Question (part of) Resolved - Possible new audio stream on Amz

[skozyla]

Okay, is there any way to differentiate between these audio streams in AnyStream because dialogue boost sounds awful IMO!

I just watched an episode of Jack Ryan that I grabbed using AnyStream without realising I had selected a dialogue boosted audio track and it honestly sounds terrible. Sure, dialogue is left untouched but anything else, especially gunfire or explosions have no power to their sound. It's also quite jarring because one second the sound can be loud and the next it will be soft because the characters went from talking to fighting.

the 4 audio streams are all the same. there is no difference, except for how they sound. please read this thread where people clearly explain that no changes have been made and no new streams are available

They are all the same. Every one. It's just picked up from a different region by the scanner. One is 640kbp/s in different languages.

redfox explained it here. all 4 audio streams are the same language. the only difference is the degree of dialogue emphasis but they are the same still. If you want a different language you can pick it

So it is a possible manipulated audio track that differs from the original? I would not want any part of the dialog to be boosted. I guess the term "Night Mode" would be out of the question.

but all 4 audio streams are english? why would they be different? there is no indication that streaming services are using some sort of AI-powered system to produce alternative audio streams, so why would anystream change?

I know that. Which is precisely why I am posting here. Amazon's Dialogue Boost was applied to the audio that AnyStream downloaded. Instead of loud sounds being loud, they are now soft and sound like and only dialogue is left loud. This sounds very weird during action scenes because gunshots will sound very soft. I don't know how better to explain it other than to record the following example:

youtube is not relevant for this discussion. there is no issue as anystream clearly distinguishes between the audio streams based on language,channel configuration, codec and bitrate. there is no need for further stream identification or labelling as all streams identical

 

[DeepSpace]

as all streams identical

Yes, they are the same language. But they are NOT identical. You said it yourself, they sound different. But how should I now what I am going to download when AS does not label them correctly? I cannot see why now two people claim they are the same when they are clearly not and they should be filtered from the normal (original) stream. If there are four identical selections for four different streams, how can I know what they are when AS does not label them as even the script in #3 already does?

 

[BCloud]

I think you waste your time like me when I ask to get all the avalaible audio tracks in NF because it misses some tracks, I agree with you but I think you will wait a long time...
These kinds of discussion are endless and go nowhere each time.

 

[DeepSpace]

I think you waste your time like me when I ask to get all the avalaible audio tracks in NF because it misses some tracks, I agree with you but I think you will wait a long time...
These kinds of discussion are endless and go nowhere each time.

That's why I am now going to tag @Pete and @James (not sure if James is only for AnyDVD). Please fix this issue and label them correctly, as you already do with the descriptive track and the forced subs etc.

 

[cartman0208]

Why are you all making such a fuss about it?
That's just a minor inconvenience.
Stick with the first one in the list and you should have your original audio...

Another thing... how would the label look? The whole string would not fit in the dropdown box, even "Boost high" might be to long... so adding this would require resizing the download window and probably other things which in turn could lead to other issues.

Remember: not every new providerrs feature has to be reflected in AnyStream immediately.
Now that the dev's are aware, we should just keep it down.

 

[DeepSpace]

how would the label look? The whole string would not fit in the dropdown box

Not the drop down. In the left thing with the checkboxes. Because each version (like boost high) has multiple quality options that need to go in the drop down as they do for all other audios.

 

[BCloud]

Stick with the first one in the list and you should have your original audio...

I agree but you think as someone that already know the program and the issue, a new user uses AS and see 4 streams what the user chooses ? Of course a lot of workaround or external tools work here (like the forced subtitles) but when I use a software I generally don't need to do extra stuff or read the forum twice a day to make it work correctly. We just trying to suggest ideas to improve the software, it's not to annoy people/dev.

 

[DeepSpace]

Why are you all making such a fuss about it?
That's just a minor inconvenience.

You are right, at the moment I don't care at all as it appears to only affect english viewers. But RF1 thinks they are the same and no action needed, wich is not true. That's why I kept writing. I don't say he need to care about it, I only say he shouldn't just dismiss it because it is no issue to him.

Also I didn't know that the first one might be the the original track. If that is always the case, then yes, it is indeed only a minor issue. But I believe there are many people that don't know this forum, so it should still be fixed asap.

 

[bjjones]

Duplicates mean the same. Double or triple or quadruple of the same audio stream. Only one is in French, one is in English etc.

RF1, I respectfully disagree, when English is selected and it shows 4 640kbs streams available they are not identical other than they are all in english. I stumbled on this thread and had always wondered why there were 4 of the same thing to select so I just always chose the last one for no particular reason. In the case of the Jack Ryans the 4 are not in different languages, they are all english and having downloaded all 4 of them they do sound different. Playing them on my 2 speaker computer they sound the same in stereo but if moved to my media server and played on my 5.1 system there's a dialog change so I'm guessing it has to do with having a center channel maybe.

Not sure of a good way to distinguish them in the GUI if that's where this ultimatly goes, but for now I'm just making an assumption that the 1st of the 4 tracks available is the "original" for whatever that means and luckily that's the one that the current AS processing auto selects, the 1st of the 4 gets pre-highlighted in the dropdown

 

[Jaaaamie]

I am worried I have downloaded some of these with the alternate audio tracks. I download that much that I can't remember. What films and shows have this on? Is it a small percentage? Is there a way to check?

Sent from my SM-S918B using Tapatalk

 

[psig33]

I am worried I have downloaded some of these with the alternate audio tracks. I download that much that I can't remember. What films and shows have this on? Is it a small percentage? Is there a way to check?

Sent from my SM-S918B using Tapatalk

Same here. I don't usually mess with it since it defaults to 5.1 640kbs (when available) but I do recall a few times I did by mistake.

If the movie or show has the additional dialog boost tracks they will be listed in the the details section.

Audio languages
English Dialogue Boost: High, English, English [Audio Description], English Dialogue Boost: Medium

Not all will have all three dialogue boost, like the example above which only has two.

------------

I downloaded the dialogue boost high, medium and the "original" track from Scarface and compared them. The top track is the first track listed in AS. The channel layout is FL, FR, C, LFE, RL, RR. As you can see, there are clear differences in the levels of each channel between each track. This difference should be enough to warrant naming each track with the corresponding name. Keep "640 kbps DD+ 5.1" for the original track and something like "640 kbps DD+ 5.1 DBH/M/L" or just "640 kbps DD+ 5.1 H/M/L" for the others. I doubt adding those extra characters will break the layout, seeing how "192 kbps AAC-LC stereo" is already longer. I don't think naming different tracks is asking for much...

but all 4 audio streams are english? why would they be different? there is no indication that streaming services are using some sort of AI-powered system to produce alternative audio streams, so why would anystream change?

Per Amazon:

How Dialogue Boost works​

Dialogue Boost analyzes the original audio in a movie or series and intelligently identifies points where dialogue may be hard to hear above background music and effects. Then, speech patterns are isolated and audio is enhanced to make the dialogue clearer. This AI-based approach delivers a targeted enhancement to portions of spoken dialogue, instead of a general amplification at the center channel in a home theater system.

 

[Jaaaamie]

Same here. I don't usually mess with it since it defaults to 5.1 640kbs (when available) but I do recall a few times I did by mistake.

If the movie or show has the additional dialog boost tracks they will be listed in the the details section.


Not all will have all three dialogue boost, like the example above which only has two.

------------

I downloaded the dialogue boost high, medium and the "original" track from Scarface and compared them. The top track is the first track listed in AS. The channel layout is LF, LR, C, LFE, RL, RR. As you can see, there are clear differences in the levels of each channel between each track. This difference should be enough to warrant naming each track with the corresponding name. Keep "640 kbps DD+ 5.1" for the original track and something like "640 kbps DD+ 5.1 DBH/M/L" or just "640 kbps DD+ 5.1 H/M/L" for the others. I doubt adding those extra characters will break the layout, seeing how "192 kbps AAC-LC stereo" is already longer. I don't think naming different tracks is asking for much...

View attachment 73650

Thank you so much!

Sent from my SM-S918B using Tapatalk

 

[el Filou]

Per Amazon:

Just as FYI, for the people wondering why Amazon is doing that when Dolby Digital already has a dialog normalization feature: AC-3 & E-AC3's dialnorm is just metadata that indicates the average dialog level, so the dynamic range compressor in every official Dolby decoder can work with that and ensure that dialog is always at the same level. The problem is that if you have sound other than dialog on the center channel, which can and does happen, that sound can still be as loud as the dialog even with the DRC on the 'midnight' setting, as the decoder has no way to differentiate the two sounds.
Besides, if the center speaker level is manually set higher than the calibration measure by people who want to hear dialogue better, the other sounds on the center channel will also be louder.

Interestingly, Dolby AC-4 has integrated that feature, as it can have dialogs on a separate stream that is mixed with the ambient sound channels by the decoder at playback time, so in addition to saving bitrate when providing multiple languages it can also mix dialog louder for people who have difficulty listening to voices (example, look at the Audio card: https://www.digitalbitrate.com/dtv.php?mux=HEVC&pid=300&live=1&sec=0&lang=en).
Unfortunately, AC-4 needs new decoders that not every device currently has, so in the meantime providers like Amazon who want to offer that option have to offer multiple different full audio streams with dialog mixed differently.

 

[Balthazar2k4]

Was watching Joe Pickett today and ran across something I had experienced before, but chalked it up to the program I was watching at the time. I downloaded the show through the AP Paramount+ channel in CBR/192kbps. Noticed throughout the third episode that much of the audio was very dialog focused and was straight up compressing loud sounds from other channels. In other words it was being dynamically compressed. So, I thought maybe it was the 192k track and re-downloaded the same episode in CBR/640kbps. Went to the same scene and sure enough the same problem exists. I decided to try the episode through Paramount+ and what do you know... it was correct. All channels were balanced. I decided to experiment a little further and used a competing program to download the show through the AP Paramount+ channel with the 640k audio and much to my surprise it did not have the issue. It seems AnyStream is grabbing the new Dialogue Boost version of the track. Is there any way of correcting this?

 

[psig33]

Take a look at the following thread: https://forum.redfox.bz/threads/possible-new-audio-stream-on-amz.87151/

It is assumed that the first track of each group in the dropdown list is the "original" and the others are the dialogue boosted tracks. I have not tested this myself.

Currently this issue is being treated as silly and not necessary to label each track with the correct name ...

 

[RedFox 1]

Currently this issue is being treated as silly

Did someone tell you this? Or is it just a guess on your part? There are 2 other threads on this subject. I read them both, I don't remember anyone saying that. Please tell me who said it.

 

[psig33]

Did someone tell you this? Or is it just a guess on your part?

A guess on my part based on the responses in that thread

 

[RedFox 1]

A guess on my part based on the responses in that thread

So you are telling other members what we think , based on a guess on your part. Do I have this correct?

 

[psig33]

So you are telling other members what we think , based on a guess on your part. Do I have this correct?

"We" as the RedFox team? I did not mention anyone, but it is not hard to come to that conclusion when you read the thread I linked to. Sorry if that offended you, it was not my intent ... Anywho, l just hope this is issue gets resolved and AS+ keeps getting better

 

[Balthazar2k4]

Take a look at the following thread: https://forum.redfox.bz/threads/possible-new-audio-stream-on-amz.87151/

It is assumed that the first track of each group in the dropdown list is the "original" and the others are the dialogue boosted tracks. I have not tested this myself.

Currently this issue is being treated as silly and not necessary to label each track with the correct name ...

Yep, that's it. It is incredibly annoying and will force me to watch streams in their respective apps for now. Spending tens of thousands on my audio system to have it hamstrung by DRC is a no-go for me.