSynth one by one.

Heh! I used it for long time and I was also involved in script and plugin development. I still hang out with some of other "old timers" (names like MarcFD or mf might ring a bell?) on IRC. I assumed Avisynth would support higher bit-depths by now, if not for actual >8bit input then at least for allowing processing chains with a higher "precision". It's a shame, because from what I have seen in current video editing software (including a lot 3rd party plugins) most of the old stuff we developed back then is still not outperformed and in some cases not even reached yet (NR for example).

[Dude]

Heh! I used it for long time and I was also involved in script and plugin development. I still hang out with some of other "old timers" (names like MarcFD or mf might ring a bell?) on IRC. I assumed Avisynth would support higher bit-depths by now, if not for actual >8bit input then at least for allowing processing chains with a higher "precision". It's a shame, because from what I have seen in current video editing software (including a lot 3rd party plugins) most of the old stuff we developed back then is still not outperformed and in some cases not even reached yet (NR for example).

Very nice!  Yeah, there are a number of examples of AviSynth functionality that still surpasses the same functionality in the professional solutions.  That's why many workflows still revolve around that tool.  It's also nice how easy it is to bake extremely complicated video processing instructions into an automated workflow with AviSynth since it's essentially just text files.

We are starting to see some extremely impressive GPU image processing come out now though that surpasses anything I've seen in AviSynth for framerate/standards conversions (29.97 interlaced -> 25 interlaced, for example).

[HoffmanMyster]

What major did you major in?

Recording Engineering.  It was actually an easy transition from digital audio to digital audio/video.

There is no straight path to my job since I'm pretty sure there still aren't college degrees in what I do.  Some of my colleagues don't have college degrees, lots have backgrounds in film, lots have audio backgrounds like myself, some were IT... most of us found this career path by accident and just stayed.

Not that you would know specifically at my university, but are there common names for an intro to recording engineering class?  I've tried searching the class list for Recording Engineering and Audio Engineering with no success, but I'm hoping that I might be searching the wrong keywords.  I came very close to studying that sort of thing, and it's still a hobby of mine, so I was hoping to throw in some sort of intro class in my final semester coming up.

[Dude]

Not that you would know specifically at my university, but are there common names for an intro to recording engineering class?  I've tried searching the class list for Recording Engineering and Audio Engineering with no success, but I'm hoping that I might be searching the wrong keywords.  I came very close to studying that sort of thing, and it's still a hobby of mine, so I was hoping to throw in some sort of intro class in my final semester coming up.

When I went to college, only a few universities offered a Bachelors in recording.  There are a number of well recognized associates and certificate programs around, but again not all schools offer them.

I'm probably the wrong person to ask about this since I went a different route and I'm a little biased, but in my experience the recording industry is a very difficult one to survive in.  There so many "dues" that you have to pay, and after that it's still hard to get a decent job (I have seen so many audio mixers get laid off or pushed into different positions...).  I suppose it's like any industry though, the higher you get the less positions there are to go around.

Even with that said though, I'd totally recommend taking some classes if you can.  I loved all my audio classes, and have some incredible memories from them.  It also gives me a really nice edge over some of my colleagues, since I have such a strong background in audio.  Then again, many of them have much stronger computer programming skills which I wish I had.

[HoffmanMyster]

Not that you would know specifically at my university, but are there common names for an intro to recording engineering class?  I've tried searching the class list for Recording Engineering and Audio Engineering with no success, but I'm hoping that I might be searching the wrong keywords.  I came very close to studying that sort of thing, and it's still a hobby of mine, so I was hoping to throw in some sort of intro class in my final semester coming up.

When I went to college, only a few universities offered a Bachelors in recording.  There are a number of well recognized associates and certificate programs around, but again not all schools offer them.

I'm probably the wrong person to ask about this since I went a different route and I'm a little biased, but in my experience the recording industry is a very difficult one to survive in.  There so many "dues" that you have to pay, and after that it's still hard to get a decent job (I have seen so many audio mixers get laid off or pushed into different positions...).  I suppose it's like any industry though, the higher you get the less positions there are to go around.

Even with that said though, I'd totally recommend taking some classes if you can.  I loved all my audio classes, and have some incredible memories from them.  It also gives me a really nice edge over some of my colleagues, since I have such a strong background in audio.  Then again, many of them have much stronger computer programming skills which I wish I had.

I'm not interested in it as a career, just a hobby at this point.    And I'm already mixing on some pretty great equipment at this point (not my own, obviously) and I love it, I was just hoping to get some more experience in an educational setting.  So I'll keep digging into it and see what I can find.

[Dude]

I'm not interested in it as a career, just a hobby at this point.    And I'm already mixing on some pretty great equipment at this point (not my own, obviously) and I love it, I was just hoping to get some more experience in an educational setting.  So I'll keep digging into it and see what I can find.

Awesome, that's great to hear.  If there are no programs at your school, you should talk to any local recording studios and see if you could intern there for a little while.  There's nothing like real-world experience for a crash course in something new.

[TheSoulhunter]

Heh! I used it for long time and I was also involved in script and plugin development. I still hang out with some of other "old timers" (names like MarcFD or mf might ring a bell?) on IRC. I assumed Avisynth would support higher bit-depths by now, if not for actual >8bit input then at least for allowing processing chains with a higher "precision". It's a shame, because from what I have seen in current video editing software (including a lot 3rd party plugins) most of the old stuff we developed back then is still not outperformed and in some cases not even reached yet (NR for example).

Very nice!  Yeah, there are a number of examples of AviSynth functionality that still surpasses the same functionality in the professional solutions.  That's why many workflows still revolve around that tool.  It's also nice how easy it is to bake extremely complicated video processing instructions into an automated workflow with AviSynth since it's essentially just text files.

We are starting to see some extremely impressive GPU image processing come out now though that surpasses anything I've seen in AviSynth for framerate/standards conversions (29.97 interlaced -> 25 interlaced, for example).

GPU processing has big potential indeed, at least for parallel processing (sequential processing is another story).
I'd really like to see (or hear) some rough figures for such processing on current gen GPUs, because the last time I coped with it was like... in the GeForce 6 era XD