

Fromthedeep
Members-
Posts
264 -
Joined
-
Last visited
Content Type
Profiles
Forums
Events
Everything posted by Fromthedeep
-
reported earlier jdam still inaccurate beyond the 5m CEP
Fromthedeep replied to Sinclair_76's topic in Bugs and Problems
Is the tag 'PM evidence' placed there by a CM? This behaviour is intended? -
Why so much negativity? A Phantard Speaks.
Fromthedeep replied to Aussie_Mantis's topic in DCS: F-4E Phantom
All of this sounds nice in theory but what you're doing is destructive flame baiting and when you show up to 'discuss the naval Phantoms' it will inevitable end up being a pointless mud slinging contest. -
Why so much negativity? A Phantard Speaks.
Fromthedeep replied to Aussie_Mantis's topic in DCS: F-4E Phantom
You stated your opinion and no one cares, you can move on. Heatblur already decided to deliver two E variants regardless of what you think about it. If you don't like it, post in some other subforum, no one here cares. -
Sure, I can understand that.
-
Very possible and if I did, I apologize. Can you please explain to me what you meant to say in a bit more detail?
-
F-4E INS alignment duration + nav system question
Fromthedeep replied to Leviathan667's topic in DCS: F-4E Phantom
You'd get roll/drift indication tabs and an analog range scale. -
How is what you outlined a 90% realistic solution? You argued for making a completely made up aircraft that has virtually nothing to do with the real jet aside from maybe the flight model. Without using the NATOPS, there's no way to get even the most basic systems into the game in an accurate manner, such as engine, fuel, lighting, flight control, hydraulics and electrical systems. Using a Tomcat LANTIRN or a Legacy Hornet DDI page is also fully made up. How would you make a navigation system with realistic limitations, symbology, capability and procedures if you just either make it up or copy it from a completely different aircraft? What you describe is a 9% realistic solution, not a 90%. DCS is supposed to be a realistic recreation of aircraft with their systems. If something is already unrealistic and/or simplified why would you argue that they should make other things unrealistic as well? This is not an F-22. Just because Heatblur may not be able to use certain documents legally that doesn't mean people wouldn't read it regardless, nor does it mean that users from the US can't read them legally. There are NATOPS manuals online. There are videos about the "MFD" that they use to control the missiles. There are videos about the TRAM system. It would take a few hours at most for people to find out and Heatblur's reputation would be entirely ruined if they ever did something like this. Whether you like it or not, the developers and the customers all expect to get as close to the real aircraft as possible within the legal boundaries. If a certain module cannot be recreated faithfully even in the most basic manner (and this is what's being suggested), I'd much rather not have it at all. Heatblur also agrees with this, that's why we don't have a fictional, made up F-14D in the game. If someone wants to have the A-6 because they like the idea of dropping bombs from a module that looks like an A-6, they can find other products that cater to that fantasy. If an A-6 is not possible to be recreated realistically, developers should choose a different aircraft to model. There's no point in a completely made up aircraft. The current modules do have aspects and certain systems that are classified or cannot be recreated, but the fundamental systems, system logic and MFD pages are close. Not perfect, but definitely close, aside from a few actually classified systems or systems that are too difficult to implement. If the A-6 truly is as sensitive of a platform because of export controls as it seems, the correct solution is to make an AI version and find a module that can actually be recreated at the proper level. If you guys truly believe that 90% of DCS users don't care about realism at all and they are find with using F-14 and F-18 systems in an A-6, then you'd be proven horribly wrong if something like this ever happened.
-
F-4E INS alignment duration + nav system question
Fromthedeep replied to Leviathan667's topic in DCS: F-4E Phantom
Even the old DSCG version will have dive toss, it's a much more sophisticated ground attack platform than the F-5. -
Exactly.
-
Probably the TRAM and later variants will be the new F-14B(U)/D when the Intruder releases.
-
I don't really see how this is a leap, the comparison was of rather limited scope, and from this it makes sense to me that even if all computers have the same processing power, the fact that the Super Hornet's systems can talk to one another and transfer data faster means it is a relevant advantage because the latter can utilize the fast hardware at its fullest extent. The fact that the Super Hornet is a more scalable platform means that the software that these components are running will also be constantly evolving at a much higher pace.
-
I've never said that the F-15E only has one computer. I've also never said that the F-15E didn't receive upgrades during its lifetime. As I've said, I'm not guessing about anything, there's a very extensive comparison made by a pilot who flew both platforms. And according to him, the Super Hornet does in fact have the advantage compared to the F-15E when it comes to data buses and similar aspects. If a person with flight hours in virtually every platform except for the F-35 considers that one platform is superior in one aspect, it's wise to listen to him. There really aren't that many people who can make a factual comparison. You know the Strike Eagle very well obviously, but what do you know about the Super Hornet compared to someone who works on it? Do you have enough knowledge about its systems, data buses, computing power, programming architecture and whatnot to make an assessment?
-
It has an RLG INS with Kalman filter, a software based INS-like program without filtering (mission navigator) and an EGI. With GPS available the use the EGI, without GPS or in old school scenarios the mission nav will be your primary source for navigation. The INS is still used for attitude data even in modern times.
-
The only people who actually know would be aircrew or other SMEs that worked on both platforms. There is such a description on Reddit by a test pilot who has flight hours in pretty much all relevant 4th gen platforms, including the Super Hornet and the Strike Eagle. According to their statement, the SH has much better fibre optic data buses, more roboust and easily scalable programming architecture and so on.
-
Isn't the model that we're getting pre ADCP1? The CC was about as fast as a 486. At that point the comparison is made to the Super Hornet and that's a drastically superior platform in this regard.
-
No, they fall an even less efficient flight path, they sag below the ballistic trajectory. The only LGBs in DCS that can glide are Paveway 3s, which are completely different weapons. PW2s are at their core 1970s era primitive technology that are cheap and effective but they require operator skill and know how to be effective.
-
It will try but it due to the inherent PW2 characteristics (lag, sag and energy loss) it's entirely likely that your bomb will land short if you don't do it properly. The basic employment consideration for PW2 against movers is to get a lead impact point (either manually or modern aircraft may automatically calculate it), lase it continuously and lead the laser spot just in front of the mover to account for the inherent negative characteristics. If you don't do these, your results will be very inconsistent. There's good reason why modern aircraft calculate the lead impact point and automatically utilize lead laser guidance on their own. The biggest consideration is lag. PW2 is a reactionary weapon, the laser detector is a disk with 4 quadrants and the control section is trying to keep the laser spot center. If it shifts away from the center, the bomb will compensate to recenter it but these can only take effect once the deviation already occurs. Therefore, the corrections are slightly delayed and since the bomb is chasing a constantly moving target (and since normal attack directions are from behind the target, the mover will be moving away from the bomb), in the endgame it's possible that the correction is delayed enough to land short of the target. If you lead the spot ever so slightly, the bomb has a much higher chance of still impacting the target even if the lag would result in a miss.
-
This quote seems to indicate that, but I could be misreading it.
-
Based on the previous input by Raptor it seems to me that the DCS Apache is not intended to be operated in a GPS denied environment in game.
-
There may be some miscommunication here I'm afraid. INS 'dumping' or actual INS failure is a malfunction that can happen to real aircraft. As I'm not an actual pilot, I've also never encountered that in real life, but there are plenty of first hand accounts of this happening. There's also a very harrowing tale in the Approach magazine about this happening to a Charlie Hornet pilot near the boat at night. This has never happened to me in DCS, but I haven't flown the Hornet in quite a while, which is why I asked if this is a new failure being simulated or a bug or perhaps user error. Hope this cleared up the misunderstanding.
-
It's when the INS fails entirely or "dumps" the alignment. This is a scenario when an actual in flight alignment may be necessary in most platforms.
-
The INS dumping was definitely not a thing that happened when I actively flew the Hornet. Is this a new feature (simulated failure) or a bug?
-
The DTIC document says it flies ballistically above 15 000. Is that incorrect?
-
There's a lot of confusion regarding the GBU-24. When dropped above the break (so above 15000), it bumps up then follows a ballistic trajectory until acquisition then it trajectory shapes using a g-bias guidance scheme, which is essentially going to put the path slightly above a pure ballistic profile and result in steeper impact angles. The publically documentation very clearly differentiates between the midcourse and the post acquisition guidance phases, which clearly indicates that there is no absolute need for the laser to be acquired prior to release. On top of that, there are different scan patterns that the seeker will utilize depending on the flight profile. If the seeker were to acquire the laser spot prior to release, why would it need to scan for it using either conscan or bar scan patterns? Keep in mind that delay lase vs continuous lase is a relatively simple decision matrix for PW2, for PW3, it depends on the particular attack as each target and each attack will have different specifics. Continuous lase also doesn't necessarily mean that the seeker will acquire the laser prior to release, it could simply mean that you can turn on auto lase prior to pickle and let the bomb pick it up as soon as possible once it gets in range because delay lase is not necessary. On the other hand, it's possible that you're doing a specific attack where delay lasing is necessary because you have very stringent impact velocity and impact angle requirements against a hardened target. In that situation, lasing too early could lead to a shallow profile and decreased penetration while lasing too late could lead to the bomb never acquiring the laser spot in the first place and flying off to God knows where. The point here is that delay lase vs continuous lase is a consideration when employing against hardened targets. If you have no stringent parameters to follow, the bomb will fly around in the appropriate profile and if everything works out fine, acquire the laser.
-
I've looked into the group you have mentioned and according to the administrator of the A-6 Intruder Archive, even the post 1978 NATOPS is flat out impossible for non US persons to acquire, and the TACMANs, performance charts, MTX manuals, FCF manuals, weapons checklists and all the data that would basically be required for a proper product is impossible to be exported. The same administrator also said that even non export controlled manuals should not be released because the data in them is close to the ones that are controlled. Based on this, I have very little faith that even an early A-6E is possible to be developed at a high enough standard while also complying with the regulations without acquiring an actual license by the appropriate regulatory agency. And according to Cobra, so far that hasn't happened. The F-14 is completely irrelevant, the time frame when the docs for that were acquired was essentially the period of a free for all, wild west scenario where anyone could easily buy what they wanted on EBay. On top of that, when initially asked, the HB devs said that all data needed for the Tomcat was available including plenty about the most sensitive areas. Considering that Cobra said the A-6 cannot even be started at this stage (due to the core DCS engine not supporting what they want to simulate) it's certainly possible that regulations regarding this data becomes more lax because even the most optimistic estimates shouldn't expect the A-6 in the next 5 years.