it's very sad to see people are mocking someone with a 1070 expecting to get descent fps. 1070 is a good card running every AAA title at 60 fps in 2020 with high settings. But when someone expects the same for MSFS, the arrogant people of avsim with their 2080ti mocks him. 

1 minute ago, cepact said:

it's very sad to see people are mocking someone with a 1070 expecting to get descent fps. 1070 is a good card running every AAA title at 60 fps in 2020 with high settings. But when someone expects the same for MSFS, the arrogant people of avsim with their 2080ti mocks him. 

thank you, and especially when Microsoft itself recommends GTX 970 with 4GB "as recommended" .. I thought 1070 with 8GB should be a breeze.. especially because i always fly all my sims at medium settings.. I never max out anything.. I always leave headroom to keep the sim fluid.. 

1 hour ago, devgrp said:

That's bs. My system is a gtx 1070, 16gb ram, i7 4770k and 4k monitor. With the latest alpha or beta, I get around 35 fps on high settings. I have trees on medium and lens flare off. Only time my frames dip is in dense areas like lax and NY with lots of ai traffic and I think they are still optimizing 

That's my current PC spec, but only 1080x1920 here, that's really good to read, thank you. New PC inc when 3000's come out. Until then 🙂 

Edited July 30, 20205 yr by Nyxx

yes, that's good news if you can manage steady 35 fps at high settings at 1080p with a GTX1070.. I can always go and get a RTX2080 but if the sim stutters at 10 fps with 1070, throwing a RTX2080 isnt gonna solve the problem.. but it looks like most streamers had maxed out their settings and the sim itself needs a few iterations before things settle down.. so its not entirely plug and play.. we need to spend sometime time tweaking it to get acceptable fps.. that's okay.. as long as we set the right expectations.. i was thinking i'll just treat it like a AAA title.. where you select a profile (High or Ultra) and just forget about it.. maybe its not like that and needs tweaking like we did with p3d or xp11.. 

15 minutes ago, cepact said:

it's very sad to see people are mocking someone with a 1070 expecting to get descent fps. 1070 is a good card running every AAA title at 60 fps in 2020 with high settings. But when someone expects the same for MSFS, the arrogant people of avsim with their 2080ti mocks him. 

Dude, you can see my specs, I've got a 1080, not a 2080ti. 

I'm not mocking having a 1070, I'm strongly advising anyone with a 1070 and strong expectations to get 60fps at high settings in ALL AREAS in a world-representing sim of this visual quality should probably give up on those until drivers and other optimizations have caught up.

Edited July 30, 20205 yr by scotchegg

Something to keep in mind: In the arstechnica article it was stated that CPU usage optimization will still need to be improved after release.

I'm puzzled that "performance" was never shown as an issue in all the feedback snapshots week after week.. that buoyed my optimism about the sim.. made me think those days worrying about fps are over.. oh well..

Not sure where you got your information, but the guy who created the IGN videos posted in reddit his setup:

"Also, my computer specs are:

Asus Crosshair VI HERO, Ryzen 2700X, 32GB Corsair DDR4 Vengeance LPX Red and Zotac GeForce RTX 2070 SUPER

I swapped between Ultra and High settings at 1440p, usually maintained a constant 60FPS but there were dips at times when flying through dense cities and other similar environments."

He also notes he set to vsync for recording purposes. Other YouTubers have also said the game plays higher than 30fps but recording takes a lot out of it. 

The negativity on these forums is insane! 

Looking at my settings, I have a GTX 970 card, and have 32 GB of RAM.  Do you think that will be enough to run MFS?  Should I look at upgrading?

1 minute ago, vin747 said:

I'm puzzled that "performance" was never shown as an issue in all the feedback snapshots week after week.. that buoyed my optimism about the sim.. made me think those days worrying about fps are over.. oh well..

Same here. And it amazes me how long it took before someone started talking about performance. Looks like everyone was so in hype that all those stuttery clips were just ignored.

So now we are going to be put back to reality. Sad but true!

2 minutes ago, The_sas_man said:

The negativity on these forums is insane! 

Back a few years ago, on the tail end of my flight sim days, I had to stop visiting the avsim forums because I could not take the negativity anymore. It hasn't changed that much, but somehow age has helped me tune out when needed 😉

I'd be amazed after being in development for 4 years that they are still struggling to optimise the sim for CPU usage.

Edited July 30, 20205 yr by AllRed

12 minutes ago, ThomseN_inc said:

Same here. And it amazes me how long it took before someone started talking about performance. Looks like everyone was so in hype that all those stuttery clips were just ignored.

So now we are going to be put back to reality. Sad but true!

Again, if you watch videos people are saying the stutter is there due to recording. If you have a good set up, you'll hit more than 30fps. 

A 1070 card, as good as it is, is almost two generations old now. I suspect playhi9ng this on the new 3XXX series card and you'll be fine. 

1 hour ago, vin747 said:

What bugs me the most?

1. Was expecting AAA eye candy and performance.. only got the eye candy.. seems like a slideshow from the various videos and the arstechnica article.. was hoping to throw an i7/GTX1070 at it on medium settings and get smooth 60 fps at all places.. doesn't look like we found the holy grail yet.. this is FSX all over again..

2. ASOBO mentioned that the sim was so well optimized that it was barely utilizing 5% of the CPU/GPU in some article.. doesn't look like that at all.. 

3. The long pauses that i see are probably from scenery streaming, decompressing etc. maybe i should see an offline landing to compare.. 

4. Flight model behavior (and sounds) are pretty bad for the tubeliners.. knew it wont be study level, but at least expected aerofly fs2 level...

5. I get much more fluid FPS in P3D V5 & XP11 than this.. Shouldn't a next-gen platform easily beat those old sim engines?

I'm so glad i didn't pre-order this sim.. will wait to see how PMDG 777 performs at moderate settings from an established streamer who knows how to fly tubeliners. 

It is still a great sim for VFR, no doubt about it.. but was hoping it can replace my P3D V5 for the widebodies.. doesn't look that's gonna happen anytime soon.. so do I continue investing in P3D or wait for MSFS to catch up?

Have you actually played the game to know what performance you will be able to get? An i7 and a 1070 should be no problem at med settings. I play on 4k at high settings and performance is good. Also remember the game is still being optimized 

59 minutes ago, WestAir said:

Alan, I always thought there was - with all things - a quantifiable tangible limit to the rate at which our brains interpret data. Surely the axons and dendrites, with their 200mph electro-chemical relay speed limit, can only change the frequency of the pixel each rod or cone is attempting to communicate, a certain tangible amount of times per second?

I always thought - and I've been wrong before - that the tangible limit of "this cone cell is seeing blue. No wait now it's green" can only be changed about 120 times per second with decreasing accuracy and resolution the further you get from the center. It's a figure and concept I learned in high school, but I know American school isn't the greatest.

Cheers.

Whilst it is true that everything has a limit (except the stupidity of youtube comments, which appear to have no discernible limit), the difference is that for video, there are basically two ways it can be done. Historically it was tied to the frequency of the electrical circuits for TV transmission of images, which is why we ended up with US TV being different to almost everywhere else. But this aside, essentially you have two methods, the older Interlaced method and the more modern Progressive method. Interlaced transmits half of the image per frame, which is why you can sometimes take a picture of a TV with a fast shutter camera and it will only show half the picture on the screen, Progressive transmits each full frame. So Progressive has less artifacts and tearing, but both these methods use an essentially mechanical process where the perceived motion stream is broken up into frames.

Since the eye also transmits electrical impulses to your brain, it is similar in some ways, but it is different in that there is no screen in your head, so there is no frame, it's a perception of multiple images and composites of parts of what is seen too, which so far as we know, is the result of electrical impulses in your brain being interpreted. This means that the information is transmitted at 186,000 miles per second, but not to a screen from top to bottom, or at a frame rate, so there is no individual image frame, it is a continuous stream of conscious awareness of the info. A good example of how we know this is true, is that if you actually concentrate on what your eyes are seeing, you will notice that you can see your nose, but you don't go around all day being aware of your nose being a blurry presence at the edge of each eye's field of view, no, your brain tunes that out most of the time. So we know that what you see is a perception, rather than a pure replication of the data it gets. But your eye does even more than this...

Consider the phenomenon of Saccades, which everyone does even though they are usually unaware of it. If you don't know what saccades is, read this, as it is a fundamental difference between how we as humans see things and interpret them and how a camera or computer screen depicts things.

So, imagine I place an object on a table which you have never seen before, let's say it is a small cardboard box which a product came in. Now never having seen it before, you do not know exactly how big it is, but let's imagine there are some other objects spread out on the table with which you are familiar, perhaps a pencil, a set of keys and a few coins. Now you know how big coins, keys and pencils are, so what your eye does (directed by your brain which is querying the new object) is rapidly flick about, scanning all the familiar objects, a bit like how you see the Terminator's point of view in those movies, scanning about and flashing up data about each item.

So your brain goes: keys - approximate length two inches; pencil - approximate length six inches; coin - ten pence piece, approximate diameter one inch. Now from this, it can determine the approximate size of the previously unseen cardboard box, by comparing its scale and position to the known objects, and because you have stereoscopic vision, it can also determine the range to this object by triangulating it. This is why even though you've never seen the object before and have no previous information about it, you can still reach out and accurately gauge when to clasp your fingers together to pick it up. We know this is true because we can perform the same test on young children who, without a fund of knowledge to draw upon, may not be able to compare coins, keys and pencil to the new object and accurately gauge its size. This is why if you talk to a young toddler face to face and the toddler reaches out to touch you, the child will often misjudge the distance to your face, or to an unfamiliar object you may be showing the child.

So because of that Saccades stuff, we know that humans do not project single images in their head like a projector does with film. 

Interestingly, this  phenomenon caused a problem for the designers of the Gemini, Mercury and Apollo capsules. It had been supposed that astronauts strapped down in their spacecraft during launch, and subjected to high G forces, might not be able to reach control switches easily or accurately, particularly when wearing space suit gloves. So what the control panel designers were actually considering, was trying to create controls which could be operated simply by looking at them. I think it worked off reflecting scanned light into a pupil, and they did get this method to function, but what they could not stop, was astronauts glancing about owing to this saccades phenomenon, since everyone does it and usually it is involuntary, which meant that the astronauts would have been constantly operating switches they did not intend to, so they had to scrap the idea and go with placing panels within easy reach which had conventional switches.

Edited July 30, 20205 yr by Chock