CM Black Sea - Beta Battle Report - US/UKR Side

[Bil Hardenberger]

Scott, this is an AMAZING AAR.. outstanding work!!   

 

Who knew a big galute like you could write?  

[John Kettler]

Bil,

 

Since you're here, and pnzrldr isn't, despite this being his thread, does that mean the game has been called? Seeing your striking avatar was enough to make me question whether I was on the US/UKR thread, but the remarks you made confirmed I was. Still absorbing this, though.

 

Regards,

 

John Kettler

Edited February 3, 2015 by John Kettler

[Buzz]

Big Galute can write.

I agree... an "AMAZING AAR.. outstanding work."

Solid contenders make for a great narrative... on both sides.

Thank you Scott and Bil.

Great ride and this AAR was only... round one 

Thanks,

Buzz

Edited February 3, 2015 by Buzz

[pnzrldr]

I am in fact here, but have been tied up with a variety of tasks.  Have taken time to do some posts on other topics, but just could not compile the energy to keep after this.  I will make the effort to finish but not immediately.  As reported on Bil's side, we were forced to call the scenario due to technical difficulties with the older beta build we were using.  Bil conceded defeat, though after inflicting more casualties.  Think I am down 3 Brads from where I left off.  However, Bil is down several more BMPs including one that Brytva 21 took out at near point-blank range!  Will have to post vid from Russian side, because he couldn't save that turn and send it back to me!  Thanks for your attention here, and the encouragement.  Will make the effort to compile as a .pdf, and may do something similar front to back.  Was toying with the ID of doing this sort of write up for a scenario against the AI, as it would turn faster.  We'll see.  

 

While this scenario does not appear in the release, look for it in modified form on the 'new' repository, coming soon to a BFC.com near you!  

[womble]

Thanks for your work on this, Scott. Looking forward to seeing the last episodes. Your and Bil's AAR styles were, I think highly complementary to each other.

[Michael Emrys]

Yes, Scott, thanks for taking the time and effort to provide an interesting account. I know it was a labor of love.

 

Michael

[dan/california]

You can get up early to play for a couple of days, a couple of weeks gets to be difficult.

[Vinnart]

Too bad this couldn't play out to the end, but that is the nature of beta sometimes. Thanks for taking the time it takes to put one of these AARs together Scott. It was entertaining to follow.

[LUCASWILLEN05]

Any chance of this scenario being made available in the repository please?

[pnzrldr]

Yes.  Give me some time to work it.  Bil and I had discussed.  It is a scaled down version of one that I think actually made the release copy, and we had chatted about some of the tweaks we would make to the force mix and reinforcement schema to make it more playable.  I'll see about getting after it this weekend.

[DasMorbo]

That would be great!

[LUCASWILLEN05]

The big version is good too. I am a megalomaniac. I don't recognise any scenarios from the release version though - and I have played most of them :-)

[pnzrldr]

Nah, just found out it didn't make the release.  I am working w/ Bil to modify each side a tad.  Then have to cut up the AI somewhat.  Map is the same.  Probably take a little bit though.  Will keep you posted.

[Hydaspes]

Gents,

 

Pleased to report that CM:BS Beta is up and running on the new rig, and the performance is... astonishing. Smooth like butter, even with all the setting (minus anti-alias) hard to the right. I don't do FRAPS for screen capture, but rather Bandicam. Had a weird issue w/ FRAPS on my old rig, so switched and like Bcam so far and will load it up tonight. Sent Bil his latest turn late yesterday evening. Should be able to start getting caught up on posts tonight. Never considered just how MUCH stuff I had to install to get it set up to do this! Drivers, CM:BS, BCam, PS, Office, zip utils, Adobe reader, etc... just takes time.

 

Buzz - I had heard that the 4K stuff starts to get into the arena of 'human eye can't see the difference anymore.' Since you obviously disagree, I'll go on a quest to Bestbuy or someplace similar over the holiday to take a look. I was going to do another Dell just to have paired set, but will see if I think the difference is worth the expense.

i have the same configuration and game is running bellow 30 fps or 40ish  unless i put it on Fastest models, and on best models,its unplayable  (textures are not important) , also i can put highest levels of AA in nvidia CP and it doesnt reduce performance since this game is heavy on CPU, i even unparked all cores

 

https://www.youtube.com/watch?v=IzaMs5iaX7w

Edited September 26, 2015 by Hydaspes

[alwaysfish]

i have the same configuration and game is running bellow 30 fps or 40ish  unless i put it on Fastest models, and on best models,its unplayable  (textures are not important) , also i can put highest levels of AA in nvidia CP and it doesnt reduce performance since this game is heavy on CPU, i even unparked all cores

 

https://www.youtube.com/watch?v=IzaMs5iaX7w

No, the game is not heavy on CPU. The reason we get low FPS is because rendering engine is purely optimized. There are many places in the rendering engine where CPU is just waiting for GPU to finish its job, which in turn decreases overall rendering performance. Most of the rendering is done utilizing old fixed OpenGL rendering pipeline (in other words, programmable vertex and fragment shaders are not used) which doesn't scale well with increasing power of GPUs. CM performs better on NVidia mostly because NVidia has multi-threaded OpenGL drivers. I am sure many of us, including me, have observed long map loading times on AMD GPUs. It appears, that mipmaps (smaller textures of original texture) are generated much much slower on AMD GPUs than on NVidia GPUs. Nvidia GPUs handle mipmap generation better because (it's only my guess) it's done on another thread, which in turn means less waiting time for CPU and faster map loading times.

There are many ways how BF could improve their rendering engine to get higher frames per second and deliver better user experience. Rather than create lots of intermediary blending textures for terrain, create just one atlas map. Convert all rendering to programmable pipeline. Use vertex objects, texture arrays, etc. NVidia has made many presentations where OpenGL performance could be improved 10 or even more times if some optimizations were applied. 

Maybe all this sounds a little too technical, however i just wanted to make clear, that CM is not heavy on CPU and lots of optimizations could/should be done on the rendering engine.

[Hydaspes]

No, the game is not heavy on CPU. The reason we get low FPS is because rendering engine is purely optimized. There are many places in the rendering engine where CPU is just waiting for GPU to finish its job, which in turn decreases overall rendering performance. Most of the rendering is done utilizing old fixed OpenGL rendering pipeline (in other words, programmable vertex and fragment shaders are not used) which doesn't scale well with increasing power of GPUs. CM performs better on NVidia mostly because NVidia has multi-threaded OpenGL drivers. I am sure many of us, including me, have observed long map loading times on AMD GPUs. It appears, that mipmaps (smaller textures of original texture) are generated much much slower on AMD GPUs than on NVidia GPUs. Nvidia GPUs handle mipmap generation better because (it's only my guess) it's done on another thread, which in turn means less waiting time for CPU and faster map loading times.

There are many ways how BF could improve their rendering engine to get higher frames per second and deliver better user experience. Rather than create lots of intermediary blending textures for terrain, create just one atlas map. Convert all rendering to programmable pipeline. Use vertex objects, texture arrays, etc. NVidia has made many presentations where OpenGL performance could be improved 10 or even more times if some optimizations were applied. 

Maybe all this sounds a little too technical, however i just wanted to make clear, that CM is not heavy on CPU and lots of optimizations could/should be done on the rendering engine.

then lets say its utterly not optimized and that i duuno how did the guy to whom i replayed managed to get the game to play 'smooth like butter'

[alwaysfish]

'smooth as butter' is very subjective. FPS, on the other hand, is an objective performance measure. I would bet that 'smooth as butter' meant having around 40 fps, which is really low from my point of view 

Edited September 26, 2015 by alwaysfish

[Vanir Ausf B]

Yes, i don't notice the FPS until they drop into the teens. 

[Hydaspes]

'smooth as butter' is very subjective. FPS, on the other hand, is an objective performance measure. I would bet that 'smooth as butter' meant having around 40 fps, which is really low from my point of view 

i can easily notice difference between 60 and 40, even 60 and 50, not to mention 60 and 29 or 22. this game is almost unplayable for me but its so good that i am pretending that its ok, i duuno what he meant with smooth as butter, in my mind its at the VERY LEAST steady 30 fps

[HUSKER2142]

Hi. I am have question for mechanized platoon in M2A3 Bradley.  In standart platoon 4 Bradley - that means 28 landing sites  . Standard platoon has 31 people. Where are still 3 seats ?

 

Привет. У меня вопрос по поводу механизированного взвода на M2A3 bradley . Во взводе 4 машины , что значит 28 десантных мест . Но при этом во взводе 31 солдат . Откуда берутся ещё 3 десантных места ?

[AttorneyAtWar]

Hi. I am have question for mechanized platoon in M2A3 Bradley.  In standart platoon 4 Bradley - that means 28 landing sites  . Standard platoon has 31 people. Where are still 3 seats ?

 

From the manual:

 

"Note: We have artificially increased the passenger capacity of the M2 Bradley to accommodate a full rifle squad. In reality, a full rifle squad cannot fit in a Bradley, so there is a complicated cross-loading routine"

That should answer your question.

[HUSKER2142]

From the manual:

 

"Note: We have artificially increased the passenger capacity of the M2 Bradley to accommodate a full rifle squad. In reality, a full rifle squad cannot fit in a Bradley, so there is a complicated cross-loading routine"

That should answer your question.

I figured it out . But as 31 soldiers are placed, with 28 seats.Maybe the "Fire support team" is moved by special vehicle Hummer ?

Я это понял . Но как 31 солдат помещаются при 28 посадочных мест . Может "Fire support team"  перемещаются на отдельном транспорте хаммер ?

[akd]

To accomodate mounting additional combat enablers (e.g. FO, or even organic RTO), the squads must operate understrength.  See from pg. 27 on:

http://www.rand.org/content/dam/rand/pubs/research_reports/RR100/RR184/RAND_RR184.pdf

 

In fact, a platoon’s assigned Bradleys have insufficient space to carry all the members of the platoon.45 

 

Edited December 31, 2015 by akd

[TheForwardObserver]

The optimal number of dismounts with gear in a Bradley is 4.  The maximum number of dismounts in a Bradley is limited only by the spacial creativity of the platoon involved, and can exceed 9.  This clown-car capacity is achieved primarily by depriving dismounts of elbow room but also involves the clever sourcing of additional cargo space from locations deep within these graceful chariots.  For example there is a narrow escape corridor between the driver and the troop carrying compartment referred to as the "hell hole."  Cheeky infantry platoons will use this compartment to store attachments such as Forward Observers and Medics.  

[panzersaurkrautwerfer]

What the FO said.  I'd claim six was about within the realm of "functional" with many platoons having the squad leader dual hat as both the vehicle commander and dismounted squad leader to achieve the 7 dismounts per track.  Also many units will simply be under-strength by a soldier or two so often positions for the medic and FO are gained because 2nd squad is only 7 guys instead of the desired 9.