Debunking popular myths about digital audio.

2017-10-01T15: 27

2017-10-01T15: 27

Audiophile "s Software

Note: For a better understanding of the text below, I highly recommend that you familiarize yourself with the basics of digital audio.

Also, many of the points discussed below are highlighted in my publication "Once again about the sad truth: where does good sound really come from?" ...

The higher the bitrate, the better the track is

This is not always the case. First, let me remind you what bitrey is. T(bitrate, not bitraid). In fact, this is the data rate in kilobits per second during playback. That is, if we take the track size in kilobits and divide by its duration in seconds, we get its bitrate - the so-called. file-based bitrate (FBR), usually it does not differ too much from the bitrate of the audio stream (the reason for the differences is the presence of metadata in the track - tags, embedded images, etc.).

Now let's take an example: the bitrate of uncompressed PCM audio recorded on a regular Audio CD is calculated as follows: 2 (channels) × 16 (bits per sample) × 44100 (samples per second) = 1411200 (bps) = 1411.2 kbps ... Now let's take and compress the track with any lossless codec ("lossless" - "lossless", that is, one that does not lead to the loss of any information), for example, the FLAC codec. As a result, we will get a bit rate lower than the original one, but the quality will remain unchanged - here's your first refutation.

Something else is worth adding here. The output bitrate with lossless compression can turn out to be very different (but, as a rule, it is less than that of uncompressed audio) - it depends on the complexity of the compressed signal, or rather on data redundancy. Thus, simpler signals will compress better (i.e., we have a smaller file size for the same duration => lower bit rate), and more complex signals will be worse. That is why lossless classical music has a lower bit rate than, say, rock. But it must be emphasized that the bit rate here is by no means an indicator of the quality of the sound material.

Now let's talk about lossy compression. First of all, you need to understand that there are many different encoders and formats, and even within the same format, the encoding quality for different encoders may differ (for example, QuickTime AAC encodes much better than the outdated FAAC), not to mention the superiority of modern formats (OGG Vorbis, AAC , Opus) over MP3. Simply put, out of two identical tracks encoded by different encoders with the same bitrate, some will sound better and some will sound worse.

In addition, there is such a thing as upconvert... That is, you can take a track in MP3 format with a bitrate of 96 kbps and convert it to MP3 320 kbps. Not only will the quality not improve (after all, the data lost during the previous 96 kbps encoding cannot be returned), it will even worsen. It is worth pointing out that at each stage of lossy encoding (with any bit rate and any encoder), a certain portion of distortion is introduced into the audio.

And even more. There is one more nuance. If, say, the bitrate of an audio stream is 320 kbps, this does not mean that all 320 kbps were spent on encoding that very second. This is typical for constant bitrate encoding and for those cases where a person, hoping for maximum quality, forces a constant bitrate too high (for example, setting 512 kbps CBR for Nero AAC). As you know, the number of bits allocated to a particular frame is regulated by the psychoacoustic model. But in the case when the allocated amount is much lower than the set bitrate, then even the reservoir of bits does not save (for the terms, see the article "What is CBR, ABR, VBR?") - as a result, we get useless "zero bits" that simply "finish off »The size of the frame to the required one (that is, increase the size of the stream to the specified one). By the way, this is easy to check - compress the resulting file with an archiver (7z is better) and look at the compression ratio - the more it is, the more zero bits (since they lead to redundancy), the more wasted space.

Lossy codecs (MP3 and others) are able to cope with modern electronic music, but are not able to efficiently encode classical (academic), live, instrumental music

The "irony of fate" here is that in fact everything is exactly the opposite. As you know, academic music in the overwhelming majority of cases follows melodic and harmonic principles, as well as instrumental composition. From a mathematical point of view, this leads to a relatively simple harmonic composition of the music. So the predominance of consonances produces fewer side harmonics: for example, for a fifth (the interval in which the fundamental frequencies of two sounds differ by one and a half times), every second harmonic will be common for two sounds, for a fourth, where the frequencies differ by one third - every third, and etc. In addition, the presence of fixed frequency ratios, due to the use of equal temperament, also simplifies the spectral composition of classical music. The live instrumental composition of the classics determines the absence of noises typical for electronic music, distortions, sharp jumps in amplitude, as well as the absence of an excess of high-frequency components.

The factors listed above lead to the fact that classical music is much easier to compress, first of all, purely mathematically. If you remember, mathematical compression works by eliminating redundancy (by describing similar pieces of information using fewer bits), and also by predicting (so-called. predictors predict the behavior of the signal, and then only the deviation of the real signal from the predicted one is encoded - the more accurately they match, the fewer bits are needed for encoding). In this case, a relatively simple spectral composition and harmonicity cause high redundancy, the elimination of which gives a significant degree of compression, and a small number of bursts and noise components (which are random and unpredictable signals) cause good mathematical predictability of the vast majority of information. And this is not to mention the relatively low average volume of classical tracks and the frequent intervals of silence, for the encoding of which information is practically not required. As a result, we can losslessly compress, for example, some solo instrumental music to bitrates below 320 kbps (the TAK and OFR encoders are quite capable of this).

So, firstly, the fact is that the mathematical compression underlying lossless encoding is also one of the stages of lossy encoding (read Understandably about MP3 encoding). And secondly, since lossy uses the Fourier transform (decomposition of the signal into harmonics), the simplicity of the spectral composition even makes the coder's work doubly easier. As a result, comparing the original and the encoded sample of classical music in a blind test, we are surprised to find that we cannot find any differences, even at a relatively low bitrate. And the funny thing is that when we start to completely lower the encoding bitrate, the first thing that detects the difference is the background noise in the recording.

As for electronic music, coders have a very difficult time with it: noise components have minimal redundancy, and together with sharp jumps (some sawtooth pulses) are extremely unpredictable signals (for coders who are "sharpened" for natural sounds that behave completely otherwise), the direct and inverse Fourier transform with the rejection of individual harmonics by the psychoacoustic model inevitably gives pre- and post-echo effects, the audibility of which is not always easy for the encoder to evaluate ... Add to this a high level of HF components - and you get a large number of killer -samples, which even the most advanced encoders cannot cope with at medium-low bitrates, oddly enough, it is among electronic music.

The opinions of “experienced hearers” and musicians are also amusing, who, with a complete misunderstanding of the principles of lossy coding, begin to assert that they hear how instruments in music, after coding, begin to fake, frequencies float, etc. This, perhaps, would still be true for the antediluvian cassette players with detonation, but in digital audio everything is accurate: the frequency component either remains or is discarded, there is simply no need to shift the tonality. Moreover: a person's ear for music does not at all mean that he has good frequency hearing (for example, the ability to perceive frequencies> 16 kHz, which diminishes with age) and does not make it easier for him to search for lossy coding artifacts, since distortion these have a very specific character and require the experience of blind comparison of precisely lossy audio - you need to know what and where to look for.

DVD-Audio sounds better than Audio CD (24 bits versus 16, 96 kHz versus 44.1, etc.)

Unfortunately, people usually look only at numbers and very rarely think about the influence of one or another parameter on the objective quality.

Let's first consider the bit depth. This parameter is responsible for nothing more than the dynamic range, i.e., for the difference between the quietest and loudest sounds (in dB). In digital audio, the maximum level is 0 dBFS (FS - full scale), and the minimum is limited by the noise level, i.e., in fact, the dynamic range in absolute value is equal to the noise level. For 16-bit audio, the dynamic range is calculated as 20 × log 10 2 16, which equals 96.33 wB. The dynamic range of the symphony orchestra is up to 75 dB (mostly about 40-50 dB).

Now let's imagine the real conditions. The noise level in the room is about 40 dB (do not forget that dB is a relative value. In this case, the hearing threshold is taken as 0 dB), the maximum music volume reaches 110 dB (so that there is no discomfort) - we get a difference of 70 dB. Thus, it turns out that a dynamic range of more than 70 dB in this case is simply useless. That is, at a higher range, either loud sounds will reach the pain threshold, or quiet sounds will be absorbed by the surrounding noise. It is very difficult to achieve a level of ambient noise less than 15 dB (since the volume of human breathing and other noises caused by human physiology is at this level), as a result, a range of 95 dB for listening to music is completely sufficient.

Now about the sampling rate (sampling rate, sample rate). This parameter is responsible for the sampling rate over time and directly affects the maximum signal frequency that can be described by this audio representation. By Kotelnikov's theorem, it is equal to half the sampling rate. That is, for a normal sampling frequency of 44100 Hz, the maximum frequency of the signal components is 22050 Hz. The maximum frequency. which is perceived by the human ear - just above 20,000 Hz (and even then, at birth; as we grow older, the threshold drops to 16,000 Hz).

This topic is best covered in the article 24/192 Downloads - Why They Don't Make Sense.

Different software players sound differently (e.g. foobar2000 is better than Winamp, etc.)

To understand why this is not the case, you need to understand what a software player is. In fact, this is a decoder, handlers (optional), an output plugin (to one of the interfaces: ASIO, DirectSound, WASAPI. Etc.), and of course the GUI (graphical user interface). Since the decoder in 99.9% of cases works according to the standard algorithm, and the output plug-in is just a part of the program that transmits a stream to the sound card through one of the interfaces, only handlers can be the reason for the differences. But the fact is that handlers are usually disabled by default (or should be disabled, since the main thing for a good player is to be able to transmit sound in its "original" form). As a result, the subject of comparison here can only be opportunities processing and output, in which, by the way, there is often no need at all. But even if there is such a need, then this is already a comparison of handlers, not players.

Different driver versions sound different

This statement is based on a banal ignorance of the principles of sound card operation. A driver is software necessary for the effective interaction of a device with the operating system, also usually providing a graphical user interface to be able to control the device, its settings, etc. A sound card driver ensures that the sound card is recognized as a Windows sound device, informs the OS of the supported formats, provides the transfer of uncompressed PCM (in most cases) stream to the card, and also gives access to settings. In addition, in the case of software processing (by means of the CPU), the driver can contain various DSPs (handlers). Therefore, firstly, with disabled effects and processing, if the driver does not provide accurate transfer of PCM to the card, this is considered a gross error, a critical bug. And it happens rarely... On the other hand, the differences between the drivers can be in the update of processing algorithms (resamplers, effects), although this also does not happen often. In addition, effects and any driver processing should still be omitted to achieve the highest quality.

Thus, driver updates are mainly focused on improving stability and fixing handling errors. In our case, neither the one nor the other affects the playback quality, therefore in 999 cases out of 1000 the driver does not affect the sound.

Licensed Audio CDs sound better than their copies

If during copying there were no (fatal) read / write errors and the optical drive of the device on which the copy will be played has no problems with its reading, then such a statement is erroneous and easily refuted.

Stereo encoding mode gives better quality than Joint Stereo

This misconception mainly concerns LAME MP3, as all modern encoders (AAC, Vorbis, Musepack) use only Joint Stereo mode (and this already says something)

To begin with, it's worth mentioning that Joint Stereo mode is successfully used with lossless compression. Its essence lies in the fact that the signal before encoding is decomposed into the sum of the right and left channels (Mid) and into their difference (Side), and then these signals are separately encoded. In the limit (for the same information in the right and left channels), double data savings are obtained. And since in most music the information in the right and left channels is quite similar, this method turns out to be very effective and allows you to significantly increase the compression ratio.

In lossy, the principle is the same. But here, in the constant bitrate mode, the quality of fragments with similar information in two channels will increase (in the limit, it will double), and for the VBR mode in such places the bitrate will simply decrease (do not forget that the main task of the VBR mode is to stably maintain the given encoding quality, using the lowest possible bitrate). Since during lossy encoding priority (in the allocation of bits) is given to the sum of channels to avoid degradation of the stereo panorama, dynamic switching between Joint Stereo (Mid / Side) and normal (Left / Right) frame-based stereo modes is used. By the way, the reason for this delusion was the imperfection of the switching algorithm in older versions of LAME, as well as the presence of the Forced Joint mode, in which there is no auto-switching. In the latest versions of LAME, the Joint mode is enabled by default and it is not recommended to change it.

The wider the spectrum, the better the recording (about spectrograms, auCDtect and frequency range)

Nowadays on forums, unfortunately, it is very common to measure the quality of a track with a "ruler on the spectrogram". Obviously, because of the simplicity of this method. But, as practice shows, in reality everything is much more complicated.

And the point is this. The spectrogram visually demonstrates the distribution of signal power over frequencies, but cannot give a complete picture of the sound of the recording, the presence of distortions and compression artifacts in it. That is, in fact, all that can be determined from the spectrogram is the frequency range (and partially - the spectrum density in the HF region). That is, at best, by analyzing the spectrogram, you can identify the upconvert. Comparison of the spectrograms of tracks obtained by coding by various encoders with the original is sheer absurdity. Yes, you can identify differences in the spectrum, but it is almost impossible to determine whether (and to what extent) they will be perceived by the human ear. We must not forget that the task of lossy encoding is to provide an indistinguishable result. human ear from the original (not with the eye).

The same applies to assessing the quality of encoding by analyzing the output tracks with the auCDtect program (Audiochecker, auCDtect Task Manager, Tau Analyzer, fooCDtect are just shells for the one-of-a-kind console program auCDtect). The auCDtect algorithm also actually analyzes the frequency range and only allows you to determine (with a certain degree of probability) whether MPEG compression was used at any of the encoding stages. The algorithm is tailored for MP3, so it is easy to "cheat" using Vorbis, AAC and Musepack codecs, so even if the program writes "100% CDDA" it does not mean that the encoded audio is 100% identical to the original one.

And returning directly to the spectra. Also popular is the desire of some "enthusiasts" to turn off the lowpass filter in the LAME encoder by all means. There is a lack of understanding of the principles of coding and psychoacoustics. First, the encoder cuts high frequencies for only one purpose - to save data and use it to encode the most audible frequency range. The extended frequency range can have a fatal impact on the overall sound quality and lead to audible encoding artifacts. Moreover, turning off the cutoff at 20 kHz is generally completely unjustified, since a person simply does not hear frequencies above.

There is a kind of "magic" EQ preset that can significantly improve the sound

This is not entirely true, firstly, because each separately taken configuration (headphones, acoustics, sound card) has its own parameters (in particular, its amplitude-frequency characteristic). And therefore, each configuration should have its own, unique approach. In simple terms, such an equalizer preset exists, but it is different for different configurations. Its essence lies in adjusting the frequency response of the path, namely, in "leveling" unwanted dips and surges.

Also, among people far from working directly with sound, setting the graphic equalizer "with a tick" is very popular, which actually represents an increase in the level of low and high frequency components, but at the same time leads to muffling of vocals and instruments, the sound spectrum of which is in the midrange region.

Before converting music to another format, you should "decompress" it to WAV

I note right away that WAV means PCM data (pulse-code modulation) in a WAVE container (file with * .wav extension). This data is nothing more than a sequence of bits (zeros and ones) in groups of 16, 24 or 32 (depending on the bit depth), each of which is a binary code of the amplitude of the corresponding sample (for example, for 16 bits in decimal notation these are values ​​from -32768 to +32768).

So, the fact is that any sound processor - be it a filter or an encoder - usually works only with these values, that is only with uncompressed data. This means that to convert audio from, say, FLAC to APE, simply necessary decode FLAC to PCM first and then encode PCM to APE. It's like repackaging files from ZIP to RAR, you must first unpack the ZIP.

However, if you are using a converter or just an advanced console encoder, intermediate conversion to PCM happens on the fly, sometimes even without writing to a temporary WAV file. This is what misleads people: it seems that the formats are converted directly to one another, but in fact such a program must have an input format decoder that performs an intermediate conversion to PCM.

Thus, manual conversion to WAV will give you absolutely nothing but wasting time.

So, you have assembled a decent computer for yourself, learned the dao of the mid lane in Dota 2 and decided to launch your first broadcast. Or comment on someone's match. On Twitch.tv, of course.

Congratulations, you have just plunged into an incredibly turbulent river with a steep channel and a bunch of pitfalls for the first time. Fortunately, all sorts of tricky questions such as broadcasting with a chroma key or commenting on large events can be dealt with later, but for now, you can deal with basic problems.

For example, by choosing the optimal bitrate.

We offer you a translation of the blog of one of the commentators of the Moonduck.TV studio Pimpmuckl, dedicated to the basic settings of the Dota 2 stream.

Bitrate is the most important aspect of streaming. We will try to squeeze out the highest possible quality for you.

The first step is to download a quality tester for your Twitch.tv traffic from the Team Liquid website. We launch the program and remove all regions from the test, except for the "home" one. A test across Europe, for example, would look like this:

Great, now we choose a server with the best bitrate / quality, take its test results and subtract 500kbps from them. We write down the resulting indicator as “maximum bitrate”. Abbreviated - max-bitrate.

If you managed to achieve the status of a Twitch.tv partner (for example, you are broadcasting a tournament), then you can safely set the same max-bitrate as the "maximum bitrate" already in the settings of your streamer program (for example, Open Broadcaster Software> Settings> Encoding ). One caveat: it makes no sense to install more than 3,500 kbps for the simple reason that Twitch will not let such a large stream pass.

If you don't have partner status, your ceiling is 2,500. Beyond this border, the stream will start lagging.

You should also think about your target audience. There are regions on the globe where everything is not as good with the Internet as in some Moscow, and if the main viewers of some local event do not have a connection faster than 2 Mbps, then it is better to “calm down” your stream to 1,500 kbps to save your audience from buffering.

Permission

We will set the video resolution based on the bitrate:

• 500-1,000 kbps: 480p 30fps
• 1,000-1,500: 540p 30fps
• 1,500-2,000: 720p 30fps
• 2,000-3,500: 720p 60fps
• 3,500: 900p 60fps

We remind you that we are talking about Dota 2 broadcasting. Don't even try to stream CS: Go or Overwatch 900p60 or any other FPS.

720p 60fps is the gold standard to strive for. And the quality will be good, and any device such as Chromecast will have no problems transmitting such a stream.

All sorts of little things

In the "advanced settings" of OBS, you can additionally play with the optimization. For example, tweak the x264 CPU preset by changing the default value from veryfast to fast. This, roughly speaking, will give an increase in video quality due to the greater load on the CPU.

True, in 90% of cases it is still better to stay on veryfast and play with this parameter only if your computer has a very good processor.

Another trick for GCN AMD CPU owners: set opencl = true in advanced settings, and the system will start working a little faster.

Nota bene: we never check the box opposite “Encode in Full Range”. Previously, this option was necessary, since some programs such as VLC could display colors incorrectly in the video, but now this problem does not exist at all.

If your computer does not have a very powerful processor, you can change the "Encoder" parameter. NVENC / AMD VCE load the computer much less, however, in this case you will have to sacrifice the quality of the video stream. It also makes sense to install Intel QuickSync: in terms of stream quality, this option is noticeably ahead of the previous two, but it still falls short of x264.

And finally, gamers playing on a map with a modified landscape (recall again, we are talking about Dota 2), on relatively weak PCs, it is better to switch to the standard map. The landscape from the latest The International's Battle Pass will stifle any streaming attempt on a budget computer.

True, if your hardware can handle this landscape, a video stream encoded with x264 will turn out to be even slightly better in quality. The point is that the Battle Pass arena itself is very light, and in x264, bright colors "weigh" less in bitrate / quality than dark tones. Accordingly, the "light" stream will look neater than the "dark" stream with the same bitrate.

However, my main choice is still the default landscape.

Open Broadcaster Software (OBS) is a free software for live broadcasting with ability to manage audio and audio sources.

You can download Open Broadcaster Software using this link: http://obsproject.com/download

Getting Started and Setup

After downloading OBS, navigate to your settings by clicking Settings> Settings or clicking Settings on the bottom right button menu. A screenshot demonstrating this is below:

"General" settings.

1. Select your language, and set a profile name.

Encoding Settings

1. Check "Use CBR"

2. Check "Enable CBR padding"

3. Max bitrate should be 3300 or 80% of your upload throughput, whichever is lower. Recommended bitrates for different resolutions are listed below.

4. Buffer Size is recommended to be equal to the max bitrate. Setting this lower will have the encoder closer to the targeted bitrate. We do not recommend changing this unless you know what you are doing.

• Recommended bitrate for 1080p: 3000-3500

• Recommended bitrate for 720p: 1800-2500

• Recommended bitrate for 480p: 900-1200

• Recommended bitrate for 360p: 600-800

• Recommended bitrate for 240p: Up to 500

Audio Encoding:

1.We recommend AAC with a bitrate of 64-128, although this is up to personal preference and bandwidth constraints. This is an example of what it should look like after you are done:

Broadcast Settings

1. Mode: Live stream

2. Streaming Service: Custom

3. Server: rtmp: / /live.site/live

4. Play Path / Stream Key:

1. Autoreconnect: Recommended checked.

2. Auto-Reconnect Timeout: 10 seconds

3. Delay: 0, but if you need delay set locally you can do this to prevent "ghosting." We do NOT recommend delay.

4. Minimize Network Impact: Unchecked. If you are an advanced user, or are having issues with your network settings, use this setting.

5. Save to file: We HIGHLY recommend you keep local recordings on your computer as we make changes to our VOD storage, to ensure you always have easy access to your broadcasts.

6. File Path: Select a file path for where you want to save your local files. Not needed if you do not save a local file.

7. Start Stream Hotkey: Custom key to start stream with.

8. Stop Stream Hotkey: Custom key to stop stream with. An example of what this would look like is below:

Video settings

1. Video Adapter should be set by default. If you have more than one, select the adapter you are playing your game on.

2.Base resolution typically is your monitors resolution. You can alternatively select a monitor to default this.

3. Resolution Downscale is the resolution that you send our servers. Lower resolutions will consume less bandwidth overall, and use much less processing power.

4. Filter should be "Bilinear" unless you have issues with blurring in your downscaling. Bicubic and Lanczos are both supported, but will take additional processing.

5. FPS is recommended to be 30. Note that 720p at 60 frames per second for some games will look better than low bitrate 1080p at 30 FPS.

6. Aero is recommended to be disabled only if you are using monitor or screen capture.

Audio settings

1. Desktop Audio Device: We recommend that this be set to your "Default" playback device. To change this, right click on your volume slider, then click playback devices. Then, right click on the audio device you "d like to make default and select" Set as Default Device. "Two images will show that process below:

2. Microphone / Auxiliary device: Set this to your headset or microphone if you have one.

3. Use Push to talk: Set this if you want push to talk set to a custom key.

4. Push to talk delay: Time after key is released and OBS is still recording your mic.

5. Mute / Unmute mic hotkey: User preferred hotkey to toggle mute settings for the Microphone / Auxiliary device.

6. Mute / Unmute Desktop Hotkey: User preferred hotkey to toggle mute settings for the Desktop Audio Device.

7. Force Microphone / Auxiliary to Mono: If you want this to only use one channel. We do not recommend this.

8. Desktop Boost (multiple): Force OBS to boost your desktop audio. 1 is "100%"

9. Mic / Aux Boost (multiple): Force OBS to boost your microphone audio. 1 is "100%"

10. Mic Time Offset (ms): Default 0. Use this if you have sync issues. An example of this page filled out is below:

Advanced tab

1. Use multithreaded Optimizations: Checked

2. Process Priority Class: Normal. Changing this higher will make OBS get CPU before other programs and can cause lag on many systems. Scene Buffering Time (ms): 400

3. Disable encoding while previewing: Unchecked unless you have lag while previewing your stream.

4. Allow other modifiers on hotkeys: Checked Video

5.x264 CPU Preset: This will set the encoding level. We recommend "veryfast" unless you have no bandwidth and beastly computer. Then, set it to be slower. Warning: setting your stream to a lower setting when at a high resolution is very CPU intensive.

6.x264 Encoding profile: This setting changes what profile you record on. Some devices (notably tablets and phones) may have issues with decoding streams with "high" profiles, so we recommend main if you want to have the highest compatibility at the sacrifice of some quality.

7. Use CFR: Checked

8.Custom x264 Encoder Settings: Default (blank)

9. Keyframe Interval: Set this to 2

10. Allow 61-120 FPS entry in video settings: Unchecked. We don "t recommend users going above 60FPS for any game.

11. Use Quicksync: If you have certain Intel processors (Sandy Bridge / Ivy Bridge), you can use this alternative method of encoding to use less CPU (it will use the hardware video encoder on your integrated GPU). There are quality differences due to the change of encoding. This nullifies x264 presets, but you can set the custom encoding settings if you want by checking "use custom x264 settigns for Quick Sync"

12. Use Nvidia NVENC: Similar to quicksync, this uses an alternative encoding method, with quality differences (usually lower at the same bitrate) due to the change of encoding. There are several presets you can choose from with this using the NVENC Preset dropdown.

13. Sub-options of Use Quick Sync and Use custom x264 settings for QSV should remain unchecked.

1.Force desktop audio to use timestamps as a base for audio time: Check this if you are having problems with syncing only.

2. Global audio Time Offset (ms): Set this to the number of ms you "d like to offset this to. We recommend 0 unless having issues with sync.

3. Use Mic QPC timestamps: Use this only if having sync issues.

Network

1. Bind to Interface: Default. You can select another network adapter here if you need to.

2. Automatic low latency mode: Check this only if you "ve talked to a OBS developer or Twitch staff as very few users would need this.

3. Latency tuning factor: Set this only if you "ve talked to a OBS developer or Twitch staff as very few users would need this. An example of what this would look like for a user is below:

Microphone noise gate settings

This setting allows users to set an automatic threshold for their mic being turned on and off. You can select the decibel level of the Close and Open thresholds here.

1. Attack Time: This is the time it takes for your mic to "spin up" to reach hold to output.You generally do not modify this.

2. Hold time (ms) how long the gate will stay open after it falls below threshold. You generally do not modify this.

3. Release time: Inverse of attack time. You generally do not modify this. Our recommended "off" settings are found below:

Now, you are ready to add scenes, and then sources to those scenes.

Scenes and Sources

Scenes and sources within OBS are fairly simple to add, and highly customizable. Generally, we recommend that you add as little dynamic content such as screen regions as possible due to the fluctuating nature of whats on your screen. Broadcasting a game directly or through a window is the recommended setup, although this is not compatible with every game or system.

1. To add a scene, right click the blank space under "Scenes" in the main OBS window, then click "Add Scene"

2. Enter a descriptive name such as "League of Legends in Game"

Next, make sure you have your scene selected, and right click the white space under the Sources header.

Source: Window Capture

1. Make sure that Aero is enabled, as this will not capture a specific layered window if not enabled.

2. Add a descriptive name like "Microsoft Paint"

3. Window: Under the window dropdown, select the correct program. In our example, it will be Untilted - Paint.

4. Innter / Outer Window: We want to select the header / title as well as whats inside the box, so we will select Entire Window. If you don "t want the outside edge of the window, select Inner window.

5. Capture Mouse Pointer: Check if you want a mouse pointer in the source.

6. Compatibility mode: This generally is for when you don "t want certain programs being caught with your stream. Leave this unchecked unless you have programs with 3rd party programs like Stream Privacy or performance issues.

7. Gamma: You can adjust the gamma of the scene. Recommended to leave this at 1.

8. Use point filtering: Use this if you are wanting to upscale your source in OBS only.

9. Opacity: You can set the transparency / opacity of this source in OBS. Sub Region:

10.Sub-Region: This is an option to capture only a certain portion of a window. Check this if you only want part of a window, such as the drawing portion. We will leave this unchecked since we want the whole window as previously noted.

11. If you do want it, you can select the region using the mouse pointer, or select the coordinates manually.

Color Key:

1. This is also known as a chroma key. Use this to select parts of windows based on their color.

2. Color: Select the color you want to select in your window

3. Similarity 1-100: Select the similarity. IE blue to sky blue, royal blue, turquoise, etc.

4. Blend: Select the sharpness of the chroma key edges. Finishing up and looking at it: Click "Ok". Click "Preview stream" on the bottom right button menu in the main OBS screen. Note that the window is not taking up the entire screen. Click "Edit Scene" on the bottom right button menu in the main OBS screen.

Example of our Paint Window Capture:

Source: Monitor Capture

Monitor Capture is a capturing tool that allows you to capture a monitor "s output. This is useful for quickly getting started streaming, but is generally not recommended outside of ease of use due to several security (information shown) and production issues. Do NOT use Aero when using this as a primary capture means, as it has significant performance drawbacks when using Aero.

1. Monitor: Select the monitor you "d like to capture, listed by number.

2. Capture Mouse Pointer: Check if you want a mouse pointer in the source.

3. Compatibility mode: This generally is for when you don "t want certain programs being caught with your stream. Leave this unchecked unless you have programs with 3rd party programs like Stream Privacy or performance issues.

4. Gamma: You can adjust the gamma of the scene. Recommended to leave this at 1.

5. Use point filtering: Use this if you are wanting to upscale your source in OBS only.

6. Opacity: You can set the transparency / opacity of this source in OBS.

Sub Region:

1. Sub-Region: This is an option to capture only a certain portion of a window. Check this if you only want part of a window, such as the drawing portion. We will leave this unchecked since we want the whole window as previously noted.

2. If you do want it, you can select the region using the mouse pointer, or select the coordinates manually.

Source: Image Slideshow

An Image slideshow is a set of pictures that change periodically. This is usually used for advertisements, although serves many different purposes.

1. Time between images (seconds): Time between changes in photos.

2. Disable fading: Check this if you want a cut to the next image rather than a fade.

3. Fade in Only: Uncheck if you want it to fade out and in.

4. Randomize: Check this if you want the next picture shown to be random.

5. Add Button: Add to select pictures

6. Remove Button: Remove images from queue

7. Move up: Change order of a picture upwards in queue

8. Move Down: Change order of a picture downwards in queue

A screenshot below demonstrates the slide show in action and the settings used.

Source: Text

Text sources are about what they sound like: text with manipulation.

1. Font: Font for your text

2. Color: Color of your text

3. Opacity: Transparency / Opacity of text

4. Scroll Speed: How fast text will scroll across the screen. Higher speeds are hard to read!

5. Background color: Set a background color for the text to be on (rectangular). Background opacity: Transparency / Opacity of background color

6. Use Outline: Set a outline for the text (not background box). Outline color: Color of the outline around text. Thickness: Thickness of the outline. Opacity: Transparency / Opacity of outline

7. Font Size: Set size of the text. We recommend larger fonts downscaled as opposed to upscaled small fonts.

8. Bold / Italic / Underline General text settings

9.Vertical: Place letters vertically stacked as opposed to horizontally across screen

10. Use custom text extents: Check this if you want your long text to be constrained to a smaller source size, or if you want your small text. We recommend this only for dynamic (scrolling or dynamic files) text. Size is size of extent in pixels. Wrap checked is if text wraps in side extent. For example: (start of extent) testing 123 testin (end of extent). Align: Align text to left / center / right. Only supported if using extent and wrapping.

11. Use text from File: Select this to have text come from a file. Useful for chatting capture or other dynamic data. This is an example (and properties) of a text source.

Source: Video Capture Device

These are webcams, DV cameras, and notably most capture cards will have output in this source.

1. Device: Select device here

2. Flip Image Vertically / horizontally: Flip image either mirrored down, or sideways

3. Deinterlacing: If your source has lines or computorized disortion in it, select this and a method. Methods are dependent on the video source, and vary widely. Top Field / Bottom Field first: Start deinterlacing from top or bottom. We recommend top generally.

4. Custom Resolution: Select only part of a camera or video source here.

5. FPS: We recommend 29.97 for most source inputs. This depends on your capture card or camera "s settings however. Please use native framerates.

6. Use output formats: Change how the video source is output. Some cameras have additional functions if set to certain output formats.

7. Use Buffering (ms): Buffer video. Recommended not turned on unless you are having major issues and are technically inclined.

8. Chroma Key: Use this to select parts of windows based on their color. Color: Select the color you want to select in your window. Similarity: Select the similarity. IE blue to sky blue, royal blue, turquoise, etc. Blend: Select the sharpness of the chroma key edges. Spill Reduction: Use this if you "re getting reflections off of objects onto normally not color keyed objects such as a background tinting someone" s hair.

9. Audio: Audio Input Device: Have a secondary audio feed from your video source. Generally recommended for capture cards if only if you are capturing console output. Not recommended on for video and webcams or PC output. Output Audio to stream only: Set it to capture internally to stream. Output audio to desktop: Externally capture it to your desktop audio devices.

10. Gamma: Set this to adjust gamma. We recommend playing with this to make sure that it "s correct, as video sources sometimes need adjusting vs other sources.

This is an example of a video taken from a super old DV camera connected via i1394 and it "s settings.

Source: Game Capture

This is a source that directly captures your video output. It is very efficient when using Aero, and only captures the game itself. If you have difficulty getting your game to capture, try to use 32-bit OBS as this sometimes will solve it. Note: You should not have to do this anymore, but we recommend to test before submitting any bug reports to them.

1. Select Application: Select the game you want to capture.

2. Use / set hotkey: Use hotkey to capture to capture current game. Some settings below alter how this is done. Stretch image to screen: Check if you want the game to be fullscreen on stream, but not on your own screen. Ignore aspect ratio: Ignore games aspect ratio. Capture mouse cursor: Uncheck if you do not want a mouse cursor. Useful for games played with keyboard only. Invert cursor on click: Invert the color of the cursor when you click something.

This guide is dedicated to the program Open Broadcaster Softwar e (hereinafter OBS) and its settings for streaming on Twitch.tv and Cybergame.tv... So let's start in order.
1. First you need the program itself OBS- to do this, go to the site http://obsproject.com/ go to the section Download and download the distribution kit. Install it following the instructions of the installer.
2. Run the program. And we will make the settings for the stream on Twitch.tv
2.1. Next, we need to go to the program settings - Settings -> Settings

2.2. In the menu that appears, we can change the Language, we can also immediately call our profile (Profiles are some kind of presets of settings, for example, you can create a profile for stream to Twitch in 720p quality, and create a profile for stream to Cyberheim in 1080p, and switch between them just with a couple of mouse clicks). First, let's create our first profile. To do this, you need to click in the window to the right of the inscription " Profile: "erase everything that is written there and write your name, for example I will write" 720p Twitch ", and press the button Add.

Let's also look at the steps required to delete a profile right away. When installing the program, a profile is automatically created for you " Untitled", now we will delete it with you. For this, to the right of the line" Profile: "there is an arrow down (drop-down menu) select a profile there" Untitled"and press the button" Delete".

2.3. Go to the " Coding". In this window, one of the most important settings for your stream is set, in most cases the picture quality will depend on them in dynamic scenes.
From 1st September Twitch.tv began to demand from streamers to set the Constant bitrate, respectively, we put a daw next to CBR (constant bitrate) we also check the presence of a jackdaw at CBR padding(in the absence - order!).
To stream on Twitch.tv with permission 1280x720 I would advise using a bitrate in the 2000-2500 range (at 2000 there will be a less clear picture, but less viewers will complain about freezes, at 2500, on the contrary, the picture will be of higher quality, but viewers may start complaining about more frequent freezes of the picture). For example, let's take something in between - 2200
Below we see the Audio settings, everything is simple, we put Codec: AAC and Bit rate: 128.

2.4. Broadcast... In this tab, we must select the broadcasting service and specify the channel key in it. In our case it will be Twitch.tv... So we expose:
Mode: Live
Broadcast service: Twitch / Justin.tv
Server: EU: London, UK(another one starting with EU is possible :)
Play Path / Stream Key (if available): here we must insert the key from our channel. To get it, you need to go to the Twitch website, create an account / log in and follow the following link http://ru.twitch.tv/broadcast On the right you will see the button " Show Key"

click on it, and copy the key that appears (starts with live_). Be VERY careful and copy the ENTIRE key, an error in 1 character will prevent you from starting the stream.
Auto reconnect: Checkmark
Auto-reconnect delay: 10(As small as possible, this number determines how many seconds after the fall of the stream, OBS will try to start it again.)
Delay (sec): 0(As a rule, the delay is set on the stream of Company or Special Battles, the delay is set in seconds, for example, to set the delay to 10 minutes need to write 600 )

Please note that OBS writes in red, this is just due to the new requirements Twitch.tv which entered into force on 01.09.2013. (below we will fix it)
2.5. Tab Video... Here we choose the resolution in which viewers will see our picture. V Base Resolution: choose Custom: and enter 1280 and 720.
Frames Per Second (FPS): put 30

2.6. Audio... Microphone and general sound settings. We choose playback device sound (usually this Loudspeakers) also choose Microphone if you want to use the Push To Talk system (so that what you said could be heard on the stream only when you press a certain button) then put a daw next to Use "Click to Talk" and to the right, select the window and press the button to which we want to assign this function (for example, I assigned it to Q)
NIG Delay (ms): 200(if viewers complain that the endings of your phrases often disappear, then you can increase this value (but do not overdo it, I advise you to increase it by 200 and conduct tests. For me personally, everything is fine and with a value of 200)
Microphone on / off hotkey and Hotkey On / Off sound- you can set hotkeys for these actions (they will mute the microphone and sound on the stream)
Application gain (multiplier): 1(this setting increases the sound of all applications, I advise you to leave 1, but if suddenly setting the sound to maximum in the game, the audience complains that they do not hear the sound, you can change this value (I advise you to add 1) (I am fine even with a value of 1)
Microphone gain (multiplier): 1(this setting increases the sound of the microphone, I advise you to leave 1, but if suddenly turning the volume of the microphone, the audience complains that they cannot hear you, you can change this value (I advise you to add 1) (I am fine even with a value of 1)

2.7. Advanced tab.
Multithreaded optimization checkmark
Process Priority Medium
Scene Buffering Time (ms): 400
Preset x264 CPU: Veryfast(for owners of heavy-duty processors, you can put faster or fast, I do not advise, tk. the load on the CPU will grow very much)
Keyframe Interval (sec, 0 = auto): 2(Twitch's demand)
CFR (Constant Frame Rate) checkmark
Match the sound to the timing of the video checkmark(there is a rare bug that the sound lags behind the video and this checkbox fixes it, one of our streamers came across this)

3. Settings for Cybergame.tv
3.1. We create a profile - for this we go to the tab Are common... to the right of Profile: write the name of the profile, for example: " 1080p Cybergame"and click Add.

Note! If you had a profile selected (for example 720p Twitch) and you create a new one, then it completely copies all the settings of the previous profile, and you only need to adjust it a little.

3.2. Coding... To stream on Cybergame.tv, you don't need to use CBR (constant bitrate) but we still use because we are streaming on Twitch.tv.
Max Bitrate (Kbps): 3700(For 1080p streaming on Cybergame.tv I advise you to use the bitrate 3500-4000 (since the service Cybergame.tv broadcast servers are located in Of Russia(at Twitch.tv coming in Europe) then the bitrate can be set higher, for example, if you stream 720p on Twitch - use a bitrate of 2000-2500, then for the same stream on Cybergame.tv you can use a bitrate of 2500-3000))
Audio: AAC - 128

3.3. Broadcast
Mode: Live
Broadcast Service: Custom
Server: In order to find out the server - you need to log in / register on the Cybergame.tv website - go to your account at the link http://cybergame.tv/cabinet.php select the "Channel" tab and copy what is next to Broadcast settings:(For example rtmp: //st.cybergame.tv: 1953 / live)
Play Path / Stream Key (if available): And here we copy what is next to Stream Name (Path):(but first you need to click the Show button to make the numerous asterisks disappear) usually starts with your nickname. (copy from the same page from which the Server was copied)

3.4. Video
since we plan to stream in 1080p then write to Custom: 1920 1080
Frames Per Second (FPS): 30

3.5. Settings Audio and Extended you can take exactly the same as for stream on Twitch.tv.

4. Settings Scenes and Sources
First, let's figure out what a Scene is and what a Source is.
A scene is a certain profile that contains one or more source (s). Those. for convenience, we can create scenes with the name of the games: "WoT" "WoWP" "CS", etc. and already in each scene its sources will be configured, for example, in the "WoT" scene there will be a source with game capture, a source with your webcam, etc. those. Sources are some kind of layers, and the source that is higher in the list will be in the foreground, and the one below will be in the background. Well, let's get down to business.
4.1. Initially we have Scene let's rename it to "WOT" to do this, right-click on it and select "Rename"

we write "WOT" Click ok. we get the scene with the name WOT
4.2. Next, let's add to this scene a source with a picture of the game. To do this, the game must be running!
Right-click in an empty window Sources: and choose Add -> The game

Enter a name, for example WOT.
A window appears. V Appendix: we should find our game in the drop-down menu : WoT Client
also put a tick "Stretch image to full screen" and "Capture the mouse" we press OK

Also in the sources you can add Slide show(several pictures periodically changing) Image(static picture or gif animation) Text(any text) Device(Webcam).
You can view the result of the picture by clicking on the button "Preview"

You will have a video with your layers. As I wrote above, the source that is above is in the foreground, and the one below is in the background. If you plan to overlay pictures / text on top of the game, then the game should be at the very bottom of the Sources list.

In order to adjust this or that layer (its size or position on the screen) - WITHOUT leaving the Preview mode, click on Scene change and click on the source you want to edit. A red frame will appear around the selected source, by dragging which you can resize the source itself. You can also move the source to any place.

We also see red "bars" that will help you adjust the volume balance between the microphone and the rest of the sounds (here I am not an advisor to you, this is very individual and needs to be coordinated with the audience.)

Well, the finish line, in order to start the broadcast - stop the preview and press Start broadcast.

It is very important that when streaming you do not have Loss of frames. If you have a loss of frames, then you may have problems with the Internet or you simply do not have enough of your channel for the current stream settings. Try to lower your bitrate.

Hyde prepared neRRReQuCb specially for viewers of ACES TV.

07. 09.2017

Dmitry Vassiyarov's blog.

What is bitrate? Or a tale about the quality of the video stream

Hello dear readers.

The topic of our conversation - what is video bitrate - will be interesting both to those who record videos to discs or upload them to the network, and to those who watch them. After all, the quality of the picture depends on this parameter.

In this article, you will not only get acquainted with the term, but also find out what types of bitrate are and what is its optimal value for video recordings in different situations.

Explanation of the term

Bit rate is used to calculate the number of bits that are contained in a second of a video stream. This concept is used to determine the efficiency of data transmission over a channel, that is, what its minimum size should be in order for a video to be played without delays.

For you to better understand what this term means, I will tell you about it without technical words. So, any video is a sequence of frames. For normal perception by the human eye, the optimal frame rate is 24 frames per second.

If, when recording video to the hard disk, you leave each frame in its original size, then there will not be enough space; not to mention how long it will take to put it online.

Let's estimate together: 1 frame of standard resolution 1920 x 1080 will weigh 2,073,600 bytes, that is, almost 2 MB. There are 24 such frames in 1 second - it turns out 48 MB. How Much Per Minute? Multiply 48 MB by 60 seconds - the size of a minute video is 2880 MB, which is practically 3 GB. What can you say about a 1.5 hour film?

The solution to this problem is to encode the file using codecs, that is, compression. Its degree and displays the bitrate, which is responsible for the optimal ratio of picture quality and video size. After all, if you squeeze it, then you will get an unpleasant graininess of the image, that is, the video will be light, but the picture is all in pixels.

Bitrate types

When compressing video, you are given a choice of 3 modes: constant, variable and average. Let's start in order:

• Constant bitrate (CBR). You set the desired value and it does not change throughout the entire video. The advantage of this option is that you know in advance what the final file size will be.
  But there is also a drawback, especially in relation to the sound. It can grow during playback, which may require changing the bit rate. Since he does not get what he wants, quality will suffer.

• Variable (VBR). In this case, you are working in tandem with a codec. Your task is to set the maximum bitrate, and the programs are to select the required value for each scene. Thus, the "minus" of the previous regime has been eliminated. In addition, the size of the file may turn out to be even less than expected, but it is impossible to predict the result.
• Average (ABR). From the name it is clear - this is a cross between the first and second modes. Here you set not only the maximum, but also the minimum bitrate, and the codec itself selects it within these limits, based on the dynamics of the video. Its quality is better than with the variable option, because the bit rate does not reach below the value you set.

Bitrate measurement

This parameter is measured in bits per second. Are you used to calculating in bytes? Be aware that there are 8 bits in one byte. If the number turns out to be large, the prefixes "kilo" (1 includes 1024 bit / s), "mega" (the same number, only kilobits), "giga" (the same number in megabytes) or "tera" (1024 gigs in 1 Tbit / s). Instead of the designation "bit / s", you can often find another variant on the Internet - bps.

The effect of bitrate on video quality is such that as the former grows, so does the latter. But keep in mind that as more bits are added, the file size will grow as well, since the codec does not have to compress the recording too much.

Average values

Of course, each file needs to be approached individually when setting the bitrate, but I will give you average examples:

• For uploading videos to YouTube or Vimeo, a value of 10-16 mbps is suitable.

• Want the best quality and average file weight? You can reach the bitrate up to 18-25 mbps.
• The maximum quality will be preserved if you set the number to 50 mbps.

Another important thing: the limit for recording a Blu-ray disc is 35 mbps, and the optimal figure for a DVD is 9 mbps.

How to set the bitrate correctly?

You need to rely on the value of the original version. For example, if the original was recorded at a bit rate of 10 mbps, then raising the value to 30, you will only achieve an increase in the file size, but the picture will remain the same.

Where can I see how many kilobits are there in a second of a video? We open its properties through the menu of the right mouse button.

Also keep in mind that lower video resolutions require lower bitrates.

We carry out counting

You can calculate the bitrate yourself. For example, let's say you're going to encode a 2-hour movie of excellent quality to burn to DVD. The volume of the drive is 4482 MB, and the length of the film is 7200 seconds. We carry out the calculation according to the following formula: (4482 \ 7200) x8x1000 = 4980 kbps.

You should also leave about 200 kbps for audio encoding and 100 kbps for creating a menu. In general, always cut the bitrate by about 7% for these tasks. It turns out that the optimal value in this case is 4700 kbps.

Don't want to bother with calculations? Use the Bitrate Calculator program. Moreover, there is a free version both for installation on a computer and online.

So you got to know the bitrate. Any other computer related questions? In our articles, you will find the answers you need.