Video Transcoding

Practical Mobile Video Workflow

2009-10-07T07:43:00.000-07:00

Here is an interesting article about delivering video over 3G networks. It outlines a very common work flow.
http://mobiledevdesign.com/tutorials/ensure-video-quality-your-mobile-delivery-100209/

Open Video Conference

2009-06-18T08:52:00.000-07:00

It’s going to start tomorrow June 19th in NYC. If you can’t make it in person, you can view it on your computer as it is being streamed live.

What is open video? (Excerpt from the website)

Open Video is a movement to promote free expression and innovation in online video. Join us for three days of inspiring talks, awesome video and film, open hacking sessions, parties, and cutting edge tech from around the world.

Check it out.

http://openvideoconference.org/2009/06/follow-the-open-video-conference-from-home/

The order of processing affects the result.

2009-05-19T07:38:00.000-07:00

When pre-processing videos for encoding, the order of the filters does have an effect on image quality and the processing time.
I always try to follow the order below for best results.

1) De-interlace if needed.
2) Frame rate conversion if smaller than original.
3) Cropping if needed.
4) Resize if reducing image size.
5) Color space conversion.
6) De-noise.
7) Resize if enlarging image.
8) Frame rate conversion if greater than original.
9) Sharpen or other enhancements as needed.

The idea is that first you start with a progressive image to then reduce the size and number of frames to work with smaller data sets. Perform color space conversions and apply any noise reduction if needed. If you are enlarging the image, do it after the color space conversion and noise reduction. Then apply frame rate conversions.
Finally perform any sharpening or other image enhancing filters to the final image.

Preprocessing

2009-05-14T06:15:00.001-07:00

There are great benefits from processing videos prior to encoding. The right preprocessing can gain you time, bandwidth and increase the image quality of your final video.
There are six important processes that you should consider applying to your source video prior to encoding:

1) De-interlacing. Will avoid creating line crawling artifacts.

2) Noise reduction. Motion adaptive noise reduction can greatly improve the performance of your encoder as random noise in images produce artifacts in the encoded image.

3) Color Space Conversion. Use of high quality color space conversion algorithms can reduce color bands and “blotches” as bit reduction and color plane decimation (going from RGB or 4:4:4 to 4:2:0 for example.)

4) Resampling and resizing. Will greatly reduce encoding time and improve quality. Specially true if you are down sampling or reducing sizes.

5) Cropping. Same as above.

6) Frame Rate Conversion. Will allow you to use high quality frame rate conversion algorithms. If you are reducing the frame rate, you’ll have the added benefit of faster encodes.

Video Encoding Parameter Calculator

2009-04-28T11:00:00.000-07:00

I wrote a free utility iPhone application to help video encoding professionals and consumers calculate bitrates and file sizes as well as aspect ratios and dimensions. It is called the VidCalc.

You can get it at the app store for free. Please be sure to give me some feedback. Visit my website for the app at http://www.valenzj.com/ for more info.

I would like to know if it is usefull at all and what features you would like me to add to the application.

AviSynth

2009-04-21T10:43:00.000-07:00

AviSynth is a frame server –it feeds uncompressed frames to other applications directly- that can be scripted to automate image processing and editing of videos.

It is not a video editing or special effects workstation. Its strengths lie on allowing the video professional to pre-process videos to resize, crop or remove noise and to conform video sources with simple cuts. Basically you create a script that indicates AviSynth which filters to apply in what order. If you are not comfortable with writing scripts in text form, you could use VirtualDubMod or MeGUI graphic user interfaces to create them.

AviSynth comes very handy when preparing videos for compression in an automated and streamlined fashion. In particular, you can perform noise reduction, resizing and color conversions to ensure that the compressed video looks as best as possible.
It is an open source project that has many followers. Check it out:

http://avisynth.org/mediawiki/Main_Page

Fonts

2009-04-17T20:20:00.000-07:00

Fonts of different types and sizes look different across video displays and resolutions. Therefore, choosing the right font size is critical to ensure readability on the final encoded video.

Some broadcasters recommend that as a rule of thumb font sizes to be displayed on an HD 1920 x 1080 video should be about 80% of the sizes used in SD. For example, you have been using a font that is 23 scan lines high on SD NTSC. First we consider the ratio of the HD / SD heights: 1080 / 486 = 2.22 and 80% of 2.22 is 1.776. Use 1.776 as the font scaling factor. That means that our new font size is 23 x 1.776 = 41 scan lines high (with rounding.)

Always check how your titles look in the target player and display to be sure that the text is readable and that aesthetics translate.

Audio Quality

2009-04-08T10:04:00.000-07:00

Some studies have pointed out that the human brain puts together both audio and images as a single message and therefore audio has a tremendous impact on viewer experience. Some research has even pointed out that viewers are less likely to completely view a video clip if the audio quality is poor.

Below is a list of audio encoding artifacts to check for as well as some tips.

Lip-sync. Most people find "out-of-sync" audio as unacceptable.
Audio drop outs and pre- echo are perceived as “very annoying” artifacts by most people. They tend to be more prevalent in material that is percussive like a gunshot or a drum hit. Increase the bit rate as much as possible to compensate.
Narrow audio program bandwidth. If high audio frequencies have been removed, the audio will sound “nasal” or “muddy”. If Low audio frequencies have been removed, the audio will sound “thin” and “harsh”. Make sure the resulting audio covers as much frequency spectrum as the original. Keep the audio sampling rate the same as the source or above 44.1 KHz for best results.
“Garbled” audio due to over compression tends to have sort of an underwater sound. These artifacts will be more noticeable in material that has random characteristics like rain, applause or city traffic. Give enough bits to the audio stream. Try to keep audio bitrates above 96 kbs for best results.
Summing some stereo material to mono will cause phasing artifacts and audio dropouts. Experiment by taking only the left or only the right channel if you hear this artifact. You might get better results.
Audio re-sampling or clock errors often have a metallic / ringing sound and can produce loud ticks. If you are encoding video from tape or other hardware playback device, make sure the audio is properly clocked and sync-locked. When using AES/EBU or SPDIF audio, clock your input device to the source’s digital signal as a rule of thumb.
Make sure the audio is noise (hiss) free. Be careful of not going overboard if you have to pre-process the audio with noise suppression as you might induce more serious artifacts. Also, make sure that your audio has the required bit resolution. Today, 16 bit audio is common place. Avoid a lower value.

De-interlacing algorithms

2009-04-02T10:14:00.000-07:00

De-interlacing algorithms can be sub divided into two broad categories:

Spatial Based
Motion Based

Some popular spatial based algorithms are:

Simply weaving the odd field and doubling each line. (Produces aliasing)
Averaging or blending both fields. (Produces ghosting)
Edge dependent interpolation techniques between adjacent lines in odd-even fields. (Few artifacts if edge direction is inaccurate)

Some motion based algorithms are:

Selective blending (or motion adaptive blending) basically only blend areas where there is motion and weaving where there is none. (Few artifacts if motion calculation is correct)
Motion and edge adaptive algorithms combine edge interpolation techniques and motion adaptive blending. (Fewer artifacts at computational expense)
Motion compensated algorithms can use the information to find the matching blocks in the neighboring fields. (Fewer artifacts at computational expense)

Be sure to check what algorithms your encoding software implements, it might have a great impact in your final video quality.

De-interlacing

2009-03-26T10:16:00.000-07:00

To compensate for CRT artifacts, an image frame has been traditionally broken up into two fields that contain even and odd scan lines. Each field thus contains half of the vertical resolution of a frame. Most standard definition video cameras record in this field manner (in NTSC they record 59.94 fields per second) which looks great in a CRT display.

To reproduce the image in its original resolution on a “progressive” type display –LCD, Plasma, etc.- both fields can be combined. However, because each field is recorded at slightly different time (1/59.94 = 16.68 ms) some artifacts may show up in the resulting picture. These artifacts, particularly visible in scenes where there is fast motion, appear as a kind of line crawling over the picture.

Most video original material is interlaced, but some original film material that has been converted to video (Telecine) might also be interlaced. For that reason, when transcoding video for playback in computer systems or on “progressive” type displays, always check if the video source is interlaced and if so, apply de-interlacing. If the source is a Telecine of a film, use an inverse telecine setting.
The link below gives some great examples of interlacing. Check it out.

http://scanline.ca/deinterlacing/visual.html

VirtualDub

2009-03-16T09:28:00.000-07:00

VirtualDub is one of many very useful open source programs that allows you to quickly convert and process video.

VirtualDub provides you with a framework for basic video editing, compression and filtering. It comes with a few plug-ins for encoding, frame rate conversion and image processing, but there are many third party plug-ins in the open source community that can be added very easily.

http://www.thedeemon.com/VirtualDubFilters/

With this powerful application you can stitch videos together, change audio or add overlays very quickly. This is one tool that video compression professionals must have in their arsenal. You can get it here:

http://www.virtualdub.org/

More technical video professionals can create or adapt their own plug-ins for VirtualDub using its open source filter SDK.

http://www.virtualdub.org/filtersdk.html

As with any open source solution, the main issue is support. However, with VirtualDub there is a very large community of developers and users that you can tap to find answers.

Video stream analyzers

2009-03-02T16:15:00.000-08:00

I looked around for video stream analyzers and got a couple of trial versions to test. Here is a brief description:

Elecard StreamEye

An intuitive interface that lets you navigate and display the stream picture by picture while it determines the real bit rate and compares it to the header values. It will tell you the streams real peak bit rate and how it varies as well as frame data sizes. It gives you access to macroblock information and its file address as it is displayed on the corresponding image.

http://www.elecard.com/products/products-pc/professional/streameye-studio/

MTS4EA Tektronix Compressed video elementary stream analyzer.

A much more powerful tool that can help codec developers debug and optimize codec code. The learning curb is steep around their user interface as it offers lots of features .
It verifies stream compliance against standard. It gives you the ability to perform a differentiating analysis against a reference video. (.yuv file) and calculates PSNR, RMSE, MSE and other measurements. It gives you macroblock statistics. The video view selection may show you the hex code for a macroblock or a macroblock set. It includes a bit stream editor for debugging applications and a batch mode with logging capability. Their image inspector lets you view pixel data for individual macroblocks and separates them into their component channels for YUV or RGB images. They have a macroblock type overlay that lets you see, in color, if a macroblock is an intra 4x4 or inter type block. It also overlays the motion vectors as color coded arrows for B or P frames. It is capable of producing a bit usage histogram and give you buffer analysis.

http://www.tek.com/Measurement/applications/video/mpeg_derivatives.html?WT.srch=1&WT.mc_id=ppc,ggl,vid_aw_us_us_ngt,k3AF2,s,1520098873&

Interra Vega-Media analyzers

I wasn’t able to get a trial version of this product so I can only paraphrase what they have on their website.
It supports H264, VC1, MPEG2, Audio and compares with other encoded streams. It has a YUV Dif utility to evaluate quality against an original video. It provides with syntax analysis tools and frame statistics for analysis of interpolation and prediction.

http://www.interrasystems.com/dms/dms_vega.php

HPA Retreate Notes

2009-02-23T12:42:00.000-08:00

There was an interesting session at the HPA (Hollywood Post Alliance) retreat last week where many studios as well as a representative of the EBU (European Broadcasting Union) talked about their real world experiences with video encoding and distribution.
Here are my notes.

The EBU has published their recommendations for HDTV video compression for acquisition, production and distribution in this document: http://tech.ebu.ch/docs/r/r124.pdf
The full report is only available to EBU members, but we can take note that:

Progressive pictures look better than interlaced.
Acquisition is recommended at 4:2:2 where 8bit is sufficient.
To manintain quality after 7 cycles of encoding, Long GOP, CBR, MPEG-2 at 50Mbit/sec at the least is recommended (post and production).
Distribution using H.264/AVC, CBR encoding requires 50% less bit rate than MPEG-2.
Interlaced video requires 20% more bit rate than progressive video for the same quality.
The dominant image quality impairments are determined by the distribution codec.

In that panel, there were engineers form three mayor broadcast studios who revealed their own conclusions:

MPEG-4 @ 20Mbits/sec, MPEG-4 @25Mbits/sec and MPEG-2 @45Mbits/sec are used in their facilities to maintain reasonably constant quality.

Automated Transcoding and Publishing

2009-02-16T19:03:00.000-08:00

The panacea of original video publishing automation would be to go from final edited version to the asset management system to the web and other distribution channels. And along the way collect all relevant metadata, generate thumbnails and create videos that suit players that can capture relevant advertising based on the content.

A few companies have made great efforts to automate this kind of work flow, but one that stood out for me recently is Entriq’s Dayport solution: http://www.entriq.com/.

Their solution allows content creators to integrate their workflow with FinalCut (and carry relevant metadata to the content management system), live capture off air or tape (which can be scheduled to trigger automatically), thumbnail generation, create rules for security and usage policy, DRM, syndication and CDN integration for delivery. Companies that syndicate content to multiple partners can automate the process and save lots of time re-formatting metadata, transcoding, etc.

Entriq provides a UGC platform that incorporates an approval process along with download controls and players that make the automation of video workflows from creation to publishing possible.

Cloud Video Transcoding

2009-02-11T07:37:00.000-08:00

Encoding.com offers video encoding services through a web service. Transcoding jobs are submitted using their XML API where video parameters, source, destination and logo insertion information are specified. They use Amazon.com EC2 to provide scalable transcoded nodes. They have a great cost estimate tool and case study here: http://www.encoding.com/pricing/.

Then there is Panda: http://pandastream.com/. It is an open source solution that runs on Amazon EC2, S3 web services. The main issue with this is licensing as it is an open source application based on ffmpeg.

Zencoder, http://zencoder.tv, in beta also runs on EC2 or your own Linux hardware. It works with On2 flix engine and ffmpeg.

Sorenson squish: http://www.sorensonmedia.com/products/?pageID=1&ppc=12&p=41 It is a distributed, Java based, client side encoding solution that uses the submitter’s computer as the transcoding node.

Now for the plug, Framecaster iNCoder Pro: http://framecaster.com/products.html. It is a distributed, client side, web plug-in. Like squish, it performs all video encoding on the submitter’s computer directly from a DV Cam, Webcam, or a Bluetooth enabled cell phone. It offers a Javascript API to integrate with an asset management system with ease. It can output H.264, FLV (VP6), MPEG4, MPEG2, or WMV.

A couple of questions about Panda and Zencoder are that since they use and deploy ffmpeg: How does one manage the licensing of codecs across all the different IP owners? Is that a potential liability on patent infringement? Maybe someone out there has an answers to those questions.

Standard VIdeo Players

2009-02-09T09:17:00.001-08:00

When it comes to transcoding, having a standard video player will make sure that videos look as they were intended. It will help the encoding professionals deliver what the audience expects and provide an engaging user experience.

Akamai has spearheaded the Open Video Player initiative to provide a set of Flash and Silverlight classes, sample code and other documentation that serves as a media framework and a path to standardize video players.

The application structure from the engineering perspective is described in figure 1 below.

The sample code will give you a head start in creating and fine tuning your own player that adheres to these set of best practices. There are many samples on how to integrate with advertizing platforms and create custom skins. You’ll also get samples on how to connect to Akamai’s servers and manage different types of streams and playlists.

Check it out: http://www.openvideoplayer.com/

Video encoding basics

2009-02-07T10:38:00.000-08:00

I came across a great video post by Lisa Larson that talks about the basic encoding concepts as well as very useful tips to make your videos look great. For those of you who use the Adobe Media Encoder, there is a great tutorial on how to do an encode and tweak parameters.

Check it out: http://www.flashconnections.com/?p=66

Resizing Affects Color?

2009-02-04T07:22:00.000-08:00

Resizing YUV type video might cause undesired color artifacts. These artifacts manifest as color blotches and sometimes color banding in areas of the image that has very smooth color or gradients.

4:2:2, 4:2:0, 4:0:0, etc. often refers to a YCbCr or YUV (Y’UV) image where Y is the Luma (intensity or detail) and UV are the chrominance components (color.) You can think of these three components as three planes Y, U and V that when super imposed create a color image. More on the subject here: http://en.wikipedia.org/wiki/YUV.

In the case of 4:2:2 images, the horizontal size of the chrominance planes is reduced by half, usually by discarding every other pixel. 4:0:0 images have both horizontal and vertical their UV planes' size reduced by half. When recreating the original image, the UV planes are enlarged to their original size by using simple linear interpolation methods. This process in itself causes artifacts that are compounded when the video is resized (see previous articles on resizing algorithms.) These errors in the UV planes are more obvious in smooth color or slow gradients.

Whenever possible, it is better to resize your images in the RGB or 4:4:4 space and then convert to a compressed YUV type format to minimize artifacts. Try to follow this rule especially with computer animation or other computer generated images.

Non Square Pixels

2009-02-01T20:54:00.000-08:00

I found a really good document that describes aspect ratios and non-sqare pixels. It is worth a read.

http://www.artbeats.com/assets/articles/pdf/aspect_ratio_1.pdf

Resizing SD to HD

2009-01-29T10:50:00.000-08:00

Let’s now take a look at up rezing from SD to HD and a couple of different available options. We are going to consider NTSC and assume square pixels for simplicity.

SD has an aspect ratio of 4:3 with video frame with horizontal dimension of 640 and a vertical dimension of 480 pixels or 640 x 480 (we always count horizontal first.) The two flavors of HD are 1280 x 720 and 1920 x 1080 with an aspect ratio of 16:9. Figure 2 shows frames with the same proportions to illustrate.

The simplest approach is to stretch the SD image to fill the HD frame. But this results with very fat looking people. Not very pretty. The next easy option is to enlarge the SD image in the vertical dimension to match HD first. For that we compute the ratios of the vertical sizes. For 720 we have 720/480 = 1.5. So we need to enlarge the horizontal size by the same ratio to maintain the picture proportions: 640 x 1.5 = 960. That means that we are left with 1280 – 960 = 320 in width without picture. We could split the 320 in half to create two vertical black bands of 160 pixels in width each.

Now for the 1080 case we have 1080/480 = 2.25 and 640 x 2.25 = 1440. So we are left with a 1920 – 1440 = 480 wide area of the 16:9 picture that can be split in two bands, one on each side of the SD image of 240 pixels in width.

In both cases (720 and 1080) the resulting video looks like figure 3.

There are some algorithms that stretch the image horizontally to fill the entire HD frame at the expense of distorting the original image. Some algorithms evenly stretch the image horizontally and others keep the center of the original image intact and incrementally stretch out to the sides (kind of like a fish eye effect.)

Next post: Resizing affects color?

Under $1000 Encoding Software Reviews

2009-01-26T12:01:00.001-08:00

Below are two links to really good reviews for under $1000 video conversion software for the mac and pc.

They contain samples and compare not only quality but also speed.

Check them out:

http://digitalcontentproducer.com/videoencodvd/revfeat/encoder_shootout_1208/index.html

http://digitalcontentproducer.com/videoencodvd/revfeat/expertise_encoder_shootout_0109/index3.html

Resizing HD to SD

2009-01-22T10:31:00.000-08:00

Let’s look at a very common problem these days: Down rezing from high definition (HD) to standard definition (SD) in NTSC.

HD comes in two flavors, 720 or 1080, which refers to the vertical dimension of the video frame. Considering that the image has square pixels (as in the previous post) the dimensions are:
Horizontal = 1280 and Vertical = 720 where the aspect ratio is 1280/720 = 1.78 or more commonly known as 16:9 (16/9 = 1.78 .)

And the other flavor where Horizontal = 1920 and Vertical = 1080 that also gives an aspect ratio of 16:9. The full NTSC frame dimensions are Horizontal = 640 and Vertical = 480 with an aspect ratio of 4:3. (Figure 3)

If we want to shrink the image without distortion we have to make a “Letterbox” 4:3 meaning that it will have black horizontal bands at the top and at the bottom of the frame. Otherwise people would look really skinny. (Figure 4)

Let’s say that we are working with 1080. We know what the horizontal dimension should be in SD NTSC; 640 and that the HD horizontal dimension is 1920. To find the resizing factor we compute the ratio of the horizontal dimensions 640/1920 = 1/3. We are shrinking the image by a third. And to keep the proportions in the HD image unaltered we need to also shrink the vertical dimensions by one third. That means that the new vertical dimension is 1080/3 = 360 pixels high. But since the full NTSC frame is 480 we have 480 – 360 = 120 pixels that are unused. That means that each black band above and below the letterbox image is 60 pixels high.

The most common method to fill an entire 4:3 target image with a 16:9 source is to do a Pan and Scan. That means that we use a 4:3 area of interest inside the 16:9 picture and move it around to where the action is in the image. Of course the draw back is that we miss a good portion of the original image. It is a costly and time consuming process that is performed manually to a great extent (figure 5.)

Next post I will cover the "Up Rezing" algorithms and tips to up convert SD to HD.

Resizing Algorithms

2009-01-15T06:27:00.000-08:00

In this article I will describe the most common algorithms for resizing images and in particular how to convert images to a smaller size (“Down Rezing.”)

Resizing bitmaps or pixel based images to a smaller dimension consists of a mapping of a larger group of pixels in the original picture to a smaller group of pixels or even a single pixel. This seems pretty obvious; you start with a larger image (more pixels) and end up with a smaller one (less pixels.) The important question is how you go about it. What algorithm will produce the best quality picture?

When you are shrinking an image to half its size you could simply decimate or discard every other pixel. But you end up with a very bad looking picture.

Let’s say that you divide the target image in a grid. Now, each cell in this grid corresponds to one pixel. We’ll assume that the dimensions of the source image are multiples of the target image dimensions to make things easier. You then make the same grid division on the source image so that you end up with the same number of cells in the grid. The cells in the source image grid will now contain more pixels than the target image. (Figure 1)

In nearest neighbor interpolation (Figure 1) you choose the pixel that is closest to the center of the grid without any regard to the other pixels in the grid. The images produced by this algorithm are of very poor quality as much of the detail is lost because none of the surrounding pixels are taken into account when computing pixels for the new image.

Bi-linear is another fast algorithm that is very common that consists of entering four neighboring pixel values into a bi-linear type equation (it computes the weighted average of the four pixels). The result of this equation gives you the value of the pixel in the target image.

Bi-cubic interpolation produces much better quality pictures than bi-linear, but it takes much more processing time as the equations for calculating the value of pixels in the target picture are more complex. It takes into account the values of neighboring pixels and how they change.

There are many other methods such as fractals and super sampling that produce even better quality images. They usually take longer to compute also, but if quality is what you are after, you should consider using them. The link below shows some very good samples of the each method. The samples on the page show enlarging, but I hope you get the idea.

http://www.dpreview.com/learn/?/key=interpolation

Next post will cover the specifics of resizing High Definition content to Standard Definition.

CES 2009

2009-01-12T17:48:00.000-08:00

First, I want to thank all of you for taking the time to give me such valuable feedback for the blog.

Now, I will deviate a bit from what I usually write to report what I thought was notable at the CES show in Las Vegas last week.

In general, there were many products for consumers to view video on the internet. There were many devices that connected directly to the internet and offered direct access to yahoo! or YouTube videos. The internet is really making its way into the living room with monitors that are networkable as well as very inexpensive network appliances (under $200) that let you view videos on the internet using a simple remote control.

Monitor resolutions, contrast and size have greatly improved. I saw some 4K monitors that looked fantastic, but for which there is no content for. High Definition, solid state drive video cameras are now the norm from every manufacturer. Many of the cameras use the AVCHD codec.

The definite trend is that video on-demand and on the internet will be the norm for years to come.

Next post I will try cover some key concepts for resizing videos and how it affects quality.

Generation Loss. Avoid it!

2009-01-07T07:49:00.000-08:00

Happy new year to all. I hope that everybody had a nice time over the holydays.

Generation loss is a term used to describe the degradation of video or audio quality when making copies. It is usually the primary cause for bad quality video. In this article I will give some tips to avoid it and encoding tips when it is unavoidable.

Let’s look at how generation loss happens. For example, you shoot a video with your firewire camcorder and produce a camera original that it is stored on a tape. You then transfer that video to your windows computer using movie maker. The result is a windows media file that is the second generation from the camera original. Depending on the settings that you choose, there will be varying degrees of quality degradation. Now suppose that you would like to send the video to YouTube, but the file is too big for YouTube to accept it. You are now forced to convert the video into a smaller file, so you open the second generation windows media file with movie maker one more time. Now, you play with the parameters to compress the video further and manage to produce a file that is now small enough to be accepted by YouTube. This file is now the third generation from the camera original with more quality degradation. So you upload it to YouTube and a few days later it shows up on the website, but it has now been converted to Flash format and it doesn’t look very good. This is now the fourth generation from the camera original.

The obvious first tip to avoid generation loss is not to cause it in the first place. In the above example, when you get ready to transfer the video from the camera to the computer, you could choose video encoding parameters that will produce a file that is smaller than the maximum required by YouTube. You would go straight to second generation before uploading to YouTube and thus saving one generation loss. If that sounds too complicated you can go to a free site www.compressmyvideos.com where you can transfer the videos from your DV camcorder, phone, webcam or even use a file and make the output file less than 100MB.

But you could also transfer the video from the camcorder to a lossless format or in the case of digital cameras, to a format that is an exact copy of what is on the tape without any loss. Some software packages allow you to do that, including movie maker. Now, the size of the file will be pretty big. For a DV AVI the file size is roughly 215MB per minute. For uncompressed YUV video files (4:2:0) it is roughly 1.25GB per minute. For full RGB uncompressed video you need about 1.87GB per minute in NTSC (29.97 frames per second.) Always try to make a lossless copy of the original video. If it came from a DV camcorder, use a DV AVI or “bump it up” to YUV or RGB.

With tapeless cameras it is easier to obtain a lossless copy of the camera original as you can just copy the file to your computer via USB. Just make sure that you have the appropriate codec to “decode” this file installed in the computer other wise you might be forced to use some kind of software utility to convert it to another format that your computer knows how to “decode.” This can potentially induce a generation loss.

The next tip is not so obvious, but it has a great impact in image quality when making several copies or transcodes. Try not to resize your video. Always try to keep the size of your video picture the same through the conversion process. If you really have to change the size, it is always better to go from big to small.

My last tip is to recommend that if you need to make the video file size small, you need to convert your video using a high quality codec. Some codecs are much better than others. MPEG2 is better than MPEG1 and MPEG4 is better than MPEG2, but VC-1 (windows media) or H.264 or VP6 are better than all the MPEG’s. Always keep looking for new and improved video codecs as technology in this field changes rapidly.

Tomorrow I'll be traveling to Vegas to check out CES. I'll be back with a report of the newest, coolest gadgets for video transcoding presented at the show.

User Generated Video Workflows

2008-12-18T10:55:00.000-08:00

The explosion of YouTube and user generated video (UGV) have introduced a new kind of workflow in the video space. Its popularity really got started with American Funnies Home Videos, where the public would submit a video to compete in a contest for the funniest video. The show is still on the air and enjoys great ratings. In its origins, the public would send their videos in the mail in the form of a video tape, a DVD, and now a digital file through the internet.

Today, videos are sent to a plethora of video aggregators and sharing sites that collect in mass from the public. Further, video sharing sites are not only interested in funny videos, but also serve as outlets for independent news, video productions, viral marketing and promotions among other uses.

A typical UGV work flow is described in the figure below.

Today users, the public, create videos using many different devices such as mobile phones, DV Cameras or WebCams and store these videos on tape or digital files. They use many different video tools across platforms to extract their videos from their devices or tapes and then send them to a video site. These videos are then stored and processed to meet the video site’s streaming server’s specifications. Videos are then made available for viewers to download or stream.

To limit bandwidth and storage use, users are asked to limit the size of the video file they are submitting. Some video sharing sites have a file limit of 100MB others go as high as 2GB. Video encoding parameters can be adjusted to generate a file of a particular size using the formula below.

(videoBitrate+audioBitrate)=(fileSizeLimit*8)/videoDuration

Just fix the audio bit rate to 64 kilo bits per second (Kbs), or a higher value if you would like better audio quality, and you can obtain the video bit rate settings.
For example:

A video sharing site has a 100MB file size limit and my video is 10 minutes long. I set my audio bit rate to 64 Kbs. The video bit rate setting should then be:

(100MB*8bits)/(10min*60sec)- 64000bps=1269.3Kbs

Although there is nothing magical about the above formula, most users will have a hard time doing all the unit conversions required to figure out the parameters. In my experience, users produce files smaller than the file restriction through trial an error. Besides the video and audio bit rate settings on the encoding software, frame size or resolution has a great impact in video quality and size. The smaller the resolution or picture size the smaller the file size. Lossy encoders will not necessarily allocate more bandwidth (larger bit rate) if the picture is small because there is simply not enough data in the resulting picture. (In my opinion, resizing images should be avoided using consumer applications as it may degrade the quality of the video tremendously. )

Users send the resulting video file to the video sharing site where it is stored and usually it is reviewed for content. Videos are then sent to a queue in video aggregator’s Transcoding system where it eventually gets converted and posted for streaming or user download.

An alternative way of processing user generated videos is through a distributed encoding application as illustrated below:

The most prominent competitors in this space are On2’s flix publisher, Framecaster’s iNCoderPro, Sorenson Squish and the late VideoEgg. These systems allow the video sharing site or aggregator to set up a file format and video setting to receive videos through a web plug-in or widget embedded in their submission page. The plug-in or widget is really a video encoder that processes the user’s video to the video aggregator’s specifications. This option can improve the resulting video quality as videos are transcoded from the original source. This systems offer a varying degree of device support. From no device support to Webcam, and DVCam support as in the case of VideoEgg and Squish, but the only product that also supports mobile phones via Bluetooth is iNCoderPro. The supported video formats also vary. Flix Publisher and VideoEgg only support. VP6 (flash 8). Squish only outputs Flash video (maybe the Spark codec or flash 7. It’s not clear on the Sorenson website,) iNCoderPro supports VP6, h.264, MPEG4, MPEG2, Windows Media and 3GP (Try it for free at www.compressmyvideos.com.)

This alternative kind of workflow for UGV has many advantages for both the user and the video sharing site such as:

Better user experience as users are not required to use third video tools.
Better video quality as videos are converted from the original video and the right settings are applied automatically by the software.
Minimizes the need for an in-house Transcoding system.
Saves bandwidth as smaller video files are sent to the aggregator.
Saves storage space as only one file is stored.
Less time to post as videos don’t need to go through a queue to be formatted for streaming.
Self scalable Transcoding. Video sharing sites won’t have to worry about increasing their Transcoding capacity for higher volumes.
Lower startup cost for video sharing sites as they don’t have to buy hardware and only minimal programming or scripting is required.

Next time I will discuss Generation Loss and how to minimize it.

Transcoding Automation

2008-12-11T13:00:00.000-08:00

Below are some of the most important questions and concepts you need to keep in mind when designing or choosing a video Transcoding system. They are described in general terms and more detail is needed to really optimize a Transcoding system for your particular workflow. Keep in mind that there are systems that require some level of programming or scripting and that might have a big impact on your budget.

In deciding how to automate your video Transcoding workflow, you should ask your self the following questions:

1) What is the volume of videos you need to process?

2) How fast do you need the output videos?

3) Where do the input videos reside? (Single or multiple locations.)

4) What is the input format(s)? (Device, file format.)

5) What is the output file format(s)?

6) Where do you need to deliver the output videos?

The answer to the first couple of questions will help you determine how large and complex your system is going to be. If your volume is very low, from 1 to a couple of dozen videos a week, you could set up a simple software-based, batch encoding system in a single computer. As your video demands begin to grow, you could batch process the videos in more than one computer, and manage the queue manually.

There is a point where manually managing a queue not only becomes a full time job, but also introduces too many human errors. You should consider setting up a system with automated queue management. The queue system should provide you with different error handling choices. How you handle encoding errors will ensure that your Transcoding system continues to operate on subsequent files. It is critical to prevent your queue from jamming up due to an expected error. They queue system should allow you to set and change priority levels for your Transcoding jobs and ultimately, it should let you dictate how you allocate both processing and storage resources. It should let you bring online and take off line additional hardware without shutting down the system.

As your volume and throughput demand increases, you need to begin to consider what kind of hardware you need to use with your software solution or whether you need a hardware accelerated platform. If you want to stay with a software based solution you need to look at multiple CPU computer systems, what kind of file storage (disk I/O is usually the slowest part of the system) is available, and Network connectivity. Keep in mind that the Transcoding process takes a good portion of the time in processing your videos, but it is network latencies and disk I/O bottlenecks what often brings down productivity. A good system makes sure that more than enough bandwidth is allocated for file I/O.

The location of your input videos is an important consideration on how you design your system. Do you have a library of content that resides in a single location? You should have then a Transcoding system that is located closest to the content and make sure that the data throughput from your sources has enough bandwidth to avoid I/O bottlenecks.

Does the content reside at multiple locations? You might increase your throughput by having a encoding system at a few of the source locations. Your queue system and asset management system should be designed in a way that it allows you to distribute the load across locations if necessary.

Is the content being sent to you from many locations? Here, you have different factors at play and a few options. First, you must make sure that you have enough bandwidth to receive the input videos. Here you can leverage a CDN to increase your bandwidth and take advantage of the geographical proximity to the submitter. Second, you must make sure that your file storage (for input files) is adequate and scalable. Third, you might be better off setting up a distributed encoding solution where videos are converted at a point closest to the submitter, or even at the submitter end. Fourth, content management must be able to track all the incoming videos and move them to the output location(s) with flexibility.

Input formats or devices that need to be supported are important constrains in designing the Transcoding system. Do you need tape support? Do you need analog or digital video inputs? (SDI?) If so, you need a capture card with a serial machine control interface and make sure that it supports the right kind of video inputs. You also need to make sure that the input file format is supported by your hardware of software Transcoding system.

The output file format(s) is also an important consideration. Do you need to output videos in one single format? What format? Some systems charge more for certain output formats or doesn’t support them at all. Don’t assume, read the specifications or ask the vendor.

Where and how you are going to deliver your videos is the last piece of the Transcoding system. Again, you need to make sure that you have enough bandwidth to support your desired output rate. Consider a CDN when you need to deliver to multiple geographic locations.

Next post will cover user generated video transcoding workflows.

Video Mastering Transcoding Workflows

2008-12-04T16:28:00.000-08:00

Encoding of videos for mastering is driven by quality. The objective is to achieve a video file that adheres to strict industry standards.
Usually, a single or a series of masters are created to be to be replicated or to be used as sources in making copies of the video to other formats.
As an example, a video master for a DVD is created with a professional MPEG-2 video encoder from a video tape (a master video tape itself, usually.) Great attention is placed on creating a pristine, free of artifacts MPEG 2 file that goes through a quality control process (QC) by a human that watches the video in a room equipped with video monitors that have been calibrated for color, brightness, etc. The QC process also involves looking at the video signal on a waveform monitor and vector scope to verify its integrity. There are also QC software programs that can analyze the file for anomalies in the stream.
It is not uncommon that file candidates to be masters are rejected. Video material varies greatly and that makes the video encoding for masters challenging. Parameters in mastering type video encoders can be adjusted for scenes with fast motion, noise, or scene changes. Some encoding packages allow the encoding operator to adjust these parameters on a scene by scene basis. And as you can imagine, this process becomes very time consuming.
A good video mastering workflow is illustrated in the figure below.

To avoid storing RGB or YUV files, some facilities and studios have been creating high quality compressed files like MPEG-2 @ 50Mbs I frame only to be stored and used as masters to create other files down stream. They call them “mezzanine” files as they are an intermediate master in the creation of other videos. These mezzanine files should also follow the same type of workflow as they are a master on their own. Mezzanine files are being used today to create podcast, streaming files for web delivery in many formats as FLV, H.264, WMV or several flavors of Quicktime. The great advantage of using a mezzanine file as the source is that the encoding of new target files can be automated, and managed by a computer. Also, it uses less bandwidth on the network as compared to uncompressed videos or using tapes as sources.
In my opinion, H.264 should be used to create mezzanine files as greater video quality can be obtained while producing smaller files.

Next post will cover transcoding automation

Free and Open Source Transcoding Software

2008-12-01T18:18:00.000-08:00

First of all I would like to thank all of you for the great topic suggestions. I'll try to get to them in a timely manner.

To conclude a cursory review of the different Transcoding options, I will examine some of the most popular free or open source solutions available.

As with any free software or open source package, if there are any problems, you won’t have any body to complain to or to help you. Also, some free codec packs and encoding packages may contain spy ware. Definitely Caveat Emptor.
There are some very good video Transcoding tools available as open source or free. Probably the most popular open source software package is FFMPEG.

FFMPEG:
This is an open source video Transcoding platform that covers pretty much every video codec out there with the exception of On2’s VP6 Flash-8 codec.
There are many graphic user interfaces to FFMPEG’s command line interface. I have tried the SUPER package and found the UI intuitive and easy to use.

http://www.theinquirer.net/en/inquirer/news/2006/07/31/dont-buy-any-video-converter-before-trying-this-freebie
http://www.erightsoft.com/SUPER.html
http://ffmpeg.mplayerhq.hu/

Riva FLV Encoder
Converts to Flash FLV 1.1 from many input formats including AVI, MPEG, DV, MOV and WMV.

http://www.download.com/Riva-FLV-Encoder/3000-2140_4-10320097.html?cdlPid=10381392

Windows Media Encoder
Converts to wmv files. It accepts many input video files as long as the codecs are installed in your system. Free from Microsoft. If your input file doesn’t work, you might need to install a codec pack.

http://www.microsoft.com/windows/windowsmedia/forpros/encoder/default.mspx

DivX Encoder
Converts to DivX format. The input support also depends on the codecs you have installed in your computer. You can get it free at the DivX website

http://www.divx.com/divx/windows/

TubeTilla YouTube downloader
Downloads a YouTube video from a URL and converts it to mp4 or wmv. A couple of people seem to have liked the software.

http://www.snapfiles.com/Freeware/gmm/fwvideconvert.html

Next week I will be talking about some of the best video transcoding practices for mastering and how to optimize that workflow.

Professional Hardware Encoding Systems

2008-11-20T10:49:00.000-08:00

Below is a short description of the most popular video Transcoding appliances available in the market today.

Inlet technologies:

Armada: $14,000 + $7,500 additional node.

Windows Media VC-1, Flash 8 (VP6), H.264, MPEG-4 Part 2, MPEG-2 and AC3 Stereo.
Nice management system. You can define pre encoding tasks such as file verification and metadata extraction and post encoding tasks such as quality control, encryption and even publishing. Multi-CPU system.
No formal reviews are available yet.

http://www.inlethd.com/encoding/65/47/Automated-Transcoding-and-Encoding-Workflow-Management/

Fathom:

H.264, MPEG-4, Windows Media, VC-1, MPEG-2, AVI, and Flash
Watch folders, accepts from tape, servers or editing workstations. Both HD and SD are offered.
http://www.xyhd.tv/2006/11/reviews/review-of-inlets-fathom-vc-1-hardware-accelerator/
http://www.inlethd.com/encoding/18/16/Fathom/

Spinnaker: $9,995

Live encoding and streaming. VC-1, VP6 and H.264.
Price ranges between 30k to 70K.
http://www.streamingmedia.com/article.asp?id=10182
http://www.inlethd.com/encoding/20/18/Spinnaker/

Media Excel:

HERA Line of appliances.

The HERA u is designed with web and mobile delivery of user generated content while the HERA f is targeted to Broadcasting.
HERA Series outputs to MPEG-2, MPEG-4, H.264, DivX, xVID, QuickTime, Flash 9, WM9, 3GPP. And is configurable to output on Gigabit Ethernet, Fiber Channel and may accept HD-SDI as input.
Metadata extraction and insertion. Scalable and high speed.
No reviews as of the writing of this post. This system is being unveiled at IBC 2008

http://www.mediaexcel.com/hera_4000u.php

HMS File management

The HMS File management system is designed to manage all the Transcoding and delivery of files across many HERA encoding appliances.

http://www.mediaexcel.com/hera_management_system.php

ViewCast:

Niagara Streaming Encoders:

Capture, encode and stream workflows. Depending on the mode it can compress to Mpeg-4 and Flash, wmv and Real. Designed to stream live. Offer SDK.
http://www.viewcast.com/product_GoStreamSURF.asp

Digital Rapids:

Transcode Manager with StreamZ: $20,000 + $5,000 additional node.

Digital Rapids Transcode Manager with StreamZ

A very sophisticated and with unlimited scalability. It offers a full API via XML. It is a very robust system that can be configured to handle specific outputs if needed. You need a server ($20,000) and you can add nodes for $5,000 to $8,000 depending on the output types. They offer a Lite system that can support up to 10 engines. The light version server costs $5,000 and each node is $5,000. The lite API is has some restrictions and limitations.
http://www.streamingmedia.com/article.asp?id=9423&page=1
http://www.digital-rapids.com/Products/IndividualProducts/Transcode%20Mgr.aspx
http://www.digital-rapids.com/Products/IndividualProducts/StreamZ.aspx

Ripcode:

Video Transcoding Appliance

Formats: WMV9 / VC-1, MPEG-1, MPEG-2 MP@ML, MPEG-1 Program Stream, H.264 BP and MP, Flash Video (On2 VP6), MPEG-4 Part 2 Simple Profile, 3GPP, AVI, M-JPEG
This is a different take on video Transcoding. The concept is to provide an array of DSP’s that can simultaneously transcode up to 8 different video streams. I saw a demo at streaming media west where the system was simultaneously Transcoding 8 mobile streams.
http://www.ripcode.com/productOverview.php

Consumer Software Transcoding Applications

2008-11-13T10:18:00.000-08:00

Below are some of the most popular consumer video transcoding applications in the market. They often are bundled with DVD authoring and DV Cam capture modules that come in handy. Some of these software packages don’t allow you to change or customize encoding parameters like the bit rate. Often they only offer a limited amount of pre-defined profiles for users to choose from.

AVS Video Converter: $39.95
Movavi Video Converter 6: $29.99
Pinnacle Studio 12: $49.99
Nero: $59.99
QuickTime Pro: $29.99
TMPGEnc 4 XPress: $99.95
Roxio Crunch: $39.99

AVS Video Converter:
Good value for your buck. You can convert to many formats, AVI, MPEG (including for DVD, iPod, and Mobile) QT, WMV, RM and Flash. It offered some simple yet handy editing features. They have a trial version that watermarks the output file. Output video quality is good.

http://video-converter-software-review.toptenreviews.com/avs-video-tools-review.html
http://www.avsmedia.com/VideoTools/index.aspx

Movavi Video Converter 6:
It doesn’t accept flash video files. Output formats cover MPEG, MOV, FLV, 3GP, AVI and WMV. Real.
You can select a profile and then go into “Settings” to change bit rates, resizing parameters, audio settings and more. It also offers some editing tools that come in handy. The Evaluation version didn’t work for me at all so I can’t comment on the output quality.

http://video-converter-software-review.toptenreviews.com/movavi-video-converter-review.html
http://movavi.com/videoconverter/whats-new/v60.html

Pinnacle Studio 12:
No trial version, so just read the review below.

http://computershopper.com/reviews/pinnacle-studio-12-review
http://www.pinnaclesys.com/PublicSite/us/Products/Consumer+Products/Home+Video/Studio+Family/Studio.html

Nero:
Good application. Mainly targets users that want to burn video to DVD or CD. Supports AVCHD on regular DVD’s.

http://www.softpedia.com/reviews/windows/Nero-7-Premium-Review-11730.shtml
http://www.nero.com/enu/nero8-introduction.html?NeroSID=7e4d874f7793823d9a17d6d5055b0141

Quicktime Pro:
Nice Product. The user interface for video conversion is not very good and the video quality is not the greatest, but you can’t beat the price. Output formats are limited. There is no support to convert to WMV, FLV or Real.

http://www.pcmag.com/article2/0,2817,1875130,00.asp
http://store.apple.com/us/product/D3380Z/A

TMPGEnc 4 Xpress:
Easy to use. Has a batch encode utility. It offers basic editing features and filters for video and audio. It lets users set a great deal more parameters within a template profile than other consumer applications. There is no conversion to Real or Flash. You can save to DVD and VCD formats. Output video quality is good.

http://dvdcreation.digitalmedianet.com/articles/viewarticle.jsp?id=48204
http://tmpgenc.pegasys-inc.com/en/product/te4xp.html

Roxio Crunch:
They didn’t offer a trial version, so I couldn’t give it a try myself.

http://www.itreviews.co.uk/software/s511.htm
http://www.roxio.com/enu/products/crunch/mac/overview.html

Next time I will review hardware encoding solutions.

Professional Video Transcoding Systems

2008-11-06T14:15:00.000-08:00

In this blog I will be covering real world encoding issues with the most current tools available in the market.
To start, I wanted to list some of the software professional transcoding tools that are most common in the industry today. Follow the links to the company websites as well as product reviews when available. The prices were gathered at their websites or reviews for the products and might have changed by the time you read this article.
This are software only solutions, you have to get your own hardware to run them.

Rhozet Carbon Server: $14,995 (minimum requirement)
Rhozet Carbon Coder: $4,995 (additional node)
Anystream: $10,000 (base system)
Flip factory : $5,495
ProCoder 3: $470
Sorenson squeeze 5: $499
Flip4Mac Episode: $ 499
Flip4Mac Episode Pro: $ 995

Rhozet Carbon Server / Coder

Probably the most popular application for video transcoding today is Rhozet. And for good reason. It is, very robust, scalable with a very flexible API.
The system requires that you buy the Carbon Server at a minimum, which is itself an encoding node, and you can buy additional Carbon Coders or nodes to extend the system's capacity.
http://www.streamingmedia.com/article.asp?id=9543
http://www.rhozet.com/

Anystream Agility System

This is another excellent product with a good API. Very reliable and robust. The system can be extended to suit your needs.
http://www.streamingmedia.com/article.asp?id=10177
http://www.anystream.com/

Telestream FlipFactory

A good system, but with a bit less straight forward interface and approach.
http://www.telestream.net/news/article_streamingmedia_FF12_18_01.htm
http://www.telestream.net/products/flipfactory.htm

ProCoder 3

Good bang for your buck. Not scalable but can easily automate simple workflows.
http://digitalcontentproducer.com/videoencodvd/revfeat/grass_valley_procoder/
http://desktop.thomsongrassvalley.com/products/ProCoderSW/closeup.php

Telestream Flip4Mac and Flip4Mac Pro

Supports many of the semi-pro and pro formats (Pro version.) You can do batch encoding. If you use a Mac, you need this tool.
http://www.macworld.com/article/46733/2005/09/flip4mac.html
http://www.flip4mac.com/

Sorenson Squeeze 5

Also a good bang for your buck. Not scalable, but you can batch encode with ease.
http://www.streamingmedia.com/article.asp?id=10439
http://www.sorensonmedia.com/products/?pageID=1&ppc=3

Next post will cover some of the most popular consumer applications available in the market today.