reported GM EXP incorrectly blacked out

[Arctic Fox]

In Wags' latest video on the F-16's A-G radar, the GM EXP mode is shown as blacked out on the area directly in front of the nose of the aircraft, as when doppler beam sharpening modes are in use. From my understanding this is not correct behaviour: EXP should simply zoom in the normal ground map and doesn't rely on doppler variation to display the image, so it should not be affected by the angle of the scan in the same way.

For example, in this video you can see the transition from NORM to EXP works just fine with the cursor in front of the nose of the aircraft, and the offset maneuver is only performed when they transition to DBS:

https://youtu.be/qg1Ojydzv8U?t=1308

Additionally, at least in the MLU M1 manual, when DBS modes are in use the radar image should gradually revert to real beam resolutions when at too low of an angle from the aircraft's nose, rather than blacked out.

Edited October 17, 2021 by Arctic Fox

[Frederf]

Yeah, a few oddities and notes about the video:

• The display is not range/azimuth (B-scope) as verbally described. Instead it is true position (PPI). If it was was range/azimuth then all horizontal slices would be equal range and all vertical slices would be equal azimuth.
• MAN/AUTO as shown are backward. When "MAN" is displayed the system is in manual mode.
• It is possible to manually change range in automatic scale despite what is verbally suggested. Changing scale by the arrows perform a range change and simultaneously kick scale mode from auto to manual.
• SEA is actually less sensitive than GM mode due to a filter. Its purpose is to suppress sea state clutter which would otherwise make GM difficult to use over rough water. Over calm seas this clutter rejection is unnecessary and GM mode is preferable to SEA.
• EXP is a 4:1 expansion of the normal radar image with no improvement in resolution.
• EXP does not have a black zone nor use Doppler info. It is simply a dumb "zoom" like blowing up a JPEG in Photoshop.
• Resolution improvement and the geometrical expansion of the display are independent concepts. EXP is geometrically 4:1 resolution 1:1. DBS1 expands 4:1 but resolution 8:1 (I think this is correct for v5/v7, 24:1 for APG-66v2?). DBS2 expands variable with resolution 64:1 compared to NORM/EXP.
• DBS also does not have a black zone. DBS uses a technique which sharpens the image in the azimuthal axis using Doppler data. The technique simply produces no improvement over the standard NORM/EXP radar data near the nose but it is displayed all the same.
• The expansion tick marks on the DBS1 crosshairs are missing. They are showing the area that will be displayed in DBS2 (variable) in much the same way as the NORM tick marks describe the view area of EXP. There would be tick marks shown in EXP but since DBS1 is the same size the tick marks would be at the edge of the display already.
• Quality difference between RBM and EGM is perhaps exaggerated. EGM narrows the radar beam in azimuth which helps azimuth resolution about 4:1. Range resolution would be the same. EGM can have some undesirable effects in some situations.
• Good to mention that the range knob on the throttle has partial control of the radar gain in conjunction with the gain rocker.
• Freeze shows the ownship white circle flying over the expanded display as if it's not an expanded display showing a false relationship between ownship and radar image.
• Manner description suggests that SPI set by radar requires freeze and/or TMS forward. SPI is always at the FCR cursor crosshairs regardless of entering any kind of track mode or use of freeze. Natural FZ and any sort of designation as shown would also work.
• The radar antenna continues motion in freeze.

I'm curious about designating a FTT from FZ. That's a new one to me. The radar isn't radiating so it can't really track with real radar beams. I would expect that would either do nothing or kick the radar back into radiation to perform a FTT. I'm interested to see how antenna elevation control changes things in GM.

I know it looks like a lot of whining and nitpicking but really the overall impression was very good. That's a lot of work and ED should be proud.

[falcon_120]

I wonder if A2G radar technology used for the F16 is exactly the same as on the hornet, or if improvements have been made along the way.

Just saw the video and I got the impression that Exp resolution on the airfiled was incredible; clearly discerning aircrafts and surrondings, have not been able to get so good resolution with the APG73 radar of the hornet in DCS. It could be just my memory playing tricks on me ofc. I'm looking forward to receiving the new A2G mode for the viper in the upcoming patch, that way we will be able to compare apples with apples.

Edited October 18, 2021 by falcon_120

[BIGNEWY]

13 hours ago, Arctic Fox said:

In Wags' latest video on the F-16's A-G radar, the GM EXP mode is shown as blacked out on the area directly in front of the nose of the aircraft, as when doppler beam sharpening modes are in use. From my understanding this is not correct behaviour: EXP should simply zoom in the normal ground map and doesn't rely on doppler variation to display the image, so it should not be affected by the angle of the scan in the same way.

For example, in this video you can see the transition from NORM to EXP works just fine with the cursor in front of the nose of the aircraft, and the offset maneuver is only performed when they transition to DBS:

https://youtu.be/qg1Ojydzv8U?t=1308

Additionally, at least in the MLU M1 manual, when DBS modes are in use the radar image should gradually revert to real beam resolutions when at too low of an angle from the aircraft's nose, rather than blacked out.

 

Based on documents we have and sme feedback it is correct as is. 

Thanks

For the other comments can I suggest waiting for the patch and if you think something is wrong post 1 bug per thread and include track replays and any evidence. 

Thanks

[TLTeo]

9 minutes ago, BIGNEWY said:

Based on documents we have and sme feedback it is correct as is. 

 

Uh, interesting since the video above clearly shows different behaviour. Could this be due different variants of the APG-68?

[OperatorDerp]

1 hour ago, BIGNEWY said:

Based on documents we have and sme feedback it is correct as is. 

Thanks

For the other comments can I suggest waiting for the patch and if you think something is wrong post 1 bug per thread and include track replays and any evidence. 

Thanks

Want to share those documents or atleast snippets that prove your point? The above mentioned is quite literary video proof. 

[BIGNEWY]

Just now, OperatorDerp said:

Want to share those documents or atleast snippets that prove your point? The above mentioned is quite literary video proof. 

let me know which block viper is in the video. We have to be very careful with video evidence. 

[BIGNEWY]

update:

The team will investigate further doppler blanking in EXP.

Thanks for the PM's

[CZQS]

Why is there a black area in the middle when the radar is zoomed in when flying directly on the target? Can a professional pilot explain it?

[Raptor9]

4 minutes ago, CZQS said:

Why is there a black area in the middle when the radar is zoomed in when flying directly on the target? Can a professional pilot explain it?

It's due to the requirement to use Doppler Beam Sharpening to provide a fine picture with high detail and resolution. But this can only be done when not flying directly at an area, which is on the nose. I would just google it to read up on it, since a quick google can provide graphics as well.

And most of us will probably provide a slightly less than accurate or clear explanation anyway.

Edited October 21, 2021 by Raptor9

[CZQS]

7 minutes ago, Raptor9 said:

这是因为需要使用多普勒波束锐化来提供具有高细节和分辨率的精美图片。但这只能在不直接飞到鼻子上的区域时才能完成。我只是谷歌它阅读它，因为一个快速的谷歌也可以提供图形。

无论如何，我们大多数人可能会提供稍微不太准确或清晰的解释。

 

oh,thank you.I will look for the information myself.

[Frederf]

36 minutes ago, CZQS said:

Why is there a black area in the middle when the radar is zoomed in when flying directly on the target? Can a professional pilot explain it?

I don't think it's actual realistic. I've read that DBS has degrading sharpness from 15° off the nose to equal to normal processing at 5° off the nose. It looks to be a sort of copy-paste from the F-18 radar which shows the same thing. But I can explain why it might be the like that due to the theory and physics of Doppler beam sharpening (DBS).

Normally, the radar is gathering information which are points of: azimuth, elevation, range, and gain. This places a return in space and tells how bright it is. The result is not that high resolution because the radar beam is rather large and fuzzy. Mapping with it is sort of like doing surgery with a toilet brush. But, there is a technique which can use extra information from the return to artificially sharpen the image using the fact that the airplane is moving, DBS.

In addition to azimuth, elevation, range, and gain there is a fifth piece of information the radar can use: Doppler shift. When the radar pulse hits the ground and reflects its frequency is increased according to how fast the ground is coming at the airplane. This effect only depends on the radial velocity or component coming toward or going away. Coming straight at you the frequency is increased, going away decreased, and sideways no change. Because the radar is looking at different angles relative to the velocity vector of the airplane the amount of frequency shift should correlate with various angles of radar.

During DBS image processing the computer looks at the frequency of the returns in different directions and compares it to the frequency it should have based on that look angle. When the frequency of a return doesn't match what it should be the computer nudges the pixel in the (azimuth) direction it should have come from. The idea is that the original angle information was imprecise and wrong and the nudged pixel is closer to what it should have been. The result is a sharper image, at least in the tangential direction.

That sounds great so why the poor (or non-existent) performance near the nose? The DBS technique relies upon the relationship between angle and frequency. The graph of angle vs. frequency should have a steep slope so variations in frequency can be turned into good angle corrections. The graph of angle vs. frequency is a cosine curve which has good steepness centered around 45° but poor slope at 0° and 90°. The color difference between 44.999° and 45.001° is large compared to the color difference between 0.001° and 0.002°. Near the "flat spots" of the graph the DBS technique vanishes in performance so that means close to the nose and if the radar could scan that wide near 90 degrees off axis it's useless.

DBS has another issue besides poor to none performance along particular directions: what if the object illuminated is moving? A car driving down a highway, a tank moving, helicopter blades spinning, trees swaying in the wind. All of these will provide frequency shift that's incorrect for its angular position. The DBS processor will look at the color of the return and try to fix its position based on that color. This nudge will often smear out moving objects until they are completely unnoticeable or often offset from their true position. A radar image of a highway might have half the cars shown as driving on the dirt on one side of the road and the other half of the cars on the other side. In a particular geometry of highway, cars might have a purely radial velocity in which case they would be unaltered by DBS which only works perpendicular to radial.

Edited October 21, 2021 by Frederf

[CZQS]

1 hour ago, Frederf said:

我认为这不太现实。我读到 DBS 的锐度从离鼻子 15° 降低到等于离鼻子 5° 处的正常处理。它看起来是 F-18 雷达的一种复制粘贴，显示了相同的内容。但我可以解释为什么它可能是这样的，因为多普勒波束锐化 (DBS) 的理论和物理学。

通常，雷达正在收集以下信息：方位角、仰角、距离和增益。这将返回空间并告诉它有多亮。结果不是那么高分辨率，因为雷达波束相当大且模糊。用它映射有点像用马桶刷做手术。但是，有一种技术可以使用来自返回的额外信息，利用飞机正在移动的事实人为地锐化图像，DBS。

除了方位角、仰角、距离和增益之外，雷达还可以使用第五条信息：多普勒频移。当雷达脉冲撞击地面并反射时，其频率会根据地面到达飞机的速度而增加。这种效果仅取决于接近或远离的径向速度或分量。直奔你的频率增加了，离开的频率减少了，横向没有变化。因为雷达相对于飞机的速度矢量从不同的角度观察，所以频移量应该与雷达的不同角度相关。

在 DBS 图像处理期间，计算机查看不同方向的返回频率，并将其与基于该视角的频率进行比较。当返回的频率与它应该是的频率不匹配时，计算机会在它应该来自的（方位角）方向上轻推像素。这个想法是原始角度信息不精确和错误，并且微调的像素更接近它应该的样子。结果是更清晰的图像，至少在切线方向。

这听起来很棒，那么为什么鼻子附近的表现不佳（或根本不存在）？DBS 技术依赖于角度和频率之间的关系。角度与频率的关系图应该有一个陡峭的斜率，这样频率的变化就可以转化为良好的角度修正。角度与频率的关系图是一条余弦曲线，其以 45° 为中心具有良好的陡度，但在 0° 和 90° 处具有较差的斜率。44.999°和45.001°之间的色差比0.001°和0.002°之间的色差大。在图表的“平坦点”附近，DBS 技术在性能上消失了，这意味着靠近机头，如果雷达可以在离轴 90 度附近扫描那么宽，它就没有用了。

除了沿特定方向的性能不佳甚至没有性能外，DBS 还有另一个问题：如果被照亮的物体正在移动怎么办？一辆汽车在高速公路上行驶，一辆坦克在行驶，直升机的叶片在旋转，树木在风中摇曳。所有这些都将提供与其角位置不正确的频移。DBS 处理器将查看返回的颜色并尝试根据该颜色确定其位置。这种轻推通常会抹掉移动的物体，直到它们完全不明显或经常偏离它们的真实位置。高速公路的雷达图像可能显示一半的汽车在道路一侧的泥土上行驶，而另一半的汽车在另一侧。在高速公路的特定几何形状中，

 

Wow，Compared to F18,i feel F16 is better. Why is that?

[Furiz]

42 minutes ago, CZQS said:

Wow，Compared to F18,i feel F16 is better. Why is that?

It has been reported already

 

 

[llOPPOTATOll]

6 minutes ago, Furiz said:

It has been reported already

 

 

They are looking into the incorrect Doppler shift in EXP, not DBS as a whole afaik