Bouben

Nice to know you are aware of it. Thanks.

Automatic AI healing would need a carefully thought-out set of conditions when to allow and when to disallow AI healing. Easy conditions like "heal only if safe" are probably a matter of course. Healing in combat? That is a different story. Sometimes you don't want them to heal because you need them to run or whatever. Those things would have to be solved in order to have it working properly.

Excellent. Wish you the best of luck!

You will have to go in details in order to make your opinion relevant.

No. Audio card is not the culprit in audio spatial issues (unless you have one that is really malfunctioning or has buggy drivers). You have to have a decent headphones/decent speakers (correctly positioned), sufficient room acoustics (if you use speakers) and you have to make sure you disabled any software audio FX in your audio driver. Most of those in-box FXs are cheap gimmicks that do more harm than good and have no use for consumers. ---------- Post added at 17:22 ---------- Previous post was at 17:17 ---------- If I am not mistaken, there is no actual shrapnel simulation in Arma 3. It is simply radius-related damage stuff. Correct me if I am wrong.

Cool, I am glad we have found a consensus :-) @ RobertHammer EDIT: undecided about JSRS and Arma Vanilla

If you talk to me, you completely missed my point. I was talking about quality of youtube videos in general and its compression format sound quality. I guess it is quite clear that if you upload garbage to youtube, it will sound like garbage. Anyway, 1) 192 MP3 is completely fine sound quality for deciding if a sound is generally good or not from a sound-design/mixing standpoint (even 128 could be enough in many situations for rough decision making). You will miss some subtle highs and some sonic nuances here and there but that is about it really. A shit-sounding sound in 192 will still sound shitty in lossless. 2) Even mono sounds carry a lot of fundamental portion of information from sound design/mixing standpoint (but it is true that some transient-heavy stereo sounds can sound drastically different in mono). 3) 80hz HPF is nothing drastic that would really change the first impressions of a overall quality sounding sound (a lot of "consumer" people don't notice/hear these frequencies anyway or have no appropriate system for playback). So it is, again, perfectly usable as a reference for a rough decision-making from a sound-design/mixing standpoint. 4) There is not much information above 17.000 Hz and many people don't even hear/don't notice that high. So, again, nothing that would drastically change the reference from sound-design/mixing standpoint. 5) The only relevant point you have posted is the "average camera or mobile phone" as it is very likely these will have a really curvy overall frequency response and will dramatically affect the final emotion you get from the sounds it recorded. Yes, bitrate (up to some point or rather down to some point) is one of the least important things when considering sound quality from sound-design/mixing standpoint. I am completely serious. Your color reference is completely missing the point (not to mention that even very old paintings with really washed-out/damaged colors are still very valuable from artistic standpoint as they still show the most fundamental parts of the artist's skill and his/her work). Just pick a bitrate/samplerate VST and apply it to some audio. You will have to use really drastic reductions in order to really change the sounds and their fundamental design qualities. To sum it up. The most important quality factors from sound-design/mixing standpoint are: - overall frequency spectrum of a sound - macro and micro dynamics of a sound (and its transients) - length of a sound

Youtube videos can have perfectly acceptable audio quality that would be undistinguishable in blind tests for 99.9% listeners. I would like to ask local "audio engineers" and "audiophiles" to stop overreacting about it and to stop spreading myths. Bitrate (to some point) is one of the least important things when deciding about production quality and design of a sound.

Large dynamic range is very problematic in games. You need some control over the dynamic range in order to make sound useful and safe. Classic music has a big dynamic range and you cannot really enjoy the quiet parts unless you play it really loud or you have an exceptionally quiet environment in your room. Not really practical and safe for peak/transient-heavy action game sounds. Big dynamic range is not about "audiophilia" only. It is much more complex problem than most "audiophiles" realize (most of them know nothing about sound in general).

That would be nice. However, units should move in the normal position in such circumstances and take a knee only when stationary, as you said.

Yes, and AT guys should be more specific when out of rockets. "Out of AT ammo" would be more informative.

No, it does not necessarily have to be the basic volume of sounds. It can be a masking prolem in combination with poor TS audio quality (bad mics and noise floor, bad TS EQ making voices muffled, lot of echo in user's room leaking into the mic...). In order to have reliable voice readability in the game, you have to have a limited dynamic range (both for the game and chat audio) to have a predictable loudness to work with and a lot of EQ involved so that the game sounds don't mask the main frequency band of human voice frequency spectrum. In real life, in combat, there is so much masking and noise occurring that it is often impossible to hear properly what the radio or the squad-mate is saying. That means that clear readability of sounds in such situations is an unnatural feature requiring unnatural audio processing (and yes, this is the sound-design problem we might be talking about). It is a very complex mixing task and TS audio limitations don't help it at all.

The stereo width should be fully dynamic, if possible. In the distance, the source sound should be very narrow with a very wide echo around the listener. From close distance, the source sound should be wider and echoes not too prominent.

In which state I should examine the sound? Close helicopter or distant...?

Speaker placement can, of course, cause a distinct lack of bass because it could cause nulls in your room and your sweet-spot. EDIT: Anyway, my main point was to be careful with making "definitive" statements as there is a lot of variables with sound. That is all. Please, point me to the helicopter and reproduction steps so I can check the sound. And yes, I would actually like to see the frequency spectrum. Thanks.

I would not want to be an audio engineer for a realistic milsim game where its community demands hollywood-like sounds :-) Also, I would be very careful about describing specific frequency problems unless you have a well acoustically treated room, correct listening position and accurate monitoring. Just saying...

"Improved: Commander-gunner communication (commander will assign targets faster)" This is a worrying change as it will probably make crew-manned weapons even more overpowered then they currently are. For more info, see this ticket: http://feedback.arma3.com/view.php?id=22531

Yes, it would be nice to have this sorted but I suppose it is not easy.

You could be surprised that various seemingly non-AI-related mods indeed have "something" to do with the AI. It can be a config of a weapon, for example, that AI has to use. We appreciate your effort to debug stuff. Just to be sure, please, try to reproduce it without mods at all. Thanks, mate.

You seem to be using some mods. Of course, I don't know if they affect your AI but you should always test without mods at all.

I want to congratulate you, Ryd. Great job. I have always considered you to be one of the key members of the community. I am glad it is now "official". :-) Cheers!

Guys, remember that there is also the EQs and not just faders. EQs can be dynamic, too...

Exactly, guys. Proper AI logic is one thing. Believable, non-immersion-breaking visual feedback is another thing.

Oukej, I just want to thank you and all the guys programming AI that you took this action and started to work on it again more visibly. Very excited to see what comes next. Cheers!

And then you took a shot and AI immediately knew where you, the sniper, are. ---------- Post added at 01:39 ---------- Previous post was at 01:36 ---------- Ha ha, I expected this to happen. This is going to be difficult to get right. Also, suppression should not be only about aiming abilities. It should, most importantly, force units to take cover.