When can we talk about the Ka-50?

[fudabidu]

We've seen issues with the Ka-50 for years now and after testing something I think it's really urgent we talk about it.

A while ago I was having issues with DL targets being saved with wrong altitude information (approx +30m), even when received from an AI wingman. Couldn't remember the exact mission, so I just took a wild guess and ended up with "Clear Tkvarcheli". It wasn't the mission I was looking for, but I did find problems. Oh boy ....
 

Quick mission summary:
Take off, turn E, destroy single AH-1, destroy ground targets

Attempt 1:
I kill the AH-1 with a pair of Vikhrs. No issues.
I then send my WM to scout ahead 8km, exactly the range to the targets. Perfect weather, 8:30am, targets already in sight. WM goes NOE, doesn't spot a single target and gets shot down.

Attempt 2:
AH-1 gets hit, trailing black smoke and goes into a steep dive. AH-1 surprisingly continues the attack and knocks out some of my systems. Attempt aborted.

Attempt 3:
AH-1 coming straight at me, AA mode on, laser on, targeting box right over the target, target clearly visible, "lock" ... I'm getting a 6km range but no tracking. How often do we have to ask for a Shkval update???
After killing it I realize my WM went RTB and shut down for absolutely no reason.

Attempt 4:
AH-1 gets hit only slightly and breaks lock by maneuvering hard. I reset the Shkval to cancel the drift, re-enable AA mode, uncage, laser on, clear line of sight, way above any obstacles. I completely 'miss' the AH-1 with my first locking attempt, hit the ground behind it (good range information) and the Shkval just goes nuts.
Eventually I send my WM to scout. He goes NOE which means he no longer has LOS to the targets, but he somehow sends me DL targets anyways (paused and checked in F2).

At least all those targets were stored with the correct altitude, so it might depend on the area of the map, but I'm done with testing for the next few weeks.

Let's recap:
- AI has no sense for self-preservation
- AI scouting is completely unreliable and more of a "kill wingman" function
- AI can randomly break and RTB
- AI LOS is inconsistent at best
- Shkval tracking range is still artificially limited which can make it trash even in seemingly excellent conditions
- The Shkval can start drifting randomly regardless of any obstacles and with valid range information
- Even SP track files can have massive desyncs

Yes, ALL of the track files were broken. Shkval looking in weird directions, missiles going dumb, AH-1 crashing when it shouldn't have ... To say that I was disappointed would be like saying water is slightly moist. I would have loved to provide all the tracks.

If resources are limited and the Hornet, Viper and Hind are priority, that's cool. I understand. Not even reacting properly to long known issues I don't, unless you want people to think the Ka-50 has been abandoned.

So, what's your take on the current state of the Ka-50 and what kind of genuine improvements are planned for BS3?

"thanks"

[Volk.]

TBH I've got both Republic & Oil War campaigns, but have yet to try them after especially the deployment campaign. It'd be one thing if the AI weren't necessary in the campaigns, but the way they were designed, their fuel, munitions and scouting were definititely considered as a given in the campaign design. But as it stands, their AI is lacking as it. And while DCS doesn't use contrast (yet) as method of locking targets for the Shkval, the artificial limitations from lighting weather fx and time of day are excessive from what you can see on the IT-23 vs. what it can actually lock. Combined with the other choices they made with making it single-seat autoamtion & helo dynamics, it really doesn't help selling the very competent Ka-50 to new players.

 

BS3 news, even if just saying it's months away, would be welcome, though the patches needed to the introductory campaigns, AI behaviour and Shkval limitations are additional issues.

 

Make no mistake, I love the helo and the FM, but it needs a LOT of selling to newcomers believing it's all borked.

[Taz1004]

You can read the patch notes to see what the dev's are working on.  It's normal that programmers will work on one area (module) at a time as shifting focus is time consuming.

Problem is that in expanding game like DCS, unless they expand their staff, they will always work on new content than go back to old content.  I'm afraid things like AI will not be fixed until BS3.

 

The community is capable enough to fix most of these issues if ED would open little more of the SDK and relax the scope of IC.

[fudabidu]

1 hour ago, Taz1004 said:

You can read the patch notes to see what the dev's are working on.  It's normal that programmers will work on one area (module) at a time as shifting focus is time consuming.

Problem is that in expanding game like DCS, unless they expand their staff, they will always work on new content than go back to old content.  I'm afraid things like AI will not be fixed until BS3.

I understand how software development goes, but thanks for reminding me.
2.5 released years ago and we still have issues with DL targets, AI, broken missions, Shkval, ... you name it. You think this is acceptable if we don't even get feedback from the developers?
Just to pick one example: was there any reply addressing the hardcoded Shkval slew axis deadzone? NO! Every 'veteran' with an analogue TDC-stick knows about it, it's easily reproduced by just flying the Ka-50, but ED chose to ignore it. It has been an issue from the very start (12 years ago) and we still have to deal with it. Good luck finding an excuse for that.

 

2 hours ago, Volk. said:

BS3 news, even if just saying it's months away, would be welcome, though the patches needed to the introductory campaigns, AI behaviour and Shkval limitations are additional issues.

Make no mistake, I love the helo and the FM, but it needs a LOT of selling to newcomers believing it's all borked.

I'm already telling everybody the Ka-50 is borked beyond believe and nobody should buy it in its current state. It would be a great module if you weren't required to have a list of "bugs" at hand just to see if you did something wrong.

Thanks to the (lack of) communication this is what I predict for the future:
1. the Ka-50 will get some new toys, but long standing issues will not be addressed
2. every module older than ~4 years will be neglected in favor of new modules, especially helicopters (despite advertising Syria to be "perfect for helicopters")

Where is ED now to tell me I'm wrong? They are in the F-16 forum section, telling people to properly align the INS or boresight the MAVs. Amazing support!

[Taz1004]

3 minutes ago, fudabidu said:

You think this is acceptable if we don't even get feedback from the developers?

 

Huh... completely missed the point but oh well.

[3WA]

I agree, Fudabidu.  They seem more interested in coming out with new modules, (i.e. mo $ ), than fixing their World or existing aircraft, which have MANY bugs.  I can understand them working on the F-16 and F-18 right now, because that will add more functionality to the "World", and they have been promised for many years.  However, I want the old "World" and the old modules FIXED.    I want an engine that runs on more than one core.  I want an engine that actually utilizes the GPU.  I want AI that does more than just sit there.  I can't in good faith recommend this "sim" to anyone at this point, either.

 

But for now, it's what there is.

 

I'm hoping that someday, someone will come out with a flight sim on a modern engine.  As I don't think DCS is ever going to do this.

Edited December 4, 2020 by 3WA

[dfpoor]

Let's talk gunships.  The game is not broken.  It's not borked.

Last night I flew a mission.  After 5 minutes, an Apache shot me down.  I flew another sortie and changed tactics.  A huge melee was taking place between 6 Apaches and 3 Hinds, red and blue armor, and 2 red mobile SAMs.  Using terrain as cover, I engaged the closest Apache at 2.7nm.  My first missile missed but the second scored.

https://youtu.be/9c57Mw_hcQ4

I lost all SA engaging the Apache, so I quickly egressed to safety.  After taking a deep breath I realized 4 Apaches had been killed and 2 were RTB with damage.  The coast was clear.  I next flew a few km and engaged 2 M1 tanks.  My first kill was at 3.2nm.

https://youtu.be/Dqt7ZzE7i3w

Blue armor was making a breakout 8km to the south.  I flew halfway and engaged 3 more M1s.  My last kill was at 2.8nm.

https://youtu.be/ybwnbHOywQc

An M2 was flanking rapidly to my right.  My first concern was Blue tanks, I was going to let friendlies deal with the Brad.  But then I realized there were MTLBs in the area and they were going to get toasted.  I ordered the MTLBs to hide in some woods.  I searched for the M2 in the Shkval but couldn't find him.  I finally spotted him visually, he was farther right and closer than I was expecting.  As I was turning to face him, the laser attack warning went off.  Fearing the TOW, I dived wildly into terrain and increased speed rapidly.  I'm almost certain I saw the puff of smoke when the Brad launched at me.  I flew so wildly I almost hit a tree and house.  I thought I might have spoofed the missile but suddenly I died.  Sadly, Tacview recorded my death but not the TOW.

https://youtu.be/ft5b_8W71PE

While my mission was a failure, it was tons of fun.  You can fly, fight, and win in the Kamov.  When you RTB winchester, and you have destroyed 2 platoons of armor, and your efforts have saved the battle for Redfor, it is a very satisfying feeling.

The lists of things which work in DCS is miles long.  The list of things which are broken is short.  ED is a boutique publisher.  I don't think these Russian programmers are making 100K.  I don't think they are making 50K.  ED is a minor miracle.  We should be thankful for what they have given us.

The KA50 is the only hardcore gunship simulator in existence.  Hopefully, we will get an Apache, Cobra, Eurocopter, Alligator, and Havoc eventually.  Let's hope and pray the HInd is a success.

 

[Scofflaw]

I wish they would at least let the community fix the campaigns.  First mission of the deployment campaign, and your AI partner can't even do half of the introduction mission without crashing into the ground for no reason.

 

I would and have fixed some of these things in the past, but each patch I have to reapply my changes.

[fudabidu]

On 12/5/2020 at 4:27 AM, dfpoor said:

The game is not broken.  It's not borked.

Sorry, but you just don't know any better yet.

I played the same mission 4 times 'as intended' (or: like most people would approach the mission) and ran into several issues we've known about for years. Some issues even date back to the original BS release in 2008 (like the analogue Shkval axis deadzone). Ask some experienced Ka-50 pilots and they will all tell you similar things...or maybe just browse the forum for 30 minutes.

The Ka-50 is fun when it works. Running into some of the same issues on nearly every flight is not. Getting no response about these issues is slowly driving people away.
But please continue telling us how amazing the module is.

[Mars Exulte]

It's got some issues, yes. I've been flying it quite a bit and encountered some of them. Calling it ''unusable'' is just being a drama queen, but I get this is the social media oriented world we live in. Drama and clicks @@

Edited December 8, 2020 by zhukov032186

[Cheetah7798]

Having flown the Ka-50 as my primary module for the best part of the last two years, and occasionally flew it for many years prior. Here's a comprehensive list of of 'bugs' that I've encountered which are consistently present:

1) Most recently, the Vikhrs pylons don't depress, requiring whole-aircraft depression. (I'm on stable, may have changed in beta)

2) Manual inputs in the PVI-800 don't account for altitude.

3) Datalink targets are (very) occasionally stored incorrectly. I've witnessed it once in the last couple weeks.

4) Shkval lock is archaic compared to other DCS modules.

5) Shkval can occasionally drift.

 

I'd consider the AI issues as something altogether separate, as well as the issues with the campaigns, which for the record, are from a time prior to the Caucasus remaster.

 

Despite this, the module is functioning as intended, albeit a little dated compared to more recent modules, as can be expected; and any other bugs I've missed must be so insignificant that I can't recall them readily. Now, despite owning numerous other modules, be they western, eastern, fixed-wing or rotor-craft, I keep coming back to the Ka-50, because I genuinely find it among the most entertaining and capable of the lot.

 

It is as far from unusable as the majority of other modules.

[Volk.]

Cheetah - one of the more recent OpenBeta fixed the Vikhr reticule not dropping and external-view launchers not lowering - that was a more recent bug and has been fixed (in Openbeta).

As for the Shkval drift - if there's no laser ranging then the ground stabilisation/internal tracking doesn't kick in, so in certain flight regimes it will drift. If you laser lock it generally stays on target unless you run the target outside the gimbal limits, or bank/yaw too hard. That said it may still lose a lock on a moving target, especially a fast mover - though that doesn't seem like a bug. If I recall you can't store datalink targets without a laser-lock (ie. correct ranging), but if you could, or you corrupting the lasing by refiring it early then that datalink store would be innaccurate. Other than that the ABRIS is supposed to be accurate to ~20m, though I've read if you have insufficient satelites it may be less accurate - haven't tested that.

The Shkval itself isn't on par with other modules with a similar locking mechanism - that's just what it is. That said they could fix the curves/deadzone on the axis slewing and make it proper contrast lock system. Otherwise try the helmet to lock things up at certain ranges.

 

Can't disagree with your other statements.

[Taz1004]

28 minutes ago, Cheetah7798 said:

I'd consider the AI issues as something altogether separate, as well as the issues with the campaigns, which for the record, are from a time prior to the Caucasus remaster.

 

I would disagree with this.  Somehow we are expected to treat campaign and AI as separate element but they are important part of the game.  If this were to happen on any other game, there's no doubt this is considered bug.  In racing sim, if a career mode is broke so you can't finish a race, wouldn't that be a bug just because the car handles and drives well?  If the AI are broken so they constantly run into you and crash you, wouldn't it be a bug?

Edited December 9, 2020 by Taz1004

[Cheetah7798]

37 minutes ago, Volk. said:

As for the Shkval drift...

I should have specified, Minor drift around the point of lock I don't consider a bug; Merely accumulated tracking errors. The more radical ones, like suddenly tracking some trees in the foreground, or sudden, substantial changes to range data strike me more as a bug than INS error.

 

My understanding is, even with a laser lock, the ranging information is instantaneous, since the Ka-50's laser doesn't like continuous usage. That is, after the range has been found and the Shkval aligned, it enters the normal INS stabilisation that it would be in with any other contrast lock. It's accurate at first, but I think we've all seen it suddenly return some ridiculous range reading after either minor slew adjustments or helicopter attitude changes. Sometimes even flying straight and level triggers it. Just the other day, I'd locked (via laser) a piece of terrain near some infantry and upon closing with the target, the range suddenly spiked form 2.5km to 8+km. This seems a bit too drastic to be some INS error accumulation. It's just never really bugged me, since re-ranging will immediately fix the issue.

 

44 minutes ago, Taz1004 said:

I would disagree with this...

Let's agree to disagree, then. The way I see it, DCS is an entirely different beast to Racing simulators or anything else, really. The work goes into making an accurate, simulated replica of the real aircraft, more than anything else. The campaign and AI, while important additions, don't come close (in my eyes) to the importance of the model's implementation as an isolated factor. Though, I'm willing to say my view might be a little skewed, as I'm quite proficient with the mission editor, and don't mind getting involved with the lua, when required. Content in terms of the campaigns feels like a gimmicky addition, more than anything. Enjoyable when they work; Not the end of the world when they don't.

 

However, more to the point, the reason I didn't want to associate the AI and Campaign problems with the Ka-50 is because they are not unique to the Ka-50. They plague every module, so dissuading people from buying the Ka-50 because of a global DCS issue seems dishonest to say the least.

[Volk.]

Cheetah, I have found that if you refire the laser (ie. lock with laser) after slewing while the laser is already firing, like trying to tap the lock button in trying to lock up a fast moving target, it's almost guaranteed to corrupt the range, usually coming up short once the laser stops firing. Think my second vid on sniping & locking shows this better than I can explain in words. No idea if that's a bug or simulation of the real thing, given re-firing/ranging isn't protocol from the manual.

The Range also gets funky if you scroll 'upwards' and it goes past the 15km limit - then it keeps going into infinity. I've also seen it do this over time with tracking a helo at 6.5km - eventually it lost sync and kept increasing the ranging. Also unsure if bug or what it's supposed to do. As far as I've read there's isn't INS error accumulation/drift in the K-041/PVI-800 in the sim at the moment (i.e. like with the Tomcat where even without violent maneuvers your stuff becomes misaligned), though they did code the alignment procedures on the panels.

Lastly, it's only accurate when you've ranged for a distance - ie. pressed lock. At any other time, barring HMS usage the laser isn't firing and it's just guessing the range - same with pressing laser reset - from then on you'll see the number adjust subtly as it adapts that number, guessing it by angle I guess, but it's rarely exactly accurate unless you laser-lock again - that I believe is correct. Whether it should overshoot / undershoot to a crazy number, I don't know, but then again not all this tech from 20-40 years ago was accurate/solid on every use.

[DthFrmAbv]

I think there is no DCS module which "really works". We have to live with that and I don't expect it to ever change.

 

What really annoys me in the KA-50 is Shkval's which just don't lock or having a clean lock and the Shkval decides to make a nice little trip to wherever but not to the target.
Had one of those just yesterday. Static target (truck), about 3km distance, fired Shkval but it just went down in some -30 degrees from where fired, hitting the ground.
Unlock, lock again, fire, Shkval doing the same...

[Volk.]

@DthFrmAbv outside of extreme bank/pitch throwing out the lock there are two other events I can think that might do that - if you have a Target Point (Nav TGT) lit up on your PVI-800, or Datalink set to DL Ingress lit up, then when Uncaging the Shkval will always default to trying to target that point first... Although you're describing Locking/Designate not Uncage, so not sure what that's about.

[Cheetah7798]

11 hours ago, DthFrmAbv said:

...fired Shkval but it just went down in some -30 degrees from where fired...

Sounds like you messed with the automatic tracking switch and were attempting to use rockets. The Ka-50 has an odd-ball CCIP/loft calculator for bombs and rockets, achieved by placing the automatic tracking switch in the down position, lasing a target and selecting a ballistic weapon. It pans the Shkval down 'x' number of degrees to predict the impact point.

 

Lock:

 

Auto Tracking off:

 

Select rockets (S-13 in my case)

 

Note that the Skhval automatically pitched down to the point indicated by the left box highlight in the last image. Aligning the (now fixed) shkval with the target will provide an approximate firing solution.

Edited December 11, 2020 by Cheetah7798

[DthFrmAbv]

That's quite interesting, thanks.
 

But no, I only hat Shkvals with me and was attacking a (static) truck convoi. I already had killed three or four when those Skvals failed.
I then switched to another target which I could destroy without any issues.

 

I have seen this before in similar situations. The reason is unclear to me.

[3WA]

Lol, I think you mean Vhikers ( the actual laser homing missile ).  The shkval is the screen.

[Cheetah7798]

You know what, in hindsight that makes a lot of sense. I read "Fired Shkval" in DthFrmAbv's message, thinking he meant firing the laser.

 

In that case, if you're having trouble with your Vikhrs falling short, or going nuts, here's the list of things to check (Feel free to ignore if I'm just telling you what you already know):
- Laser overheated (Requires aircraft repair):

Text appears on the bottom-centre of the HUD (Forgotten what it said, exactly. It's been a while since I've seen it).

Attempting to lase something results in the cool-down timer (Bottom-left of HUD) appearing then disappearing immediately.

 

- Poor Alignment of the reticles:

If the Vikhr's reticle isn't close to the Shkval reticle (within ~1-2 deg) their attempt to re-enter the laser beam after launch can be so harsh as to shoot right through it and off into infinity. You want to slot the Shkval reticle right in the centre of the Vikhr's reticle if possible. Do note, while the Vikhr pylons do pitch down, they don't yaw, so you'll have to manually rotate the platform.

 

- Out of range:

If the range to the target is 8+ km, then there is the chance that they lose control authority just before impact (~8.5-9km) or drop off the Shkval screen entirely (9+ km). This largely depends on altitude of launch, speed of launch platform and whether or not you're manually compensating for the drop, etc.

[Volk.]

Responding to @DthFrmAbv's issue specifically on the Shkval dropping down when attempting to lock - I might have experienced this once before - just trying to lock up civilian cars. Somewhere along the line (usual automatic tracking etc switches) it seemed to 'hop' back to a previous target, rather than lock the new thing. I thought it might have had to do with the civvy vehicles often despawning and respawning, and maybe sharing some kind of behind-the-scenes target ID. That said, I didn't pay enough attention to record that, and I don't know if I had a target point brought up or DL ingress on, though that should only affect uncaging, not locking.

Edited December 13, 2020 by Volk.

[fudabidu]

BIGNEWY ... since you're having an eye on me now, how about some answers?

Shkval tracking range, Shkval axis response, broken missions and terrible AI we all know about. Are we really supposed to provide track files, even though we've known about this for years? What kind of new knowledge are you expecting to find?

You're always asking about tracks, but for a while now ALL my tracks are broken, SP included. So, what am I supposed to do? Not post any reports at all, because I can't provide a track?

How about a pinned announcement saying there is no point in reporting anything, because that's how it is. Maybe also add a line that BS3 is on hold.

[BIGNEWY]

Please report the issues one per thread with a track replay.

 

Track replays need to be short and to the point if you want to report any issue, I work with track replays all day every day and most of them work. 

 

If they are from heavily scripted servers they are more likely to create problems, so if a problem can be reproduced in single player it is better. 

 

 

 

 

 

[Taz1004]

7 hours ago, BIGNEWY said:

Please report the issues one per thread with a track replay.

 

Track replays need to be short and to the point if you want to report any issue, I work with track replays all day every day and most of them work. 

 

If they are from heavily scripted servers they are more likely to create problems, so if a problem can be reproduced in single player it is better. 

 

I think what everyone would appreciate is a bit of transparency.  Why are these being delayed?  Some things we're asking for are really quick changes like rocket recoil.  Is everything on BS2 on hold too until BS3?  But instead of explanation, I do get the feeling you're just asking for track files on issues everyone knows about to delay working on the Shark.  This probably isn't true.  But at least that is how it feels when there's no transparency.

 

Below was such issue.  It's something so prevalent that anyone who ever flown a mission encountered it.  But you acted like you never heard of it and asked for track file.  Which is actually somewhat insulting.  But I spend time to provide the track file anyway.  Then nothing.

 

 

 

Edited December 30, 2020 by Taz1004